Graduate Statistics in Excel Manual 1 S

Manual 1 of 4
Practical and Clear Graduate Statistics

in Excel
Normal Distribution
t-Distribution
Binomial Distribution
Combinations & Permutations
The Excel Statistical Master

(that’ll be you!)
By Mark Harmon
Copyright © 2014 Mark Harmon
No part of this publication may be reproduced
or distributed without the express permission
of the author.
mark@ExcelMasterSeries.com
ISBN: 978-1-937159-20-7
Table of Contents
Histograms and Basic Graphing in Excel

Histograms in Excel................................................................................ 17
Overview .................................................................................................................... 17
Steps To Take Before Creating a Histogram .............................................................. 18
a) Sorting the Data ................................................................................................ 20
b) Standardizing the Data ..................................................................................... 21
c) Creating the Bins .............................................................................................. 23
The Two Methods To Create a Histogram in Excel .................................................... 24
Creating a Histogram With the Excel Histogram Data Analysis Tool.......................... 25
Creating a Histogram With Excel Formulas and a Bar Chart ..................................... 27
Step 1 – Highlight the Y Data................................................................................ 29
Step 2 – Select Chart Type ................................................................................... 30
Step 3 – Insert X-Axis Data ................................................................................... 32
Step 4 – Delete the Legend ................................................................................... 36
Step 5 – Insert Chart Title ..................................................................................... 37
Step 6 – Insert X-Axis Label ................................................................................. 40
Step 7 – Insert Y-Axis Label ................................................................................. 42
Scatterplot Graph in Excel ..................................................................... 44

Overview .................................................................................................................... 44
Step 1 – Highlight the X-Y Data and Labels ........................................................ 45
Step 2 – Select the Chart Type ............................................................................. 46
Step 3 – Select the Chart Layout.......................................................................... 47
Step 6 – Highlight the Legend and Delete It ........................................................ 49
Step 7 – Insert the Chart Title ............................................................................... 51
Step 8 – Configure the X and Y Axis Labels ....................................................... 55
Bar Chart in Excel ................................................................................... 58

Overview .................................................................................................................... 58
Step 1 – Highlight the Y Data................................................................................ 59
Step 2 – Select Chart Type ................................................................................... 60
Step 3 – Insert X-Axis Data ................................................................................... 62
Step 4 – Delete the Legend ................................................................................... 66
Step 5 – Insert Chart Title ..................................................................................... 67
Step 6 – Insert X-Axis Label ................................................................................. 70
Step 7 – Insert Y-Axis Label ................................................................................. 72
Combinations and Permutations in Excel............................................. 74

Combinations Overview ............................................................................................. 74
Permutations Overview .............................................................................................. 74
Simultaneous vs. Sequential Selection ............................................................... 75
Excel Functions For Combinations ............................................................................. 75
Excel Functions For Permutations.............................................................................. 76
Combination Problems in Excel ................................................................................. 78
Problem 1: Combinations of Investment Proposals .......................................... 78
Problem 2: Combinations of Newly-Opened Offices.......................................... 78
Problem 3: Multiple Combinations of Newly-Opened Offices ........................... 78
Problem 4: Combinations of Committees ........................................................... 79
Problem 5: Combinations of Sub-Groups ........................................................... 79
Permutation Problems in Excel: ................................................................................. 80
Problem 6: Permutations of Delivery Routes...................................................... 80
Problem 7: Permutations of Seating Arrangements .......................................... 80
Problem 8: Permutations of Executive Groups .................................................. 80
Problem 9: Permutations of Book Arrangements .............................................. 81
Problem 10: Permutations of Letter Groups ....................................................... 81
Normal Distribution in Excel .................................................................. 82

Overview .................................................................................................................... 82
Uses of the Normal Distribution .................................................................................. 83
Useful Normal Approximations of Other Distributions ................................................ 84
Normal Approximation of the Binomial Distribution .......................................... 84
Normal Approximation of the Poisson Distribution ........................................... 84
Statistical Tests That Require Normality of Data or Residuals ................................... 85
History of the Normal Distribution ............................................................................... 85
Normal Distribution’s PDF (Probability Density Function) .......................................... 87
Example in Excel ................................................................................................... 88
The Normal Distribution’s CDF (Cumulative Distribution Function) ............................ 90
Example in Excel ................................................................................................... 93
Problems Using NORM.DIST() To Calculate F(X,µ,σ) ............................................... 95
Problem 1 ............................................................................................................... 95
Problem 2 ............................................................................................................... 95
Problems Using NORM.INV() To Calculate X Given F(X,µ,σ) .................................... 97
Problem 1 ............................................................................................................... 97
Problem 2 ............................................................................................................... 97
Standard Normal Distribution ..................................................................................... 99
Calculating the PDF of a Standardized Normal Distribution ........................... 101
Calculating the CDF of a Standardized Normal Distribution ........................... 102
The t-Distribution’s Convergence to the Standard Normal Distribution ........ 103
Properties of the Normal Distribution........................................................................ 106
Calculating How Much of the Data Is a Certain Distance From the Mean ................ 108
For Normally-Distributed Data, Use the Empirical Rule ................................... 108
Empirical Rule Problem in Excel ................................................................................................ 108
For Data of Unknown Distribution, Use Chevbyshev’s Theorem.................... 108
Chebyshev’s Theorem Problem in Excel ................................................................................... 108
The Central Limit Theory in Excel ....................................................... 109

Overview .................................................................................................................. 109
Why the Central Limit Theorem Is So Important to Statistics ................................... 109
A Demonstration of the Central Limit Theorem Using Excel .................................... 110
Population (Sample size n = 1) .......................................................................... 114
Sample Size n = 2 ................................................................................................ 116
Sample Size n = 5 ................................................................................................ 118
Sample Size n = 10 .............................................................................................. 120
The Resulting Robustness of the t-Test ................................................................... 121
Derivation of the Central Limit Theorem ................................................................... 121
Hypothesis Tests Using the Normal Distribution in Excel ................. 122

Overview .................................................................................................................. 122
Null Hypothesis ........................................................................................................ 122
The Null Hypothesis is Either Rejected or Not Rejected But Is Never Accepted
.............................................................................................................................. 122
Alternative Hypothesis.............................................................................................. 123
One-Tailed Test vs. a Two-Tailed Test .................................................................... 123
Level of Certainty ..................................................................................................... 124
Level of Significance (Alpha) .................................................................................... 124
Region of Acceptance .............................................................................................. 124
Region of Rejection .................................................................................................. 124
Critical Value(s) ........................................................................................................ 125
Test Statistic ............................................................................................................. 125
Critical t Value or Critical z Value ............................................................................. 125
Relationship Between p Value and Alpha ................................................................ 126
Critical z Values........................................................................................................ 126
Critical z Value for a one-tailed test in the right tail: ........................................ 126
Critical z Value for a one-tailed test in the left tail:........................................... 126
Critical z Values for a two-tailed test: ................................................................ 127
p Value ..................................................................................................................... 127
The 3 Equivalent Reasons To Reject the Null Hypothesis ....................................... 127
Independent Samples vs. Dependent Samples ....................................................... 128
Pooled vs. Unpooled Tests ...................................................................................... 128
Type I and Type II Errors .......................................................................................... 128
Power of a Test ........................................................................................................ 128
Effect Size ................................................................................................................ 128
Nonparametric Alternatives ...................................................................................... 129
Hypothesis Test of Mean vs. Proportion................................................................... 129
Hypothesis Tests of Mean – Basic Definition ................................................... 129
Hypothesis Tests of Proportion – Basic Definition .......................................... 129
Hypothesis Tests of Mean ........................................................................................ 130
t-Test versus z-Test............................................................................................. 130
Normal Distribution of Means of Large Samples.............................................. 130
Requirements for a z-Test .................................................................................. 130
Requirements for a t-Test ................................................................................... 130
Basic Steps of a Hypothesis Test of Mean ....................................................... 131
Uses of Hypothesis Tests of Mean .................................................................... 133
Types of Hypothesis Tests of Mean .................................................................. 133
1) One-Sample z-Test in Excel ............................................................. 134
Overview .................................................................................................................. 134
Example of a One-Sample, Two-Tailed z-Test in Excel ........................................... 134
Summary of Problem Information...................................................................... 135
Question1) Type of Test? ................................................................................... 136
a) Hypothesis Test of Mean or Proportion? .............................................................................. 136
b) One-Sample or a Two-Sample Test? ..................................................................................... 136
c) Independent (Unpaired) Test or a Dependent (Paired) Test? .............................................. 136
d) One-Tailed or Two-Tailed Hypothesis? ................................................................................. 136
e) t-Test or z-Test? ....................................................................................................................... 137
Question 2) Test Requirements Met? ............................................................... 137
a) Normal Distribution of Test Statistic...................................................................................... 137
1) Population standard deviation, σ, is known ............................................................................ 137
2) Sample size is large (n > 30) .................................................................................................. 137
Step 1 – Create the Null and Alternative Hypotheses ...................................... 138
Step 2 – Map the Distributed Variable to Normal Distribution......................... 139
Step 3 – Map the Regions of Acceptance and Rejection ................................. 140
Calculate Critical Values ............................................................................................................. 140
Step 4 – Determine Whether to Reject Null Hypothesis ................................... 141
1) Compare x-bar With Critical Value ......................................................................................... 141
2) Compare the t Value With Critical t Value ............................................................................. 141
3) Compare the p Value With Alpha ........................................................................................... 142
Excel Formula Shortcut to Performing a One-Sample z-Test.......................... 144
2) Two-Independent-Sample, Unpooled z-Test in Excel .................... 145

Overview .................................................................................................................. 145
Example of 2-Sample, 2-Tailed, Unpooled z-Test in Excel ...................................... 146
Question 1) What Type of Test Should Be Done? ............................................ 147
b) One-Sample or Two-Sample Test? ........................................................................................ 147
c) Independent (Unpaired) Test or Dependent (Paired) Test? ................................................. 147
) One-Tailed or Two-Tailed Test? ............................................................................................... 147
e) t-Test or z-Test? ....................................................................................................................... 147
f) Pooled or Unpooled t-Test? ..................................................................................................... 148
Question 2) Test Requirements Met? ................................................................ 148
a) Normal Distribution of Both Sample Means .......................................................................... 148
1) Both Population Standard Deviations, σ1 and σ2, Are Known .............................................. 148
2) Both samples sizes are large (n > 30). ................................................................................... 148
Step 2 – Map the Distributed Variable on a Normal Distribution Curve ......... 150
Calculate the Critical Values ....................................................................................................... 151
Two-Tailed Critical Values .......................................................................................................... 151
1) Compare Sample Mean With Critical Value ........................................................................... 152
2) Compare the z Score with the Critical z Value ...................................................................... 153
Excel Data Analysis Tool Shortcut .................................................................... 155
3) Paired (Two-Sample Dependent) z-Test in Excel ........................... 157

Overview .................................................................................................................. 157
Example of Paired, 1-Tailed (Left-Tail) z-Test in Excel ............................................ 157
d) One-Tailed or Two-Tailed Test? ............................................................................................. 160
e) t-Test or z-Test? ....................................................................................................................... 160
a) Test Statistic Distributed According to Normal Distribution .............................................. 160
Step 2 – Map Distributed Variable to Normal Distribution Curve .................... 162
Calculate the Critical Value ......................................................................................................... 163
1) Compare x-bardiff, with Critical Value ..................................................................................... 164
2) Compare z Score with Critical z Value ................................................................................... 165
3) Compare p Value to Alpha. ..................................................................................................... 165
z-Based Confidence Intervals of a Population Mean in Excel ........... 168

Overview .................................................................................................................. 168
Example of a Normal-Distribution-Based Confidence Interval in Excel .................... 169
Question 1) What Type of Confidence Interval? ............................................... 170
a) Confidence Interval of the Population Mean or Proportion? .............................................. 170
b) Based on t-Distribution or Based on Normal Distribution? ................................................ 170
a) Normal Distribution of Sample Mean ..................................................................................... 170
b) Population Standard Deviation Is Known ............................................................................. 170
Step 1) Calculate Half-Width of Confidence Interval ........................................ 171
Step 2 Confidence Interval = Sample Mean ± C.I. Half-Width ......................... 171
Min Sample Size to Limit Width of Confidence Interval ............................................ 173
Requirements ...................................................................................................... 173
1) Minimum Sample Size is 30 .................................................................................................... 173
2) Population Standard Deviation Known.................................................................................. 173
Example in Excel ................................................................................................. 174
t-Distribution in Excel ........................................................................... 175

Overview .................................................................................................................. 175
History of the t-Distribution ....................................................................................... 177
Properties of the t-Distribution .................................................................................. 178
t-Distribution’s PDF .................................................................................................. 179
Example in Excel With Sample Data .................................................................. 181
The t-Distribution’s CDF ........................................................................................... 187
Example With Sample Data ................................................................................ 188
A Big Difference Between t and Normal Distributions .............................................. 191
Normal Distribution ............................................................................................. 192
t-Test: t-Distribution-Based Hypothesis Test ............................................................ 193
Null Hypothesis ................................................................................................... 193
Null Hypothesis - Rejected or Not But Never Accepted ................................... 193
Alternative Hypothesis ....................................................................................... 194
One-Tailed Test vs. Two-Tailed Test ................................................................. 194
Level of Certainty ................................................................................................ 195
Level of Significance (Alpha) ............................................................................. 195
Region of Acceptance ......................................................................................... 195
Critical Value(s) ................................................................................................... 196
Test Statistic ........................................................................................................ 196
Critical t Value or Critical z Value ...................................................................... 196
p Value ................................................................................................................. 197
Critical t Value For 1-Tailed Test in Right Tail: ................................................. 197
Critical t Value For 1-Tailed Test in Left Tail: .................................................... 197
Critical t Values For a 2-Tailed Test:.................................................................. 198
3 Equivalent Reasons To Reject Null Hypothesis ............................................ 198
1) Sample Statistic Beyond Critical Value ................................................................................. 198
2) Test Statistic Beyond Critical t or z Value ............................................................................. 198
3) p Value Smaller Than α (1-Tailed) or α/2 (2-Tailed) .............................................................. 198
Independent vs. Dependent Samples ................................................................ 198
Pooled vs. Unpooled Tests ................................................................................ 199
Type I and Type II Errors .................................................................................... 199
Power of a Test .................................................................................................... 199
Effect Size ............................................................................................................ 199
Nonparametric Alternatives for t-Tests in Excel............................................... 199
Hypothesis Test of Mean vs. Proportion ........................................................... 200
Hypothesis Tests of Mean – Overview ....................................................................................... 200
Hypothesis Tests of Proportion – Overview ............................................................................. 200
Hypothesis Tests of Mean .................................................................................. 200
t-Test vs. z Test ................................................................................................... 200
Means of Large Samples Are Normally Distributed ......................................... 201
Requirements of a z-Test.................................................................................... 201
Requirements of a t-Test .................................................................................... 201
Basic Steps of a Hypothesis Test of Mean ....................................................... 202
Uses of Hypothesis Tests of Mean .................................................................... 204
Types of Hypothesis Tests of Mean .................................................................. 205
1) One-Sample t-Test in Excel .............................................................. 206

Example of a 1-Sample, 2-Tailed t-Test in Excel ..................................................... 206
Question 1) Type of Test? .................................................................................. 208
c) Independent (Unpaired) Test or a Dependent (Paired) Test? .............................................. 208
d) One-Tailed or Two-Tailed Hypothesis? ................................................................................. 208
e) t-Test or z-Test? ....................................................................................................................... 208
a) t-Distribution of Test Statistic ................................................................................................. 209
When Sample Size Is Large ....................................................................................................... 209
When Sample Size is Small ....................................................................................................... 209
Evaluating the Normality of the Sample Data ........................................................................... 210
Nonparametric Alternatives in Excel ................................................................. 211
Step 1 – Create the Null and Alternate Hypotheses ......................................... 212
Step 2 – Map the Distributed Variable to t-Distribution ................................... 213
Calculate Critical Values ............................................................................................................. 215
1) Compare the x-bar With Critical Value................................................................................... 216
2) Compare the t Value With Critical t Value ............................................................................. 216
Excel Shortcut to Performing a One-Sample t-Test ......................................... 219
2) Two-Independent-Sample, Pooled t-Test in Excel ......................... 220

Example of 2-Sample, 1-Tailed, Pooled t-Test in Excel ........................................... 221
c) Independent a Dependent (Paired) Test? .............................................................................. 224
e) t-Test or as a z-Test? ............................................................................................................... 224
Question 2) Requirements Met? ........................................................................ 224
1) Sample Size of Both Samples Greater Than 30 .................................................................... 225
2) Both Populations Are Normally Distributed ............................................................................ 225
3) Both Samples Are Normally Distributed ................................................................................. 225
Evaluating the Normality of the Sample Data ............................................................................. 225
b) Similarity of Sample Variances .............................................................................................. 228
c) Independence of Samples ....................................................................................................... 228
Nonparametric Alternatives ........................................................................................................ 228
Step 1 – Create Null and Alternate Hypotheses................................................ 229
Step 2 – Map Distributed Variable on a t-Distribution Curve ........................... 230
One-Tailed Critical Values .......................................................................................................... 232
1) Compare the Sample Mean With Critical Value .................................................................... 233
2) Compare t Value With Critical t Value .................................................................................... 234
Excel Statistical Function Shortcut ................................................................... 241
3) Two-Independent-Sample, Unpooled t-Test in Excel ..................... 242

Example of 2-Sample, 2-Tailed, Unpooled t-Test in Excel ....................................... 243
e) t-Test or z-Test? ....................................................................................................................... 246
Nonparametric Alternatives in Excel ........................................................................................... 247
b) Significantly Different Sample Variances .............................................................................. 249
Nonparametric Alternatives ........................................................................................................ 249
Step 2 – Map the Distributed Variable on a t-Distribution Curve .................... 251
One-Tailed Critical Value ............................................................................................................ 254
1) Compare Sample Mean With Critical Value ........................................................................... 254
4) Paired (Two-Sample Dependent) t-Test in Excel ............................ 264

Overview .................................................................................................................. 264
Example of Paired, 1-Tailed t-Test in Excel ............................................................. 264
e) t-Test or z-Test? ....................................................................................................................... 267
a) Test Statistic Distributed According to t Distribution .......................................................... 268
When Difference Sample Size Is Large...................................................................................... 268
When the Difference Sample Size is Small ................................................................................ 268
1) Population of Paired Differences Is Normally Distributed ...................................................... 269
2) Difference Sample Is Normally Distributed ............................................................................ 269
Evaluating Normality of the Difference Sample .......................................................................... 269
b) Significantly Different Sample Variances .............................................................................. 270
Nonparametric Alternatives in Excel.......................................................................................... 270
Step 2 – Map Distributed Variable to t Distribution Curve ............................... 272
Calculate the Critical Value ......................................................................................................... 275
1) Compare x-bardiff With Critical Value ..................................................................................... 276
3) Compare p Value With Alpha .................................................................................................. 276
t-Based Confidence Intervals of a Population Mean .......................... 284

Overview .................................................................................................................. 284
Example of a t-Distribution-based Confidence Interval............................................. 284
a) Normal Distribution of Sample Mean ..................................................................................... 288
1) Sample Size of Each Sample Greater Than 30 ..................................................................... 289
2) Population Normally Distributed ............................................................................................. 289
3) Sample Normally Distributed .................................................................................................. 289
Evaluating Normality of Sample Data ......................................................................................... 289
Step 2 Confidence Interval = Sample Mean ± C.I. Half-Width ......................... 291
Excel Shortcut For t-Based Confidence Interval .............................................. 292
Requirements ...................................................................................................... 294
1) Minimum Sample Size is 30 .................................................................................................... 294
2) Population Standard Deviation Known.................................................................................. 294
Example ............................................................................................................... 295
Binomial Distribution in Excel ............................................................. 296

Overview .................................................................................................................. 296
Each Trial is a Bernoulli Trial ............................................................................. 297
Maximum Sample Size Rule ............................................................................... 297
Binomial Distribution Has Two Parameters n and p ........................................ 298
Population Parameters of the Binomial Distribution ....................................... 298
Binomial Distribution’s PDF and CDF....................................................................... 299
Binomial Distribution’s PDF - Probability Density Function ........................... 299
Binomial Distribution’s CDF - Cumulative Distribution Function ................... 302
Problems Solved With the Binomial Distribution in Excel ......................................... 304
Problem 1 – Solving With the Binomial’s PDF in Excel ........................................................... 304
Problem 2 – Solving With the Binomial’s CDF in Excel ........................................................... 304
Problem 3 – Calculating a Range of Binomial Probabilities in Excel ..................................... 304
Problem 4 – Sampling With and Without Replacement in Excel ............................................. 305
Normal Approximation of the Binomial Distribution .................................................. 307
Continuity Correction ................................................................................................ 308
Hypothesis Testing on Binomial Data ................................................. 309

Overview .................................................................................................................. 309
Null Hypothesis ........................................................................................................ 309
Null Hypothesis is Either Rejected or Not Rejected But Is Never Accepted .. 309
Alternative Hypothesis.............................................................................................. 310
One-Tailed Test vs. a Two-Tailed Test .................................................................... 310
Level of Certainty ..................................................................................................... 311
Level of Significance (Alpha) .................................................................................... 311
Region of Acceptance .............................................................................................. 311
Region of Rejection .................................................................................................. 311
Critical Value(s) ........................................................................................................ 312
Test Statistic ............................................................................................................. 312
Critical t Value or Critical z Value ............................................................................. 312
Relationship Between p Value and Alpha ................................................................ 313
The 3 Equivalent Reasons To Reject the Null Hypothesis ....................................... 313
Type I and Type II Errors .......................................................................................... 313
Power of a Test ........................................................................................................ 314
Effect Size ................................................................................................................ 314
Hypothesis Test of Mean vs. Proportion................................................................... 314
Uses of Hypothesis Tests of Proportion ................................................................... 317
Types of Hypothesis Tests of Proportion.................................................................. 317
1) One-Sample Hypothesis Test of Proportion in Excel..................... 318

Overview .................................................................................................................. 318
Example of a One-Sample, Two-Tailed Hyp. Test of Proportion in Excel ................ 319
Null and Alternative Hypotheses ................................................................................................ 319
Standard Error .............................................................................................................................. 319
Critical Values ............................................................................................................................... 320
Region of Rejection...................................................................................................................... 320
z Value ........................................................................................................................................... 320
Critical z Values ............................................................................................................................ 320
p Value ........................................................................................................................................... 320
Determine Whether to Reject Null Hypothesis ................................................. 321
1) Compare p_bar with Critical Value ......................................................................................... 321
2) Compare z Value with Critical z Value ................................................................................... 321
3) Compare the p Value to Alpha ................................................................................................ 321
2) Two-Sample, Pooled Hypothesis Test of Proportion ..................... 322

Example of a Two-Sample, Two-Tailed Pooled Hyp. Test of Proportion in Excel .... 323
ppooled .............................................................................................................................................. 324
Standard Error .............................................................................................................................. 324
Critical Values ............................................................................................................................... 324
z Value ........................................................................................................................................... 324
Critical z Values ............................................................................................................................ 325
p Value ........................................................................................................................................... 325
1) Compare p_ba2-p_bar1 with Critical Value ............................................................................ 325
3) Two-Sample, Unpooled Hypothesis Test of Proportion ................ 327

Overview .................................................................................................................. 327
Example of a Two-Sample, One-Tailed Unpooled Hyp. Test of Proportion in Excel 328
Standard Error .............................................................................................................................. 329
Critical Values ............................................................................................................................... 329
z Value ........................................................................................................................................... 330
Critical z Value .............................................................................................................................. 330
p Value ........................................................................................................................................... 330
1) Compare p_bar with Critical Value ......................................................................................... 330
Confidence Intervals of a Population Proportion ............................... 332

Overview .................................................................................................................. 332
Example of a Confidence Interval of a Population Proportion in Excel ..................... 333
a) Normal distribution can be used to approximate the binomial distribution ...................... 334
Step 2 Confidence Interval = Sample Mean ± C.I. Half-Width .......................... 336
Example 1 in Excel – Min Number of Voters to Limit Poll Margin of Error ............................ 338
Example 2 in Excel – Min Number of Production Samples to Limit Defect Rate Margin ...... 339
Family of Closely-Related Distributions.............................................. 340

Binomial Distribution................................................................................................. 341
Hypergeometric Distribution ..................................................................................... 343
Multinomial Distribution ............................................................................................ 345
Negative Binomial Distribution ................................................................................. 347
Geometric Distribution .............................................................................................. 349
Check Out the Latest Book in the Excel Master Series! .................... 350
Meet the Author .................................................................................... 355
Advanced Statistical & Solver Optimization Project Consulting

Availability ............................................................................................. 356
Histograms in Excel
Overview
A histogram provides the counts of the number of items in groups after the items have been divided into
groups of similar items. In statistics the items are divided into groups based upon the value of a
continuous variable that can be measured for each item. The value of that variable for each item
determines which group the item will fall into.
Each group only accepts items that have variable values in a specific and unique range. The groups are
sometimes referred to as bins. A histogram provides the counts of the number of items that have fallen
into each bin. Each bin only accepts items within a specific range of variable values.
Histograms are often presented in the form of bar charts. Each bar is associated with a unique bin. The
length of each bar represents the number items in the bin associated with that bar.
Adjacent bars on the histogram bar chart are associated with bins that have adjacent ranges for
accepting items. The histogram bar chart therefore shows how all of the items are distributed or divided
up based on the variable value measured for each item. An example of a histogram is as follows:
This histogram provides count of the number of items that have z Scores in each of the five categories or
bins. The first bin on the right accepted items that have a z Score between 2.5 and 1.5. Only one item has
been observed to have a z Score in that range. The second bin to the left shows that four items have
been observed to have z Scores between 0.5 and 1.5.
The bell-shaped curve of the histogram indicates that the z Score might be normally distributed. A
histogram provides a quick visual estimation of a variable’s distribution.
The preceding histogram was taken from the following data. This histogram shows how the 26 z Scores
in the sample group are distributed. The bell-shaped histogram indicates that these z Score might be
normally distributed. The bell-shaped histogram results from the groups or bins that accept z Scores in
the middle ranges having have significantly higher counts than bins that accept z Scores in the outer
ranges.
An important component of a histogram is establishing the upper and lower boundaries of ranges that
each bin will accept. The following images shows data sample of 26 z Scores, the upper and lower
ranges of each of the five bins, the count of z Scores falling into each bin, and the bar chart histogram
that provides a graphically display of the bin counts.
17
Steps To Take Before Creating a Histogram
Data that will be evaluated with a histogram are often not provided in pre-determined bins. The bins, or
more specifically, the upper and lower boundaries of each of the bins, have not been established. Sorting
and standardizing the data greatly facilitates the development of the bin specifications.
Sorting the data makes the data’s range and any significant outliers apparent. Outliers judged to be
extreme and therefore non-representative of the data can be removed. Each significant outlier should be
evaluated on a case-by-case basis. Outliers that have been removed and the justifications for removal
should be noted.
Standardizing a data value converts that value to its z Score. A z Score is equal to the number of sample
standard deviations that the value is from the sample mean. Standardizing the data allow bin boundaries
to be based upon increments of sample standard deviation which is fairly intuitive and uncomplicated.
18
An example of sorting and standardizing a sample of raw data is performed as follows:
Unsorted Raw Data Sample
19
a) Sorting the Data
The raw data can be sorted using the sorting tool available in Excel. This is effective but an even better
way to sort the data is use the formula shown in the following diagram. Using this formula has the
advantage that the data will be automatically resorted if any of the data are changed. The sorting tool
would need to be reapplied each time any of the data have been changed. The formula can be typed into
the top cell and then quickly copied down to the bottom as follows:
20
b) Standardizing the Data
Standardizing the data simply involves subtracting the mean from the data value and then dividing by the
standard deviation. This calculation converts each data value to its z Score. For population data, the z
Score is the number of population standard deviations that the data value is from the population mean.
For sample data, the z Score is the number of sample standard deviations that a data value is from the
sample mean.
The z Scores in this example are calculated from sample data as follows:
21
The raw data sample has been sorted and standardized to produce the following sorted sample of 26 z
Scores.
22
c) Creating the Bins
Bin creation involves specifying the upper and lower boundaries of each bin into which the values will fall.
Creating optimal bins ranges depends on the skill of the user. The bins ranges should be large enough
that at least several observations are likely to be in each bin but not so large that the shape of the data’s
distribution shape cannot be ascertained because too many observations have fallen into too few bins.
Ultimately there needs to be a sufficiently large number of observations collected to enable the creation of
a reasonably clear distribution shape with a histogram.
The z Scores of the data range from -1.787 to 2.490. The bins should cover that entire range because
there are no significant outliers among the 26 total z Scores. One possible configuration of bins ranges is
the following set of five adjacent bins each having the width equal to one z Score as follows:
23
The Two Methods To Create a Histogram in Excel
There are two ways to create a histogram in Excel. Histograms will be created in this section using both of
these methods.
The first method uses Excel’s built-in Histogram tool that is part of the Data Analysis ToolPak. The Data
Analysis ToolPak ships with most versions of Excel but need to be manual activated by the user before it
is available for use. Like all of the Data Analysis tool, the Histogram tool must be manually rerun to
update the histogram if any of the input data has changed.
The second method combines Excel formulas and a bar chart to create a histogram. This is the preferred
method because the histogram will be automatically updated if any of the input data has changed. Excel
formulas and chart automatically adjust their output when input data is changed.
It is sometimes more efficient to use formulas in place of the Data Analysis tools because the formula
automatically update the output when input data is changed. Substituting formulas for Data Analysis tools
is only a good solution for simple tools such the Histogram tool but not complicated tools such as Two-
Way ANOVA With Replication tool or the Multiple Regression tool.
24
Creating a Histogram With the Excel Histogram Data
Analysis Tool
The Excel Histogram tool requires that the bin ranges be specified by providing only the upper boundary
of each bin. The Histogram tool dialogue box can be accessed in Excel under the Data tab by selecting
Data Analysis / Histogram. The Histogram dialogue box then appears. The data range, bin upper
boundaries, upper left corner of the output, and chart Output checkmark are input into this dialogue box
as follows:
25
The specified histogram output which includes the frequency chart and the histogram bar chart as shown
as follows:
Note that the frequency chart requires the creation of a bin above and below the five bins that are the
target for this histogram analysis. This is not intuitive and usually requires some experimentation to
properly create the five bin ranges that are the target of this analysis. Another shortcoming of the
Histogram tool in the Data Analysis ToolPak is that the histogram must be re-run whenever input data has
changed.
These two shortcomings of the Histogram tool can be overcome by creating the histogram with Excel
formulas and a bar chart as follows:
26
Creating a Histogram With Excel Formulas and a Bar Chart
A histogram ultimately provides the count of observations in each bin. These bin counts can be obtained
using Excel formulas as follows:
27
There are two formulas that will provide the bin counts. The COUNTIFS() formula works in Excel versions
2007 and later. The SUMPRODUCT() formula works in previous Excel versions and the current version
as well.
In this example the following histogram bar chart will be created from the X-Y data in cells I4 to J8 that
contain the bins counts.
28
Step 1 – Highlight the Y Data
Select the data in cells J4 to J8.
29
Step 2 – Select Chart Type
Make the following selections:
Insert tab / Column / 2-D Clustered Column
30
This produces the following basic bar chart:
31
Step 3 – Insert X-Axis Data
Right-click anywhere in the chart and choose Select Data as follows:
32
This brings up the following dialogue box. Under the label Horizontal (Category) Axis Labels on the
right side, select the Edit button.
This brings up a blank Axis Labels dialogue box as follows:
33
Select the X-Axis data in cells I4 to I8 as follows:
This produces the following labels for the X-Axis:
34
Format these labels by right-clicking on the X-Axis, select Bold, and then set the font size in the cell font
dialogue box as follows:
35
This produces the following X-Axis:
Step 4 – Delete the Legend

Select the legend and delete it.
36
Step 5 – Insert Chart Title
Click anywhere on the chart to bring up the Chart Tools menu. Choose the Layout tab / Chart Title /
Above Chart as follows:
37
This creates up the following basic chart title:
38
Highlight the text and type in the correct title as follows:
39
Step 6 – Insert X-Axis Label
Right-click anywhere on the chart to bring up the Chart Tools menu. Choose Layout tab / Axis Titles /
Primary Horizontal Axis / Title Below Axis as follows:
40
Highlight the basic X-Axis label that appears. Type in the correct X-Axis text, select Bold, and select the
correct font size as follows:
41
Step 7 – Insert Y-Axis Label
Primary Vertical Axis / Horizontal Title as follows:
42
After correcting the Y-Axis label by typing in the correct text, making it Bold, and setting the correct font
size as was just done for the X-Axis label, the final bar graph will appear as follows:
43
Scatterplot Graph in Excel
Overview
Only three types of Excel charts are presented in this manual. They are the area chart, the bar chart, and
the scatterplot. There are, of course, many types of charts available in Excel and many variations of each
type. This chapter will provide instructions on creating both the scatterplot and bar chart in Excel.
Instructions to create an interactive area chart of a statistical distribution are presented at the beginning of
the final chapter of this manual which covers graphing statistical distributions in Excel.
A scatterplot is a type of mathematical diagram using Cartesian coordinates to display data points that
are described by two numerical variables. Scatterplots provide a visual presentation of the relationship
between two variables associated with a single data point. Each data point’s value of one variable is its
coordinate location on the horizontal axis. Each data point’s value of the other variable is its coordinate
location on the vertical axis.
Scatterplots are most often used in statistics to determine the following aspects of the relationship
between two variables associated with each data point:
1) Whether the relationship is monotonic, i.e., one variable tends to move in a specific direction when the
other variable moves in a specific direction. An example would be a data point’s X value that tends to rise
as Y values rise.
2) Whether the two variables have a positive correlation or a negative correlation if a monotonic
relationship is shown to exist. Variables that have a positive correlation move in the same direction, i.e.,
when one increases, the other generally increases as well. A negative correlation occurs if variables tend
to move in opposite directions, i.e., when one increases, the other tends to decrease.
3) Whether the relationship between the two variables is linear or non-linear.
In this example this following scatterplot chart will be created from the X-Y data in cells B2 to C9.
44
Step 1 – Highlight the X-Y Data and Labels
45
Step 2 – Select the Chart Type
Insert tab / Scatter / Scatter with only markers
The following basic scatterplot chart will appear:
46
Step 3 – Select the Chart Layout
Click on the chart to bring up the Chart Tools menu / Design / Layout 3 of Chart Layouts
47
Layout 3 creates a Least-Square Line on the scatterplot chart as follows:
48
Step 6 – Highlight the Legend and Delete It
49
Hitting Delete removes the legend and leaves the following:
50
Step 7 – Insert the Chart Title
Click on the chart to bring up the Chart Tools menu / Layout / Chart Title / Above Chart
51
This brings up the following title above the chart:
52
Highlight the text in the title and type in the correct text as follows:
53
Resize the grid by clicking on it. The grid will automatically resize to the following:
54
Step 8 – Configure the X and Y Axis Labels
Highlight the text in each label. Type in the correct text. Right-click on the highlighted text and select the
correct font size.
55
The complete scatterplot chart appears as follows:
56
57
Bar Chart in Excel
Overview
Only three types of Excel charts are presented in this manual. They are the area chart, the bar chart, and
the scatterplot. There are, of course, many types of charts available in Excel and many variations of each
type. This chapter will provide instructions on creating both the scatterplot and bar chart in Excel.
Instructions to create an interactive area chart of a statistical distribution are presented at the beginning of
the final chapter of this manual which covers graphing statistical distributions in Excel.
A bar chart is a chart with rectangular bars with lengths proportional to the values that they represent.
Bars can be single, grouped, or stacked. Bar can be horizontal or vertical. This section will discuss how to
create a bar chart that presents single, vertical bars. Single bars present the value of one variable that is
associated with each data observation.
In this example this following bar chart will be created from the X-Y data in cells I4 to J8.
58
Step 1 – Highlight the Y Data
Select the data in cells J4 to J8.
59
Step 2 – Select Chart Type
Insert tab / Column / 2-D Clustered Column
60
This produces the following basic bar chart:
61
Step 3 – Insert X-Axis Data
Right-click anywhere in the chart and choose Select Data as follows:
62
This brings up the following dialogue box. Under the label Horizontal (Category) Axis Labels on the
right side, select the Edit button.
This brings up a blank Axis Labels dialogue box as follows:
63
Select the X-Axis data in cells I4 to I8 as follows:
This produces the following labels for the X-Axis:
64
Format these labels by right-clicking on the X-Axis, select Bold, and then set the font size in the cell font
dialogue box as follows:
65
This produces the following X-Axis:
Step 4 – Delete the Legend

Select the legend and delete it.
66
Step 5 – Insert Chart Title
Click anywhere on the chart to bring up the Chart Tools menu. Choose the Layout tab / Chart Title /
Above Chart as follows:
67
This creates up the following basic chart title:
68
Highlight the text and type in the correct title as follows:
69
Step 6 – Insert X-Axis Label
Primary Horizontal Axis / Title Below Axis as follows:
70
Highlight the basic X-Axis label that appears. Type in the correct X-Axis text, select Bold, and select the
correct font size as follows:
71
Step 7 – Insert Y-Axis Label
Primary Vertical Axis / Horizontal Title as follows:
72
After correcting the Y-Axis label by typing in the correct text, making it Bold, and setting the correct font
size as was just done for the X-Axis label, the final bar graph will appear as follows:
73
Combinations and Permutations in Excel
Both combinations and permutations represent all the possible ways to arrange k items that are randomly
and repeatedly selected from a larger set of n items.
Combinations Overview
A single combination is an unordered collection of k unique items. One combination is different from
another combination if at least one of the items in one collection is different than any of the items in all
other collections. Re-arrangement of the same items in a collection does not create a new combination.
For example, from the set of four letters [ A, B, C, D ], the following four total unique three-letter
combinations exist:
ABC
ABD
DBA
DBA
Different arrangements or orderings of these same k items in a subset do not create different
combinations. For example, [A B C] is not a different combination than [B A C] because both collections
contain the same three letters, just in a different order. The number of possible combinations is the total
number of ways that k elements can be arranged into n total elements when order of the k objects DOES
NOT matter.
Permutations Overview
A single permutation describes a unique ordering of those k items. For example, the combination that
contains the letters A, B, and C has six possible permutations (orderings) of that one combination
(unordered collection) of letters. The six total different ways that the three letters A, B, and C can be
ordered (ABC, ACB, BAC, BCA, CAB, CBA) represent the six total permutations of those three items
(letters) that exist.
The total number of possible permutations equals the total number of possible combinations times the
total number of permutations that can be derived from each combination. The preceding example shows
that the set of four letters [ A, B, C, D ] has a total of four unique three-letter combinations. Six different
permutations (orderings) can be derived from each one of these four combinations. The total number of
three-letter permutations that can be created from the set of four letters [ A, B, C, D ] is 24.
A permutation can have the same items as another permutation but those items must be arranged in a
different order to be considered a unique permutation.
Different arrangements or orderings of the same k items in a subset DO create different permutations.
The number of possible permutations is the total number of ways that all of the possible subsets of k
items can be ordered differently. The number of possible combinations is the total number of ways that k
elements can be arranged from n total elements when order of the k objects DOES matter.
Re-arrangement of the same k elements within a subset creates a new permutation but does not create a
new combination. There are always more permutations combinations of k objects taken from n total items
because each combination of k elements can be re-ordered to create k! permutations. The total number
of permutations of k items equals the following:
Total number of permutations of k items = (Total number of combinations of k items) * k!
74
Simultaneous vs. Sequential Selection
Combinations - Elements are picked simultaneously, all at once.
Permutations - Elements are picked sequentially, one after another, i.e., in a specific order.
If there is no order to the arrangement, it is a combination.
If there is any specified order to the arrangement, it is a permutation.
Excel Functions For Combinations

The Excel formula for the total possible number of combinations of k elements into n total elements is
given by the following:
is sometimes referred to as “n choose k.” This can be interpreted as follows: “Choose a subset of k
items from a total of n items available.”
is called the binomial coefficient and the is source of the name of the binomial distribution. The
binomial coefficient is equal to the total number of subsets of k items that can be taken from the larger set
of n items if it does not matter how the subset of k items is arranged.
COMBIN(n,k) = Number of Combinations of n Objects Taken k at a Time Simultaneously
FACT ( n ) = n!
----> FACT() is a Math & Trig Excel function and not a Statistical function such as COMBIN().
For example, the number of Combinations of 9 objects taken 4 at a time simultaneously
= nPk / k!
= 9P4/ 4!
= n! / [ k! * (n - k )! ]
= 9! / [ 4! * (9 - 4 )! ]
= FACT(n) / ( FACT(k) * FACT(n - k) )
= FACT(9) / ( FACT(4) * FACT(5) )
= 126
= COMBIN (n,k)
= COMBIN (9,4)
= 126
75
Excel Functions For Permutations
The Number of Permutations equals the number of times that n different objects taken k at a time
sequentially is given by the following:
PERMUT ( n, k ) = Number of Permutations of n Objects Taken k at a time Sequentially

FACT ( n ) = n!
----> FACT() is a Math & Trig Excel function, not a Statistical category function like PERMUT().
For example, the number of permutations of 9 objects taken 4 at a time sequentially =
= nPk = 9P4 = n! / (n - xk)!
= 9! / (9 - 4 )!
= FACT(n) / FACT(n - k)
= FACT(9) / FACT(5)
= 3,024
= PERMUT (n,k)
= PERMUT (9,4)
= 3,024
The Excel formulas for combinations and permutations are verified by referring back to the original
example provided in this section as follows:
From the set of four letters [ A, B, C, D ], the following four unique three-letter combinations can be
created:
ABC
ABD
DBA
DBA
This agrees with the result of the following Excel formula:
C(n,k) = COMBIN(n,k) = COMBIN(4,3) = 4
Six permutations can be derived from each one of the above three-letter combinations. The number of
permutations that can be derived from a combination of k items equals k!. In this case k = 3. k! = 6.
The six different ways that the three letters A, B, and C can be ordered (ABC, ACB, BAC, BCA, CAB,
CBA) represent six unique permutations of those three letters.
76
Total number of permutations of k items = (Total number of combinations of k items) * k!
Total number of permutations = 4 * 6 = 24
This agrees with the result of the following Excel formula:
P(n,k) = PERMUT(n,k) = PERMUT(4,3) = 24
C(n,k) = 4
P(n,k) = 24
k! = 6
77
Combination Problems in Excel
Problem 1: Combinations of Investment Proposals

A company is evaluating 6 investment proposals. If the company selects as many different groups of3
proposals simultaneously, how many different groups of three investment proposals can be selected?
This is a combination problem because the groups of three investment proposals are selected
simultaneously. The only difference between each group are the items in each groups but not how the
items are ordered.
n = 6 = total number of investment proposals available for inclusion in each combination group
x = 3 = number of investment proposals that will be simultaneously selected to fill each combination group
The number of combinations of n = 6 different investment proposals selected x = 3 at a time
(simultaneously) equals:
nPk / k! = 6P3 / 3! = COMBIN (n,k) = COMBIN (6,3) = 20
Problem 2: Combinations of Newly-Opened Offices

A consultancy wants to open 4 offices in 10 northern states. Each new office will be in a different state.
The offices will open all at the same time. How many different ways can these four offices be situated
among the 10 possible northern states?
This is a combination problem because the groups of four different states are selected as locations
simultaneously. The only difference between each group is the individual items in each group but not
the ordering of the items.
n = 10 = total number of states available for inclusion in each combination group
k = 4 = number of states that will simultaneously be selected to fill each combination group
The number of combinations of n = 10 different states available to selected at k = 4 at a time
(simultaneously) equals:
nPk / k! = 10P4 / 4! = COMBIN (n,k) = COMBIN (10,4) = 210
Problem 3: Multiple Combinations of Newly-Opened Offices

A consultancy wants to open 4 offices in 10 northern states, 3 offices in 9 southern states, and 2 offices in
8 eastern states. Each new office will be in a different state and all offices will be opened at the same
time. How many different combinations does the company have to evaluate?
Total number of combinations =

(all possible combinations of northern combinations)*(all possible combinations of southern
combinations)*(all possible combinations of eastern combinations)
= COMBIN(10,4) * COMBIN(9,3) * COMBIN(8,2) = 493,920
= 210 * 84 * 28 = 493,920
78
Problem 4: Combinations of Committees
From a group of 10 men and 8 women, a committee is formed. The committee will have 3 men and 4
women. How many different ways can this committee of 3 men and 4 women be formed from the overall
group of 10 men and 8 women? All committee members are picked at the same time.
This is combination problem because all committee members are picked at the same time. The
problem asks how many ways can all possible combinations of 3 out of 10 men be combined with all
possible combinations of 4 out of 8 women?
= (All possible combinations of men) * (All possible combinations of women)
= COMBIN(10,3) * COMBIN(8,4) = 8,400
= 120 * 70 = 8,400
Problem 5: Combinations of Sub-Groups

How many ways can a group of 12 people be divided into one group of 7 and another group of 5?
This is a combination problem because all members of any one group can be picked simultaneously.
One way to solve the problem would be to determine the total number of 7-person combinations that can
be formed from 10 people and then multiply that number by the number of 5-person combinations that
can be formed from the remaining 5 people.
= (all possible combinations of 7 out of 12) * (all possible combinations of 5 out of remaining 5)
= COMBIN(12,7) * COMBIN(5,5) = 792
= 792 * 1 = 792
79
Permutation Problems in Excel:
Problem 6: Permutations of Delivery Routes

A milkman makes 7 deliveries on his route. How many different sequences can he make to complete all 7
stops?
This is a permutation problem because the stops are done sequentially. After each stop, there is one
less stop to make.
n = 7 = total number of objects initially available for inclusion in each permutation group
k = 7 = number of objects that will sequentially fill the permutation group
The number of permutations of n = 7 different stops taken k = 7 at a time sequentially equals:
nPk = 7P7 = PERMUT(n,k) = PERMUT(7,7) = 5,040
Problem 7: Permutations of Seating Arrangements

How many ways can 5 people be seated on a sofa if only 3 seats are available and the 3 seats are filled
sequentially by the available 5 people?
This is a permutation problem because the elements of the permutation group are filled sequentially.
When each person is seated, there is one less person available to be seated.
The number of permutations of n = 5 different people seated k = 3 at a time sequentially equal:
nPk = 5P3 = PERMUT(n,k) = PERMUT(5,3) = 60
Problem 8: Permutations of Executive Groups

A group of 9 people needs to appoint 1 person to be group president, another person to be group vice
president, and a third person to be group treasurer. If the group first votes for the president, then votes for
the vice president, and finally votes for the treasurer, how many different executive groups can be created
from the original 9 people?
This is a permutation problem because the elements of the permutation group are filled up sequentially.
When each position is filled, there is one less person available for a position.
k= 3 = number of objects that will sequentially fill the permutation group
The number of permutations of n = 9 different people elected k = 3 at a time sequentially equals:
80
Problem 9: Permutations of Book Arrangements
How many ways can 3 books be placed next to each other on a shelf one at a time?
This is a permutation problem because the elements of the permutation group are filled sequentially.
When each book is placed, there is one less book to place.
n = 3 = total number of initially objects available for inclusion in each permutation group
The number of permutations of n = 3 different books placed k = 3 at a time sequentially equals:
Problem 10: Permutations of Letter Groups

From the following six letters: A, B, C, D, E, F, how many groups of 3 letters can be created if none of the
letters from the original 6 are repeated in any group?
This is a permutation problem because none of the letters can be repeated. When the first letter of
one of the permutation groups is chosen, there are only five remaining letters to choose from. Thus, the
elements of the permutation group are filled sequentially.
The number of permutations of n = 6 different letters placed k = 3 at a time sequentially equally:
81
Normal Distribution in Excel
Overview
The normal distribution is a very useful and widely-occurring distribution. The normal distribution curve
has the well-known bell shape and is symmetrical about a mean. The normal distribution is actually a
family of distributions with each unique normal distribution being fully described by the following two
parameters: the population mean, µ (Greek letter “mu”), and the population standard deviation, σ (Greek
letter “sigma”).
The following Excel-generated image is an example of the PDF (Probability Density Function) of a normal
distribution curve whose population mean, µ, equals 10 and population standard deviation, σ, equals 5.
Following is image of the PDF of a different normal distribution curve with a population mean, µ, equals 0
and population standard deviation, σ, equals 1. This is a special normal distribution known as the
Standard Normal Distribution as shown in the following Excel-generated image:
82
You may notice that the shapes of both normal distribution PDF graphs are the same but the x axis has
been shifted and scaled. The entire family of normal distributions is symmetrical about a mean and
contains approximately 68 percent of the entire area under the curve within one standard deviation of the
mean. Approximately 95 percent of the total area under the curve lies within two standard deviations of
the mean, and approximately 99.7 percent of the total curve area within three standard deviations of the
mean.
If PDF curves of different normal distributions are created with the mean and standard deviation in the
same locations on the horizontal axis, the curves will be in the same place and have exactly the same
shape. The two previous images demonstrate this.
If the PDF curves of different normal distributions are placed on the same horizontal and vertical axes,
each curve will be shifted left or right so it is symmetrical about its population mean, µ, and its width will
be scaled (widened or narrowed) depending on the size of its population standard deviation, σ.
The PDF graph of a normal distribution with a smaller population standard deviation would appear thinner
and taller than the PDF graph of a normal distribution with a larger population standard deviation, if both
curves were graphed on the same set of horizontal and vertical axes.
Uses of the Normal Distribution

Many real-valued variables are modeled using the normal distribution. An example of normally-distributed
natural phenomenon would be the velocities of the molecules in an ideal gas. Test scores for large
numbers of people follow the normal distribution. Random variation in output of a machine barring special
causes is often normal-distributed.
In other cases it is often a variable’s logarithm that is normal-distributed. Biological sciences provide many
such examples such as the length of appendages and blood pressure. A variable whose logarithm is
normal-distributed is said to have a log-normal distribution.
83
An outcome that is the result of a number of small variables that are additive, independent, and similar in
magnitude is often normal-distributed. Measurement error of a physical experiment might be one example
of such an outcome.
In some fields the changes in the logarithm of a variable are assumed to be normal-distributed. In the
financial fields, changes in the logarithms of exchange rates, price indices, and the stock market indices
are modeled by the normal distribution.
Useful Normal Approximations of Other Distributions

Several distributions can be approximated by the normal distribution under certain conditions. When the
normal approximation is valid, tools such as normal-distribution-based hypothesis testing and hypothesis
testing can be conveniently applied directly to samples from that-Distribution.
Normal Approximation of the Binomial Distribution

The normal approximation of the binomial distribution is of particular importance. Each unique binomial
distribution, B(n,p), is described with the following two parameters: n (sample size) and p (probability of
each binary sample having a positive outcome of the two possible outcomes). A binomial distribution is
approximately normal-distributed with a mean np and variance np(1-p) if n is large an p is not too close to
zero. A unique normal distribution, N(µ,σ), is completely described by the following two parameters: µ
(population mean) and σ (population standard deviation). When the normal approximation of the binomial
distribution is appropriate, normal-distribution-based hypothesis testing and confidence intervals can be
performed with binomially-distributed data by applying the following substitution: N(µ,σ) = N(np,
SQRT(np(1-p))).
The normal approximation of the binomial distribution enables the convenient analysis of population
proportions with normal-distribution-based hypothesis testing and confidence intervals. A well-known
example of this type of analysis would be the estimation of the percentage of a population along with a
percent margin of error that will vote for or against a particular political candidate. This is known as a
confidence interval of a population proportion.
The t-Distribution (sometimes called the Student’s t-Distribution) is closely related to the normal
distribution in function and appearance. The t-Distribution is used to analyze normally-distributed data
when the sample size is small (n< 30) or the population standard deviation is not known. The t-
Distribution is always centered about its mean of zero and closely resembles the standard normal curve.
The standard normal curve is the unique normal distribution curve whose mean equals 0 and standard
deviation equals 1. The t-Distribution has a lower peak and slightly thicker outer tails than the standard
normal distribution, but converges to the exact shape of the standard normal distribution as sample size
increases.
In the real world, normally-distributed data are much more frequently analyzed with t-Distribution-based
tools than normal-distribution-based tools. The reason is that the t-Distribution more correctly describes
the distribution normally-distributed data in the common occurrences of small sample size and unknown
population standard deviation. t-Distribution-based tools are also equally appropriate for analysis of large
samples of normally-distributed data because the t-Distribution converges to nearly an exact match of the
standard normal distribution when sample size exceeds 30.
Normal Approximation of the Poisson Distribution

Of lesser importance is the normal distribution’s approximation of the Poisson and chi-Square distribution
under certain conditions. Each unique Poisson distribution curve is completely described by a single
parameter λ. When λ is large, data distributed according to the Poisson distribution can be analyzed with
normal-distribution-based tools by making the following substitution:
84
N(µ,σ) = N(λ, SQRT(λ)).
Each unique Chi-Square distribution curve is completely described by the single parameter that is its
degrees of freedom, k. When k is large, data distributed according to the Chi-Square distribution can be
analyzed with normal-distribution-based tools by making the following substitution:
N(µ,σ) = N(k, SQRT(2k)).
One of the most useful modeling applications of the normal distribution is due to the fact the means of
large random samples from a population are approximately normal-distributed no matter how the
underlying population is distributed. This property is described by the Central Limit Theorem. Normal-
distribution based analysis tools such as hypothesis testing and confidence intervals can therefore be
applied to the means of samples taken from populations that are not normal-distributed.
Statistical Tests That Require Normality of Data or Residuals

Statistical tests that as classified a parametric tests have a requirement that the sample data or the
residuals are normal-distributed. For example, linear regression requires that the residuals be normal-
distributed. t-tests performed on small samples (n < 30) require that the samples are taken from a
normally-distributed population. ANOVA requires normality of the samples being compared.
When normality requirements cannot be met, nonparametric tests can often be substituted for parametric
tests. Nonparametric tests do not requirements that data or residuals follow a specific distribution.
Nonparametric tests are usually not as powerful as parametric tests.
History of the Normal Distribution

The first indirect mention of the normal distribution is credited to French mathematician and friend of
Isaac Newton, Abraham De Moivre. De Moivre developed such repute as a mathematician that Newton
often referred questions to him, saying “Go to Mr. de Moivre. He knows these things better than I do.”
In 1738 De Moivre published the second edition of his The Doctrine of Chances that contained a study of
binomial coefficients which is considered to the first reference, albeit, indirect, to the normal distribution.
That particular book became highly valued among gamblers of the time. The book noted that the
distribution of the number of times that a binary event produces a positive outcome, such as a coin toss
resulting in heads, becomes a smooth, bell-shaped curve when the number of trials is high enough. This
bell-shaped curve approaches the normal curve. DeMoivre was alluding to what we know today as the
normal distribution’s approximation of the binomial distribution. DeMoivre did not speak of the normal
distribution in terms of a probability density function and therefore does not receive full credit for
discovering the normal distribution.
De Moivre, who spent his adult life in London, remained somewhat poor because he was unable to
secure a professorship at any local university, partially due to his French origins. He earned a substantial
part of his living from tutoring mathematics and being a consultant to gamblers. One day when De Moivre
became older, he noticed that he required more sleep every night. He determined that he was sleeping
an extra 15 minutes every night. He correctly calculated the date his own death to be November 27,
1754, the date that the total required sleep time would reach 24 hours.
The first true mention of the normal distribution was made by German mathematician Carl Friedrich
Gauss in 1809 as a way to rationalize the nonlinear method of least squares. The normal distribution
curve is often referred to as the Gaussian curve as a result.
Gauss was one of the greatest mathematicians who ever lived and is sometimes referred to as “the
Prince of Mathematicians” and “the greatest mathematician since antiquity.” Gauss was a child prodigy
and made some of his groundbreaking mathematical discoveries as a teenager. Gauss was an
astonishingly prolific scientist in many fields. Here is a link to a partial list of over 100 scientific topics
named after him:
85
http://en.wikipedia.org/wiki/List_of_things_named_after_Carl_Friedrich_Gauss
Significant contributions to the normal distribution were made by another of the greatest mathematicians
of all-time, Frenchman Pierre Simon LePlace, who was sometimes referred to as the French Newton.
LePlace is credited with providing the first proof of one of statistics’ most important tenets related to the
normal distribution, the Central Limit Theorem. LePlace has an interesting life and was appointing by
Napoleon to be the French Minister of the Interior shortly after Napoleon seized power in a coup. The
appointment lasted about six weeks until nepotism took over and the post was given to Napoleon’s
brother.
The distribution’s moniker, the “normal distribution,” was made popular by Englishman Karl Pearson,
another giant in the field of mathematics. Karl Pearson is credited with establishing the discipline of
mathematical statistics and founded the world’s first university statistics department in 1911 at the
University College of London. Many of topics covered in all basic statistics courses such a p Values and
correlation are the direct result of Karl Pearson’s work.
Some of Pearson’s publishings, particularly his book The Grammar of Science, provided a number of
themes that Einstein would weave into several of his most well-known theories of relativity. Pearson
postulated that a person traveling faster than the speed of light would experience time being reversed.
Pearson also discussed the concept that the operation of physical laws depended on the relative position
of the observer. These are central themes in Einstein’s relativity theories.
86
Normal Distribution’s PDF (Probability Density Function)
The normal distribution is a family of distributions with each unique normal distribution being fully
described by its two parameters µ (“mu,” population mean) and σ (“sigma,” population standard
deviation). The population mean, µ, is a location parameter and the population standard deviation, σ, is
a scale parameter. When two different normal distribution curves are plotted on the same set of
horizontal and vertical axes, the means determine how shifted one curve is to the left of right of the other
curve. The standard deviations detail how much wider or more narrow the first normal curve is than to the
second.
2 2
The normal distribution is often denoted as N(µ,σ ). σ equals the population variance. When a random
2
variable X is normal-distributed with a population mean µ and population variance σ , it is written in the
following form:
2
X ~ N(µ, σ )
It is important to note that the two parameters of the normal distribution are population parameters, not
measurements taken from a sample. Sample statistics would provide only estimates of population
parameters. The t-Distribution is used to analyze normally-distributed data when only sample statistics
and/or population parameters are not known. In the real world it is much more common to analyze
normally-distributed data with t-Distribution based tests than normal-distribution-based tests because only
data from small samples (n<30) are available.
As with all distributions, the normal distribution has a PDF (Probability Density Function) and a CDF
(Cumulative Distribution Function).
The normal distribution’s PDF (Probability Density Function) equals the probability that sampled point
from a normal-distributed population has a value EXACTLY EQUAL TO X given the population’s mean,
µ, and standard deviation, σ.
The normal distribution’s PDF is expressed as f(X,µ,σ).
f(X,µ,σ) = the probability that a randomly-sampled point taken from normally-distributed population with a
mean µ and standard deviation σ has the value of X. It is given by the following formula:
exp refers to the value of the mathematical constant e which is the base of the natural logarithm. e is
n a
equal to 2.71828 and is the limit of (1 + 1/n) and n approaches infinity. e would be expressed in Excel as
=exp(a).
The mathematical constant π (“pi”) is equal to 3.14159 and is the ratio of a circle’s circumference to its
diameter.
In Excel 2010 and beyond, the normal distribution’s PDF can be calculated directly by the following Excel
formula:
f(X,µ,σ) = NORM.DIST(X,µ,σ,FALSE)
The Excel formula parameter “FALSE” indicates that the formula is calculating the normal distribution’s
PDF (Probability Density Function) and not its CDF (Cumulative Distribution Function)
Prior to Excel 2010, the normal distribution’s PDF was calculated in Excel by this formula:
f(X,µ,σ) = NORMDIST(X,µ,σ,FALSE)
Statistical formulas that worked in Excel versions prior to 2010 will also work in Excel 2010 and 2013.
87
The following Excel-generated graph shows the PDF of a normal distribution that has a population mean
of 10 and population standard deviation equal to 5.
Example in Excel
Determine the probability that a randomly-selected variable X taken from a normally-distributed
population has the value of 5 if the population mean equals 10 and the population standard deviation
equals 5. The preceding Excel-generated image shows a normal distribution PDF curve with the
population mean equaling 10 and the population standard deviation equaling 5.
X=5
µ = 10
σ=5
f(X,µ,σ) = NORM.DIST(X,µ,σ,FALSE)
f(X=5,µ=10,σ=5) = NORM.DIST(5,10,5,FALSE) = 0.04834
There is a 4.834 percent chance that randomly-selected X = 5 if X is taken from a normally-distributed
population with a population mean µ = 10 and population standard deviation σ = 5. The PDF diagram of
this normal distribution curve also shows the probability of X at X = 5 to that value.
88
Performing the same calculation in Excel using the full normal distribution PDF formula as shown as
follows:
f(X=5,µ=10,σ=5) =(1/(SQRT(2*3.14159*5^2)))*EXP((-1)*((5-10)^2)/(2*5^2))
= 0.04834
89
The Normal Distribution’s CDF (Cumulative Distribution
Function)
The normal distribution’s CDF (Cumulative Distribution Function) equals the probability that sampled point
from a normal-distributed population has a value UP TO X given the population’s mean, µ, and standard
deviation, σ.
The normal distribution’s CDF is expressed as F(X,µ,σ).
Unlike the normal distribution’s PDF, the CDF has no convenient closed form of its equation, which is the
integral just shown.
In Excel 2010 and beyond, the normal distribution’s CDF must be calculated by the following Excel
formula:
F(X,µ,σ) = NORM.DIST(X,µ,σ,TRUE)
The Excel formula parameter “TRUE” indicates that the formula is cumulative, i.e., it is calculating the
normal distribution’s CDF (Cumulative Distribution Function) and not its PDF (Probability Density
Function).
Prior to Excel 2010, the normal distribution’s PDF was calculated in Excel by this formula:
F(X,µ,σ) = NORMDIST(X,µ,σ,TRUE)
Statistical formulas that worked in Excel versions prior to 2010 will also work in Excel 2010 and 2013.
Note that the CDF has asymptotic values of 0 as X decreases and an asymptotic value of 1 as X
increases as shown in the following Excel-generated image:
90
The normal distribution’s CDF has a value of exactly 0.5 when X equals the population mean. This
indicates that 50 percent of the entire area under the normal distribution’s PDF is contained under the
curve before X reaches a value of the population mean. The underlying meaning is that a randomly-
sample point from a normally-distributed population has a 50 percent chance of having a value less than
or equal the population mean. This can be seen in the Excel-generated graph of the normal distribution’s
PDF as follows:
91
This CDF always has a value of 0.15866 when X is one standard deviation below the mean. This
indicates that 15.866 percent of the area under the normal distribution’s PDF curve occurs before X
reaches the value of the point that is one standard deviation below the population’s mean. The underlying
meaning is that a randomly-sample point from a normally-distributed population has a 15.866 percent
chance of having a value less than or equal the value that is one standard deviation below the population
mean.
This CDF always has a value of 0.84135 when X is one standard deviation above the mean. This
indicates that 84.135 percent of the area under the normal distribution’s PDF curve occurs before X
reaches the value of the point that is one standard deviation below the population’s mean. The underlying
meaning is that a randomly-sample point from a normally-distributed population has an 84.135 percent
chance of having a value less than or equal the value that is one standard deviation below the population
mean.
A randomly-selected point from a normally-distributed population has a 68.269 percent chance of having
a value X that is within one standard deviation above or below the mean (84.135 – 15.866 = 68.269). In
other words, 68.269 percent of normally-distributed data lie within one standard deviation of the mean.
Similar analysis shows that approximately 95 percent of all normally-distributed data lie within two
standard deviations of the mean and 99.7 percent of the data are within three standard deviations of the
mean. This rule is often referred to as the Empirical Rule or the 68-95-99.5 Rule.
92
Example in Excel
Determine the probability that a randomly-selected variable X taken from a normally-distributed
population has the value of UP TO 5 if the population mean equals 10 and the population standard
deviation equals 5.
X=5
µ = 10
σ=5
F(X=5,µ=10,σ=5) = NORM.DIST(5,10,5,TRUE) = 0.15866
There is a 15.866 percent chance that randomly-selected X equals UP TO 5 if X is taken from a normally-
distributed population with a population mean µ = 10 and population standard deviation σ = 5. The CDF
diagram of this normal distribution curve also shows the probability of X at X = 5 to that value in the
following Excel-generated image.
93
The following PDF diagram of this normal distribution curve shows 15.866 percent of the total area under
the bell-shaped curve that is to the left of X = 5.
94
Problems Using NORM.DIST() To Calculate F(X,µ,σ)
Problem 1
A certain brand of automobiles has normally-distributed fuel consumption. The brand’s mean fuel
consumption is 27 miles-per-gallon and standard deviation is 5 miles-per-gallon. What percentage of
automobiles of this brand can be expected to have fuel consumption between 25 and 30 miles-per-
gallon?
0.3812 = NORM.DIST(30,27,5,TRUE) - NORM.DIST(25,27,5,TRUE)
0.3812 = 0.7257 – 0.3446
38.12 percent of autos of the brand have fuel consumption between 25 and 30 miles-per-gallon as shown
in the following Excel-generated graph:
Problem 2
A company packages potatoes into knitted net bags for shipment to supermarkets. The weight of the filled
bags of potatoes is normal-distributed with a mean of 20 pounds and a standard deviation of 1 pound.
What is the probability that a randomly selected bag of potatoes will either weigh more than 22 pounds or
less than 18.5 pounds?
Probability that potatoes weigh less than 18.5 pounds =
NORM.DIST(18.5,20,1,TRUE)
95
Probability that potatoes weigh less than 18.5 pounds = 0.0668 = 6.68 %
Probability that potatoes weigh less than 22 pounds = NORM.DIST(22,20,1,TRUE)

Probability that potatoes weigh more than 22 pounds = 1 - NORM.DIST(22,20,1,TRUE)
Probability that potatoes weigh more than 22 pounds = 1 - 0.9773 = 0.0227 = 2.27 %
Probability that the potatoes weigh less than 18.5 pounds OR more than 22 pounds equals
Probability that the potatoes weigh less than 18.5 pounds = 0.0668
PLUS
Probability that the potatoes weigh more than 22 pounds = 0.0227
Which equals
NORM.DIST(18.5,20,1,TRUE) + [1 - NORM.DIST(22,20,1,TRUE) ]
Which equals
0.0668 + 0.0227 = 0.0896 = 8.96 %
8.96 percent of bags of potatoes have weights outside of 18.5 pounds and 22 pounds as shown in the
following Excel-generated graph:
96
Problems Using NORM.INV() To Calculate X Given F(X,µ,σ)
NORM.INV[ F(X,µ,σ), µ, σ ] = X
Problem 1
A tire company makes a tire with a normally-distributed tread life. The mean tread life is 39,000 miles and
the standard deviation of tread life is 5,300 miles. What tread life would be exceeded by only 3 percent of
all tires?
48968 = NORM.INV(0.97,39000,5300)
97 percent of all tires will wear out before they are driven 48,968 miles hours as shown in the following
Excel-generated graph:
Problem 2
A company’s package delivery time is normal-distributed and has a mean of 10 hours and a standard
deviation of 3 hours. What delivery time will be beaten by only 2.5 percent of all deliveries?
4.120 = NORM.INV(0.025,10,3)
97
Only 2.5 percent of all deliveries are made quicker than 4.120 hours as shown in the following Excel-
generated graph:
98
Standard Normal Distribution
The normal distribution is actually a family of distributions. Each unique normal distribution can be fully
described by its two parameters, which are the following:
1) Its population mean, µ, which is a location parameter.
2) Its population standard deviation, σ, which is a scale parameter.
The most basic normal distribution is called the Standard Normal Distribution. Its population mean, µ,
equals 0 and its population standard deviation, σ, equals 1.
The PDF curve of the Standard Normal Distribution appears as follows in this Excel-generated graph:
99
The CDF curve of the Standard Normal Distribution appears as follows in this Excel-generated graph:
When the population mean, µ, equals 0 and the population standard deviation, σ, equals 1, the x values
equal the number of standard deviations that each x is from the mean. In this situation, each x value
equals its z score. A data point’s z score equals the number of population standard deviations that the
point is from the population mean. The formula to calculate an x value’s z score is the following:
Every unique normal distribution becomes to the Standard Normal Distribution when the x values are
converted to z scores. The Standard Normal Distribution (with zero population mean and unit population
standard deviation) is sometimes referred to as the standard Gaussian distribution or the unit normal
distribution is denoted by the Greek letter “phi” as follows:
φ(x) (small “phi”) denotes the PDF of the standard normal distribution at x, which also equals z score(x)
because σ = 1.
Φ(x) (capital “phi”) denotes the CDF of the standard normal distribution at x, which also equals z score(x)
because σ = 1.
When µ = 0 and σ = 1, an X value is equal to its z score. If µ = 0 and σ = 1, then the Excel formulas
NORM.DIST(X,µ,σ, TRUE or FALSE) can be replaced by the simpler Excel formula NORM.S.DIST(z,
TRUE or FALSE)
100
Calculating the PDF of a Standardized Normal Distribution
X=2
µ=0
σ=1
f(X = 2, µ = 0, σ = 1) = NORM.DIST(2,0,1,FALSE) = 0.05399

f(X = 2, µ = 0, σ = 1) = NORM.S.DIST(z score(x),FALSE)
z score(x) = (2 – 0)/1 = 2
φ(2) = f(X = 2, µ = 0, σ = 1)
f(X = 2, µ = 0, σ = 1) = NORM.S.DIST(2,FALSE) = 0.05399
φ(2) = 0.05399
This is shown in the following Excel-generated graph:
101
Calculating the CDF of a Standardized Normal Distribution
X=2
µ=0
σ=1
F(X = 2, µ = 0, σ = 1) = NORM.DIST(2,0,1,TRUE) = 0.9773
F(X = 2, µ = 0, σ = 1) = NORM.S.DIST(z score(x),TRUE)
z score(x) = (2 – 0)/1 = 2
Φ(2) = F(X = 2, µ = 0, σ = 1)
F(X = 2, µ = 0, σ = 1) = NORM.S.DIST(2,TRUE) = 0.9773
Φ(2) = 0.9773
This is shown in the following Excel-generated graph:
102
Here are some properties of the CDF of the standard normal distribution
Φ(-∞) = 0 = 0%
Φ(0) = 0.5 = 50%

Φ(∞) = 1 = 100%
Φ(X) = 1 - Φ(-X) and therefore Φ(X) + Φ(-X) = 100%
The t-Distribution’s Convergence to the Standard Normal Distribution

The t-Distribution has the following two important similarities to the standard normal distribution:
1) Both the t-Distribution and the standard normal distribution are centered about means of 0. One
difference between the t-Distribution and the normal distribution is that the normal distribution can
assume any value as its mean. The t-Distribution is always symmetrical about a mean of 0, as is the
standard normal distribution.
2) The horizontal axis of the t-Distribution is measured in units of t value. The t-value of a point is the
number of sample standard errors that the point is from the mean. The horizontal axis of the standard
normal distribution is measured in units of z Value, i.e., the number of population standard deviations that
the point is from the mean. This is the result of the standard normal distribution’s population standard
deviation being set to the unit value of 1. The formulas for z score(x) and t value(x) are shown as follows:
The z score of a randomly-sampled point, x, from a normally-distributed population is calculated as
follows:
The z score of point x taken from a large sample (n > 30) from a population of unknown distribution is
calculated as follows:
The t value of a point x taken from a small sample (n < 30) of a normally-distributed population or a large
sample from a population of unknown distribution is calculated as follows:
103
When sample size is large, the means of large, similar-sized random samples are normal-distributed
regardless of the distribution of the underlying population as per the Central Limit Theorem.
The Standard Error (SE) that is calculated for the t value using the sample standard deviation, s, is an
estimate of actual SE that would be calculated with the population standard deviation, σ.
The z Value (z score) is the unit of measure of the horizontal axis of the standard normal distribution and
the t value is the unit of measure of the horizontal axis of the t-Distribution.
z score(x) is the number of population standard deviations that a point x is from the population mean. t
value(x) is the number of standard errors that point x is from the sample mean.
It is important to note that z scores are created using population parameters µ (population mean) and σ
(population standard deviation). t values are created using the sample statistics x_bar (sample mean), s
(sample standard deviation), and n (sample size).
The underlying reason for this is that the normal distribution is used to analyze normally-distributed data
only when population parameters µ and σ are known. The t-Distribution is used to analyze normally-
distributed data when only sample statistics x_bar, s, and n are known. It is much more common in the
real world that only sample statistics are known so the t-Distribution is often the tool of choice for
analyzing normally-distributed data.
The t-Distribution has only a single parameter: v, the degrees of freedom, which is equal to v = n – 1. The
t-Distribution’s shape changes as sample size, n, changes. Very low values of n (very small sample sizes)
produce a t-Distribution PDF graph with wider tails and a lower peak. The follow Excel-generated graph
shows a t-Distribution’s PDF curve with a sample size of n = 3:
As sample increases, the t-Distribution’s shape changes as its peak rises and less weight remains in the
outer tails. The t-Distribution converges to exactly resemble the standard normal distribution when the
sample size is large enough. The follow Excel-generated graph shows a t-Distribution’s PDF curve with a
sample size n approaches infinity:
104
The PDF curve of the standard normal distribution shows an exact match:
It should be noted that a number of texts incorrectly state the t-Distribution converges to the normal
distribution as sample size increases. The normal distribution represents a family of distribution curves
each having a unique combination of µ and σ. The specific normal curve needs to be specified. The
correct statement would be that the t-Distribution converges to the standard normal distribution as sample
size increases.
105
Properties of the Normal Distribution
This graph of the standard normal distribution’s PDF is once again presented to assist in the
understanding of each listed property of the normal distribution.
The normal distribution’s probability density function, f(x), has the following properties:
1) It is symmetric about its population mean µ. Half of the values of a normally-distributed population will
be less than (to the left of) the population mean and the other half of the population’s value will be greater
than (to the right of) the population mean.
2) Its mode and median are equal to the population mean µ.
3) It is unimodal. This means that it has only one peak, i.e., only one point that is a local maximum.
4) The total area under the normal distribution’s PDF is equal to 1.
5) Each unique normal distribution curve is entirely defined by its two parameters population mean µ and
population standard deviation σ.
6) The density of the normal distribution’s PDF is the highest at its mean and always decreases as
distance from the mean increases.
7) 50 percent of values of a normally-distributed population are less than the population mean and 50
percent of the values are greater than the mean.
8) Approximately 68 percent of the total area under the PDF curve resides within one σ from the mean,
approximately 95 percent of the total resides within two σs, and approximately 99.7 percent of the total
area resides within three σs from the mean. This is sometimes known as the Empirical Rule or the 68-
95-99.7 Rule.
9) f(x) is infinitely differentiable.
106
10) The first derivative of f(x) is positive for all x < µ and negative for all x > µ.
11) The second derivative of f(x) has two inflection points which are located one population standard
deviation above and below the population mean. These inflection points are located at x = µ ± σ. An
nd
inflection point occurs at the point that the 2 derivative equals zero and changes sign as x continues.
12) It is log-concave. A function f(x) is log-concave if its natural log, ln[f(x)], is concave. A log-concave
function does not have multiple separate maxima and its tails are not “too thick.” Other well-known
distributions that are log-concave include the following:
- exponential distribution
- uniform distribution over any convex set
- logistic distribution
- gamma distribution if its shape parameter is >=1
- Chi-Square distribution if the number of degrees of freedom >=2
- beta distribution if both shape parameters are >=1
- Weibull distribution if the shape parameter is >=1
The following well-known distributions are non-log-concave for all parameters:
- t-Distribution
- log-normal distribution
- F-distribution
107
Calculating How Much of the Data Is a Certain Distance From
the Mean
There are two rules that can be used to calculate the proportion of data values that will be within a
specified number of standard deviations from the mean. They are as follows:
For Normally-Distributed Data, Use the Empirical Rule

The Empirical Rule, a.k.a., the 68-95-99.7 Rule, states that when data is normal-distributed, the following
is true:
68 percent of all data points have values that are within one standard deviation of the mean.
95 percent of all data points have values that are within two standard deviations of the mean.
97.5 percent of all data points have values that are within three standard deviations of the mean.
Empirical Rule Problem in Excel

Calculate the percent of values in a large normally-distributed data set of unknown distribution that will fall
between 12 and 22 if the data’s set’s mean is 16 and its standard deviation is 2.
0.9759 = NORM.DIST(22,16,2,TRUE) - NORM.DIST(12,16,2,TRUE)
97.59 percent of all data values of a large, normally-distributed data set will fall between 12 and 22 if the
data set’s mean is 16 and its standard deviation is 2.
For Data of Unknown Distribution, Use Chevbyshev’s Theorem

2
Chebyshev’s Theorem states that (1 – 1/z ) of the data values will fall within z standard deviations of the
mean as long as z is any value greater than 1.
According to Chebyshev’s Theorem, the following are true:
75 percent of data values will be within two standard deviations of the mean.
89 percent of all data points have values that are within three standard deviations of the mean.
94 percent of all data points have values that are within four standard deviations of the mean.
Chebyshev’s Theorem Problem in Excel

Use Chebyshev’s Theorem to calculate the percent of values in a large data set of unknown distribution
that will fall between 12 and 22 if the data’s set’s mean is 16 and its standard deviation is 2.
12 is |12 – 16|/2 = 2 standard deviations to the left of the mean.
There will be at least 0.75 / 2 or 37.5 percent of all data values falling between the mean and 8.
22 is |22 – 16|/2 = 3 standard deviations to the left of the mean.
There will be at least 0.89 / 2 or 44.5 percent of all data values falling between the mean and 22.
82 percent = 37.5 percent + 44.5 percent
Chebyshev’s Theorem states that 82 percent of all data values will fall between 12 and 22 if of a large
data set of unknown distribution if the data set’s mean is 16 and its standard deviation is 2.
108
The Central Limit Theory in Excel
Overview
The Central Limit Theorem, or CLT, is one of statistics’ most basic principles. The CLT states the
following points:
When an entire population is grouped into random samples of size n, the following are true:
1) As sample size n increases, the distribution of the sample means approaches the normal distribution
regardless of how the population is distributed. A sample size of 30 is often quoted as being the minimum
sample size necessary to ensure that the sample means will be normal-distributed. In the real world,
sample means will be normal-distributed when the sample size is much smaller. This will be
demonstrated with Excel shortly.
2) The sample standard deviation will equal the population standard deviation divided by the square root
of the sample size. This will also be demonstrated in Excel shortly.
Why the Central Limit Theorem Is So Important to Statistics

The Central Limit Theorem is very important to statistical work because it enables convenient normal-
distribution-based and t-Distribution-based hypothesis testing and confidence-interval creation to be
performed based upon a single large sample of data without having to verify that the population from
which the sample was taken is normal-distributed. Samples are considered to be large if the size of each
sample, n, is greater than or equal to 30.
Performing normal-distribution-based and t-Distribution-based hypothesis testing and creating confidence
intervals based upon a small sample of data requires verification that the population from which the
sample was taken is normal-distributed. The normality of the population is confirmed if the data within the
data sample is verified to be normal-distributed. This step is not necessary when the sample is large.
Samples are considered to be small if the size of each sample, n, is smaller than 30.
Hypothesis testing of a population mean and confidence intervals of a population mean are discussed in
much greater detail in their own sections of this book. Hypothesis testing and confidence intervals of a
population mean are calculated based upon a single sample of data. The fundamental assumption that
underlies normal-distribution-based and t-Distribution-based hypothesis testing and confidence interval
creation of a population mean is that the sample mean is normal-distributed.
The mean of a single sample is normal-distributed if the means of other similar samples are also normal-
distributed. In other words, if the means of a number of other same-sized samples randomly taken from
the same population were normal-distributed, then the mean of the single sample is normal-distributed.
When the single sample taken is large enough (sample size n is greater than or equal to 30), the mean of
that sample is assumed to be normal-distributed as per the Central Limit Theorem. When the size of the
single sample is less than 30, the sample’s mean cannot be assumed to be normal-distributed unless one
of the following is true:
1) The population from which the sample was drawn is known to be normal-distributed.
2) The data within the single sample is verified through normality testing to be normal-distributed.
Quite often it is not possible to confirm the normality of the population. In this case normality testing
should be performed on the data within the single sample taken before normal-distribution-based or t-
Distribution-based hypothesis testing or the confidence interval creation can be performed based upon
that single sample of data.
109
A Demonstration of the Central Limit Theorem Using Excel
The Central Limit Theorem is not immediately intuitive can be more quickly understood when an example
of how it works is shown in graphical form. The following example will be performed in Excel as follows:
A population of 5,000 randomly-generated data points whose distribution is highly skewed will be created
in Excel. The population of 5,000 will be randomly into samples of sample size n = 2, sample size n = 5,
and sample n = 10.
The distribution of the sample means will be shown in an Excel histogram created for each of the three
sample sizes.
The histograms will demonstrate that the distribution of the sample means become closer and closer to
the normal distribution. This exercise will also show that the sample standard deviation equals the
population standard deviation divided by the square root of the sample size. This means that the
distribution’s shape become tighter as the sample size increases.
Here are the steps in demonstrating how the Central Limit Theorem works using Excel.
The first step is to use Excel’s random-number generator to generate 5,000 random numbers that are
highly skewed. In this case, the data will be highly skewed to the right and will following this model, which
is an Excel-generated histogram:
This population of 5,000 skewed random numbers by creating the following sets of random numbers in
Excel:
● 1,500 random numbers between 0 and 100. Each of these numbers was created with the Excel formula
RANDBETWEEN(0,100)
110
● 1,100 random numbers between 100 and 200. Each of these numbers was created with the Excel
formula RANDBETWEEN(100,200)
● 700 random numbers between 200 and 300. Each of these numbers was created with the Excel
● 100 random numbers between 900 and 1,000. Each of these numbers was created with the Excel
formula RANDBETWEEN(900,1,000)
These 5,000 numbers need to be randomly placed in a table so that samples can be taken. One way of
doing that is to initially place all 5,000 numbers in a single Excel column. These numbers need to be
randomly scrambled in the column. This is done as follows:
The columns of numbers on the right in column D are the randomly-generated numbers from several of
the ranges previously created. The column of numbers on the left in column C are randomly-generated
numbers from a single range using the Excel formula RANDBETWEEN(0,1000).
111
The objective is to randomly scramble the numbers in column D. The numbers in Column D will be
scrambled randomly be performing a single sort of both column based upon an ascending sort of the
numbers in Column C. Prior to the sort, the numbers in the two columns appear as follows:
112
After the sort according to Column C, the numbers appear as follows:
The next step is to take the numbers from the column on the right and place them into a table. This is
necessary for taking samples and creating histograms in Excel. Converting column data into a table can
be done in Excel using the OFFSET() command. As soon as any cell from C2 to E4 has the correct
OFFSET() formula typed into it, that cell can be copied to all of the other cells in the table. The result is
the following:
113
The data in this exercise was arrange into a table of dimensions 20 rows by 500 columns as follows:
Population (Sample size n = 1)
A histogram was created in Excel of this population of 5,000 data points. It population parameters are as
follows:
Population size = N = 5,000 (Note that capital N is used to describe population size)
Population mean = µ = 260
Population standard deviation = σ = 230
Note that the following Excel histogram shows the population’s distribution to be highly skewed to the
right.
114
115
Sample Size n = 2
A table of samples of sample size n = 2 was created by taking the average of every 2 data points going
down each column. A table with these samples appears as follows:
Total number of samples = 2,500

Sample size = n = 2
Sample mean = x_bar = 260
Sample standard deviation = 163 = σ / SQRT(n) (with rounding error)
116
Note that the shape of the distribution of the sample means when the sample size is 2 is still highly
skewed to the right. The bell shape of the normal distribution is already starting to form at a sample size
of 2 even though the population from which the samples were taken is highly skewed.
Note how much more narrow the distribution’s shape has become as sample size increase from n = 1 to n
= 2.
117
Sample Size n = 5
Total number of samples = 1,000

Sample size = n = 5
118
Note that the shape of the distribution of the sample means when the sample size is 5 already has a
strong resemblance to the bell-shape normal distribution’s PDF curve even though the population from
which the samples were taken is highly skewed. A mild skew to the right can still be seen.
= 5. The reduction in the sample standard deviation indicates how much thinner the distribution’s shape
has become.
119
Sample Size n = 10
Total number of samples = 500

Sample size = n = 10
Note that the shape of the distribution of the sample means when the sample size is 10 now nearly
exactly resembles the bell-shape normal distribution’s PDF curve even though the population from which
the samples were taken is highly skewed.
= 10. The reduction in the sample standard deviation indicates how much thinner the distribution’s shape
has become.
120
The most important point of the Central Limit Theorem is that the distribution of sample means become
closer and closer to the normal distribution as sample size increases regardless of the distribution of the
population from which the samples were taken.
Most statistics texts state that the sample size must reach at least 30 before the sample means are nearly
certain to be normal-distributed. In the real world, the distribution of the sample means converges to
normality at significantly smaller sample sizes. The example just provided shows a reasonable
resemblance to the normal distribution’s PDF curve when the sample size is only 5 and the population
was highly skewed. A sample size of 10 shows an almost exact resemblance to the normal distribution’s
PDF curve.
The Resulting Robustness of the t-Test

The quick convergence of sample mean distribution to normality as sample size increases means that the
t–test is relatively robust on non-normally-distributed data as long as the sample size is not too small and
the data is not too skewed.
The example just provided shows that even highly-skewed data will still have near-normal distribution of
sample means when the sample size is as low as 10. Data that is not skewed will have sample means
achieving near-normality at smaller sample sizes than 10.
The t-test is therefore very robust in the face of non-normal data. The t-test can produce a reasonably
accurate result for sample sizes as low as 5 to 10 for data that is not skewed. Many statistics texts state
that sample size must be large (n > 30) for a t-test to be reliable but that is not the case. The example
provided in this section shows the sample means of highly-skewed data converging to near-normality at a
sample size of 10.
Derivation of the Central Limit Theorem

The derivation of the Central Limit Theorem is readily available on the Internet. The CLT’s derivation is
not trivial and involves a significant amount of calculus. Understanding this complicated derivation does
not significantly add to one’s understanding of when and how to apply the CLT. For that reason the
derivation of the CLT is not discussed here.
121
Hypothesis Tests Using the Normal Distribution in
Excel
Overview
A hypothesis test evaluates whether a sample is different enough from a population to establish that the
sample probably did not come from that population. If a sample is different enough from a hypothesized
population, then the population from which the sample came is different than the hypothesized
population.
Null Hypothesis
A hypothesis test is based upon a Null Hypothesis which states that the sample did come from that
population. A hypothesis test compares a sample statistic such as a sample mean to a population
parameter such as the population’s mean. The amount of difference between the sample statistic and the
population parameter determines whether the Null Hypothesis can be rejected or not.
The Null Hypothesis states that the population from which the sample came has the same mean or
proportion as a hypothesized population. The Null Hypothesis is always an equality stating that the
means or proportions of two populations are the same.
An example of a basic Null Hypothesis for a Hypothesis Test of Mean would be the following:
H0: x_bar = Constant = 5
This Null Hypothesis would be used to state that the population from which the sample was taken has a
mean equal to 5. The Constant (5) is the mean of the hypothesized population that the sample’s
population is being compared to. The Null Hypothesis states that the sample’s population and the
hypothesized population have the same means. The Alternative Hypothesis states that they are different.
An example of a basic Null Hypothesis for a Hypothesis Test of Proportion would be the following:
H0: p_bar = Constant = 0.3
proportion equal to 0.3. The Constant (0.3) is the proportion of the hypothesized population that the
sample’s population is being compared to. The Null Hypothesis states that the sample’s population and
the hypothesized population have the same proportions. The Alternative Hypothesis states that they are
different.
The Null Hypothesis is Either Rejected or Not Rejected But Is Never

Accepted
A hypothesis test has only two possible outcomes: the Null Hypothesis is either rejected or is not rejected.
It is never correct to state that the Null Hypothesis was accepted. A hypothesis test only determines
whether there is or is not enough evidence to reject the Null Hypothesis. The Null Hypothesis is rejected
only when the hypothesis test result indicates a Level of Certainty that the Null Hypothesis is not valid at
least equals the specified Level of Certainty.
If the required Level of Certainty for a hypothesis test is specified to be 95 percent, the Null Hypothesis
will be rejected only if the test result indicates that there is at least a 95 percent probability that the Null
Hypothesis is invalid. In all other cases, the Null Hypothesis would not be rejected. This is not equivalent
to stating that the Null Hypothesis was accepted. The Null Hypothesis is never accepted; it can only be
rejected or not rejected.
122
Alternative Hypothesis
The Alternative Hypothesis is always in inequality stating that the means or proportions of two populations
are not the same. The Alternative Hypothesis can be non-directional if it states that the means or
proportions of two populations are merely not equal to each other. The Alternative Hypothesis is
directional if it states that the mean or proportion of one of the populations is less than or greater than the
mean of proportion of the other population.
An example of a non-directional Alternative Hypothesis for a Hypothesis test of Mean would be the
following:
H1: x_bar ≠ 5
This Alternative Hypothesis would be used to state that the population from which the sample was taken
has a mean that is not equal to 5.
An example of a directional Alternative Hypothesis would be the following:
H1: x_bar > 5
or
H1: x_bar < 5
These Alternative Hypotheses would be used to state that the population from which the sample was
taken has a mean that is either greater than or less than 5.
An example of a non-directional Alternative Hypothesis for a Hypothesis test of Proportion would be the
following:
H1: p_bar ≠ 0.3
has a proportion that is not equal to 0.3.
H1: p_bar > 0.3
or
H1: p_bar < 0.3
taken has a proportion that is either greater than or less than 0.3.
One-Tailed Test vs. a Two-Tailed Test

The number of tails in a hypothesis test depends on whether the test is directional or not. The operator of
the Alternative Hypothesis indicates whether or not the hypothesis test is directional. A non-directional
operator (a “not equal” sign) in the Alternative Hypothesis indicates that the hypothesis test is a two-
tailed test. A directional operator (a “greater than” or “less than” sign) in the Alternative Hypothesis
indicates that the hypothesis test is a one-tailed test.
The Region of Rejection (the alpha region) for a one-tailed test is entirely contained in the one of the
outer tails. A “greater than” operator in the Alternative Hypothesis indicates that the test is a one-tailed
test in the right tail. A “less than” operator in the Alternative Hypothesis indicates that the test is a one-
tailed test in the left tail. If α = 0.05, then one of the outer tails will contain the entire 5-percent Region of
Rejection.
The Region of Rejection (the alpha region) for a two-tailed test is split between both outer tails. Each
outer tail will contain half of the total Region of Rejection (alpha/2). If α = 0.05, then each outer tail will
contain a 2.5-percent Region of Rejection if the test is a two-tailed tailed.
123
Level of Certainty
Each hypothesis test has Level of Certainty that is specified. The Null Hypothesis is rejected only when
that Level of Certainty has been reached that the sample did not come from the population. A commonly
specified Level of Certainty is 95 percent. The Null Hypothesis would only be rejected in this case if the
sample statistic was different enough from the population parameter that at least 95 percent certainty was
achieved that the sample did not come from that population.
Level of Significance (Alpha)

The Level of Certainty for a hypothesis test is often indicated with a different term called the Level of
Significance also known as α (alpha). The relationship between the Level of Certainty and α is the
following:
α = 1 – Level of Certainty
An alpha that is set to 0.05 indicates that a hypothesis test requires a Level of Certainty of 95 percent that
the sample came from a different population to be reached before the Null Hypothesis is rejected.
Region of Acceptance
A Hypothesis Test of Mean or Proportion can be performed if the Test Statistic is distributed according to
the normal distribution or the t-Distribution. The Test Statistic is derived directly from the sample statistic
such as the sample mean. If the Test Statistic is distributed according to the normal or t-Distribution, then
the sample statistic is also distributed according to normal or t-Distribution. This will be discussed is
greater detail shortly.
A Hypothesis Test of Mean or Proportion can be understood much more intuitively by mapping the
sample statistic (the sample mean or proportion) to its own unique normal or t-Distribution. The sample
statistic is the distributed variable whose distribution is mapped according its own unique normal or t-
Distribution.
The Region of Acceptance is the percentage of area under this normal or t-Distribution curve that equals
the test’s specified Level of Certainty. If the hypothesis test requires 95 percent in order to reject the Null
Hypothesis, the Region of Acceptance will include 95 percent of the total area under the distributed
variable’s mapped normal or t-Distribution curve.
If the observed value of the sample statistic (the observed mean or proportion of the single sample taken)
falls inside of the Region of Acceptance, the Null Hypothesis is not rejected. If the observed value of the
sample statistic falls outside of the Region of Acceptance (into the Region of Rejection), the Null
Hypothesis is rejected.
Region of Rejection
The Region of Rejection is the percentage of area under this normal or t-Distribution curve that equals the
test’s specified Level of Significance (alpha). It is important to remember the following relationship:
Level of Significance (alpha) = 1 – Level of Certainty.
If the required Level of Certainty to reject the Null Hypothesis is 95 percent, then the following are true:
Level of Certainty = 0.95
Level of Significance (alpha) = 0.05
124
The Region of Acceptance includes 95 percent of the total area under the normal or t-Distribution curve
that maps the distributed variable, which is the sample statistic (the sample mean or proportion).
The Region of Rejection includes 5 percent of the total area under the normal or t-Distribution curve that
maps the distributed variable, which is the sample statistic (the sample mean or proportion). The 5-
percent alpha region is entirely contained in one of the tails if the test is a one-tailed test. The 5-percent
alpha region is split between both of the outer tails if the test is a one-tailed test.
falls inside of the Region of Rejection (outside the Region of Acceptance), the Null Hypothesis is rejected.
If the observed value of the sample statistic falls inside of the Region of Acceptance, the Null Hypothesis
is not rejected.
Critical Value(s)
Each hypothesis test has one or two Critical Values. A Critical Value is the location of boundary between
the Region of Acceptance and the Region of Rejection. A one-tailed test has one critical value because
the Region of rejection is entirely contained in one of the outer tails. A two-tailed test has two Critical
Values because the Region of Rejection is split between the two outer tails.
The Null Hypothesis is rejected if the sample statistic (the observed sample mean or proportion) is farther
from the curve’s mean than the Critical Value on that side. If the sample statistic is farther from the
curve’s mean than the Critical value on that side, the sample statistic lies in the Region of Rejection. If the
sample statistic is closer to the curve’s mean than the Critical value on that side, the sample statistic lies
in the Region of Acceptance.
Test Statistic
Each hypothesis test calculates a Test Statistic. The Test Statistic is the amount of difference between
the observed sample statistic (the observed sample mean or proportion) and the hypothesized population
parameter (the Constant on the right side of the Null Hypothesis) which will be located at the curve’s
mean.
This difference is expressed in units of Standard Errors. The Test Statistic is the number of Standard
Errors that are between the observed sample statistic and the hypothesized population parameter. The
Null Hypothesis is rejected if that number of Standard Errors specified by the Test Statistic) is larger than
a critical number of Standard Errors. The critical number of Standard Errors is determined by the required
Level of Certainty.
The Test Statistic is either the z Score or the t Value depending on whether a z-Test or t-Test is being
performed. This will be discussed in greater detail shortly.
Critical t Value or Critical z Value

Each hypothesis test calculates Critical t or z Values. A Critical t Value is calculated for a t-Test and a
Critical z Value is calculated for a z-Test. A Critical t or z Value is the amount of difference expressed in
Standard Errors between the boundary of the Region of Rejection (the Critical Value) and hypothesized
population parameter (the Constant on the right side of the Null Hypothesis) which will be located at the
curve’s mean.
A one-tailed test has only one Critical t or z Value because the Region of Rejection is entirely contained in
one outer tail A two-tailed test has two Critical z or t Values because the Region of Rejection is split
between the two outer tails.
125
The Test Statistic (the t Value or z Score) are compared with the Critical t or z Value on that side of the
mean. If the Test Statistic is farther from the standardized mean of zero than the Critical t or z Value on
that side, the Null Hypothesis is rejected.
The Test Statistic is the number of Standard Errors that the sample statistic is from the curve’s mean. The
Critical t or z Value on the same side is the number of Standard Errors that the Critical Value (the
boundary of the Region of Rejection) is from the mean. If the Test Statistic is farther from the
standardized mean of zero than the Critical t or z value, the sample statistic lies in the Region of
Rejection.
Relationship Between p Value and Alpha

Each hypothesis test calculates a p Value. The p Value is the area under the curve that is beyond the
sample statistic (the observed sample mean or proportion). The p Value is the probability that a sample of
size n with the observed sample mean or proportion could have occurred if the Null Hypothesis were true.
If, for example, the p Value of a Hypothesis Test of Mean or Proportion were calculated to be 0.0212, that
would indicated that there is only a 2.12 percent chance that a sample of size n would have the observed
sample mean or proportion if the Null Hypothesis were true. The Null Hypothesis states that the
population from which the sample came has the same mean as the hypothesized population. This mean
is the Constant on the right side of the Null Hypothesis.
The p Value is compared to alpha for a one-tailed test and to alpha/2 for a two-tailed test. The Null
Hypothesis is rejected if p is smaller than α for a one-tailed test or if p is smaller than α/2 for a two-tailed
test. If the p Value is smaller than α for a one-tailed test or smaller than α/2 for a two-tailed test, the
sample statistic is in the Region of Rejection.
Calculations of the Critical z Value(s) and the p Value are as follows:
Critical z Values
Critical z Value for a one-tailed test in the right tail:

Excel 2010 and beyond
Critical z Value = NORM.S.INV(1-α)
Prior to Excel 2010

Critical z Value = NORMSINV(1-α)
Critical z Value for a one-tailed test in the left tail:

Critical z Value = NORM.S.INV(α)
Prior to Excel 2010

Critical z Value = NORMSINV(α)
126
Critical z Values for a two-tailed test:
Critical z Values = ±NORM.S.INV(1-α/2)
Prior to Excel 2010

Critical z Values = ±NORMSINV(1-α/2)
p Value
p Value =MIN(NORM.DIST(x_bar,µ,SE,TRUE),1-NORM.DIST(x_bar,µ,SE,TRUE))
x_bar represents one of the following:
- the sample mean if this is a one-independent sample z-Test
- the difference between the sample means of a two-independent sample z-Test
- the mean difference between the paired values if this is a paired z-Test.
If the z Score (Test Statistic) is known, the p Value can be calculated more simply as follows:
p Value =MIN(NORM.S.DIST(z Score,TRUE),1-NORM.S.DIST(z Score,TRUE))
Prior to Excel 2010

p Value =MIN(NORMDIST(x_bar,µ,SE,TRUE),1-NORMDIST(x_bar,µ,SE,TRUE))
If the z Score (Test Statistic) is known, the p Value can be calculated more simply as follows:
p Value =MIN(NORMSDIST(z Score),1-NORMSDIST(z Score))
The 3 Equivalent Reasons To Reject the Null Hypothesis

The Null Hypothesis of a Hypothesis Test of Mean or Proportion is rejected if any of the following
equivalent conditions are shown to exist:
1) The sample statistic (the observed sample mean or proportion) is beyond the Critical Value. The
sample statistic would therefore lie in the Region of Rejection because the Critical Value is the boundary
of the Region of Rejection.
2) The Test Statistic (the t value or z Score) is farther from zero than the Critical t or z Value. The
Test Statistic is the number of Standard Errors that the sample statistic is from the curve’s mean. The
Critical t or z Value is the number of Standard Errors that the boundary of the Region of Rejection is from
the curve’s mean. If the Test Statistic is farther from farther from the standardized mean of 0 than the
Critical t or z Value, the sample statistic lies in the Region of Rejection.
3) The p value is smaller than α for a one-tailed test or α/2 for a two-tailed test. The p Value is the
curve area beyond the sample statistic. α and α/2 equal the curve areas contained by the Region of
Rejection on that side for a one-tailed test and a two-tailed test respectively. If the p value is smaller than
α for a one-tailed test or α/2 for a two-tailed test, the sample statistic lies in the Region of Rejection.
127
Independent Samples vs. Dependent Samples
A sample that is independent of a second sample has data values that are not influenced by any of the
data values within the second sample. Dependent samples are often referred to as paired data. Paired
data are data pairs in which one of the values of each pair has an influence on the other value of the data
pair. An example of a paired data sample would be a set of before-and-after test scores from the same
set of people.
Pooled vs. Unpooled Tests

A two-independent-sample Hypothesis Test of Mean can be pooled or unpooled. A pooled test can be
performed if the variances of both independent samples are similar. This is a pooled test because a
single pooled standard deviation replaces both sample standard deviations in the calculation of the
Standard Error. An unpooled test must be performed when the variances of the two independent samples
are not similar.
Type I and Type II Errors

A Type I Error is a false positive and a Type II Error is a false negative. A false positive occurs when a
test incorrectly detects of a significant difference when one does not exist. A false negative occurs when a
test incorrectly fails to detect a significant different when one exists.
α (the specified Level of Significance) = a test’s probability of a making a Type I Error.
β = a test’s probability of a making a Type II Error.
Power of a Test
The Power of a test indicates the test’s sensitivity. The Power of a test is the probability that the test will
detect a significant difference if one exists. The Power of a test is the probability of not making a Type II
Error, which is failing to detect a difference when one exists. A test’s Power is therefore expressed by the
following formula:
Power = 1 – β
Effect Size
Effect size in a t-Test or z-Test is a convention of expressing how large the difference between two
groups is without taking into account the sample size and whether that difference is significant.
Effect size of Hypotheses Tests of Mean is usually expressed in measures of Cohen’s d. Cohen’s d is a
standardized way of quantifying the size of the difference between the two groups. This standardization of
the size of the difference (the effect size) enables classification of that difference in relative terms of
“large,” “medium,” and “small.” A large effect would be a difference between two groups that is easily
noticeable with the measuring equipment available. A small effect would be a difference between two
groups that is not easily noticed.
128
Nonparametric Alternatives
Nonparametric tests are not substituted for z-Tests because a z-Test (a Hypothesis test of Mean that is
performed using the normal distribution) can only be performed on large samples (n > 30). The sample
mean and therefore the Test Statistic will always be normal-distributed as per the Central Limit Theorem.
Nonparametric tests are sometimes substituted for t-Tests because normality requirements cannot be
met. A t-Test is a Hypothesis Test of Mean that can be performed if the sample statistic (and therefore the
Test Statistic) is distributed according to the t-Distribution under the Null Hypothesis. The sample statistic
(the sample mean) is distributed according to the t-Distribution if any of the following three conditions
exist:
1) Sample size is large (n > 30). The sample taken for the hypothesis test must have at least 30 data
observations.
2) The population from which the sample was taken is verified to be normal-distributed.
3) The sample is verified to be normal-distributed.
If none of these conditions can be met or confirmed, a nonparametric test can often be substituted for a t-
Test. A nonparametric test does not have normality requirements that a parametric test such as a t-Test
does.
Hypothesis Test of Mean vs. Proportion

Hypothesis Test covered in this section will either be Hypothesis Tests of Mean or Hypothesis Test of
Proportion. A data point of a sample taken for a Hypothesis Test of Mean can have a range of values. A
data point of a sample taken for a Hypothesis Test of Proportion is binary; it can take only one of two
values.
Hypothesis Tests of Mean – Basic Definition

A Hypothesis Test of Mean compares an observed sample mean with a hypothesized population mean to
determine if the sample was taken from the same population. An example would be to compare a sample
of monthly sales of stores in one region to the national average to determine if mean sales from the
region (the population from which the sample was taken) is different than the national average (the
hypothesized population parameter). As stated, a sample taken for a Hypothesis Test of Mean can have
a range of values. In this case, the sales of a sample sampled store can fall within a wide range of values.
Hypothesis Tests of Proportion – Basic Definition

A Hypothesis Test of Proportion compares an observed sample proportion with a hypothesized
population proportion to determine if the sample was taken from the same population. An example would
be to compare the proportion of defective units from a sample taken from one production line to the
proportion of defective units from all production lines to determine if the proportion defective from the one
production line (the population from which the sample was taken) is different than from the proportion
defective of all production lines (the hypothesized population parameter). As stated, a sample taken for a
Hypothesis Test of Proportion can only have one of two values. In this case, a sampled unit from a
production line is either defective or it is not.
Hypothesis Test of Proportion are covered in detail in a separate section in this manual. They are also
summarized at the end of the binomial distribution section.
129
Hypothesis Tests of Mean
Hypothesis Tests of Mean require that the Test Statistic is distributed either according to the normal
distribution or to the t-Distribution. The Test Statistic in a Hypothesis Test of Mean is derived directly from
the sample mean and therefore has the same distribution as the sample mean.
t-Test versus z-Test

A Hypothesis Test of Mean will either be performed as a z-Test or as a t-Test. When the sample mean
and therefore the Test Statistic are distributed according to the normal distribution, the hypothesis test is
called a z-Test and the Test Statistic is called the z Score. When the sample mean and therefore the
Test Statistic is distributed according to the t-Distribution, the hypothesis test is called a t-Test and the
Test Statistic is called the t Value. The Test Statistic is the number of Standard Errors that the observed
sample mean is from the hypothesized population mean.
t-Tests are covered in detail is a separate section in this manual. They are also summarized at the end of
the t-Distribution section.
z-Tests are covered in detail is a separate section in this manual. They are also summarized at the end of
this normal distribution section.
Normal Distribution of Means of Large Samples

According to the Central Limit Theorem, the means of large samples will be normal-distributed no matter
how the population from which the samples came is distributed. This is true as long as the samples are
random and the sample size, n, is large (n > 30). n equals the number of data observations that each
sample contains.
If the single sample taken for a Hypothesis Test of Mean is large (n > 30), then the means of a number of
similar samples taken from the same population would be normal-distributed as per the Central Limit
Theorem. This is true no matter how the population or the single sample are distributed.
If the single sample taken for a Hypothesis Test of Mean is small (n < 30), then the means of a number of
similar samples taken from the same population would be normal-distributed only if the population was
proven to be normal-distributed or if the sample was proven to be normal-distributed.
Requirements for a z-Test

A z-Test can be performed only if the sample mean (and therefore the Test Statistic, which is derived
from the sample mean) is normal-distributed. The sample mean and therefore the Test Statistic are
normal-distributed only when the following two conditions are both met:
1) The size of the single sample taken is large (n > 30). The Central Limit Theorem states that means of
large samples will be normal-distributed. When the size of the single sample is small (n < 30), only a t-
Test can be performed.
2) The population standard deviation, σ (sigma), is known.
Requirements for a t-Test

A t-Test can be performed only if the sample mean (and therefore the Test Statistic, which is derived from
the sample mean) is distributed according to the t-Distribution. The sample mean and therefore the Test
Statistic are distributed according to the t-Distribution when both of these conditions are met:
1) The sample standard deviation, s, is known.
130
2) Either the sample or the population has been verified for normality.
A t-Test can be performed when the single sample is large (n > 30) but is the only option when the size of
the single sample is small (n < 30). A z-Test can only be performed when the size of the single sample is
large (n > 30) and the population standard deviation is known.
As mentioned, a Hypothesis Test of Mean requires that the sample mean and therefore the Test Statistic
is distributed either according to the normal distribution or to the t-Distribution. The sample mean and the
Test Statistic are distributed variables that can be graphed according to the normal or t-Distribution.
The Test Statistic, which represents the number of Standard Errors that the sample mean is from the
hypothesized population mean, could be graphed on a standard normal distribution curve or a
standardized t-Distribution curve. Both these two distribution curves have their means at zero and the
length of one Standard Error is set to equal 1
Basic Steps of a Hypothesis Test of Mean

The major steps of the simple Hypothesis Test of Mean, a one-sample z-Test, are described as follows:
1) A sample of data is taken. The sample statistic which is the sample mean is calculated.
2) A Null Hypothesis is created stating the population from which the sample was taken has the same
proportion as a hypothesized population proportion. An Alternative Hypothesis is constructed stating that
sample population’s proportion is not equal to, greater than, or less than the hypothesized population
proportion depending on the wording of the problem.
3) The sample proportion is mapped to a normal curve that has a mean equal to the hypothesized
population proportion and a Standard Error calculated based upon a formula specific to the type of
Hypothesis Test of Proportion.
4) The Critical Values are calculated and the Regions of Acceptance and Rejection are mapped on the
normal graph that maps the distributed variable. The Critical Values represent the boundaries between
the Region of Acceptance and Region of Rejection.
5) Critical z Values, the Test Statistic (the z Score) and p Value are then calculated.
6) The Null Hypothesis is rejected if any of the following three equivalent conditions are shown to exist:
a) The observed sample mean, x_bar, is beyond the Critical Value.
b) The z Score (the Test Statistic) is farther from zero than the Critical z Value.
c) The p Value is smaller than α for a one-tailed test or α/2 for a two-tailed test.
131
The following graph represents the final result of a typical one-sample, two-tailed z-Test. In this case the
Null Hypothesis was rejected.
This z-Test was a two-tailed test as evidenced by the yellow Region of Rejection split between the both
outer tails. In this t-Test the alpha was set to 0.05. This 5-percent Region of Rejection is split between the
two tails so that each tail contains a 2.5 percent Region of Rejection.
The mean of this non-standardized normal distribution curve is 186,000. This indicates that the Null
Hypothesis is as follows:
H0: x_bar = 186,000
Since this is a two-tailed t-Test, the Alternative Hypothesis is as follows:
H1: x_bar ≠ 186,000
This one-sample z-Test is evaluating whether the population from which the sample was taken has a
population mean that is not equal to 186,000. This is a non-directional z-Test and is therefore two-tailed.
The sample statistic is the observed sample mean of this single sample taken for this test. This observed
sample mean is calculated to be 200,000.
The boundaries of the Region of Rejection occur at 176,703 and 195,297. Everything outside of these two
points is in the Region of Rejection. These two Critical Values are 1.959 Standard Errors from the
standardized mean of 0. This indicates that the Critical z Values are ±1.959.
The graph shows that the sample statistic (the sample mean of 200,000) falls beyond the right Critical
value of 195,257 and is therefore in the Region of Rejection.
The sample statistic is 2.951 Standard Errors from the standardized mean of 0. This is further from the
standardized mean of 0 than the right Critical t value which is 1.959.
The curve area beyond the sample statistic consists of 2.4 percent of the area under the curve. This is
smaller than α/2 which is 2.5 percent of the total curve area because alpha was set to 0.05.
132
As the graph shows, all three equivalent conditions have been met to reject the Null Hypothesis. It can be
stated with at least 95 percent certainty that the mean of the population from which the sample was taken
does not equal the hypothesized population mean of 186,000.
Uses of Hypothesis Tests of Mean

1) Comparing the mean of a sample taken from one population with the another population’s
known mean to determine if both populations have the different means. An example of this would be to
compare the mean monthly sales of a sample of retail stores from one region to the national mean
monthly store sales to determine if the mean monthly sales of all stores in the one region are different
than the national mean.
2) Comparing the mean of a sample taken from one population to a fixed number to determine if
that population’s mean is different than the fixed number. An example of this might be to compare the
mean product measurement taken a sample of a number of units of a product to the company’s claims
about that product specification to determine if the actual mean measurement of all units of that
company’s product is different than what the company claims it is.
3) Comparing the mean of a sample from one population with the mean of a sample from another
population to determine if the two populations have different means. An example of this would be to
compare the mean of a sample of daily production totals from one crew with the mean of a sample of
daily production totals from another crew to determine if the two crews have different mean daily
production totals.
4) Comparing successive measurement pairs taken on the same group of objects to determine if
anything has changed between measurements. An example of this would be to evaluate whether there is
mean difference in before-and-after tests scores of a small sample of the same people to determine if a
training program made a difference to all of the people who underwent it.
5) Comparing the same measurements taken on pairs of related objects. An example of this would
be to evaluate whether there is mean difference in the incomes of husbands and wives in a sample of
married couples to determine if there is a mean difference in the incomes of husbands and wives in all
married couples.
It is important to note that a hypothesis test is used to determine if two populations are different, The
outcome of hypothesis test is to either reject or fail to reject the Null Hypothesis. It would be incorrect to
state that a hypothesis test is used to determine if two populations are the same.
Types of Hypothesis Tests of Mean

Hypothesis Tests of Mean are either t-Tests or z-Tests.
The 4 types of t-Tests discussed here are the following:
One-sample t-Test
Two-Independent-Sample, Pooled t-Test
Two-Independent-Sample, Unpooled t-Test
Two-Dependent-Sample (Paired) t-Test
The 3 types of z-Test discussed here are the following:
One-sample z-Test
Two-independent-Sample, Unpooled z-Test
Two-Dependent-Sample (Paired) z-Test
133
Detailed discussions of each of the three types of z-Tests along with examples in Excel are as follows:
1) One-Sample z-Test in Excel
Overview
This hypothesis test determines whether the mean of the population from which the sample was taken is
equal to (two-tailed test) or else greater than or less than (one-tailed test) than a constant. This constant
is often the known mean of a population from which the sample may have come from. The constant is the
constant on the right side of the Null Hypothesis.
x_bar = Observed Sample Mean
Null Hypothesis H0: x_bar = Constant

The Null Hypothesis is rejected if any of the following equivalent conditions are shown to exist:
1) The observed x_bar is beyond the Critical Value.
2) The z Score (the Test Statistic) is farther from zero than the Critical z Value.
3) The p value is smaller than α for a one-tailed test or α/2 for a two-tailed test.
Example of a One-Sample, Two-Tailed z-Test in Excel

This problem is very similar to the problem solved in the t-test section for a one-sample, one-tailed t-test.
Similar problems were used in each of these sections to show the similarities and also contrast the
differences between the one-sample z-Test and t-test as easily as possible.
This problem compares average monthly sales from one fast food chain’s retail stores in one region with
the average monthly sales of all of the fast food chain’s retails in the entire country. The region being
evaluated has more than 1,000 very similar stores. The national mean monthly retail store sales equals
$186,000. The standard deviation of the monthly sales for the entire population of stores is $30,000.
Determine with at least 95% certainty whether the average monthly sales of all of the fast food chain’s
stores in the one region is different than the national average monthly sales of all of the fast food chain’s
stores.
134
The data sample of sales for the month for a random sample of 40 retail stores in a region is as follows:
Summary of Problem Information

x_bar = sample mean = AVERAGE() = 200,000
µ = national (population) mean = 186,000
α = 1-Level of Certainty Required = 1 – 0.95 = 0.05
s = sample standard deviation = Not Known and not needed for a z-Test
σ (Greek letter “sigma”) = population standard deviation = 30,000
n = sample size = COUNT() = 40
135
SE = Standard Error = σ / SQRT(n) = 30,000 / SQRT(40) = 4,743
Note that this calculation of the Standard Error using the population standard deviation, σ, is the true
Standard Error. If the sample standard error, s, were used in place of σ, the Standard Error calculated
would be an estimate of the true Standard Error. The z-Test requires the population standard deviation
but the t-test uses the sample standard deviation as an estimate of the population standard deviation.
As with all Hypothesis Tests of Mean, we must satisfactorily answer these two questions and then
proceed to the four-step method of solving the hypothesis test that follows.
The Initial Two Questions That Must be Answered Satisfactorily

What Type of Test Should Be Done?
Have All of the Required Assumptions For This Test Been Met?
The Four-Step Method For Solving All Hypothesis Tests of Mean

Step 1) Create the Null Hypothesis and the Alternative Hypothesis
Step 2 – Map the Normal or t-Distribution Curve Based on the Null Hypothesis
Step 3 – Map the Regions of Acceptance and Rejection
Step 4 – Perform the Critical Value Test, the p Value Test, or the Critical t Value Test
The Initial Two Questions That Need To Be Answered Before Performing the Four-Step Hypothesis Test
of Mean are as follows:
Question1) Type of Test?

a) Hypothesis Test of Mean or Proportion?
This is a Hypothesis Test of Mean because each individual observation (each sampled monthly retail
store sales figure) within the sample can have a wide range of values. Data points for Hypothesis Tests of
Proportion are binary: they can take only one of two possible values.
b) One-Sample or a Two-Sample Test?

This is a one-sample hypothesis test because only one sample containing monthly sales figures from
forty stores has been taken and is being compared to the national monthly retail store average for the
same month.
c) Independent (Unpaired) Test or a Dependent (Paired) Test?

It is neither. The designation of “paired” or “unpaired” applies only for two-sample hypothesis tests.
d) One-Tailed or Two-Tailed Hypothesis?

The problem asks to determine whether the forty-store monthly average is simply different than the
national average. This is a non-directional inequality making this hypothesis test a two-tailed test. If the
problem asked whether the forty-store average was either higher or was lower, the inequality would be
directional and the resulting hypothesis test would be a one-tailed test. A two-tailed test is more stringent
than a one-tailed test.
136
e) t-Test or z-Test?
A hypothesis test of means can be performed if the distribution of the Test Statistic under the Null
Hypothesis can be approximated by either the normal distribution or the t-Distribution.
A z-Test is a statistical test in which the distribution of the Test Statistic under the Null Hypothesis can be
approximated by the normal distribution. A t-test is a statistical test in which the distribution of the Test
Statistic under the Null Hypothesis can be approximated by the t-Distribution.
This hypothesis test of mean can be performed as z-Test because sample size is large (n = 40) and the
population standard deviation (σ = 30,000) is known. The large sample size and known population
standard deviation ensure that the distribution of the sample mean (and therefore Test Statistic, which is
derived from the sample mean) can be approximated by the normal distribution under the Null
Hypothesis.
It should be noted that a one-sample t-Test can always be used in place of a one-sample z-Test. All z-
Tests can be replaced be their equivalent t-Tests. As a result, some major commercial statistical software
packages including the well-known SPSS provide only t-Tests and no direct z-Tests.
This hypothesis test is a t-Test that is one-sample, two-tailed hypothesis test of mean as long as
all required assumptions have been met.
Question 2) Test Requirements Met?

a) Normal Distribution of Test Statistic
The normal distribution can be used to map the distribution of the sample mean (and therefore the Test
Statistic, which is derived from the sample mean) only if the following conditions exist:
1) Population standard deviation, σ, is known

Population standard deviation, σ, is one of the two required parameters needed to fully describe a unique
normal distribution curve and must therefore be known in order to perform a z-Test (which uses the
normal distribution).
and
2) Sample size is large (n > 30)

The Central Limit Theorem states that if a number of large, random samples of the same size were taken
from the same population, the means of the samples would be normal-distributed. If the sample mean is
normal-distributed, the Test Statistic, which equals (sample mean – Constant ) / SE, will also be normal-
distributed because it is derived from the sample mean.
If sample size is large, neither the normality of the population nor the normality of the sample data has to
be confirmed.
The sample man and therefore the Test Statistic are normal-distributed because sample size is large (n =
40) and the population standard deviation (σ = 30,000) is known.
137
We now proceed to complete the four-step method for solving all Hypothesis Tests of Mean. These four
steps are as follows:
Step 4 – Determine Whether to Accept or Reject theNull Hypothesis By Performing the Critical
Value Test, the p Value Test, or the Critical t Value Test
Proceeding through the four steps is done is follows:
Step 1 – Create the Null and Alternative Hypotheses

The Null Hypothesis is always an equality that states that the items being compared are the same. In this
case, the Null Hypothesis would state that the average monthly sales of all stores in the region (the
population from which the forty-store sample was taken) is not different than the national monthly store
average sales, µ, which is $186,000. We will use the variable x_bar to represent the sample mean of the
forty stores. The Null Hypothesis is as follows:
H0: x_bar = Constant = 186,000
The Constant is quite often the known population mean, µ, to which the sample mean is being compared.
The Alternative Hypothesis is always in inequality and states that the two items being compared are
different. This hypothesis test is trying to determine whether the average monthly sales of all stores in the
region (the population from which the forty-store sample was taken) is merely different than the national
monthly store average sales, µ, which is $186,000.
The Alternative Hypothesis is as follows:
H1: x_bar ≠ Constant
H1: x_bar ≠ 186,000
The Alternative Hypothesis is non-directional (“not equal” instead of “greater than” or “less than”) and the
hypothesis test is therefore a two-tailed test. It should be noted that a two-tailed test is more rigorous
(requires a greater differences between the two entities being compared before the test shows that there
is a difference) than a one-tailed test.
It is important to note that the Null and Alternative Hypotheses refer to the means of the populations from
which the samples were taken. A one-sample t-Test determines whether to reject or fail to reject the Null
Hypothesis that states that that population from which the sample was taken (the entire region) has a
mean equal to the Constant. The Constant in this case is equal to known national average.
Parameters necessary to map the distributed variable, x_bar, are the following:
σ (Greek letter “sigma”) = population standard deviation = 30,000
SE = Standard Error = σ / SQRT(n) = 30,000 / SQRT(40) = 4,743

138
These parameters are used to map the distributed variable, x_bar, to the normal distribution are as
follows:
Step 2 – Map the Distributed Variable to Normal Distribution

A z-Test can be performed if the sample mean, and the Test Statistic (the z Score) are distributed
according to the normal Distribution. If the sample size is large and the population standard deviation is
known, the sample mean and closely-related Test Statistic are distributed according to the normal
Distribution.
The sample mean x_bar is distributed according to the normal distribution. The distributed variable would
in this case be the sample mean, x_bar.
Mapping this distributed variable to a normal Distribution curve is shown as follows:
This non-standardized normal Distribution curve has its mean set to equal the Constant taken from the
Null Hypothesis, which is:
This non-standardized normal Distribution curve is constructed from the following parameters:
Mean = 186,000 (This x_bar is hypothesized by the Null Hypothesis to be the curve’s mean.)
Standard Error = 4,743
Distributed Variable = x_bar
139
The goal of a hypothesis test is to determine whether to reject or fail to reject the Null Hypothesis at a
given level of certainty. If the two things being compared are far enough apart from each other, the Null
Hypothesis (which states that the two things are not different) can be rejected. In this case we are trying
to show graphically how different the sample mean, x_bar = $200.000, is from the national average of
$186,000.
The non-standardized normal distribution curve can be divided up into two types of regions: the Region of
Acceptance and the Regions of Rejection. A two-tailed test has the Region of Rejection split between the
two outer tails. A boundary between a Region of Acceptance and a Region of Rejection is called a Critical
Value.
If the sample mean’s value of x_bar = 200,000 falls into a Region of Rejection, the Null Hypothesis is
rejected. If the sample mean’s value of x_bar = 200,000 falls into a Region of Acceptance, the Null
Hypothesis is not rejected.
The total size of the Region of Rejection is equal to Alpha. In this case Alpha, α, is equal to 0.05. This
means that the Region of Rejection will take up 5 percent of the total area under this t-Distribution curve.
This 5 percent is divided up between the two outer tails. Each outer tail contains 2.5 percent of the curve
that is the Region of Rejection.
The boundary between Regions of Acceptance and Regions of Rejection are called Critical Values. The
locations of these Critical values need to be calculated.
Calculate Critical Values

A Critical Value is the boundary between a Region of Acceptance and a Region of Rejection. In the case
of a two-tailed test, the Region of rejection is split between two outer tails. There are therefore two Critical
Values.
The Critical Value is the boundary on either side of the curve beyond which 2.5 percent of the total area
under the curve exists. In this case both Critical Values can be found by the following:
Critical Values = Mean ± (Number of Standard Errors from Mean to Region of Rejection) * SE
Critical Values = Mean ± NORM.S.INV(1-α/2) * SE
Critical Values = 186,000 ± NORM.S.INV(1 - 0.05/2) * 4,743
Critical Values = 186,000 ± NORM.S.INV(0.975) * 4,743
Critical Values = 186,000 ± 9,296
Critical Values = 176,703 and 195,297
The Region of Rejection is therefore everything that is to the right of 195,297 and everything to the left of
176,703.
It should be noted that the Mean in the Critical Values formula refers to the mean of the mapped normal
distribution curve. The mean of this curve is the Constant from the Null Hypothesis. The Mean in the
Critical values formula is therefore the hypothesized mean from the Null Hypothesis. The Null Hypothesis
is as follows:
The hypothesized mean is the hypothesized x_bar which is 186,000.
140
The following Excel-generated distribution curve with the blue Region of Acceptance and the yellow
Regions of Rejection is shown is as follows:
Step 4 – Determine Whether to Reject Null Hypothesis

The object of a hypothesis test is to determine whether to accept or reject the Null Hypothesis. There are
three equivalent-Tests that determine whether to accept or reject the Null Hypothesis. Only one of these
tests needs to be performed because all three provide equivalent information. The three tests are as
follows:
1) Compare x-bar With Critical Value
Reject the Null Hypothesis if the sample mean, x_bar = 200,000, falls into the Region of Rejection.
Equivalently, reject the Null Hypothesis if the sample mean, x_bar, is further from the curve’s mean of
186,000 than the Critical Value.
The Critical Values have been calculated to be 176,703 on the left and 195,297 on the right. X_bar
(200,000) is further from the curve mean (186,000) than the right Critical Value (195,297). The Null
Hypothesis would therefore be rejected.
2) Compare the t Value With Critical t Value

The z Score corresponds to the standardized value of the sample mean, x_bar = 200,000. The z Score is
the number of Standard Errors that x_bar is from the curve’s mean of 186,000.
The Critical z Value is the number of Standard Errors that the Critical Value is from the curve’s mean.
Reject the Null Hypothesis if the z Score is farther from the standardized mean of zero than the Critical z
Value.
141
Equivalently, reject the Null Hypothesis if the z Score is farther from the standardized mean of zero than
the Critical z Value.
The Constant is the Constant from the Null Hypothesis (H0: x_bar = Constant = 186,000)
Z Score (Test Statistic) = (200,000 – 186,000)/4,743
Z Score (Test Statistic) = 2.951
This means that the sample mean, x_bar, is 2.951 standard errors from the curve mean (186,000).
Two-tailed Critical z Values = ±NORM.S.INV(1-α/2)

Two-tailed Critical z Values = ±NORM.S.INV(1-0.05/2)
Two-tailed = ±NORM.S.INV(0.975) = ±1.9599
This means that the Region of Rejection for this two-tailed hypothesis test in either tail begins at 1.9599
standard errors from (to the left of and to the right of) the standardized mean of zero.
This means that the boundaries of the Region of Rejection are 1.9599 standard errors from the curve
mean (186,000) on each side since this is a two-tailed test.
The Null Hypothesis is rejected because the z Score (+2.951) is further from the standardized mean of
zero than the Critical z Values (±1.9599). This is another indication that x_bar (200,000) is in the Region
of Rejection.
3) Compare the p Value With Alpha

The p Value is the percent of the curve that is beyond x_bar (200,000). If the p Value is smaller than
Alpha/2, the Null Hypothesis is rejected.
p Value =MIN(NORM.S.DIST(2.951,TRUE),1-NORM.S.DIST(2.951,TRUE))
p Value = 0.0016
142
The p Value (0.0016) is smaller than Alpha/2 (0.025) Region of Rejection in the right tail and we therefore
reject the Null Hypothesis. The following Excel-generated graph shows that the red p Value (the curve
area beyond x_bar) is smaller than the yellow Alpha, which is the 5 percent Region of Rejection split
between both outer tails.
143
Excel Formula Shortcut to Performing a One-Sample z-Test
This problem could also be quickly solved with the following Excel z-Test formula:
p Value =MIN(Z.TEST(array,Constant,σ),1-Z.TEST(array,Constant,σ))
It should be noted that when the Constant is positive, the p Value Excel formula is
p Value = Z.TEST(array,Constant,σ),
When the Constant is negative, the p Value Excel formula is
p Value = 1-Z.TEST(array, Constant,σ))
The Constant is taken from the Null Hypothesis and is equal to 0.
The Null Hypothesis is as follows:
The Excel z-Test formula produces the p Value as follows:
Note that the array can be spread across two columns as is done here. The array does not have to be
entirely contained in a single column in this case.
144
2) Two-Independent-Sample, Unpooled z-Test in Excel
Overview
This hypothesis test evaluates two independent samples to determine whether the difference between the
two sample means (x_bar1 and x_bar2) is equal to (two-tailed test) or else greater than or less than (one-
tailed test) than a constant.
This is an unpooled test. An unpooled test can always be used in place of a pooled test. An unpooled test
must be used when population variances are not similar. An unpooled test calculates the Standard Error
using separate standard deviations instead of combining them into a single, pooled standard deviation as
a pooled test does.
The t-Test is nearly always used to compare two independent samples. For this reason, only the
unpooled, two-independent-sample z-Test will be covered. The pooled version of this z-Test will not be
covered.
In the real world, the only the sample variances are known but the population variances are usually not
known and therefore t-tests are nearly always used to perform a two-independent-sample hypothesis test
of mean. For this reason, only the unpooled, two-independent-sample z-Test will be explained. This z-
Test can always be used in place of the pooled z-Test that could be used if population variances were
known to be similar enough.
x_bar1 - x_bar2 = Observed difference between the sample means
Note that this is the same formula for SE for the two-independent-sample, unpooled t-test except that the
variance for the z-Test is the population variance as follows:
2
var1 = σ1
2
var2 = σ2
and not the sample variance used for the t-test as follows:
2
var1 = s1
2
var2 = s2
Null Hypothesis H0: x_bar1 - x_bar2 = Constant

1) The observed x_bar1 - x_bar2 is beyond the Critical Value.
2) The z Score (the Test Statistic) is farther from zero than the Critical t Value.
145
Example of 2-Sample, 2-Tailed, Unpooled z-Test in Excel
This problem is very similar to the problem solved in the t-test section for a two-independent-sample, two-
tailed t-test. Similar problems were used in each of these sections to show the similarities and also
contrast the differences between the two-independent-sample z-Test and t-test as easily as possible.
Two shifts on a production are being compared to determine if there is a difference in the average daily
number of units produced by each shift. The two shifts operate eight hours per day under nearly identical
conditions that remain fairly constant from day to day. A sample of the total number of units produced by
each shift on a random selection of days is taken. Determine with a 95 percent Level of Confidence if
there is a difference between the average daily number of units produced by the two shifts.
Note that when performing two-sample z-tests in Excel, always designate Sample 1 (Variable 1) to be the
sample with the larger mean.
The results of the two-independent-sample z-Test will be more intuitive if the sample group with the larger
mean is designated as the first sample and the sample group with the smaller mean is designated as the
second sample.
Details about both data samples are shown as follows:

Sample Group 1 – Shift A (Variable 1)
x_bar1 = sample1 mean = 46.55
µ1 (Greek letter “mu”) = population mean from which Sample 1 was drawn = Not Known
σ1 (Greek letter “sigma”) = population standard deviation from which Sample 1 was drawn = 25.5
2
Var1 = population1 variance = σ1 = 650.25
n1 = sample1 size = 40
Sample Group 2 – Shift B (Variable 2)

x_bar2 = samples mean = 42.24
σ2 (Greek letter “sigma”) = population standard deviation from which Sample 2 was taken = 11.2
2
Var2 = population2 variance = σ2 = 125.44
n2 = sample1 size = 36
x_bar1 - x_bar2 = 46.55 – 42.24 = 4.31
Alpha = 1 - Level of Certainty = 1 – 0.95 = 0.05
As mentioned, always designate Sample 1 (Variable 1) to be the sample with the larger mean when
performing two-sample z-Tests in Excel.
The results of the Unpooled z-Test will be more intuitive if the sample group with the larger mean is
designated as the first sample and the sample group with the smaller mean is designated as the second
sample.
Another reason for designating the sample group with the larger mean as the first sample is to obtain the
correct result from the Excel data analysis tool for two-independent-sample, unpooled z-Tests called the
z-Test:Two-Sample for Means. The test statistic (z in the Excel output, which stands for z Score) and
146
the Critical z value (z Critical in the Excel output) will have the same sign (as they always should) only if
the sample group with the larger mean is designated the first sample.


The initial two questions that need to be answered before performing the Four-Step Hypothesis Test of
Mean are as follows:
Question 1) What Type of Test Should Be Done?

This is a test of mean because each individual observation (each sampled shift’s output) within each of
the two sample groups can have a wide range of values. Data points for tests of proportion are binary:
they can take only one of two possible values.
b) One-Sample or Two-Sample Test?

This is a two-sample hypothesis test because two independent samples are being compared with each
other. The two sample groups are the daily units produced by Shift A and the daily units produced by Shift
B.
c) Independent (Unpaired) Test or Dependent (Paired) Test?

It is an unpaired test because data observations in each sample group are completely unrelated to data
observations in the other sample group. The designation of “paired” or “unpaired” applies only for two-
sample hypothesis tests.
) One-Tailed or Two-Tailed Test?

The problem asks to determine whether there is a difference in the average number of daily units
produced by Shift A and by Shift B. This is a non-directional inequality making this hypothesis test a two-
tailed test. If the problem asked to determine whether Shift A really does have a higher average than Shift
B, the inequality would be directional and the resulting hypothesis test would be a one-tailed test. A two-
tailed test is more stringent than a one-tailed test.
A z-Test is a statistical test in which the distribution of the Test Statistic under the Null Hypothesis can be
approximated by the normal distribution.
147
The Test Statistic is distributed by the normal distribution if both samples are large and both population
standard deviations are known. Both samples considered to be large samples because both sample sizes
(n1 = 40 and n2 = 36) exceeds 30. Both population standard deviations (σ1 = 25.5 and σ2 = 11.2) are
known.
Because both sample sizes (n1 = 40 and n2 = 36) exceeds 30, both sample means are therefore normal-
distributed as per the Central Limit Theorem. The difference between two normally-distributed sample
means is also normal-distributed. The Test Statistic is derived from the difference between the two means
and is therefore normal-distributed. A z-Test can be performed if the Test Statistic is normal-distributed.
It should be noted that a two-independent-sample, unpooled t-Test can always be used in place of a two-
independent-sample, unpooled. All z-Tests can be replaced be their equivalent t-Tests. As a result, some
major commercial statistical software packages including the well-known SPSS provide only t-Tests and
no direct z-Tests.
f) Pooled or Unpooled t-Test?

A pooled z-Test can be performed if the variances of both populations are similar, i.e., one population’s
standard deviation is no more than twice as large as the other population’s standard deviation. An
unpooled z-Test must be performed otherwise.
An unpooled z-Test can always be performed in the place of a pooled z-Test. Excel only provides a tool
and formula for an unpooled z-test but not a pooled z-Test. For this reason the only type of two-
independent-sample z-Test covered in this section will be the unpooled one.
t-Tests can always be performed in place of z-Tests. Excel does have separate tools and formulas for
pooled and unpooled, two-independent-sample t-Tests.
This hypothesis test is a z-Test that is two-independent-sample, unpooled two-tailed hypothesis
test of mean as long as all required assumptions have been met.

a) Normal Distribution of Both Sample Means
The normal distribution can be used to map the distribution of the difference of the sample means (and
therefore the Test Statistic, which is derived from this difference) only if the following conditions exist:
1) Both Population Standard Deviations, σ1 and σ2, Are Known

Those values are σ1 = 25.5 and σ2 = 11.2. Population standard deviation, σ, is one of the two required
parameters needed to fully describe a unique normal distribution curve and must therefore be known in
order to perform a z-Test (which uses the normal distribution).
and
2) Both samples sizes are large (n > 30).

Because both sample sizes (n1 = 40 and n2 = 36) exceeds 30, both sample means are therefore normal-
distributed as per the Central Limit Theorem. The difference between two normally-distributed sample
means is also normal-distributed. The Test Statistic is derived from the difference between the two means
and is therefore normal-distributed.
The distributions of both samples and populations do not have to be verified because both sample means
are known to be normal-distributed as a result of the large size.
The difference between the sample means and therefore the Test Statistic are normal-distributed
because both samples are large and both population standard deviations are known.
148
We can now proceed to complete the four-step method for solving all Hypothesis Tests of Mean. These
four steps are as follows:
Step 4 – Determine Whether to Accept or Reject the Null Hypothesis By Performing the Critical
Value Test, the p Value Test, or the Critical z Value Test
Proceed through the four steps as follows:

The Null Hypothesis is always an equality and states that the items being compared are the same. In this
case, the Null Hypothesis would state that the average optimism scores for both sample groups are the
same. We will use the variable x_bar1-x_bar2 to represent the difference between the means of the two
groups. If the mean scores for both groups are the same, then the difference between the two means,
x_bar1-x_bar2, would equal zero. The Null Hypothesis is as follows:
H0: x_bar1-x_bar2 = Constant = 0
different. This hypothesis test is trying to determine whether the first mean (x_bar 1) is different than the
second mean (x_bar2). The Alternative Hypothesis is as follows:
H1: x_bar1-x_bar2 ≠ Constant
H1: x_bar1-x_bar2 ≠ 0
The parameters necessary to map the distributed variable, x_bar1-x_bar2, to the normal distribution are
calculated as follows:
149
Step 2 – Map the Distributed Variable on a Normal Distribution Curve
n1 = 40
n2 = 36
2 2
Var1 = σ1 = (25.5) = 650.25
2 2
Var2 = σ2 = (11.2) = 125.44
Unpooled Population Standard Error
SE = SQRT[ (Var1/n1) + (Var2/n2) ]

SE = SQRT[ (650.25/40) + (125.44/36) ]
SE = 4.443
This non-standardized normal distribution curve has its mean set to equal the Constant taken from the
Null Hypothesis, which is:
This non-standardized normal distribution curve is constructed from the following parameters:
Mean = 0 (This x_bar1-x_bar2 is hypothesized by the Null Hypothesis to be the curve’s mean.)
Standard Error = 4.443
Distributed Variable = x_bar1-x_bar2
150
to show graphically how different x_bar1 is from x_bar2 by showing how different x_bar1-x_bar2 (4.31) is
from zero.
The non-standardized t-Distribution curve can be divided up into two types of regions: the Region of
Acceptance and the Region of Rejection. A boundary between a Region of Acceptance and a Region of
Rejection is called a Critical Value.
If the difference between the sample means, x_bar1-x_bar2 (4.31), falls into a Region of Rejection, the
Null Hypothesis is rejected. If the difference between the sample means, x_bar1-x_bar2 (4.31), falls into a
Region of Acceptance, the Null Hypothesis is not rejected.
This 5 percent Alpha (Region of Rejection) is entirely contained in the outer right tail. The operator in the
Alternative Hypothesis whether the hypothesis test is two-tailed or one-tailed and, if one tailed, which
outer tail. The Alternative Hypothesis is the follows:
A “not equal” operator indicates that this will be a two-tailed test. This means that the Region of Rejection
is split between both outer tails.
The boundaries between Regions of Acceptance and Regions of Rejection are called Critical Values. The
locations of these Critical Values need to be calculated.
Calculate the Critical Values
Two-Tailed Critical Values

Critical Values = Mean ± NORM.S.INV(1-α/2) * SE
Critical Values = 0 ± NORM.S.INV(1 - 0.05/2) * 4.443
Critical Values = 0 ± NORM.S.INV(0.975) * 4.443
Critical Values = 0 ± 8.708
Critical Values = -8.708 and 8.708
The Region of Rejection is therefore everything that is to the right of 8.708 and everything to the left of -
8.708.
is as follows:
H0: : x_bar1-x_bar2 = Constant = 0
The hypothesized mean is the hypothesized x_bar1-x_bar2 which is 0.
151
The following Excel-generated distribution curve with the blue Region of Acceptance and the yellow
Regions of Rejection is shown is as follows:

The object of a hypothesis test is to determine whether to accept or reject the Null Hypothesis. There are
follows:
1) Compare Sample Mean With Critical Value

Reject the Null Hypothesis if the sample mean, x_bar1-x_bar2 = 4.31, falls into the Region of Rejection.
Fail to reject the Null Hypothesis if the sample mean, x_bar1-x_bar2 = 4.31, falls into the Region of
Acceptance.
Equivalently, reject the Null Hypothesis if the sample mean, x_bar1-x_bar2, is further from the curve’s
mean of 0 than the Critical Value. Fail to reject the Null Hypothesis if the sample mean, x_bar1-x_bar2, is
closer than the curve’s mean of 0 than the Critical Value.
The Critical Values have been calculated to be +8.708 on the left and -8.708 on the right. x_bar1-x_bar2
(4.31) is closer to the curve mean (0) than the right Critical Value (+8.708). The Null Hypothesis would
therefore not be rejected.
152
2) Compare the z Score with the Critical z Value
The z Score is the number of Standard Errors that x_bar1-x_bar2 (4.31) is from the curve’s mean of 0.
The Critical z Value is the number of Standard Errors that the Critical Value is from the curve’s mean.
Reject the Null Hypothesis if the z Score is farther from the standardized mean of zero than the Critical z
Value. Fail to reject the Null Hypothesis if the z Score is closer to the standardized mean of zero than the
Critical z Value.
Equivalently, reject the Null Hypothesis if the z Score is farther from the standardized mean of zero than
the Critical z Value. Fail to reject the Null Hypothesis if the z Score is closer to the standardized mean of
zero than the Critical z Value.
The Constant is the Constant from the Null Hypothesis (H0: x_bar1-x_bar2 = Constant = 0)
Z Score (Test Statistic) = (4.31 – 0)/4.443
Z Score (Test Statistic) = 0.97
This means that the sample mean, x_bar1-x_bar2 (4.31), is 0.97 standard errors from the curve mean (0).
Two-tailed Critical z Values = ±NORM.S.INV(1-α/2)
Two-tailed Critical z Values = ±NORM.S.INV(1-0.05/2)
Two-tailed = ±NORM.S.INV(0.975) = ±1.9599
This means that the boundaries between the Region of Acceptance and the Region of Rejection are
1.9599 standard errors from the curve mean on each side since this is a two-tailed test.
The Null Hypothesis is not rejected because the z Score (+0.97) is closer to the standardized mean of
zero than the Critical z Value on the right side (+1.9599).

The p Value is the percent of the curve that is beyond x_bar 1-x_bar2 (4.31). If the p Value is smaller than
Alpha/2, the Null Hypothesis is rejected. If the p Value is larger than Alpha/2, the Null Hypothesis is not
rejected.
p Value =MIN(NORM.S.DIST(0.97,TRUE),1-NORM.S.DIST(0.97,TRUE))
p Value = 0.1660
The p Value (0.1660) is larger than Alpha/2 (0.025) Region of Rejection in the right tail and we therefore
do not reject the Null Hypothesis.
153
The following Excel-generated graph shows that the red p Value (the curve area beyond x_bar1-x_bar2) is
larger than the yellow Alpha, which is the 5 percent Region of Rejection split between both outer tails.
154
Excel Data Analysis Tool Shortcut
This two-independent-sample, unpooled z-Test can be solved much quicker using the following Excel
data analysis tool:
z-Test: Two Sample For Means. This tool uses the formulas for an unpooled, two-sample z-Test as are
shown above. This tool can be accesses by clicking Data Analysis under the Data tab. The entire Data
Analysis Toolpak is an add-in that ships with Excel but must first be activated by the user before it is
available. This tool calculates the z Score and p Value using the same equations as shown.
Note that this tool requires that all data in each sample group be placed in a single column. In the
following image, only the first 19 data points of each sample are showing.
The completed dialogue box for this tool which produced the preceding output is as follows:
155
156
3) Paired (Two-Sample Dependent) z-Test in Excel
Overview
This hypothesis test determines whether the mean of a sample of differences between pairs of data
(x_bardiff) is equal to (two-tailed test) or else greater than or less than (one-tailed test) than a constant.
Before-and-after fitness levels of individuals undergoing a training program would be an example of
paired data. The sample evaluated would be the group of differences between the before-and-after
scores of the individuals. This is called the difference sample.
The t-test is nearly always used instead of a z-Test to perform a two-dependent-sample (paired)
hypothesis test of mean. The z-Test requires the population standard deviation of the differences
between the pairs be known. The sample standard deviation of the difference sample is readily available
but the population standard deviation of the differences is usually not known. The t-test requires only the
sample standard deviation of the sample of paired differences be known.
x_bardiff = difference sample mean
Null Hypothesis H0: x_bardiff = Constant
Example of Paired, 1-Tailed (Left-Tail) z-Test in Excel

This problem is very similar to the problem solved in the t-test section for a paired, one-tailed t-test.
differences between the paired z-Test and t-test as easily as possible.
A new clerical program was introduced to a large company with the hope that clerical errors would be
reduced. 5,000 clerical workers in the company underwent the training program. 50 Clerical employees
who underwent the training were randomly selected. The average number of clerical errors that each of
these 50 employees made per month for six months prior to the training and also for six months following
the training were recorded. Each of the 50 employees had a similar degree of clerical experience within
the company and performed nearly the same volume and type of clerical work in the before and after
months. The standard deviation of the after-before differences for all 5,000 employees who underwent the
training is known to be 6.4.
Based upon the results of the 50 sampled clerical employees, determine with 95 percent certainty
whether the average number of monthly clerical mistakes was reduced for the entire 5,000 clerical
employees who underwent the training.
It is the difference that we are concerned with. A hypothesis test will be performed on the sample of
differences. The distributed variable will be designated as x_bar diff and will represent that average
difference between After and Before samples.
157
x_bardiff was calculated by subtracting the Before measurement from the After measurement. This is the
intuitive way to determine if a reduction in error occurred.
This problem illustrates why the t-test is nearly always used instead of a z-Test to perform a two-
dependent-sample (paired) hypothesis test of mean. The z-Test requires the population standard
deviation of the differences between the pairs be known. This is rarely ever the case, but will be given for
this problem so that a paired z-Test can be used. The t-test requires only the sample standard deviation
of the sample of paired differences be known.
intuitive way to determine if a reduction in error occurred
158
x_bardiff = sample mean =AVERAGE() = -2.14
σdiff = population standard deviation = 6.4
n = sample size = number of pairs = COUNT() = 40
SEdiff = Standard Error = σdiff / SQRT(n) = 6.4 / SQRT(40) = 1.01
Note that this calculation of the Standard Error, SEdiff, using the population standard deviation, σdiff, is the
true Standard Error. If the sample standard error, sdiff, were used in place of σdiff, the Standard Error
calculated would be an estimate of the true Standard Error. The z-Test requires the population standard
deviation of the paired differences but the t-test uses the sample standard deviation as an estimate of the
population standard deviation of the paired differences.
The Excel data analysis tool Descriptive Statistics in not employed when the z-Test is used. Descriptive
Statistics should only be used if a t-Test will be performed. The Standard Deviation and Standard Error
calculated by Descriptive Statistics is based upon the sample standard deviation. the z-Test uses the
population standard deviation instead of the sample standard deviation used by the t-Test.


159
of Mean Are As Follows:

This is a test of mean because each individual observation (each sampled difference) within the sample
can have a wide range of values. Data points for tests of proportion are binary: they can take only one of
two possible values.

This is a two-sample hypothesis test because the data exists in two groups of measurements. One
sample group contains Before measurements and the other sample group contains After measurements.

This is a paired (dependent) hypothesis test because each Before observation has a related After
observation made on the same person.
d) One-Tailed or Two-Tailed Test?

The problem asks to determine whether there has been a reduction in clerical mistake from Before to
After. This is a directional inequality making this hypothesis test a one-tailed test. This one-tailed test will
be in the left tail because the Alternative Hypothesis, which will be created shortly, will use the “less than”
operator.
If the problem asked whether Before and After were simply different, the inequality would be non-
directional and the resulting hypothesis test would be a two-tailed test. A two-tailed test is more stringent
than a one-tailed test.
A z-Test can be performed if the Test Statistic’s distribution can be approximated by the normal
distribution under the Null Hypothesis. The Test Statistic’s distribution can be approximated by the normal
distribution only if the difference sample size is large (n > 30) and the population standard deviation, σ, is
known. A t-Test must be used in all other cases.
Sample size, n, equals 40 and population standard deviation, σ, equals 6.4 so both conditions are met for
the z-Test.
It should be noted that a paired t-Test can always be used in place of a paired z-Test. All z-Tests can be
replaced be their equivalent t-Tests. As a result, some major commercial statistical software packages
including the well-known SPSS provide only t-Tests and no direct z-Tests.
This hypothesis test is a z-Test that is two-sample, paired (dependent), one-tailed hypothesis test
of mean.

a) Test Statistic Distributed According to Normal Distribution
A z-Test can be performed if the distribution of the Test Statistic can be approximated under the Null
Hypothesis by the normal distribution. The Test Statistic is derived from the mean of the difference
sample and therefore has the same distribution that the difference sample mean would have if multiple
similar samples were taken from the same population of differences between data sample pairs.
160
The difference sample size indicates how to determine the distribution of the difference sample mean and
therefore the distribution of the Test Statistic. As per the Central Limit Theorem, as the difference sample
size increases, the distribution of the difference sample means converges to the normal distribution.
In actuality, the sample mean converges toward the t-Distribution as sample size increases. The t-
Distribution converges to the standard normal distribution as sample size increases. The t-Distribution
nearly exactly resembles the standard normal distribution when sample size exceeds 30. The sample
mean’s distribution can therefore be approximated by the normal distribution. The Test Statistic’s
distribution can therefore be approximated by the normal distribution because the Test Statistic is derived
from the sample mean.
As per the Central Limit Theorem, the Test Statistic’s distribution can be approximated by the normal
distribution when the difference sample size is large regardless of the distribution of population from
which the sample was drawn. There is also no need to verify the normality of the difference sample, as
would be the case with a t-Test when population distribution is not known.

case, the Null Hypothesis would state that the there is no difference between before and after data. We
will use the variable x_bardiff to represent the mean between the before and after measurements. The
Null Hypothesis is as follows:
H0: x_bardiff = Constant = 0
different. This hypothesis test is trying to determine whether there has been a reduction in clerical errors,
i.e., the After measurements are, on average, smaller than the Before measurements. The Alternative
H1: x_bardiff < Constant
H1: x_bardiff < 0
The Alternative Hypothesis is directional (“greater than” or “less than” instead of “not equal,” which is non-
directional) and the hypothesis test is therefore a one-tailed test. The “less than” operator indicates that
this is a one-tailed test with the Region of Rejection (the alpha region) entirely contained in the left tail. A
“greater than” operator would indicate a one-tailed test focused on the right tail.
It should also be noted that a two-tailed test is more rigorous (requires a greater differences between the
two entities being compared before the test shows that there is a difference) than a one-tailed test.
It is important to note that the Null and Alternative Hypotheses refer to the means of the population of
paired differences from which the difference samples were taken. A population of paired differences
would be the differences of data pairs in a population of data pairs.
A paired z-Test determines whether to reject or fail to reject the Null Hypothesis that states that that
population of paired differences from which the difference sample was taken has a mean equal to the
Constant. The Constant in this case is equal to 0. This means that the Null Hypothesis states that the
average difference between data pairs of an entire population from which the sample of data pairs were
drawn is zero.
161
Parameters necessary to map the distributed variable, x_bardiff , to the normal distribution are the
following:
σdiff = population standard deviation = 6.4
SEdiff = Standard Error = σdiff / SQRT(n) = 6.4 / SQRT(40) = 1.01
These parameters are used to map the distributed variable, x_bardiff, to the Excel-generated normal
distribution curve as follows:
Step 2 – Map Distributed Variable to Normal Distribution Curve

A z-Test can be performed if the difference sample mean, and the Test Statistic (the t Value) are
distributed according to the normal distribution. If the difference sample has passed a normality test, then
the difference sample mean and closely-related Test Statistic are distributed according to the normal
distribution.
The variable x_bardiff is distributed according to the normal d Distribution. Mapping this distributed
variable to a t-Distribution curve is shown as follows:
This non-standardized t-Distribution curve has its mean set to equal the Constant taken from the Null
Hypothesis, which is:
This non-standardized normal distribution curve is constructed from the following parameters:
Curve Mean = Constant = 0 (This x_bardiff is hypothesized by the Null Hypothesis to be the curve’s
mean.)
Standard Errordiff = 1.01
Distributed Variable = x_bardiff
162
The goal of a hypothesis test is to determine whether to accept or reject the Null Hypothesis at a given
level of certainty. If the two things being compared are far enough apart from each other, the Null
to show graphically how different x_bardiff (-2.14) is from the hypothesized mean of 0.
The non-standardized normal distribution curve can be divided up into two types of regions: the Region of
The above distribution curve that maps the distribution of variable x_bar diff can be divided up into two
types of regions: the Region of Acceptance and the Region of Rejection.
If x_bardiff’s value of -2.14 falls in the Region of Acceptance, we must accept the Null Hypothesis. If
x_bardiff’s value of -2.14 falls in the Region of Rejection, we can reject the Null Hypothesis.
This 5 percent is entirely contained in the outer left tail. The outer left tail contains the 5 percent of the
curve that is the Region of Rejection.
locations of these Critical values need to be calculated as follows.
Calculate the Critical Value

One-tailed, Left tail Critical Value = Mean + (Number of Standard Errors from Mean to Region of
Rejection) * SEdiff
Note that the Mean = the Constant from the Null Hypothesis, which is 0.
One-tailed, Left tail Critical Value = Mean + NORM.S.INV(α) * SEdiff
One-tailed, Left tail Critical Value = 0 + NORM.S.INV(0.05) * 1.01
One-tailed, Left tail Critical Value = 0 + (-1.6449) * 1.01
One-tailed, Left tail Critical Value = -1.66
The Region of Rejection is therefore everything that is to the left of -1.66.
is as follows:
The hypothesized mean is the hypothesized x_bardiff which is 0.
163
The distribution curve with the blue 95-percent Region of Acceptance and the yellow 5-percent Region of
Rejection entirely contained in the left tail is shown is as follows:

The object of a hypothesis test is to determine whether to accept of reject the Null Hypothesis. There are
follows:
1) Compare x-bardiff, with Critical Value

Reject the Null Hypothesis if the sample mean, x_bardiff = -2.14, falls into the Region of Rejection.
Equivalently, reject the Null Hypothesis if the sample mean, x_bardiff, is further the curve’s mean of 0 than
the Critical Value.
The Critical Values have been calculated to be -1.66 on the left. x_bardiff (-2.14) is further from the curve
mean (0) than left Critical Value (-1.66). The Null Hypothesis would therefore be rejected.
164
2) Compare z Score with Critical z Value
The z Score is also known as the Test Statistic in a z-Test and is the number of Standard Errors that
x_bardiff is from the mean (mean = Constant = 0).
The Critical z Value is the number of Standard Errors that the Critical Value is from the mean.
If the z Score is further from the standardized mean of zero than the Critical z Value, the Null Hypothesis
can be rejected.
Z Score (as called the Test Statistic) = (x_bardiff – 0) / SEdiff

Z Score (Test Statistic) = (-2.14 – 0) / 1.01
Z Score (Test Statistic) = -2.11
This indicates that x_bardiff is 2.11 standard errors to the left of the mean (mean = Constant = 0).
Critical z Value = NORM.S.INV(α)

If α = 0.05, the Critical z Value for a one-tailed hypothesis test in the left tail is calculated as follows:
Critical z Value = NORM.S.INV(0.05) = NORM.S.INV(0.05) = -1.6449
This means that the Region of Rejection for a one-tailed hypothesis test in the left tail begins at 1.66
standard errors from (to the left of) the standardized mean of zero.
The z Score (-2.11) is farther from the standardized mean of zero than the Critical z Value (-1.6449) so
the Null Hypothesis is rejected.
3) Compare p Value to Alpha.

The p Value is the percent of the curve that is beyond x_bardiff (-2.14). If the p Value is smaller than
Alpha, the Null Hypothesis is rejected. The p Value in this case is calculated by the following Excel
formula:
p Value =MIN(NORM.S.DIST(-2.11,TRUE),1-NORM.S.DIST(-2.11,TRUE))
p Value = 0.0174
165
The p Value (0.0174) is smaller than Alpha (0.05) and we therefore reject the Null Hypothesis. The
following Excel-generated graph shows that the red p Value (the curve area beyond x_bardiff) is smaller
than the yellow Region of Rejection in the left tail.
Excel does not provide any formulas or tools in the Data Analysis ToolPak add-in that directly perform the
paired z-Test. The easy work-around is to perform a one-sample z-Test on the difference data sample.
This formula is as follows:
p Value =MIN(Z.TEST(array,Constant,σdiff),1-Z.TEST(array,Constant,σdiff))
It should be noted that when the Constant is positive, the p Value Excel formula is
p Value = Z.TEST(array,Constant,σdiff),
When the Constant is negative, the p Value Excel formula is
p Value = 1-Z.TEST(array, Constant,σdiff))
The Constant is taken from the Null Hypothesis and is equal to 0.
The Null Hypothesis is as follows:
166
Applying the Excel one-sample z-Test formula to the sample of difference data would give the following p
Value for this paired z-Test:
167
z-Based Confidence Intervals of a Population Mean in
Excel
Overview
This confidence interval of a population mean is based upon the sample mean being normal-distributed.
A 95-percent confidence interval of a population mean is an interval that has a 95-percent chance of
containing the population mean.
The sample mean is normal-distributed if the sample size is large (n > 30) as per the Central Limit
Theorem. The CTL states that the means of large samples randomly taken from the same population will
be normal-distributed no matter how the population is distributed. Confidence intervals of a population
mean can be based upon the normal distribution only if the sample size is large (n > 30).
In addition to the required large sample size that ensures the normal distribution of the sample mean, the
population standard deviation must be known as well.
Margin of Error = Half Width of C.I. = Z Valueα/2 * Standard Error

Margin of Error = Half Width of C.I. = NORM.S.INV(1 – α/2) * σ/SQRT(n)
A confidence interval of a population mean that is based on the normal distribution is z-based. A
confidence interval of a population mean that is based on the t-Distribution is t-based.
It is much more common to use the t-Distribution than the normal distribution to create a confidence
interval of the population mean. Requirements for t-based confidence intervals are much less restrictive
than the requirements for a z-based confidence interval.
A confidence interval of a population mean can be based on t-Distribution if only the sample standard
deviation is known and any of the following three conditions are met:
1) Sample size is large (n > 30). The Central Limit Theorem states that the means of large, similar-sized,
random samples will be normal-distributed no matter how the underlying population is distributed.
2) The population from which the sample was drawn is proven to be normal-distributed.
3) The sample is proven to be normal-distributed.
A confidence interval of the mean can be created based on the normal distribution only if the sample size
is large (n >30) and the population standard deviation, σ, is known. For this reason, confidence intervals
are nearly always created using the t-Distribution in the professional environment.
This example will demonstrate how to create a confidence Interval of the mean using the normal
distribution
168
Example of a Normal-Distribution-Based Confidence Interval
in Excel
In this example a 95 percent Confidence Interval is created around a sample mean using the normal
distribution.
A company received a shipment of 5,000 steel rods of unknown tensile strength. All rods originated from
the same source. The company randomly selected 100 rods from the shipment and tested each for
tensile strength. The average tensile strength of the 100 rods tested was found to be 250 MPa
(megapascals). The tensile strength of steel rods of this exact type is known to have a standard deviation
of 30 MPa.
Calculate the endpoints of the interval that 95 percent certain to contain the true mean tensile strength of
all 5,000 rods in the shipment. In other words, calculate the 95 percent confidence interval of the
population (entire shipment) mean tensile strength

x_bar = sample mean = AVERAGE() = 250 MPa
µ = (population) mean tensile strength of entire shipment = Unknown
σ (Greek letter “sigma”) = population tensile strength standard deviation = 30 MPa
SE = Standard Error = σ / SQRT(n) = 30 / SQRT(100)
SE = 3
As when creating all Confidence of Mean, we must satisfactorily answer these two questions and then
proceed to the two-step method of creating the Confidence Interval.

What Type of Confidence Interval Should Be created?
Have All of the Required Assumptions For This Confidence Interval Been Met?
The Two-Step Method For creating Confidence Intervals of Mean are the following:
Step 1 - Calculate the Half-Width of the Confidence Interval (Sometimes Called the Margin of Error)
Step 2 – Create the Confidence Interval By Adding to and Subtracting From the Sample Mean Half
the Confidence Interval’s Width
169
The initial two questions are answered as follows:
Question 1) What Type of Confidence Interval?

a) Confidence Interval of the Population Mean or Proportion?
This is a Confidence Interval of a population mean because each individual observation (each sampled
rod’s tensile strength) within the entire sample can have a wide range of values. Most of the sample
values are spread out between 200 MPa and 300 MPa.
Sampled data points used to create a Confidence Interval of a population proportion are binary: they can
take only one of two possible values.
b) Based on t-Distribution or Based on Normal Distribution?

A confidence interval can be created that is based on the normal distribution can only if both of the
following conditions are met:
Sample size is large (n > 30)

and
The population standard deviation, σ, is known.
In this case sample size is large (n = 100) and the population standard deviation is known (σ = 30 MPa).
This Confidence Interval can be created using either the t-Distribution or the normal distribution. In this
case, the normal distribution will be used to create this Confidence Interval of a population mean. This
Confidence Interval of a population mean will be z-based.
This confidence interval will be a confidence interval of a population mean and will be created
using the normal distribution.

a) Normal Distribution of Sample Mean
As per the Central Limit Theorem, the large sample size (n = 100) guarantees that the sample mean is
normal-distributed.
b) Population Standard Deviation Is Known

σ is known to equal 30 MPa.
We now proceed to the two-step method for creating all Confidence intervals of a population mean.
These steps are as follows:
Step 1) Calculate the Width of Half of the Confidence Interval
Step 2 – Create the Confidence Interval By Adding and Subtracting the Width of Half of the
Confidence Interval from the Sample Mean
170
Proceeding through the two-step method of creating a confidence interval is done is follows:
Step 1) Calculate Half-Width of Confidence Interval

Half the Width of the Confidence Interval is sometimes referred to the Margin of Error. The Margin of Error
will always be measured in the same type of units as the sample mean is measured in, which in this case
was MPa (megapascals). Calculating the Half Width of the Confidence Interval using the normal
distribution would be done as follows in Excel:
Margin of Error = Half Width of C.I. = Z Valueα/2 * Standard Error
Margin of Error = Half Width of C.I. = NORM.S.INV(1 – α/2) * σ/SQRT(n)
Margin of Error = Half Width of C.I. = NORM.S.INV(1 – 0.05/2) * 30/SQRT(100)
Margin of Error = Half Width of C.I. = NORM.S.INV(0.975) * 30/10
Margin of Error = Half Width of C.I. = 1.96 * 3
Margin of Error = Half Width of C.I. = 5.88 MPa
The Half Width of z-based Confidence Interval can also be calculated by the following Excel formula:
Margin of Error = Half Width of C.I. = CONFIDENCE.NORM(α, σ, n)
Margin of Error = Half Width of C.I. = CONFIDENCE.NORM(0.05, 30, 100)
Margin of Error = Half Width of C.I. = 5.88 MPa
Step 2 Confidence Interval = Sample Mean ± C.I. Half-Width

Confidence Interval = Sample Mean ± (Half Width of Confidence Interval)
Confidence Interval = x_bar ± 5.88
Confidence Interval = 250 ± 5.88
Confidence Interval = [ 244.12 MPa, 255.88 MPa ]
171
A graphical representation of this Confidence Interval is shown as follows:
172
Min Sample Size to Limit Width of Confidence Interval
The larger the sample taken, the smaller the Confidence Interval becomes. That makes intuitive sense
because the more sample information that is gathered, the more tightly the position of the population
mean can be defined. The Confidence Interval is an interval believed to contain the population mean with
specific degree of certainty.
As sample size increase, the Confidence Interval shrinks because greater certainty has been attained.
The margin of error, which is equal to half the width of the Confidence Interval, therefore shrinks as well.
During the design phase of a statistical experiment, sample size should be determined. Sampling has a
cost and additional sampling beyond what is necessary to attain a desired level of certainty is often
undesirable. One common objective of the design phase of a statistical test involving sampling is to
determine the minimum sample size required to obtain a specified degree of certainty.
Calculating the minimum sample size necessary to limit the size of a confidence interval of a population
mean can be done using the normal distribution but not the t-Distribution. A t-based confidence interval
requires specifying the degrees of freedom, which is derived from the sample size that is unknown.
A z-based confidence interval (a confidence interval based upon the normal distribution) requires that the
sample mean be normal-distributed and the population standard deviation be known. These requirements
are met if both of the following are true:
Requirements
1) Minimum Sample Size is 30
This ensures that the sample mean is normal-distributed as per the Central Limit Theorem. If the
calculated minimum sample size is less than 30, the sample or the population must be confirmed to be
normal-distributed.
and
2) Population Standard Deviation Known

Sample standard deviation cannot be used because a sample has not been taken.
The minimum sample size, n, to limit the width of a z-based confidence interval of a population proportion
to a specific size can be derived with the following algebra:
Confidence Interval = Sample mean ± z Scoreα,two-tailed * SE
Confidence Interval = x_bar ± NORM.S.INV(1 – α/2) * σ/SQRT(n)
Confidence Interval = x_bar ± NORM.S.INV(1 – α/2) * σ /SQRT(n)
(Half-width of C.I.) = NORM.S.INV(1 – α/2) * σ /SQRT(n)
Squaring both sides gives the following:

2 2 2
(Half-width of C.I.) = NORM.S.INV (1 – α/2) * σ /n
173
Further algebraic manipulation produces the following:
2 2 2
n = [NORM.S.INV (1 – α/2) * σ ] / (Half-width of C.I.)
or, equivalently because Half-width of C.I. = Margin of Error,
2 2 2
n = [NORM.S.INV (1 – α/2) * σ ] / (Margin of Error)
The count of data observations in a sample, n, must be a whole number so n must be rounded up to the
nearest whole number. This is implemented in Excel as follows:
2 2 2
n = Roundup( ( NORM.S.INV (1 – α/2) * σ ) / (Half-width of C.I.) ,0)
Example in Excel
A survey was taken of the monthly salaries full-time employee of the California Department of
Transportation. The standard deviation of monthly salaries throughout the entire California DOT is known
to be $500.
What is the minimum number of employees that would have to be surveyed to be at least 95% certain
that the sample average monthly salary is within $50 of the true average monthly salary of all employees
in the California DOT?
In other words, what is the minimum sample size needed to create a 95-percent confidence interval about
the population mean that has a margin of error no larger than $50?
Another way to state the problem is to ask how large must the sample size be to create a 95-percent
confidence interval about the population mean that has a half-width of no more than $50?
σ = Population standard deviation = $500
Half-width of Confidence Interval = Margin of Error = $50
(The confidence interval must be specified in the same units as the population standard deviation is.)
α = 1 – Level of Certainty = 1 – 0.95 = 0.05
2 2 2
2 2 2
n = Roundup( ( NORM.S.INV (1 – 0.05/2) * (500) ) / (50) ,0)
2 2 2
n = Roundup( ( NORM.S.INV (0.975) * (500) ) / (50) ,0)
2 2 2
n = Roundup( ( (1.96) * (500) ) / (50) ,0)
n = Roundup( 384.1459 ,0)
n = 385
A minimum of 385 employees must be surveyed to be 95 percent certain that the average salary of the
sample is no more than $50 from the true average salary within the entire California DOT.
174
t-Distribution in Excel
Overview
The t-Distribution is used much more often than the normal distribution to perform several basic
parametric statistical tests such as hypothesis tests of a population mean and confidence intervals of a
population mean. Requirements for statistical tests are generally less rigorous when a statistical test can
be based upon the t-Distribution instead of the normal distribution.
The t-Distribution (also called the Student’s t-Distribution) describes the distribution of a sample taken
from a normally-distributed population when the population standard deviation is unknown. The t-
Distribution closely resembles the standard normal distribution (the normal distribution when the mean
equals zero and the standard deviation equals one) except that the t-Distribution’s outer tails have more
weight (are thicker) and its mean has a lower peak than the standard normal distribution.
As sample size increases, the t-Distribution converges to (more closely resembles) the standard normal
distribution. When the sample size becomes large (n > 30), the t-Distribution almost exactly resembles
the standard normal distribution.
Following is an Excel-generated image of the PDF (Probability Density Function) of the t-Distribution with
a very low degrees of freedom. Sample size, n, equals 3 and degrees of freedom, df, equals n – 1 = 2.
The PDF of the t-Distribution has a lower peak and thicker tails when its degrees of freedom is small.
As the degrees of freedom increases, the PDF of the t-Distribution converges toward (resembles more
and more) the standard normal distribution. The standard normal distribution is a normal distribution curve
with its mean, μ, equal to zero and its standard deviation, σ, equal to one. the standard normal curve is a
special case of the t-Distribution with a sample size, n, equal to infinity. Note how the shape of the PDF
curve of the t-Distribution changes as the sample size increases from n = 3 to n = ∞. The height of the
peak over the mean has risen significantly and the tails are quite a bit thinner. This is shown in the
following Excel-generated image:
175
These differences are also reflected the CDF (Cumulative Distribution Function) graphs of the t-
Distribution with the same degrees of freedom. The t-Distribution’s CDF approaches its asymptotic values
of 0 and 1 much further from t =0 at smaller degrees of freedom than for larger degrees of freedom.
The t-Distribution CDF graph with only 2 degrees of freedom is still a significant distance from the
asymptotic values of 0 and 1 at 3 standard errors above and below t = 0. This is shown in the following
Excel-generated graph:
176
The t-Distribution CDF graph with the degrees of freedom equaling its highest possible value of ∞ has
nearly reached the asymptotic values of 0 and 1 quite a bit closer than 3 standard errors above and below
t = 0. This is shown in the following Excel-generated graph:
The t-Distribution’s CDF is simply a graph of the accumulation of its PDF as the t Value goes from - ∞ to
∞
+ . The t-Distribution’s PDF is bell-shaped and symmetrical about a t Value of 0. The t-Distribution’s
CDF will therefore show that 50 percent of the area under the PDF curve, F(t,v) = 0.50, occurs at the t
Value of 0.
History of the t-Distribution

One of the most colorful and well-known stories in the annals of statistics is the origin of the name of the
Student’s t-Distribution, which the t-Distribution is often called. This distribution was first presented in the
English language by William Sealy Gosset under the pseudonym “Student” in his article “The probable
error of a mean” in the scientific journal Biometrika in March 1908. At the time Gosset was employed at
the Guinness Brewery in Dublin, Ireland and was studying the nature of small samples of brewery
ingredients such as barley. Gosset published his article under the pen name “Student” because his
employer either did not allow staff to publish scientific papers or did not want competitors to know that the
Guinness Brewery was using this test on small samples of raw materials.
The name “Student’s distribution” was conferred on the distribution by Ronald Fisher in his 1925 article
“Applications of Student’s distribution.” This article also assigned the label “t” to value of the Test Statistic
for this distribution.
Prior to Gosset’s English-language introduction, the t-Distribution was first described by German
mathematicians Friedrich Helmert and Jacob Lüroth in 1876.
177
Properties of the t-Distribution
- The mean of the t-Distribution equals 0. The t-Distribution forms a bell-shaped curve about a mean of 0.
This differs from the normal distribution because the normal distribution can be symmetrical about a mean
of any real number.
- The variance of the t-Distribution is equal to v / (v – 2) when v (Greek letter “nu”) exceeds 2. v equals the
degrees of freedom, which equals the sample size minus 1. The variance is always greater than 1 but
converges to 1 as the sample size gets larger. The t-Distribution converges to the standard normal
distribution (which has a variance equaling 1) as sample size approaches infinity.
- The standard error of t-Distribution is equal to the sample standard deviation divided by the square root
of the sample size.
- The t-Distribution has only one parameter which is the degrees of freedom. Degrees of freedom is
usually designated as df or ν (Greek letter “nu”). This differs from the normal distribution because the
normal distribution is described by the following two parameters: mean (μ - Greek letter “mu”) and
standard deviation (σ - Greek letter “sigma”).
- The graph of the t-Distribution is symmetrical about a mean of 0 and the units on its horizontal axis
describing the distance from the mean of 0 are units of standard errors. This differs from the normal
distribution because the normal distribution can be symmetrical about a mean of any real number and the
units of its horizontal axis describing the distance from its mean use the same real number scale in which
the mean was measured. The t-Distribution’s PDF or CDF at any real number X requires that the X value
be converted to the number of standard errors that the X value is from the sample mean. The standard
error is equal to the sample standard deviation divided by the square root of the sample size.
The t-Distribution can be used to describe any statistic that has a bell-shaped distribution, i.e., unimodal,
symmetrical, and without significant outliers. The t-Distribution is used to analyze samples taken from a
normally-distributed population when either of the following is true:
1) Small size is small (n < 30).
2) The population standard deviation is not known, which is often the case.
The t-Distribution is used much more often than the normal distribution when performing hypothesis tests
or creating confidence intervals based upon samples taken from a normally-distributed population. If a
sample is found to be normally distributed, its population is assumed to be normally distributed.
The t-Distribution more closely describes the distribution of a small sample (n < 30) taken from a
normally-distributed population than the normal distribution does. Small samples taken from a normally-
distributed population have a slightly higher probability that sample values will occupy the outer tails than
do larger samples. The t-Distribution has slightly thicker tails and a lower peak than does the normal
distribution. The t-Distribution is therefore used to describe the distribution of small samples taken from a
normally-distributed population.
The extra weight in the outer tails of the t-Distribution accounts for the additional uncertainty of having to
use the sample standard deviation to estimate the population standard deviation. This estimate becomes
more uncertain as sample size decreases. the t-Distribution’s shape reflects that as its outer tails become
thicker as sample size decreases.
The t-Distribution can be used to perform hypothesis tests or create confidence intervals of a normally-
distributed population when the population standard deviation is not known. The normal distribution
should not be used for these types of analysis when the population standard deviation is not known. In
the real world it is much more common scenario that the standard deviation of the population from which
the sample was drawn is not known.
178
t-Distribution’s PDF
As with all distributions, the t-Distribution has a PDF (Probability Density Function) and a CDF
(Cumulative Distribution Function).
The t-Distribution’s PDF (Probability Density Function) equals the probability that sample point X taken
from a normally distributed population is a certain distance from the sample mean for the given sample
size.
More specifically, the t-Distribution’s PDF calculates the probability that the standardized distance
between sample point X and the sample mean is equal to the t Value of t for sample of size n when n = df
+ 1 = v + 1.
t equals the t Value, which is the number of standard errors that a sampled point X is from the sample
mean.
The t-Distribution’s PDF is expressed as f(t,v).
f(t,v) = the probability that a sample point is a distance of t standard errors from the sample mean if the
sample size n = v + 1 and the population is normally-distributed. It is given by the following formula:
In Excel 2010 and beyond, the t-Distribution’s PDF can be calculated directly be the following Excel
formula:
f(t,v) = T.DIST(t, v, FALSE)
Prior to Excel 2010, Excel had no formula that correctly calculated the PDF of the t-Distribution. The t-
Distribution’s PDF had to be calculated directly from the actual formula. The formula is somewhat onerous
by itself but is much more manageable if broken down into its component parts as follows:
The full formula for the t-Distribution’s PDF is the following:
This is made much more manageable if broken down into the following parts:
179
Γ(X) is the gamma distribution function and has to be calculated in Excel is a somewhat indirect fashion
as follows:
Γ(X) = EXP(GAMMALN(X))
t equals the number of standard errors that sample point X is from the sample mean.
180
Example in Excel With Sample Data
The following is a random sample of 20 data points drawn from a population. The population standard
deviation is not known. The population’s distribution is not known. It will not be assumed to be normally
distributed.
This exercise will calculate the probability that a single sample point from this sample has the value of X =
48.
If the population or the sample can be shown to be normally distributed, the t-Distribution can be used to
this probability, which would be the t-Distribution’s PDF at sample X = 48. The t-Distribution would be
used because the sample size is small (n < 30) and the population’s standard deviation is not known.
181
The t-Distribution should only be used if the sample or the population is normally distributed. The
distribution of the population is not known so the sample’s distribution needs to be evaluated for
normality. A histogram of the sample data indicates a reasonable resemblance to normality. The t-
Distribution can therefore be used to analyze the sample.
This histogram was created in Excel inputting the following information into Excel’s Histogram dialogue
box:
182
The next step is to run Excel Descriptive Statistics on the sample data. The output of this Excel data
analysis tools is as follows:
The t-Distribution is a family of distribution curves that are all symmetrical about a mean of zero. Each
unique t-Distribution curve is completely described and differentiated from all other t-Distribution curves
by only a single parameter: its degrees of freedom. Degrees of freedom equal sample size minus one as
follows:
n = 20
df = v = n – 1 = 19
Note also that the t-Distribution’s PDF formula calculates a point’s PDF at the t Value of the sample point,
not at the sample point’s actual value, X. The inputs to the t-Distribution’s PDF formula are the degrees of
freedom and the point’s t Value, t.
The t Value of a point is the distance that the point is from the sample mean expressed as the number of
standard errors. The length of one standard error is calculated from the sample as follows:
SE = Standard Error = Standard Deviation / SQRT(n)
SE = 7.434 / SQRT(20) = 1.662
183
Note that SE is also automatically calculated in the Excel Descriptive Statistics.
The t Value of point X = 48 can now be calculated since the sample mean (x_bar) and SE are known.
This is done as follows:
X = 48
t = (48 – 50) / 1.662

t = -1.20337
The two inputs required to calculate the t-Distribution’s PDF are the point’s t Value (t = -1.20337) and the
degrees of freedom (df = v = 19) from the sample.
Excel 2010 provides the following formula that enables the calculation of the t-Distribution’s PDF in one
quick step. That calculation is as follows:
T.DIST(t, df, FALSE) =
T.DIST(-1.20337, 19, FALSE) = 0.18888
This results states that there is an 18.888 percent chance the one of the data observations from the
sample group of 20 samples taken from a normally distributed population will have a t Value of -1.20337,
i.e., will be 1.20337 standard errors to the left of the sample mean.
Given that the sample mean is 50, the length of one standard error = 1.662, and the d point is 1.20337
standard errors to the left of the mean, the point being evaluated is equal to the following:
X = x_bar + t Value*SE
X = 50 + (-1.20337)*(1.662) = 48
The t-Distribution’s PDF formula calculated that the probability that a single point of the 20 sampled points
has a value of X = 48 (or equivalently has a t Value of t = -1.20337) is 18.888 percent.
Prior to Excel 2010, there was no Excel formula that calculated the correct PDF value for the t-
Distribution. This calculation has to be performed manually in any version of Excel prior to 2010. This is
done as follows (it’s really not as bad as it looks):
a = 1 / SQRT(19 * 3.14159)
a = 0.12943
184
b = EXP(GAMMALN((19+1)/2)) = 362,880
c = EXP(GAMMALN(19/2)) = 119,292
2
d = 1 + (-1.20337) /19 = 1.0762
e = -( (19+1) / 2) = -10
-10
f(t=-1.20337,v=19) = f(-1.20337,19) = 0.12943 * (362,880/119,292) * 1.0762
f(t=-1.20337,v=19) = 0.18888
This is the same result that is arrived at in the following single calculation from Excel 2010:
T.DIST(t, df, FALSE) =
T.DIST(-1.20337, 19, FALSE) = 0.18888
An Excel-generated graphical representation of this example is shown in the following diagram. The t
Value of t = -1.20337 represents the point that is 1.20337 standard errors from the t-Distribution’s mean of
zero.
185
Note that the t Value of t = -1.20337 has a PDF value equal to f(t=-1.2037,v=19) = 0.18888 on the Y axis.
This indicates that there is an 18.888 percent probability that one of the 20 samples has a t Value of
1.20337. This is shown in the following Excel-generated graph:
186
The t-Distribution’s CDF
The t-Distribution’s CDF (Cumulative Distribution Function) equals the probability that sample point X
taken from a normally distributed population is up to a certain distance from the sample mean for the
given sample size.
More specifically, the t-Distribution’s PDF calculates the probability that the standardized distance
between sample point X and the sample mean is up to the t Value of t for sample of size n when n = df +
1 = v + 1.
The t Value is the number of standard errors that a sampled point X is from the sample mean.
The t-Distribution’s CDF is expressed as F(t,v).
F(t,v) = the probability that a sample point is up to a distance of t standard errors from the sample mean if
the sample size n = v + 1 and the population is normally-distributed. It is given by the following formula:
In Excel 2010 and beyond, the t-Distribution’s CDF can be calculated directly be the following Excel
formula:
F(t,v) = T.DIST(t, v, TRUE)
Prior to Excel 2010, the t-Distribution’s CDF had to be calculated by the following Excel formula:
F(t,v) = IF(t>0,1-TDIST(ABS(t),v,1),TDIST(ABS(t),v,1))
187
Example With Sample Data
The following is the previous random sample of 20 data points drawn from a population. The population
standard deviation is not known. The population’s distribution is not known. It will not be assumed to be
normally distributed.
This exercise will calculate the probability that a single sample point from this sample has the value of up
to X = 48.
If the population or the sample can be shown to be normally distributed, the t-Distribution can be used to
calculate this probability, which would be the t-Distribution’s CDF at sample X equals up to 48. The t-
Distribution would be used because the sample size is small (n < 30) and the population’s standard
deviation is not known.
it is known from the previous example that these sample data are normally distributed so the t-Distribution
can be used to calculate the CDF.
188
Descriptive Statistics was already run on this sample data in the previous example calculating the PDF at
X = 48. The output of this Excel data analysis tools is as follows:
unique t-Distribution curve is completely described and differentiated from all other t-Distribution curves
by only a single parameter: its degrees of freedom. Degrees of freedom equal sample size minus one as
follows:
n = 20
df = v = n – 1 = 19
Note also that the t-Distribution’s CDF formula calculates a point’s PDF up to the t Value of the sample
point, not up to the sample point’s actual value, X. The inputs to the t-Distribution’s PDF formula are the
degrees of freedom and the point’s t Value, t.
The t Value of a point is the distance that the point is from the sample mean expressed as the number of
standard errors. The length of one standard error is calculated from the sample as follows:
SE = Standard Error = Standard Deviation / SQRT(n)
SE = 7.434 / SQRT(20) = 1.662
189
Note that SE is also automatically calculated in the Excel Descriptive Statistics.
The t Value of point X = 48 can now be calculated since the sample mean (x_bar) and SE are known.
This is done as follows:
X = 48
t = t Value = (X – x_bar) / SE
t = (48 – 50) / 1.662
t = -1.20337
The two inputs required to calculate the t-Distribution’s CDF are the point’s t Value (t = -1.20337) and the
degrees of freedom (df = v = 19) from the sample.
Excel 2010 provides the following formula that enables the calculation of the t-Distribution’s PDF in one
quick step. That calculation is as follows:
F(t,v) = T.DIST(t, df, TRUE) =
F(t=-1.20337, v=19) = T.DIST(-1.20337, 19, TRUE) = 0.12181
This results states that there is an 12.181 percent chance the one of the data observations from the
sample group of 20 samples taken from a normally distributed population will have a t Value of up to -
1.20337, i.e., will be up to 1.20337 standard errors to the left of the sample mean.
Given that the sample mean is 50, the length of one standard error = 1.662, and the d point is 1.20337
standard errors to the left of the mean, the point being evaluated is equal to the following:
X = x_bar + t Value*SE
X = 50 + (-1.20337)*(1.662) = 48
The t-Distribution PDF formula calculated that the probability that a single point of the 20 sampled points
has a value of up to X = 48 (or equivalently has a t Value of t = -1.20337) is 12.181 percent.
Prior to Excel 2010, there was no Excel formula that calculated the correct CDF value for the t-Distribution
in every situation. The following If-Then-Else statement correctly calculates the t-Distribution’s CDF in all
situations for Excel versions prior to 2010:
F(t,v) = IF(t>0,1-TDIST(ABS(t),v,1),TDIST(ABS(t),v,1))
= IF(-1.20337>0,1-TDIST(ABS(-1.20337),19,1),TDIST(ABS(-1.20337),19,1))
F(t=-1.20337,v = 19) = 0.12181
This is the same result that is arrived at in the following calculation from Excel 2010:
T.DIST(t, df, TRUE) =
T.DIST(-1.20337, 19, TRUE) = 0.12181
190
Below is an Excel-generated graphical representation of this example.
A Big Difference Between t and Normal Distributions

unique curve of the t-Distribution is completely described and fully differentiated from all other t-
Distribution curves by its degrees of freedom, which is designated by either df or v (Greek letter “nu”).
Degrees of freedom is related to sample size by df = n – 1.
The t-Distribution’s PDF formula calculates the PDF at the t Value of point X. The t Value of point X is the
number of standard errors that point X is from the sample mean. When calculating the t-Distribution’s
PDF (or CDF) at point X, the t Value of point X must be computed for that point X. The t Value of point X
is the required input of the t-Distribution’s PDF (and CDF) formula.
The t Value of point X must be determined as follows:
before the t-Distribution’s PDF can be calculated by the following formula:
191
The Excel formula to calculate the t-Distribution’s PDF is shown here as well.
f(t,v) = T.DIST(t, df, FALSE)
The t Value of the point X must also be calculated before calculating the t-Distribution’s CDF, which can
be found with the following Excel formula:
F(t,v) = T.DIST(t, df, TRUE)
Normal Distribution
The normal distribution is also a family of distribution curves. Each distribution curve is fully described and
completely differentiated from all other normal curves by only two parameters: its mean, μ, and standard
deviation, σ.
The normal distribution’s PDF at point X is calculated directly by using point X as one of the inputs of the
normal distribution’s PDF formula as follows:
The Excel formula to calculate the normal distribution’s PDF is shown here as well.
f(X,μ,σ) = NORM.DIST(X, μ, σ, FALSE)
The t-Distribution requires that point X be converted to a t Value before calculating the PDF or CDF at
point X. The normal distribution does not require such a conversion because point X is one of the inputs
to and part of its PFD and CDF formulas.
The normal distribution’s CDF formula in Excel is as follows:
F(X,μ,σ) = NORM.DIST(X, μ, σ, TRUE)
192
t-Test: t-Distribution-Based Hypothesis Test
population.
Null Hypothesis
A hypothesis test is based upon a Null Hypothesis which states that the sample did come from that
population. A hypothesis test compares a sample statistic such as a sample mean to a population
parameter such as the population’s mean. The amount of difference between the sample statistic and the
population parameter determines whether the Null Hypothesis can be rejected or not.
different.
Null Hypothesis - Rejected or Not But Never Accepted

193
following:
H1: x_bar ≠ 5
H1: x_bar > 5
or
H1: x_bar < 5
following:
H1: p_bar ≠ 0.3
An example of a directional Alternative Hypothesis would be the following:H 1: p_bar > 0.3
or
H1: p_bar < 0.3
One-Tailed Test vs. Two-Tailed Test

Rejection.
194
Level of Certainty

following:
the normal distribution or the t distribution. The Test Statistic is derived directly from the sample statistic
such as the sample mean. If the Test Statistic is distributed according to the normal or t distribution, then
the sample statistic is also distributed according to normal or t distribution. This will be discussed is
sample statistic (the sample mean or proportion) to its own unique normal or t distribution. The sample
statistic is the distributed variable whose distribution is mapped according its own unique normal or t
distribution
The Region of Acceptance is the percentage of area under this normal or t distribution curve that equals
variable’s mapped normal or t distribution curve.
Region of Rejection
The Region of Rejection is the percentage of area under this normal or t distribution curve that equals the
The Region of Acceptance includes 95 percent of the total area under the normal or t distribution curve
The Region of Rejection includes 5 percent of the total area under the normal or t distribution curve that
195
is not rejected.
Critical Value(s)
Test Statistic
mean.
Level of Certainty
Test Statistic
The Test Statistic is either the z Score or the t Value depending on whether a z Test or t-Test is being

Each hypothesis test calculates Critical t or z Values. A Critical t Value is calculated for a t-Test and a
Critical z Value is calculated for a z Test. A Critical t or z Value is the amount of difference expressed in
curve’s mean.
between the two outer tails.The Test Statistic (the t Value or z Score) are compared with the Critical t or z
Value on that side of the mean.
If the Test Statistic is farther from the standardized mean of zero than the Critical t or z Value on that side,
the Null Hypothesis is rejected. The Test Statistic is the number of Standard Errors that the sample
statistic is from the curve’s mean.
196
The Critical t or z Value on the same side is the number of Standard Errors that the Critical Value (the
boundary of the Region of Rejection) is from the mean. If the Test Statistic is farther from the
standardized mean of zero than the Critical t or z value, the sample statistic lies in the Region of
Rejection.
p Value
Calculations of the Critical t Value(s) and the p Value are as follows:
Critical t Value For 1-Tailed Test in Right Tail:

Critical t Value = T.INV(1-α,df)
Prior to Excel 2010

Critical t Value = TINV(2*α,df)
Critical t Value For 1-Tailed Test in Left Tail:

Critical t Value = T.INV(α,df)
Prior to Excel 2010

Critical t Value = -TINV(2*α,df)
Note that the negative sign has to be manually inserted into this pre-2010 formula to calculate the Critical
t Value in the left tail for a one-tailed test.
197
Critical t Values For a 2-Tailed Test:
Critical t Values = ±T.INV(1-α/2,df)
Prior to Excel 2010

Critical t Values = ±TINV(α,df)p Value
The p Value is calculated using the same formulas whether the test is a one-tailed test or a two-tailed
test.

p Value = T.DIST.RT(ABS(t Value), df)
Prior to Excel 2010

p Value = TDIST(ABS(t Value), df)
3 Equivalent Reasons To Reject Null Hypothesis

1) Sample Statistic Beyond Critical Value

The sample statistic (the observed sample mean or proportion) would therefore lie in the Region of
Rejection because the Critical Value is the boundary of the Region of Rejection.
2) Test Statistic Beyond Critical t or z Value

The Test Statistic (the t Value of z Score) is the number of Standard Errors that the sample statistic is
from the curve’s mean. The Critical t or z Value is the number of Standard Errors that the boundary of the
Region of Rejection is from the curve’s mean. If the Test Statistic is farther from farther from the
standardized mean of 0 than the Critical t or z Value, the sample statistic lies in the Region of Rejection.
3) p Value Smaller Than α (1-Tailed) or α/2 (2-Tailed)

The p Value is the curve area beyond the sample statistic. α and α/2 equal the curve areas contained by
the Region of Rejection on that side for a one-tailed test and a two-tailed test respectively. If the p value is
smaller than α for a one-tailed test or α/2 for a two-tailed test, the sample statistic lies in the Region of
Rejection.
Independent vs. Dependent Samples

A sample that is independent of a second sample has data values that are not influenced by any of the
data values within the second sample. Dependent samples are often referred to as paired data. Paired
data are data pairs in which one of the values of each pair has an influence on the other value of the data
pair. An example of a paired data sample would be a set of before-and-after test scores from the same
set of people.
198
Pooled vs. Unpooled Tests
A two-independent-sample Hypothesis Test of Mean can be pooled or unpooled. A pooled test can be
performed if the variances of both independent samples are similar. This is a pooled test because a
single pooled standard deviation replaces both sample standard deviations in the calculation of the
Standard Error. An unpooled test must be performed when the variances of the two independent samples
are not similar.

Power of a Test
following formula:
Power = 1 – β
Effect Size
Nonparametric Alternatives for t-Tests in Excel

Nonparametric tests are sometimes substituted for t-Tests because normality requirements cannot be
met. A t-Test is a Hypothesis Test of Mean that can be performed if the sample statistic (and therefore the
Test Statistic) is distributed according to the t distribution under the Null Hypothesis. The sample statistic
(the sample mean) is distributed according to the t distribution if any of the following three conditions
exist:
1) Sample size is large (n > 30). The sample taken for the hypothesis test must have at least 30 data
observations.
2) The population from which the sample was taken is verified to be normally distributed.
3) The sample is verified to be normally distributed.
If none of these conditions can be met or confirmed, a nonparametric test can often be substituted for a t-
Test. A nonparametric test does not have normality requirements that a parametric test such as a t-Test
does.
199
values.
Hypothesis Tests of Mean – Overview

Hypothesis Tests of Proportion – Overview

Hypothesis Test of Proportion are covered in detail in a separate section in this manual. They are also
summarized at the end of the binomial distribution section.
Hypothesis Tests of Mean

Hypothesis Tests of Mean require that the Test Statistic is distributed either according to the normal
distribution or to the t distribution. The Test Statistic in a Hypothesis Test of Mean is derived directly from
the sample mean and therefore has the same distribution as the sample mean.
t-Test vs. z Test

A Hypothesis Test of Mean will either be performed as a z Test or as a t-Test. When the sample mean
and therefore the Test Statistic are distributed according to the normal distribution, the hypothesis test is
called a z Test and the Test Statistic is called the z Score. When the sample mean and therefore the
Test Statistic is distributed according to the t distribution, the hypothesis test is called a t-Test and the
Test Statistic is called the t Value. The Test Statistic is the number of Standard Errors that the observed
sample mean is from the hypothesized population mean.
t-Tests are covered in detail is a separate section in this manual. They are also summarized at the end of
this t distribution section.
z Tests are covered in detail is a separate section in this manual. They are also summarized at the end of
the normal distribution section.
200
Means of Large Samples Are Normally Distributed
According to the Central Limit Theorem, the means of large samples will be normally distributed no matter
how the population from which the samples came is distributed. This is true as long as the samples are
random and the sample size, n, is large (n > 30). n equals the number of data observations that each
sample contains.
If the single sample taken for a Hypothesis Test of Mean is large (n > 30), then the means of a number of
similar samples taken from the same population would be normally distributed as per the Central Limit
Theorem. This is true no matter how the population or the single sample are distributed.
If the single sample taken for a Hypothesis Test of Mean is small (n < 30), then the means of a number of
similar samples taken from the same population would be normally distributed only if the population was
proven to be normally distributed or if the sample was proven to be normally distributed.
Requirements of a z-Test
A z Test can be performed only if the sample mean (and therefore the Test Statistic, which is derived from
the sample mean) is normally distributed. The sample mean and therefore the Test Statistic are normally
distributed only when the following two conditions are both met:
1) The size of the single sample taken is large (n > 30). The Central Limit Theorem states that means of
large samples will be normally distributed. When the size of the single sample is small (n < 30), only a t-
2) The population standard deviation, σ (sigma), is known.
Requirements of a t-Test
A t-Test can be performed only if the sample mean (and therefore the Test Statistic, which is derived from
the sample mean) is distributed according to the t distribution. The sample mean and therefore the Test
Statistic are distributed according to the t distribution when both of these conditions are met:
1) The sample standard deviation, s, is known.
2) Either the sample or the population has been verified for normality.
A t-Test can be performed when the single sample is large (n > 30) but is the only option when the size of
the single sample is small (n < 30). A z Test can only be performed when the size of the single sample is
large (n > 30) and the population standard deviation is known.
As mentioned, a Hypothesis Test of Mean requires that the sample mean and therefore the Test Statistic
is distributed either according to the normal distribution or to the t distribution. The sample mean and the
Test Statistic are distributed variables that can be graphed according to the normal or t distribution.
The Test Statistic, which represents the number of Standard Errors that the sample mean is from the
hypothesized population mean, could be graphed on a standard normal distribution curve or a
standardized t distribution curve. Both these two distribution curves have their means at zero and the
length of one Standard Error is set to equal 1.
201
Basic Steps of a Hypothesis Test of Mean
The major steps of the simple Hypothesis Test of Mean, a one-sample t-Test, are described as follows:
1) A sample of data is taken. The sample statistic which is the sample mean is calculated.
normal graph that maps the distributed variable. The Critical Values represent the boundaries between
the Region of Acceptance and Region of Rejection.
5) Critical t Values, the Test Statistic (the t Value) and p Value are then calculated.
6) The Null Hypothesis is rejected if any of the following three equivalent conditions are shown to exist:
a) The observed sample mean, x_bar, is beyond the Critical Value.
b) The t Value (the Test Statistic) is farther from zero than the Critical t Value.
202
The following graph represents the final result of a typical one-sample, two-tailed t-Test. In this case the
Null Hypothesis was rejected. This result is represented as follows:
This t-Test was a two-tailed test as evidenced by the yellow Region of Rejection split between the both
The mean of this non-standardized t-distribution curve is 186,000. This indicates that the Null Hypothesis
is as follows:
H0: x_bar = 186,000
H1: x_bar ≠ 186,000
This one-sample t-Test is evaluating whether the population from which the sample was taken has a
population mean that is not equal to 186,000. This is a non-directional t-Test and is therefore two-tailed.
The sample statistic is the observed sample mean of this single sample taken for this test. This observed
sample mean is calculated to be 200,000.
The boundaries of the Region of Rejection occur at 172,083 and 199,916. Everything beyond these two
point is in the Region of Rejection. These two Critical Values are 1.95 Standard Errors from the
standardized mean of 0. This indicates that the Critical t Values are ±1.95.
The graph shows that the sample statistic (the sample mean of 200,000) falls beyond the right Critical
value of 199,916 and is therefore in the Region of Rejection.
The sample statistic is 2.105 Standard Errors from the standardized mean of 0. This is further from the
203
smaller than α/2 which is 2.5 percent of the total curve area because alpha was set to 0.05.
As the graph shows, all three equivalent conditions have been met to reject the Null Hypothesis. It can be
stated with at least 95 percent certainty that the mean of the population from which the sample was taken
does not equal the hypothesized population mean of 186,000.
Uses of Hypothesis Tests of Mean

1) Comparing the mean of a sample taken from one population with the another population’s
mean to determine if both populations have the different means. An example of this would be to
compare the mean monthly sales of a sample of retail stores from one region to the national mean
monthly store sales to determine if the mean monthly sales of all stores in the one region are different
than the national mean.
2) Comparing the mean of a sample taken from one population to a fixed number to determine if
that population’s mean is different than the fixed number. An example of this might be to compare
the mean product measurement taken a sample of a number of units of a product to the company’s
claims about that product specification to determine if the actual mean measurement of all units of that
company’s product is different than what the company claims it is.
3) Comparing the mean of a sample from one population with the mean of a sample from another
population to determine if the two populations have different means. An example of this would be to
compare the mean of a sample of daily production totals from one crew with the mean of a sample of
daily production totals from another crew to determine if the two crews have different mean daily
production totals.
anything has changed between measurements. An example of this would be to evaluate whether there
is mean difference in before-and-after tests scores of a small sample of the same people to determine if a
training program made a difference to all of the people who underwent it.
be to evaluate whether there is mean difference in the incomes of husbands and wives in a sample of
married couples to determine if there is a mean difference in the incomes of husbands and wives in all
married couples.
It is important to note that a hypothesis test is used to determine if two populations are different. The
204
Types of Hypothesis Tests of Mean
Hypothesis Tests of Mean are either t-Tests or z-Tests.
The 4 types of t-Tests discussed here are the following:
One-sample t-Test
Two-Independent-Sample, Pooled t-Test
Two-Independent-Sample, Unpooled t-Test
Two-Dependent-Sample (Paired) t-Test
The 3 types of z-Test discussed in this manual are the following:
One-sample z-Test
Two-independent-Sample, Unpooled z-Test
Two-Dependent-Sample (Paired) z-Test
A detailed description of each of the 4 types of t-Tests along with examples in Excel are as follows:
205
1) One-Sample t-Test in Excel
This hypothesis test determines whether the mean of the population from which the sample was taken is
equal to (two-tailed test) or else greater than or less than (one-tailed test) than a constant. This constant
is often the known mean of a population from which the sample may have come from. The constant is the
constant on the right side of the Null Hypothesis.
df = n - 1
Null Hypothesis H0: x_bar = Constant
1) The observed x_bar is beyond the Critical Value.
2) The t Value (the Test Statistic) is farther from zero than the Critical t Value.
Example of a 1-Sample, 2-Tailed t-Test in Excel

This problem is very similar to the problem solved in the z-test section for a one-sample, two-tailed t-test.
differences between the one-sample z-Test and t-test as easily as possible.
This problem compares average monthly sales from one fast food chain’s retail stores in one region with
the average monthly sales of all of the fast food chain’s retails in the entire country. The region being
evaluated has more than 1,000 very similar stores. The national mean monthly retail store sales equals
$186,000.
Determine with at least 95% certainty whether the average monthly sales of all of the fast food chain’s
stores in the one region is different than the national average monthly sales of all of the fast food chain’s
stores.
The data sample of sales for the same month for a random sample of 20 fast food retail stores in a region
is as follows:
$240,000 $180,000 $200,000 $260,000 $200,000
$220,000 $180,000 $200,000 $200,000 $180,000
$160,000 $180,000 $200,000 $220,000 $220,000
$140,000 $220,000 $200,000 $240,000 $160,000
206
Running the Excel data analysis tool Descriptive Statistics will provide the Sample Mean, the Sample
Standard Deviation, the Standard Error, and the Sample Size. The output of this tool appears as follows:

x_bar = sample mean = AVERAGE() = 200,000
µ = national (population) mean = 186,000
s = sample standard deviation =STDEV.S() = 29735.68
σ (Greek letter “sigma”) = population standard deviation = Not Known
SE = Standard Error = s / SQRT(n) = 29735.68 / SQRT(20)

Note that this calculation of the Standard Error using the sample standard deviation, s, is an estimate of
the true Standard Error which would be calculated using the population standard deviation, σ.
SE = 6649.10
df = degrees of freedom = n – 1 = 20 – 1 = 19
207


Step 1) Create the Null Hypothesis and the Alternate Hypothesis
Step 2 – Map the Normal or t Distribution Curve Based on the Null Hypothesis
Question 1) Type of Test?

This is a Hypothesis Test of Mean because each individual observation (each sampled monthly sales
figure) within the sample can have a wide range of values. Data observations for Hypothesis Tests of
Proportion are binary: they can take only one of two possible values.

This is a one-sample hypothesis test because only one sample containing monthly sales figures from
twenty stores has been taken and is being compared to the national monthly retail store average for the
same month.
c) Independent (Unpaired) Test or a Dependent (Paired) Test?

It is neither. The designation of “paired” or “unpaired” applies only for two-sample hypothesis tests.
d) One-Tailed or Two-Tailed Hypothesis?

The problem asks to determine whether the twenty-store monthly average is simply different than the
national average. This is a non-directional inequality making this hypothesis test a two-tailed test. If the
problem asked whether the twenty-store average was greater than or less than the national average, the
inequality would be directional and the resulting hypothesis test would be a one-tailed test. A two-tailed
test is more stringent than a one-tailed test.
Assuming that the population or sample can pass a normality test, a hypothesis test of mean must be
performed as a t-Test when the sample size is small (n < 30) or if the population variance is unknown.
In this case the sample size is small as n = 20. This Hypothesis Test of Mean must therefore be
performed as a t-Test and not as a z Test.
The t Distribution with degrees of freedom = df = n – 1 is defined as the distribution of random data
sample of sample size n taken from a normal population.
208
The means of samples taken from a normal population are also distributed according to the t
Distribution with degrees of freedom = df = n – 1.
The Test Statistic (the t Value, which is based upon sample mean (x_bar) because it equals (x_bar –
Constant)/(s/SQRT(n)) will therefore also be distributed according to the t Distribution. A t-Test will be
performed if the Test Statistic is distributed according to the t Distribution.
The distribution of the Test Statistic for sample taken from a normal population is always described by the
t Distribution. The shape of the t Distribution converges to (very closely resembles) the shape of the
standard normal distribution when sample size becomes large (n > 30).
The Test Statistic’s distribution can be approximated by the normal distribution only if the sample size is
large (n > 30)
and
the population standard deviation, σ, is known. A z Test can be used if the Test Statistic’s distribution can
be approximated by the normal distribution. A t-Test must be used in all other replaced be their equivalent
t-Tests. As a result, some major commercial statistical software packages including the well-known SPSS
provide only t-Tests and no direct z Tests.cases.
It should be noted that a one-sample t-Test can always be used in place of a one-sample z Test. All z
Tests can be replaced with the equivalent t-Tests.
This hypothesis test is a t-Test that is one-sample, two-tailed hypothesis test of mean as long as
all required assumptions have been met.

a) t-Distribution of Test Statistic
A t-Test can be performed if the distribution of the Test Statistic (the t value) can be approximated under
the Null Hypothesis by the t Distribution. The Test Statistic is derived from the mean of the sample taken
and therefore has the same distribution that the sample mean would have if multiple similar samples were
taken from the same population.
The sample size indicates how to determine the distribution of the sample mean and therefore the
distribution of the Test Statistic as follows:
When Sample Size Is Large

When the sample size is large (n > 30), the distribution of means of similar samples drawn from the same
population is described by the t Distribution. As per the Central Limit Theorem, as sample size increases,
the distribution of the sample means converges to the normal distribution as does the t Distribution. When
sample size approaches infinity, the t Distribution converges to the standard normal distribution.
When sample size is large, the distribution of the sample mean, and therefore the distribution of the Test
Statistic, is always described by the t Distribution. A t-Test can therefore always be used when sample
size is large, regardless of the distribution of the population or sample.
When Sample Size is Small

The data in a sample taken from a normally-distributed population will be distributed according to the t
Distribution regardless of sample size.
The means of similar random samples taken from a normally-distributed population are also distributed
according to the t Distribution regardless of sample size.
The sample mean, and therefore the Test Statistic, are distributed according to the t Distribution if the
population is normally distributed.
209
The population is considered to be normally distributed if any of the following are true:
1) The population from which the sample was taken is shown to be normally distributed.
2) The sample is shown to be normally distributed. If the sample passes a test of normality then the
population from which the sample was taken can be assumed to be normally distributed.
The population or the sample must pass a normality test before a t-Test can be performed. If the only
data available are the data of the single data taken, then sample must pass a normality test before a t-
Evaluating the Normality of the Sample Data

There are a number of normality tests that can be performed on the data of this example. The quickest
way to check the sample data for normality is to create an Excel histogram of the data as shown below, or
to create a normal probability plot of the data if you have access to an automated method of generating
that kind of a graph.
To create this histogram in Excel, fill in the Excel Histogram dialogue box as follows:
210
The sample group appears to be distributed reasonably closely to the bell-shaped normal distribution. It
should be noted that bin size in an Excel histogram is manually set by the user. This arbitrary setting of
the bin sizes can has a significant influence on the shape of the histogram’s output. Different bin sizes
could result in an output that would not appear bell-shaped at all. What is actually set by the user in an
Excel histogram is the upper boundary of each bin.
Nonparametric Alternatives in Excel

The Sign Test and Wilcoxon One-Sample Signed-Rank Test are nonparametric alternative to the one-
sample t-test when the normality assumption of sampled data is questionable. The one-sample t-test is
used to evaluate whether a population from which samples are drawn has the same mean as a known
value. The nonparametric tests evaluate whether the sample have the same median as a known value.
The Sign Test is a much less powerful alternative to the Wilcoxon One-Sample Signed-Rank test, but
does not assume that the differences between the samples and the known value is symmetrical about a
median, as does the Wilcoxon One-Sample Signed-Rank test when used as a nonparametric alternative
to the one-sample t-test. The Sign Test is non-directional and can be substituted only for a two-tailed test
but not for a one-tailed test.
211
Step 1 – Create the Null and Alternate Hypotheses

The Null Hypothesis is always an equality that states that the items being compared are the same. In this
case, the Null Hypothesis would state that the average monthly sales of all stores in the region (the
population from which the twenty-store sample was taken) is not different than the national monthly store
average sales, µ, which is $186,000. We will use the variable x_bar to represent the sample mean of the
twenty stores. The Null Hypothesis is as follows:
The Constant is quite often the known population mean, µ, to which the sample mean is being compared.
different. This hypothesis test is trying to determine whether the average monthly sales of all stores in the
region (the population from which the twenty-store sample was taken) is merely different than the national
monthly store average sales, µ, which is $186,000.
The Alternative Hypothesis is as follows:
H1: x_bar ≠ Constant, which is 186,000
H1: x_bar ≠ 186,000
It is important to note that the Null and Alternative Hypotheses refer to the means of the populations from
which the samples were taken. A one-sample t-Test determines whether to reject or fail to reject the Null
Hypothesis that states that that population from which the sample was taken (the entire region) has a
mean equal to the Constant. The Constant in this case is equal to known national average.
Parameters necessary to map the distributed variable, x_bar, to the t Distribution are the following:
SE = Standard Error = s / SQRT(n) = 29735.68 / SQRT(20)
212
Step 2 – Map the Distributed Variable to t-Distribution
The means of similar, random samples taken from a normal population are distributed according to the t
Distribution. This means that the distribution of a large number of means of samples of size n taken from
a normal population will have the same shape as a t Distribution with its degrees of equal to n – 1.
The sample mean and the Test Statistic are both distributed according to the t Distribution with degrees of
freedom equal to n – 1 if the sample or population is shown to be normally distributed. This step will map
the sample mean to a t Distribution curve with a degrees of freedom equal to n – 1.
The t Distribution is usually presented in its finalized form with standardized values of a mean that equals
zero and a standard error that equals one. The horizontal axis is given in units of Standard Errors and the
distributed variable is the t Value (the Test Statistic) as follows:
A non-standardized t Distribution curve would simply have its horizontal axis given in units of the measure
used to take the samples. The distributed variable would be the sample mean, x_bar.
213
The variable x_bar is distributed according to the t Distribution. Mapping this distributed variable to a t
Distribution curve is shown as follows:
This non-standardized t Distribution curve has its mean set to equal the Constant taken from the Null
This non-standardized t Distribution curve is constructed from the following parameters:
Mean = 186,000 (This x_bar is hypothesized by the Null Hypothesis to be the curve’s mean.)
Standard Error = 6,649.10
Degrees of Freedom = 19
Distributed Variable = x_bar
214
to show graphically how different the sample mean, x_bar = $200.000, is from the national average of
$186,000.
The non-standardized t Distribution curve can be divided up into two types of regions: the Region of
If the sample mean’s value of x_bar = 200,000 falls into a Region of Rejection, the Null Hypothesis is
rejected. If the sample mean’s value of x_bar = 200,000 falls into a Region of Acceptance, the Null
Hypothesis is not rejected.
means that the Region of Rejection will take up 5 percent of the total area under this t distribution curve.
This 5 percent is divided up between the two outer tails. Each outer tail contains 2.5 percent of the curve
that is the Region of Rejection.
locations of these Critical values need to be calculated.
Calculate Critical Values

A Critical Value is the boundary between a Region of Acceptance and a Region of Rejection. In the case
of a two-tailed test, the Region of rejection is split between two outer tails. There are therefore two Critical
Values.
The Critical Value is the boundary on either side of the curve beyond which 2.5 percent of the total area
under the curve exists. In this case both Critical Values can be found by the following:
Critical Values = Mean ± T.INV(1-α/2,df) * SE
Critical Values = 186,000 ± T.INV(0.975, 19) * 6649.1
Critical Values = 186,000 ± 13,916
Critical Values = 172,083 and 199,916
The Region of Rejection is therefore everything that is to the right of 199,916 and everything to the left of
172,083.
is as follows:
The hypothesized mean is the hypothesized x_bar which is 186,000.
215
The non-standardized t Distribution curve with the blue Region of Acceptance and the yellow Regions of
Rejection divided by the Critical Values is shown is in the following Excel-generated graph of this non-
standardized t Distribution curve:

follows:
1) Compare the x-bar With Critical Value
Reject the Null Hypothesis if the sample mean, x_bar = 200,000, falls into the Region of Rejection.
Equivalently, reject the Null Hypothesis if the sample mean, x_bar, is further the curve’s mean of 186,000
the Critical Value.
The Critical Values have been calculated to be 172,083 on the left and 199,916 on the right. X_bar
(200,000) is further from the curve mean (186,000) than right
Critical Value (199,916). The Null Hypothesis would therefore be rejected.
2) Compare the t Value With Critical t Value

The t Value is the number of Standard Errors that x_bar is from the curve’s mean of 186,000.
The Critical t Value is the number of Standard Errors that the Critical Value is from the curve’s mean.
Reject the Null Hypothesis if the t Value is farther from the standardized mean of zero than the Critical t
Value.
216
Equivalently, reject the Null Hypothesis if the t Value is closer to the standardized mean of zero than the
Critical t Value.
t Value (Test Statistic) = (200,000 – 186,000)/6,649.1

t Value (Test Statistic) = 2.105
This means that the sample mean, x_bar, is 2.105 standard errors from the curve mean (186,000).
Critical t Values = ±T.INV(1-α/2,df)
Critical t Values = ±T.INV(1-0.05/2,19)
Critical t Values = ±2.093
This means that the boundaries of the Region of Rejection are 2.093 standard errors from the curve
mean (186,000) on each side since this is a two-tailed test.
The Null Hypothesis is rejected because the t Value (2.105) is farther from the standardized mean of zero
than the Critical t Value on that side (+2.093) indicating that x_bar is in the Region of Rejection.

The p Value is the percent of the curve that is beyond x_bar (200,000). If the p Value is smaller than
Alpha/2 (since this is a two-tailed test), the Null Hypothesis is rejected.
p Value = T.DIST.RT(ABS(2.105), 19)
p Value = 0.0244
217
The p Value (0.0244) is smaller than Alpha/2 (0.025) Region of Rejection in the right tail and we therefore
reject the Null Hypothesis. A graph below shows that the red p Value (the curve area beyond x_bar) is
smaller than the yellow Alpha, which is the 5 percent Region of Rejection split between both outer tails.
This is shown in the following Excel-generated graph of this non-standardized t Distribution curve:
It should be noted that if this t-Test were a one-tailed test, which is less stringent than a two-tailed test,
the Null Hypothesis would still have been rejected because:
1) The p Value (0.0244) would still be smaller than the Alpha (0.05) Region of Rejection, which is now
entirely contained in the right tail
2) x_bar (200,000) would still be outside the Region of Acceptance, which would now have its outer right
boundary at 197,497.2 (mean + T.INV(Alpha,df)*SE).
3) The t Value (2.105) would still be larger than the critical t Value which would now be 1.73 (t Value =
T.INV(Alpha,df))
218
Excel Shortcut to Performing a One-Sample t-Test
All of the three other types of t-Tests (two-independent-sample pooled and unpooled t-Tests along with
the paired t-Test) can be solved in one step with a built-in Excel formula and also with a built-in Data
Analysis tool for each t-Test.
Excel unfortunately does not provide a formula or tool that can perform or solve a one-sample t-Test in
one step. Interestingly enough, a one-sample z-Test can be solved in Excel in one step with the following
formula:
p Value = MIN(Z.TEST(array,Constant,σ),1- Z.TEST(array,Constant,σ))
array = Set of sample data
Constant = the Constant in the Null Hypothesis
σ = Population standard deviation
There is no such method in Excel to perform a one-sample t-Test similarly in a single step. The other
three types of t-Tests each have a one-step tool and a one-step formula. One of the main reasons that
these tools and formulas are one-step is that the t Value is calculated automatically. There is no one-
sample t-Test tool or formula that automatically calculates the t Value while performing the t-Test or
calculating the p Value. The t Value must be calculated in its own step when performing a one-sample t-
Test in Excel.
The formula needed to perform a one-sample t-Test is the following as previously shown:
This formula requires that the t Value be calculated first. This must be done manually using the following
steps:
t Value = (x-bar – Constant)/SE
SE = s/SQRT(n)
The one-sample t-Test is a very common statistical test so it is surprising that Excel does not have a one-
step formula or a Data Analysis tool to directly calculate either the p Value or t Value given the array and
the Constant from the Null Hypothesis. Each of the other three types of t-Tests has its own specific
formulas and its own Data Analysis tools to perform either entire the t-Test or calculate the p Value in a
single step.
219
2) Two-Independent-Sample, Pooled t-Test in Excel
tailed test) than a constant. This is a pooled test because a single pooled standard deviation replaces
both sample standard deviations because they are similar enough.
Pooled t-Tests are performed if the variances of both sample groups are similar. A rule-of-thumb is as
follows: A Pooled t-Test should be performed if the standard deviation of one sample, s 1, is no more than
twice as large as the standard deviation in the other sample s 2. That is the case here for the following
example.
dfpooled = degrees of freedom = n1 + n2 – 2

220
Example of 2-Sample, 1-Tailed, Pooled t-Test in Excel
In this example two different brand of the same type of battery are being tested to determine if there
probably is a real difference in the average length of time that batteries from each of the two brands last.
The length of each battery’s lifetime of operation in minutes was recorded. Determine with 95 percent
certainty whether Brand A batteries have a longer average lifetime than Brand B batteries.
Here are the data samples from batteries of the two brands:
Running the Excel data analysis tool Descriptive Statistics separately on each sample group produces the
following output:
221
Note that when performing two-sample t-Tests in Excel, always designate Sample 1 (Variable 1) to be the
The results of the Pooled t-Test will be more intuitive if the sample group with the larger mean is
sample.
correct result from the Excel data analysis tool t-Test: Two-Sample Assuming Equal Variances. The
test statistic (T Stat in the Excel output) and the Critical t value (t Critical two-tail in the Excel output) will
have the same sign (as they always should) only if the sample group with the larger mean is designated
the first sample.

Sample Group 1 – Brand A (Variable 1)
x_bar1 = sample1 mean = AVERAGE() = 43.56
s1 = sample1 standard deviation =STDEV.S() = 16.92
Var1 = sample1 variance =VAR() = 286.13
σ1 (Greek letter “sigma”) = population standard deviation from which Sample 1 was drawn = Not Known
n1 = sample1 size = COUNT() = 16
222
Sample Group 2 – Brand B (Variable 2)
µ2 (Greek letter “mu”) = population mean from which Sample 2 was drawn = Not Known or needed to
solve this problem
or needed to solve this problem
x_bar1 - x_bar2 = 43.56 – 33.53 = 10.03




This is a Hypothesis Test of Mean because each individual observation (each sampled battery’s lifetime)
within each of the two sample groups can have a wide range of values. Data points for Hypothesis Tests
of Proportion are binary: they can take only one of two possible values.

other. The two sample groups are the lifetimes in minutes of 16 Brand A batteries and the lifetimes of 17
Brand B batteries.
223
c) Independent a Dependent (Paired) Test?

The problem asks to determine whether the average lifetime of Brand A batteries is greater than the
average lifetime of Brand B batteries. This is a directional inequality making this hypothesis test a one-
tailed test. If the problem asked whether the average lifetimes of both brands was simply different, the
inequality would be non-directional and the resulting hypothesis test would be a two-tailed test. A two-
tailed test is more stringent than a one-tailed test.
e) t-Test or as a z-Test?
A two-independent-sample hypothesis test of mean must be performed as a t-Test if sample size is small
(n1 + n2 < 40). In this case the sample size is small as n1 + n2 = 33. This Hypothesis Test of Mean must
be performed as a t-Test. A t-Test uses the t distribution and not the normal distribution as does a z Test.

follows: A Pooled t-Test should be performed if the standard deviation of one sample is no more than
twice as large as the standard deviation in the other sample. That is the case here as the following are
true:
s1 = sample1 standard deviation = 16.92
and
The variances of both sample groups are verified to be similar enough to permit using a Pooled test for
this two-independent sample hypothesis test.
This hypothesis test is a t-Test that is two-independent-sample, one-tailed, Pooled hypothesis test
of mean.
Question 2) Requirements Met?

the Null Hypothesis by the t Distribution. The t Value for this test is calculated as follows:
To perform a hypothesis test that is based on the normal distribution or t distribution, both sample means
must be normally distributed. In other words, if we took multiple samples just like either one of the two
mentioned here, the means of those samples would have to be normally distributed in order to be able to
perform a hypothesis test that is based upon the normal or t distributions.
For example, 30 independent, random samples of the battery lifetimes from each of the two battery
brands could be evaluated just like the single sample of the lifetimes of 15+ batteries from each of the two
battery brands as mentioned here. If the means of all of the 30 samples from one battery brand and,
separately, the means of the other 30 samples from the other battery brand are normally distributed, a
224
hypothesis test based on the normal or t distribution can be performed on the two independent samples
taken.
The means of the samples would be normally distributed if any of the following are true:
1) Sample Size of Both Samples Greater Than 30

The Central Limit Theorem states that the means of similar-sized, random, independent samples will be
normally distributed if the sample size is large (n >30) no matter how the underlying population from
which the samples came from is distributed. In reality, the distribution of sample means converges toward
normality when n is as small as 5 as long as the underlying population is not too skewed.
2) Both Populations Are Normally Distributed

If this is the case, the means of similar sized, random, independent samples will also be normally
distributed. It is quite often the case that the distribution of the underlying population is not known and
should not be assumed.
3) Both Samples Are Normally Distributed

If the sample is normally distributed, the means of other similar-sized, independent, random samples will
also be normally distributed. Normality testing must be performed on the sample to determine whether the
sample is normally distributed.
In this case the sample size for both samples is small: n1 and n2 are both less than 30. The normal
distribution of both sample means must therefore be tested and confirmed. Normality testing on each of
the samples has to be performed to confirm the normal distribution of the means of both samples.

There are a number of ways to test sample data for normality but the quickest way is to construct an
Excel histogram from the sample data.
225
Excel histograms of both sample groups are as follows:
226
Both sample groups appear to be distributed reasonably closely to the bell-shaped normal distribution. It
227
b) Similarity of Sample Variances
The two-independent sample, Pooled t-Test requires that the two independent samples have similar
variances. Samples that have similar variances are said to be homoscedastistic. Samples that have
significantly different variances are said to be heteroscedastistic. The samples in this example have the
similar variance variances. This is confirmed by the variances comparison tests that were previously
mentioned in this example.
c) Independence of Samples
This type of a hypothesis test requires both samples be totally independent of each other. In this case
they are completely independent. There is no relationship between the observations that make up each of
the two sample groups.
When normality of data cannot be confirmed for a small sample, it is necessary to substitute a
nonparametric test for a t-Test. Nonparametric tests do not have the same normality requirement that the
t-Test does. The most common nonparametric test that can be substituted for the two-independent-
sample t-Test when data normality cannot be confirmed is the Mann-Whitney U Test.
228
Step 1 – Create Null and Alternate Hypotheses
The Alternate Hypothesis is always in inequality and states that the two items being compared are
different. This hypothesis test is trying to determine whether the mean of the population from which the
first sample (x_bar1) was taken is greater than the mean of the population from which the second sample
was taken (x_bar2).
The Alternate Hypothesis is as follows:
H1: x_bar1-x_bar2 > Constant, which is 0
H1: x_bar1-x_bar2 > 0
The Alternative Hypothesis is directional (“greater than” or “less than” instead of “not equal”) and the
hypothesis test is therefore a one-tailed test. The “greater than” operator in the Alternative hypothesis
indicates that this one-tailed test occurs in the right tail. It should be noted that a two-tailed test is more
rigorous (requires a greater differences between the two entities being compared before the test shows
that there is a difference) than a one-tailed test.
The following formulas are used by the Two-Independent Sample, Pooled t-Test:
Pooled Degrees of Freedom
df = degrees of freedom = n1 + n2 - 2
df = 16 + 17 – 2 = 31
Pooled Sample Standard Deviation
2 2
sPooled = SQRT[{(n1-1)s1 +(n2-1)s2 }/df]
2 2
sPooled = SQRT[{(16-1)*(16.915) +(17-1)*(15.277) }/31]
sPooled = 16.09
Pooled Sample Standard Error
SEPooled = sPooled *SQRT(1/n1 + 1/n2)

SEPooled = 16.09 * SQRT(1/16 + 1/17)
SEPooled = 5.6046
229
2
Note that this calculation of the Standard Error using the sample variance, s , is an estimate of the true
2
Standard Error which would be calculated using the population variance, σ , of the populations from
which the samples were drawn.
These parameters are used to map the distributed variable, x_bar1-x_bar2, to the t Distribution as follows:
Step 2 – Map Distributed Variable on a t-Distribution Curve

A t-Test can be performed if the sample mean, and the Test Statistic (the t Value) are distributed
according to the t Distribution. If the sample has passed a normality test, the sample mean and closely-
related Test Statistic are distributed according to the t Distribution.
The t Distribution always has a mean of zero and a standard error equal to one. The t Distribution varies
only in its shape. The shape of a specific t Distribution curve is determined by only one parameter: its
degrees of freedom, which equals n – 1 if n = sample size.
a normal population will hw3ave the same shape as a t Distribution with its degrees of equal to n – 1.
used to take the samples. The distributed variable would be the sample mean, x_bar1-x_bar2.
230
The variable x_bar1-x_bar2 is distributed according to the t Distribution. Mapping this distributed variable
to a t Distribution curve is shown as follows:

Mean = 0, which is the Constant taken from the Null Hypothesis (This x_bar1-x_bar2 is hypothesized by
the Null Hypothesis to be the curve’s mean.)
Standard Error Pooled = 5.604
Distributed Variable = : x_bar1-x_bar2

from zero.
Null Hypothesis is rejected. If the difference between the sample means, x_bar1-x_bar2 (10.033), falls into
a Region of Acceptance, the Null Hypothesis is not rejected.
231
H1: x_bar1-x_bar2 > 0
A “greater than” or “less than” operator indicates that this will be a one-tailed test. The “greater than” sign
indicates that the Region of Rejection will be in the right tail.
One-Tailed Critical Values

A Critical Value is the boundary between a Region of Acceptance and a Region of Rejection. The entire
5-percent alpha region lies beyond the Critical Value because this is a one-tailed test. The Critical Value
can be found as follows:
Critical Value = Mean + (Number of Standard Errors from Mean to Region of Rejection) * SE
Critical Value = Mean + T.INV(1-α,df) * SE
Critical Value = 0 + T.INV(1-0.05, 31) * 5.604
Critical Value = 0 + 9.503
Critical Value = +9.503
The Region of Rejection therefore includes everything that is to the right of +9.503.
is as follows:
The distribution curve with the blue Region of Acceptance and the yellow Region of Rejection is shown is
as follows:
232
If this were a two-tailed test, the Critical values would be determined as follows:

Critical Values = 0 ± T.INV(0,975, 31) * 5.604
Critical Values = -11.43 and +11.43
The Critical Values for the two-tailed test (-11.43 and +11.43) are farther from the mean than the Critical
Value for one-tailed test (+9.503). This means that a two-tailed test is more stringent than a one-tailed
test.

tests needs to be performed because all three provide equivalent information The three tests are as
follows:
1) Compare the Sample Mean With Critical Value

Do not reject the Null Hypothesis if the sample mean, x_bar1-x_bar2 = 10.033, falls into the Region of
Acceptance
Equivalently, reject the Null Hypothesis if the sample mean, x_bar1-x_bar2 = 10.033, is further the curve’s
mean of 0 than the Critical Value
233
The Critical Value has been calculated to be 9.503. The observed x_bar1-x_bar2 (10.033) is further from
the curve mean (0) than Critical Value (9.503). The Null Hypothesis would therefore be rejected.
2) Compare t Value With Critical t Value

The t Value is the number of Standard Errors that x_bar1-x_bar2 (10.033) is from the curve’s mean of 0
The Critical t Value is the number of Standard Errors that the Critical Value (9.503) is from the curve’s
mean
Value.
The t Value, the Test Statistic in a t-Test, is the number of Standard Errors that x_bar 1-x_bar2 is from the
mean. The Critical t Value is the number of Standard Errors that the Critical Value is from the mean. If the
t Value is larger than the Critical t Value, the Null Hypothesis can be rejected.
t Value (test statistic) = (x_bar1 - x_bar2 - 0) / SE

t Value (test statistic) = (10.033)/5.604 = 1.790
One-Tailed (Right Tail) Critical t Value = T.INV(1-α,df)
One-Tailed (Right Tail) Critical t Value = T.INV(1-0.05, 31) = 1.696
This indicates that (x_bar1 - x_bar2) is 1.696 standard errors to the right of the Constant, which is 0.
The t Value (1.790) is farther from the standardized mean of zero than the Critical t Value (1.696) so the
Null Hypothesis is rejected.

The p Value is the percent of the curve that is beyond x_bar1-x_bar2 (10.033). If the p Value is smaller
than Alpha, the Null Hypothesis is rejected.
p Value = T.DIST.RT(ABS(1.790), 31)
p Value = 0.042
The p Value (0.042) is smaller than Alpha (0.05) and we therefore reject the Null Hypothesis. A graph
below shows that the red p Value (the curve area beyond x_bar1-x_bar2) is smaller than the yellow Alpha,
which is the 5 percent Region of Rejection in the outer right tail. This is shown in the following Excel-
generated graph of this non-standardized t Distribution curve:
234
The value of x_bar1-x_bar2, 10, has a t Value of 1.79 and therefore is 1.79 standard errors from the mean.
This is further from the mean than the critical value of 9.6, which is the t critical distance of 1.69 standard
errors from the mean.
It should be noted that if this t-Test were a two-tailed test, which is more stringent than a one-tailed test,
the Null Hypothesis would be accepted because:
1) The p Value (0.042) would now be larger than Alpha/2 (0.025)
2) x_bar1-x_bar2 (10.033) would now be in the Region of Acceptance, which would now have its outer
right boundary at 11.43 (mean + T.INV(1-α/2,df)*SE)

This two-independent-sample, Pooled t-Test can be solved much quicker using the following Excel data
analysis tool:
t-Test: Assuming Equal Variances
Before this test is employed, all required assumptions such as normality of data must be verified as was
done.
The two-independent sample, pooled t-Test can be quickly solved in Excel using either the Data Analysis
tool or the formula that are both specific for this test. The Excel tool can be found by clicking Data
Analysis under the Data tab. The tool is titled t-Test:Two-Sample Assuming Equal Variances. The
entire Data Analysis Toolpak is an add-in that ships with Excel but must first be activated by the user
before it is available. This tool will be applied to the following data set using the same data as the
preceding example in this section.
As mentioned, the data should be input with the sample having the largest mean being designated as the
sample group. Doing so ensures that the Excel output will have the same signs for the t Value and Critical
t Value. If the sample with the smallest mean is input is the first sample, the t Value will correctly be
negative but the Critical t Value will be incorrectly listed by Excel as being positive.
235
Following are screen shots of how the data should be entered:
The completed dialogue box for this tool is shown as follows:
236
Clicking OK will produce the following result. This result agrees with the calculations that were performed
in this section.
237
The calculations to create the preceding output were performed as follows. The individual outputs are
color-coded so it is straight-forward to match the calculations with the outputs of the tool.
238
239
240
Excel Statistical Function Shortcut
Another very quick way to perform this t-Test is to calculate the p value and compare it to Alpha (for a
one-tailed test) or Alpha/2 for a two-tailed test.
The p Value of this two-independent-sample, Pooled t-Test can be very quickly using the following Excel
statistical function:
=T.TEST(array 1,array2,1,2)
done.
The stand-alone Excel formula to perform a two-independent sample, pooled t-Test is shown as follows. If
the resulting p Value is smaller than α for a one-tailed test or α/2 for a two-tailed test, the difference
between the means of the samples is deemed to be statistically significant. This indicates that the two
samples were likely drawn from different populations.
The Null Hypothesis of the t-Test would not be rejected if the test were two-tailed because the p Value
(0.042) is greater than Alpha/2 (0.025).
The Null Hypothesis of the t-Test would be rejected if the test were one-tailed because the p Value
(0.042) is less than Alpha/2 (0.025). A one-tailed test is less stringent than a two-tailed test.
241
3) Two-Independent-Sample, Unpooled t-Test in Excel
tailed test) than a constant. This is an unpooled test because a single pooled standard deviation
CANNOT replace both sample standard deviations because they are too different.
Unpooled t-Tests are performed if the variances of both sample groups are not similar. A rule-of-thumb is
as follows: A Pooled t-Test should only be performed if the standard deviation of one sample, s 1, is no
more than twice as large as the standard deviation in the other sample s 2. An unpooled t-Test should be
performed if that condition is not met.
242
Example of 2-Sample, 2-Tailed, Unpooled t-Test in Excel
This problem is very similar to the problem solved in the z-test section for a two-independent-sample, two-
tailed z-test. Similar problems were used in each of these sections to show the similarities and also
contrast the differences between the two-independent-sample z-Test and t-test as easily as possible.
Two shifts on a production are being compared to determine if there is a difference in the average daily
number of units produced by each shift. The two shifts operate eight hours per day under nearly identical
conditions that remain fairly constant from day to day. A sample of the total number of units produced by
each shift on a random selection of days is taken. Determine with a 95 percent Level of Confidence if
there is a difference between the average daily number of units produced by the two shifts.
Here is the sampled data:
243
Running the Excel data analysis tool Descriptive Statistics separately on each sample group produces the
following output:
Note that when performing two-sample t-Tests in Excel, always designate Sample 1 (Variable 1) to be the
The results of the Unpooled t-Test will be more intuitive if the sample group with the larger mean is
sample.
correct result from the Excel data analysis tool t-Test:Two-Sample Assuming Unequal Variances. The
test statistic (T Stat in the Excel output) and the Critical t value (t Critical two-tail in the Excel output) will
have the same sign (as they always should) only if the sample group with the larger mean is designated
the first sample.

Sample Group 1 – Shift A (Variable 1)
244
Sample Group 2 – Shift B (Variable 2)
x_bar1 - x_bar2 = 46.55 – 42.24 = 4.31



The initial two questions that need to be answered before performing the Four-Step Hypothesis Test of
Mean are as follows:

This is a Hypothesis Test of Mean because each individual observation (each sampled shift’s output)
within each of the two sample groups can have a wide range of values. Data points for Hypothesis Tests
of Proportion are binary: they can take only one of two possible values.

other. The two sample groups are the daily units produced by Shift A and the daily units produced by Shift
B.
245

The problem asks to determine whether there is simply a difference in the average number of daily units
produced by Shift A and by Shift B. This is a non-directional inequality making this hypothesis test a two-
tailed test. If the problem asked to determine whether Shift A’s production is greater than or less than
than Shift B’s, the inequality would be directional and the resulting hypothesis test would be a one-tailed
test. A two-tailed test is more stringent than a one-tailed test.
A two-independent-sample hypothesis test of mean must be performed as a t-Test if sample size is small
(n1 + n2 < 40). In this case the sample size is small as n 1 + n2 = 37 This Hypothesis Test of Mean must be
performed as a t-Test. A t-Test uses the t distribution and not the normal distribution as does a z Test.

follows: A Pooled t-Test should be performed if the standard deviation of one sample is no more than
twice as large as the standard deviation in the other sample. That is definitely not the case here as the
following are true:
and
The variances of both sample groups are dissimilar enough to force using an Unpooled test for this two-
independent sample hypothesis test.
This hypothesis test is a t-Test that is two-independent-sample, two-tailed, Unpooled hypothesis
test of mean.

the Null Hypothesis by the t Distribution. The t Value for this test is calculated as follows:
To perform a hypothesis test that is based on the normal distribution or t distribution, both sample means
must be normally distributed. In other words, if we took multiple samples just like either one of the two
perform a hypothesis test that is based upon the normal or t distributions.
For example, 30 independent, random samples of the daily production from each of the two shifts could
be evaluated just like the single sample of units produced from 15+ production days from each of the two
shifts as mentioned here. If the means of all of the 30 samples from one shift and, separately, the means
of the other 30 samples from the other shift are normally distributed, a hypothesis test based on the
normal or t distribution can be performed on the two independent samples taken.
246


If this is the case, the means of similar sized, random, independent samples will also be normally
distributed. It is quite often the case that the distribution of the underlying population is not known and
should not be assumed.

In this case the sample size for both samples is small: n 1 and n2 are both less than 30. The normal
distribution of both sample means must therefore be tested and confirmed. Normality testing on each of
the samples has to be performed to confirm the normal distribution of the means of both samples.

Normal

There are many ways to evaluate normality of sample data but the quickest way is to create an Excel
histogram from the sample data.
Excel histograms of both sample groups are as follows:
247
248
Both sample groups appear to be distributed reasonably closely to the bell-shaped normal distribution. It
b) Significantly Different Sample Variances

The two-independent sample, Unpooled t-Test expects that the two independent samples have
significantly different variances. Samples that have similar variances are said to be homoscedastistic.
Samples that have significantly different variances are said to be heteroscedastistic. The samples in this
example have the significantly different variances. This is confirmed by the variances comparison tests
that were previously mentioned in this example.
This type of a hypothesis test requires both samples be totally independent of each other. In this case
they are completely independent.
t-Test does. The most common nonparametric test that can be substituted for the two-independent-
sample t-Test when data normality cannot be confirmed is the Mann-Whitney U Test.
249

different. This hypothesis test is trying to determine whether the first mean (x_bar1) is different than the
second mean (x_bar2). The Alternate Hypothesis is as follows:
H1: x_bar1-x_bar2 ≠ Constant, which is 0
The following formulas are used by the Two-Independent Sample, Unpooled t-Test:
Unpooled Degrees of Freedom
df =INT( [ { (Var1/n1) + (Var2/n2) }^2 ] / [ {(Var1/n1)^2 / (n1 - 1) } + { (Var2/n2)^2 / (n2-1) } ] )

df = INT( [ { (613.84/20) + (139.32/17) }^2 ] / [ {(613.84/20)^2 / (20 - 1) } + { (139.32/17)^2 / (17 - 1) } ] )
df = 28
Unpooled Sample Standard Error
SE = SQRT[ (Var1/n1) + (Var2/n2) ]

SE = SQRT[ (613.84/20) + (139.32/17) ] = 6.236
250
2
Note that this calculation of the Standard Error using the sample variance, s , is an estimate of the true
2
Standard Error which would be calculated using the population variance, σ , of the populations from
which the samples were drawn.
These parameters are used to map the distributed variable, x_bar1-x_bar2, to the t Distribution curve as
follows:
Step 2 – Map the Distributed Variable on a t-Distribution Curve

A t-Test can be performed if the sample mean, and the Test Statistic (the t Value) are distributed
according to the t Distribution. If the sample has passed a normality test, the sample mean and closely-
related Test Statistic are distributed according to the t Distribution.
a normal population will have the same shape as a t Distribution with its degrees of equal to n – 1.
used to take the samples. The distributed variable would be the sample mean, x_bar1-x_bar2.
251
The variable x_bar1-x_bar2 is distributed according to the t Distribution. Mapping this distributed variable
to a t Distribution curve is shown as follows:

Mean = 0, which is the constant taken from the Null Hypothesis. This x_bar1-x_bar2 is hypothesized by
the Null Hypothesis to be the curve’s mean.
Standard Error = 6.236
Distributed Variable = : x_bar1-x_bar2

from zero.
Null Hypothesis is rejected. If the difference between the sample means, x_bar1-x_bar2 (4.31), falls into a
Region of Acceptance, the Null Hypothesis is not rejected.
252
A “not equal” operator indicates that this will be a two-tailed test. This means that the Region of Rejection
is split between both outer tails.

Critical Values = 0 ± T.INV(0.975, 28) * 6.236
The Region of Rejection therefore includes everything that is to the right of +12.77 and to the left of -
12.77.
is as follows:
The distribution curve with the blue Region of Acceptance and the yellow Regions of Rejection is shown
is as follows:
253
If this were a one-tailed test, the Critical Value would be determined as follows:
One-Tailed Critical Value

The Region of Rejection would be in the right tail because Sample 1 has the higher mean.
Critical Value = Mean + T.INV(1-α,df) * SE
Critical Value = 0 + T.INV(1-0.05, 28) * 6.236
Critical Value = 0 + 10.61
Critical Value = +10.61
The Critical Values for the two-tailed test are farther from the mean than the Critical Value for one-tailed
test. This means that a two-tailed test is more stringent than a one-tailed test.

The object of a hypothesis test is to determine whether to reject or fail to reject the Null Hypothesis. There
are three equivalent-Tests that determine whether to accept or reject the Null Hypothesis. Only one of
these tests needs to be performed because all three provide equivalent information. The three tests are
as follows:
1) Compare Sample Mean With Critical Value

Fail to Reject the Null Hypothesis if the sample mean, x_bar1-x_bar2 = 4.31, falls into the Region of
Acceptance.
254
Equivalently, reject the Null Hypothesis if the sample mean, x_bar1-x_bar2, is further the curve’s mean of
0 then the Critical Value. Fail to reject the Null Hypothesis if the sample mean, x_bar1-x_bar2, is closer the
curve’s mean of 0 then the Critical Value.
The Critical Values have been calculated to be -12.77 on the left and +12.77 on the right. x_bar1-x_bar2
(4.31) is closer from the curve mean (0) than right Critical Value (+12.77). The Null Hypothesis would
therefore not be rejected.

The t Value corresponds to the standardized value of the sample mean, x_bar1-x_bar2 = 4.31. The t Value
is the number of Standard Errors that x_bar is from the curve’s mean of 0.
Value.
Critical t Value.
The t Value, the Test Statistic in a t-Test, is the number of Standard Errors that x_bar 1-x_bar2 is from the
mean. The Critical t Value is the number of Standard Errors that the Critical Value is from the mean. If the
t Value is larger than the Critical t Value, the Null Hypothesis can be rejected.
t Value (test statistic) = (x_bar1 - x_bar2 - 0) / SE

t Value (test statistic) = (4.31)/6.239 = 0.69
Two-Tailed Critical t Values = ± T.INV(1-α/2,df)

Two-Tailed Critical t Values = ±TINV(1- 0.05/2, 28) = ±2.048
Right Critical t Value = +2.048

This indicates that (x_bar1 - x_bar2) is 2.048 standard errors to the right of the mean (mean = Constant =
0).
The t Value (0.69) is much closer to the mean (mean = Constant = 0) than the Critical t Value (+2.048) so
the Null Hypothesis is accepted.
255
The p Value is the percent of the curve that is beyond x_bar 1-x_bar2 (4.31). If the p Value is larger than
Alpha/2 (since this is a two-tailed test), the Null Hypothesis is accepted. The p Value in this case is
calculated by the following Excel formula:
p Value = T.DIST.RT(ABS(0.69), 28) = 0.247
The p Value (0.247) is much larger than Alpha/2 (0.025 – because this is one tail of a two-tailed test) and
we therefore accept the Null Hypothesis. A graph below shows that the red p Value (the curve area
beyond x_bar1-x_bar2) is much larger than the yellow Alpha, which is the 2.5 percent Region of Rejection
in the outer right tail. This is shown in the following Excel-generated graph of this non-standardized t
Distribution curve:
It should be noted that if this t-Test were a one-tailed test, which is less stringent than a two-tailed test,
the Null Hypothesis would still be accepted because:
1) The p Value (0.247) is still much larger than Alpha (0.05)
2) x_bar1-x_bar2 (4.31) is still in the Region of Acceptance, which would now have its outer right boundary
at 10.61 (mean + T.INV(1-α,df)*SE)
3) the t Value (0.69) would still be smaller than the Critical t Value which would now be 1.70 (TINV(1-
α,df))
256
analysis tool:
t-Test: Assuming Unequal Variances
This Excel tool can be found by clicking Data Analysis under the Data tab. The tool is titled t-Test:Two-
Sample Assuming Unequal Variances. The entire Data Analysis Toolpak is an add-in that ships with
Excel but must first be activated by the user before it is available.
done.
As mentioned, the data should be input with the sample having the largest mean being designated as the
sample group. Doing so ensures that the Excel output will have the same signs for the t Value and Critical
t Value. If the sample with the smallest mean is input is the first sample, the t Value will correctly be
negative but the Critical t Value will be incorrectly listed by Excel as being positive.
257
This tool will be applied to the following data set using the same data as the preceding example in this
section:
258
259
in this section.
260
261
262
The stand-alone Excel formula to perform a two-independent sample, unpooled t-Test is shown as
follows. If the resulting p Value is smaller than α for a one-tailed test or α/2 for a two-tailed test, the
difference between the means of the samples is deemed to be statistically significant. This indicates that
the two samples were likely drawn from different populations.
263
4) Paired (Two-Sample Dependent) t-Test in Excel
Overview
This hypothesis test determines whether the mean of a sample of differences between pairs of data
(x_bardiff) is equal to (two-tailed test) or else greater than or less than (one-tailed test) than a constant.
Before-and-after fitness levels of individuals undergoing a training program would be an example of
paired data. The sample evaluated would be the group of differences between the before-and-after
scores of the individuals. This is called the difference sample.
x_bardiff = Observed Difference Sample Mean
df = n – 1
Null Hypothesis H0: x_bardiff = Constant
1) The observed x_bardiff is beyond the Critical Value.
Example of Paired, 1-Tailed t-Test in Excel

This problem is very similar to the problem solved in the z-test section for a paired, one-tailed z-test.
differences between the paired z-Test and t-test as easily as possible.
A new clerical program was introduced to a large company with the hope that clerical errors would be
reduced. 5,000 clerical workers in the company underwent the training program. 17 Clerical employees
who underwent the training were randomly selected. The average number of clerical errors that each of
these 17 employees made per month for six months prior to the training and also for six months following
the training were recorded. Each of the 17 employees had a similar degree of clerical experience within
the company and performed nearly the same volume and type of clerical work in the before and after
months.
Based upon the results of the 17 sampled clerical employees, determine with 95 percent certainty
whether the average number of monthly clerical mistakes was reduced for the entire 5,000 clerical
employees who underwent the training.
264
This problem illustrates why the t-test is nearly always used instead of a z-Test to perform a two-
dependent-sample (paired) hypothesis test of mean. The z-Test requires the population standard
deviation of the differences between the pairs be known. This is often not the case, but is required for a
paired z-Test . The t-test requires only the sample standard deviation of the sample of paired differences
be known.
Before and After Results and Their Differences Are As Follows:
Running the Excel data analysis tool Descriptive Statistics on the column of Difference data produces the
following output:
Running the Excel data analysis tool Descriptive Statistics on the column of Difference data will provide
the Sample Mean, the Sample Standard Deviation, the Standard Error, and the Sample Size. It will even
provide half the width of a confidence interval about the mean based on this sample for any specified
level of certainty if that option is specified.
265
The output of this tool appears as follows:

sdiff = sample standard deviation = STDEV.S() = 6.4
df = n – 1 = 16
SEdiff = Standard Error = sdiff / SQRT(n) = 6.4 / SQRT(16)

SEdiff = 1.55
266
Note that this calculation of the Standard Error using the sample standard deviation, sdiff, is an estimate of
the true Standard Error which would be calculated using the population standard deviation, σ diff.



This is a Hypothesis Test of Mean because each individual observation (each sampled difference) within
the sample can have a wide range of values. Data points for Hypothesis Tests of Proportion are binary:
they can take only one of two possible values.

This is a two-sample hypothesis test because the data exists in two groups of measurements. One
sample group contains Before measurements and the other sample group contains After measurements.

This is a paired (dependent) hypothesis test because each Before observation has a related After
observation made on the same person.

The problem asks to determine whether there has been a reduction in clerical mistake from Before to
After. This is a directional inequality making this hypothesis test a one-tailed test. If the problem asked
whether Before and After were simply different, the inequality would be non-directional and the resulting
hypothesis test would be a two-tailed test. A two-tailed test is more stringent than a one-tailed test.
Assuming that the difference population or difference sample can pass a normality test, a hypothesis test
of mean must be performed as a t-Test when the difference sample size (n = number of difference pairs)
is small (n < 30) or if the variance of differences is unknown.
267
In this case the difference sample size (the number of data pairs) is small as n = 17 data sample pairs.
This Hypothesis Test of Mean must therefore be performed as a t-Test and not as a z Test.
The t Distribution with degrees of freedom = df = n – 1 is defined as the distribution of random data
sample of sample size n taken from a normal population.
The means of samples taken from a normal population are also distributed according to the t
Distribution with degrees of freedom = df = n – 1.
The Test Statistic (the t Value, which is based upon the difference sample mean (x_bardiff) because it
equals (x_bardiff – Constant)/(SEdiff) will therefore also be distributed according to the t Distribution. A t-
Test will be performed if the Test Statistic is distributed according to the t Distribution.
The distribution of the Test Statistic for the difference sample taken from a normal population of
differences is always described by the t Distribution. The shape of the t Distribution converges to (very
closely resembles) the shape of the standard normal distribution when the difference sample size
becomes large (n > 30).
The Test Statistic’s distribution can be approximated by the normal distribution only if the difference
sample size is large (n > 30) and the population standard deviation, σ, is known. A z Test can be used if
the Test Statistic’s distribution can be approximated by the normal distribution. A t-Test must be used in
all other cases.
It should be noted that a paired t-Test can always be used in place of a paired z Test. All z Tests can be
replaced be their equivalent t-Tests. As a result, some major commercial statistical software packages
including the well-known SPSS provide only t-Tests and no direct z Tests.
This hypothesis test is a t-Test that is two-sample, paired (dependent), one-tailed hypothesis test
of mean.

a) Test Statistic Distributed According to t Distribution
the Null Hypothesis by the t Distribution. The Test Statistic is derived from the mean of the difference
sample and therefore has the same distribution that the difference sample mean would have if multiple
similar samples were taken from the same population of differences between data sample pairs.
The difference sample size indicates how to determine the distribution of the difference sample mean and
therefore the distribution of the Test Statistic as follows:
When Difference Sample Size Is Large

When the difference sample size is large (n > 30 meaning that there are more than 30 pairs of data), the
distribution of means of similar samples drawn from the same population of differences is described by
the t Distribution. As per the Central Limit Theorem, as the difference sample size increases, the
distribution of the difference sample means converges to the normal distribution as does the t Distribution.
When the difference sample size approaches infinity, the t Distribution converges to the standard normal
distribution.
When the difference sample size is large, the distribution of the distribution sample mean, and therefore
the distribution of the Test Statistic, is always described by the t Distribution. A t-Test can therefore
always be used when the difference sample size is large, regardless of the distribution of the population
of differences or the difference sample.
When the Difference Sample Size is Small

The data in a difference sample taken from a normally-distributed population of paired differences will be
distributed according to the t Distribution regardless of the difference sample size.
268
The means of similar difference samples randomly taken from a normally-distributed population of paired
differences are also distributed according to the t Distribution regardless of the difference sample size.
The difference sample mean, and therefore the Test Statistic, are distributed according to the t
Distribution if the population of paired differences is normally distributed.
The population of paired differences is considered to be normally distributed if any of the following are
true:
1) Population of Paired Differences Is Normally Distributed
2) Difference Sample Is Normally Distributed

If the difference sample passes a test of normality then the population of paired difference from which the
difference sample was taken can be assumed to be normally distributed.
The population of paired differences or the difference sample must pass a normality test before a t-Test
can be performed. If the only data available are the data of the single difference sample taken, then
difference sample must pass a normality test before a t-Test can be performed.
Evaluating Normality of the Difference Sample

Normal


There are many ways to evaluate normality of data but the quickest is to create an Excel histogram of the
data as shown in the following diagram:
269
The sample of differences appears to be distributed reasonably closely to the bell-shaped normal
distribution. It should be noted that bin size in an Excel histogram is manually set by the user. This
arbitrary setting of the bin sizes can has a significant influence on the shape of the histogram’s output.
Different bin sizes could result in an output that would not appear bell-shaped at all. What is actually set
by the user in an Excel histogram is the upper boundary of each bin.
b) Significantly Different Sample Variances

t-Test does. The most common nonparametric tests that can substituted for the paired t-Test when data
normality cannot be confirmed are the Wilcoxon Signed-Rank Test and the nonparametric Sign Test.
The Wilcoxon Signed Rank Test is the more powerful nonparametric test but requires that data be
relatively symmetrical about a median. The nonparametric Sign Test can be used when this requirement
cannot be met.
The Wilcoxon Signed Rank Test also can only be performed in ratio or interval data. The nonparametric
Sign test can be performed on ordinal data.
Ratio and interval scales have measurable differences between values. Ordinal scales do not. The
absolute temperature scale is a ratio scale because the value of zero means that there is no heat. The
Fahrenheit or Celsius temperature scales are interval scales because the zero value is arbitrarily placed.
A Likert rating scale would be an example of ordinal data.
The Sign Test is non-directional and can only be substituted for a two-tailed test but not for a one-tailed
test like the example in this section.
270

case, the Null Hypothesis would state that the there is no difference between before and after data. We
will use the variable x_bardiff to represent the mean between the before and after measurements. The
Null Hypothesis is as follows:
different. This hypothesis test is trying to determine whether there has been a reduction in clerical errors,
i.e., the After measurements are, on average, smaller than the Before measurements. The Alternate
H1: x_bardiff < Constant , which is 0
H1: x_bardiff < 0
The Alternative Hypothesis is directional (“greater than” or “less than” instead of “not equal,” which is non-
directional) and the hypothesis test is therefore a one-tailed test. The “less than” operator indicates that
this is a one-tailed test with the Region of Rejection (the alpha region) entirely contained in the left tail. A
“greater than” operator would indicate a one-tailed test focused on the right tail.
It should also be noted that a two-tailed test is more rigorous (requires a greater differences between the
two entities being compared before the test shows that there is a difference) than a one-tailed test.
It is important to note that the Null and Alternative Hypotheses refer to the means of the population of
paired differences from which the difference samples were taken. A population of paired differences
would be the differences of data pairs in a population of data pairs.
A paired t-Test determines whether to reject or fail to reject the Null Hypothesis that states that that
population of paired differences from which the difference sample was taken has a mean equal to the
Constant. The Constant in this case is equal to 0. This means that the Null Hypothesis states that the
average difference between data pairs of an entire population from which the sample of data pairs were
drawn is zero.
Parameters necessary to map the distributed variable, x_bardiff, are the following:
sdiff = sample standard deviation = STDEV.S() = 6.4
df = n – 1 = 16
SEdiff = Standard Error = sdiff / SQRT(n) = 6.4 / SQRT(16)

These parameters are used to map the distributed variable, x_bardiff, to the t Distribution curve as follows:
271
Step 2 – Map Distributed Variable to t Distribution Curve
A t-Test can be performed if the difference sample mean, and the Test Statistic (the t Value) are
distributed according to the t Distribution. If the difference sample has passed a normality test, then the
difference sample mean and closely-related Test Statistic are distributed according to the t Distribution.
The means of similar, random difference samples taken from a normal population of paired differences
are distributed according to the t Distribution. This means that the distribution of a large number of means
of difference samples of size n taken from a normal population will have the same shape as a t
Distribution with its degrees of equal to n – 1.
The difference sample mean and the Test Statistic are both distributed according to the t Distribution with
degrees of freedom equal to n – 1 if the sample or population is shown to be normally distributed. This
step will map the sample mean to a t Distribution curve with a degrees of freedom equal to n – 1.
272
used to take the samples. The distributed variable would be the sample mean, x_bar diff.
273
The variable x_bardiff is distributed according to the t Distribution. Mapping this distributed variable to a t
Distribution curve is shown as follows:
This non-standardized t Distribution curve has its mean set to equal the Constant taken from the Null
Mean = Constant = 0 (This x_bardiff is hypothesized by the Null Hypothesis to be the curve’s mean.)
Standard Errordiff = 1.55
Distributed Variable = x_bardiff

The goal of a hypothesis test is to determine whether to accept or reject the Null Hypothesis at a given
level of certainty. If the two things being compared are far enough apart from each other, the Null
to show graphically how different x_bardiff (-3.35) is from the hypothesized mean of 0.
The above distribution curve that maps the distribution of variable x_bardiff can be divided up into two
types of regions: the Region of Acceptance and the Region of Rejection.
274
If x_bardiff’s value of -3.35 falls in the Region of Acceptance, we must accept the Null Hypothesis. If
x_bardiff’s value of -3.35 falls in the Region of Rejection, we can reject the Null Hypothesis.
This 5 percent is entirely contained in the outer left tail. The outer left tail contains the 5 percent of the
curve that is the Region of Rejection.
Calculate the Critical Value

The boundary between Region of Acceptance and Region of Rejection is called Critical Value. The
location of this Critical Value need to be calculated as follows.
One-tailed, Left tail Critical Value = Mean - (Number of Standard Errors from Mean to Region of
Rejection) * SEdiff
One-tailed, Left tail Critical Value = Mean + T.INV(α,df) * SEdiff
One-tailed, Left tail Critical Value = 0 + T.INV(0.05, 16) * 1.55
One-tailed, Left tail Critical Value = -2.711
The Region of Rejection is therefore everything that is to the left of -2.711.
is as follows:
The hypothesized mean is the hypothesized x_bardiff which is 0.
The distribution curve with the blue 95-percent Region of Acceptance and the yellow 5-percent Region of
Rejection entirely contained in the left tail is shown is as follows:
275
follows:
1) Compare x-bardiff With Critical Value

Reject the Null Hypothesis if the sample mean, x_bardiff = -3.35, falls into the Region of Rejection.
Equivalently, reject the Null Hypothesis if the sample mean, x_bardiff, is further the curve’s mean of 0 than
the Critical Value.
The Critical Values have been calculated to be -2.71 on the left. x_bardiff (-3.35) is further from the curve
mean (0) than left Critical Value (-2.71). The Null Hypothesis would therefore be rejected.

The t Value corresponds to the standardized value of the sample mean, x_bar diff = -3.35. The t Value is
the number of Standard Errors that x_bardiff is from the curve’s mean of 0.
Value.
Critical t Value.
t Value (Test Statistic) = (x_bardiff) / SEdiff = (-3.35)/1.55

t Value (Test Statistic) = -2.159
This means that the sample mean, x_bardiff, is 2.159 standard errors to the left of the curve mean of 0.
One-tailed, left-tail Critical t Value = T.INV(α,df)
One-tailed, left-tail Critical t Value = T.INV(0.05, 16) = -1.76
This means that the boundary of the Region of Rejection are 1.76 standard errors to the left of the curve
mean of 0 since this is a one-tailed test in the left tail.
The Null Hypothesis is rejected because the t Value is farther from curve mean the Critical t Values
indicating that x_bardiff is in the Region of Rejection.
3) Compare p Value With Alpha

The p Value is the percent of the curve that is beyond x_bardiff (-3.35). If the p Value is smaller than
Alpha, the Null Hypothesis is rejected.
p Value = T.DIST.RT(ABS(-2.159), 16)
p Value = 0.023
276
The p Value (0.023) is smaller than Alpha (0.05) Region of Rejection in the right tail and we therefore
reject the Null Hypothesis. A graph below shows that the red p Value (the curve area beyond x_bar) is
smaller than the yellow Alpha, which is the 5 percent Region of Rejection in the left tail. This is shown in
the following Excel-generated graph of this non-standardized t Distribution curve:
277
analysis tool:
t-Test: Paired Two Sample for Means
The Excel tool can be found by clicking Data Analysis under the Data tab. The tool is titled t-Test:Paired
Two Sample For Means. The entire Data Analysis Toolpak is an add-in that ships with Excel but must
first be activated by the user before it is available.
This hypothesis test creates the sample of differences by subtracting the Before results from the After
results. If the training program has successfully reduced the average number monthly clerical errors per
employee, the resulting average difference (x_bardiff) will be negative.
If the Before data was subtracted from the After data, the After data (in column B) sample should be
designated as Variable 1, as is done here. This ensures that the t Value (T Stat in the Excel output) has
the correct sign, which would be negative in this case.
This tool will be applied to the following data set using the same data as the preceding example in this
section.
278
279
in this section.
280
281
282
The stand-alone Excel formula to perform a paired (two-dependent sample) t-Test is shown as follows. If
the resulting p Value is smaller than α for a one-tailed test or α/2 for a two-tailed test, the mean difference
between the Before and After sample pairs is deemed to be statistically significant.
The p Value is calculated to be 0.023. This is less than Alpha (0.05) or Alpha/2 (0.025) so the Null
Hypothesis for this t-Test would be rejected for both a one-tailed test and a two-tailed test if Alpha is set
to 0.05 (95 percent certainty required for the test).
283
t-Based Confidence Intervals of a Population Mean
Overview
This confidence interval of a population mean is based upon the sample mean being distributed
according to the t distribution. A 95-percent confidence interval of a population mean is an interval that
has a 95-percent chance of containing the population mean.
The sample mean is distributed according to the t distribution if any of the following sets of conditions is in
effect:
1) The sample size is large and the population standard deviation is not known.
2) The sample size is small (n < 30) and the population is shown to be normally distributed.
3) The sample size is small (n < 30) and the sample is proven tobe normally distributed.
Margin of Error = Half Width of C.I. = t Valueα/2 * Standard Error

Margin of Error = Half Width of C.I. = T.INV(1 – α/2,df) * s/SQRT(n)
A confidence interval of a population mean that is based on the normal distribution is z-based. A
confidence interval of a population mean that is based on the t distribution is t-based.
It is much more common to use the t distribution than the normal distribution to create a confidence
interval of the population mean. Requirements for t-based confidence intervals are much less restrictive
than the requirements for a z-based confidence interval.
Example of a t-Distribution-based Confidence Interval

In this example a 95 percent Confidence Interval is created around a sample mean. There is a 95 percent
chance that the population mean is contained within this Confidence Interval.
A company is evaluating whether to purchase a large number of electric-powered machines. An
importance purchase criterion is how long the machine can operate after being fully charged.
To determine how long the machine can be expected to operate on a single charge, the company
purchased 20 machines and fully charged each. Each of these machines was then operated at full speed
until the charge ran out and the machine stopped running. The number of hours that each machine was
able to operate after a full charge at full speed was recorded.
Calculate the interval in which contains that average operation length of all of the machines with 95
percent certainty. In other words, calculate the 95 percent Confidence Interval of the mean operating
length for all machines based upon the sample of 20 machines that was tested. The data provided is as
follows:
284
Running the Excel data analysis tool Descriptive Statistics will provide the Sample Mean, the Sample
Standard Deviation, the Standard Error, and the Sample Size. The output of this tool appears as follows:
285
The above Descriptive Statistics are obtained by running Excel Descriptive Statistics data analysis tool as
shown below. It is important to select the Confidence Level checkbox and specific that confidence level
desired (95 percent in this case). Doing so calculates half of the width of the 95 percent Confidence
Interval using the t distribution as this example will also do. Below is the Descriptive Statistics completed
dialogue box:
286
x_bar = sample mean = AVERAGE() = 250
µ = (population) mean of all machines = Unknown
σ (Greek letter “sigma”) = population standard deviation = Not Known
SE = Standard Error = s / SQRT(n) = 37.170 / SQRT(20) = 8.311
Note that this calculation of the Standard Error using the sample standard deviation, s, is an estimate of
the true Standard Error which would be calculated using the population standard deviation, σ.
As when creating all Confidence of Mean, we must satisfactorily answer these two questions and then
proceed to the two-step method of creating the Confidence Interval
287
What Type of Confidence Interval Should Be created?
The Two-Step Method For Creating Confidence Intervals of Mean are the following:
The Initial Two Questions That Need To Be Answered Before Confidence Interval of the Mean are as
follows:

This is a Confidence Interval of a population mean because each individual observation (each sampled
machine’s length of operation) within the entire sample can have a wide range of values. The samples
values are spread out between 175 and 300. On the other hand, sampled data points used to create a
Confidence Interval of a population proportion are binary: they can take only one of two possible values.

A Confidence Interval created using the t distribution is said to be t-based. A Confidence Interval created
using the normal distribution is said to be z-based. It is much more common to use the t distribution to
create Confidence Intervals of a population mean because the t distribution is much less restrictive. The t
distribution can always be used. The normal distribution can only be used if:
- Sample size is large (n > 30)
AND
- The population standard deviation, σ, is known.
In this case sample size is small (n < 30) and the population standard deviation is not known. The t
distribution must therefore be used to create this Confidence Interval of a population mean. This
Confidence Interval of a population mean will be t-based.
This Confidence Interval will be a Confidence Interval of a population mean and will be created
using the t distribution.

a) Normal Distribution of Sample Mean
We are attempting to create a confidence interval about the sample mean which contains the population
mean. To create a confidence interval that is based on the normal distribution or t distribution, the sample
mean must be normally distributed. In other words, if we took multiple samples just like the one
create a confidence interval that is based upon the normal or t distributions.
For example, 30 independent, random samples of 20 machines each could be tested for mean length of
operation just like the single sample of 20 machines in this example was tested. If those means of all 30
288
samples are normally distributed, a confidence interval based on the t distribution can be created around
the mean of the single sample taken.
1) Sample Size of Each Sample Greater Than 30

2) Population Normally Distributed

. If this is the case, the means of similar sized, random, independent samples will also be normally
distributed. It is quite often the case that the distribution of the underlying population is not known and the
normal distribution of a population should not be assumed until proven.
3) Sample Normally Distributed

In this case the sample size is small (n = 10) and the population’s distribution is unknown. The only
remaining way to verify normal distribution of the sample mean is to verify normal distribution of the
sample. The sample must be therefore be tested and confirmed for normality before a Confidence Interval
based on t distribution can be created.
Evaluating Normality of Sample Data

The quickest way to check the sample data for normality is to create an Excel histogram of the data as
shown below, or to create a normal probability plot of the data if you have access to an automated
method of generating that kind of a graph.
289
The sample group appears to be distributed reasonably closely to the bell-shaped normal distribution. It
We now proceed to the two-step method for creating all Confidence intervals of a population mean.
290
Proceeding through the two steps is as follows:

Half the Width of the Confidence Interval is sometimes referred to the Margin of Error. The Margin of Error
will always be measured in the same type of units as the sample mean is measured in. Calculating the
Half Width of the Confidence Interval using the t distribution would be done as follows in Excel:
Margin of Error = Half Width of C.I. = t-Valueα/2 * Standard Error
Margin of Error = Half Width of C.I. = T.INV(1-α/2, df) * s/SQRT(n)
Margin of Error = Half Width of C.I. = T.INV(0.975, 19) * 37.17/SQRT(20)
Margin of Error = Half Width of C.I. = 2.093 * 8.311 = 17.396 hours

Confidence Interval = x_bar ± 17.396
Confidence Interval = 250 ± 17.396
Confidence Interval = [ 323.604 hours, 267.396 hours ]
A graphical representation of this Confidence Interval is shown as follows:
291
Excel Shortcut For t-Based Confidence Interval
The formula for calculating the Confidence Interval is the following:
Descriptive Statistics in Excel instantly calculates the following:
Sample mean
Half the width of the Confidence Interval at the specified level of confidence using the t distribution
Here, once again, is the Descriptive Statistics for this data set:
This is created with the following information filled in the Excel dialogue box for the Descriptive Statistics
data analysis tool:
292
Note the following from the Descriptive Statistics:
Half the Width of the Confidence Interval = 17.396 hours
These numbers can simply be plugged into the Confidence Interval formula below to obtain the t-based
C. I. as long as the sample mean has been proven to be normally distributed.
The half-width of a t-based confidence interval can also be quickly found by the following Excel formula:
Half-width of a t-based confidence interval = CONFIDENCE.T(α,s,n)
Half-width of a t-based confidence interval = CONFIDENCE.T(0.05,37.17,20)
Half-width of a t-based confidence interval = 17.396
293
Calculating the minimum sample size necessary to limit the size of a confidence interval of a population
mean can be done using the normal distribution but not the t distribution. A t-based confidence interval
requires specifying the degrees of freedom, which is derived from the sample size that is unknown.
A z-based confidence interval (a confidence interval based upon the normal distribution) requires that the
sample mean be normally distributed and the population standard deviation be known. These
requirements are met if both of the following are true:
Requirements
1) Minimum Sample Size is 30
This ensures that the sample mean is normally distributed as per the Central Limit Theorem. If the
calculated minimum sample size is less than 30, the sample or the population must be confirmed to be
normally distributed.
AND
2) Population Standard Deviation Known

Sample standard deviation cannot be used because a sample has not been taken.
The minimum sample size, n, to limit the width of a z-based confidence interval of a population proportion
to a specific size can be derived with the following algebra:
Confidence Interval = Sample mean ± z Scoreα,two-tailed * SE
Confidence Interval = x_bar ± NORM.S.INV(1 – α/2) * σ/SQRT(n)
Confidence Interval = x_bar ± NORM.S.INV(1 – α/2) * σ /SQRT(n)
(Half-width of C.I.) = NORM.S.INV(1 – α/2) * σ /SQRT(n)

2 2 2
(Half-width of C.I.) = NORM.S.INV (1 – α/2) * σ /n
Further algebraic manipulation produces the following:
2 2 2
n = [NORM.S.INV (1 – α/2) * σ ] / (Half-width of C.I.)
or, equivalently because Half-width of C.I. = Margin of Error,
2 2 2
n = [NORM.S.INV (1 – α/2) * σ ] / (Margin of Error)
294
2 2 2
Example
A survey was taken of the monthly salaries full-time employee of the California Department of
Transportation. The standard deviation of monthly salaries throughout the entire California DOT is known
to be $500.
What is the minimum number of employees that would have to be surveyed to be at least 95% certain
that the sample average monthly salary is within $50 of the true average monthly salary of all employees
in the California DOT?
In other words, what is the minimum sample size needed to create a 95-percent confidence interval about
the population mean that has a margin of error no larger than $50?
Another way to state the problem is to ask how large must the sample size be to create a 95-percent
confidence interval about the population mean that has a half-width of no more than $50?
σ = Population standard deviation = $500
Half-width of Confidence Interval = Margin of Error = $50
(The confidence interval must be specified in the same units as the population standard deviation is.)
α = 1 – Level of Certainty = 1 – 0.95 = 0.05
2 2 2
n = Roundup( ( NORM.S.INV (1 – α/2) * σ ) / (Half-width of C.I.) , 0)
2 2 2
n = Roundup( ( NORM.S.INV (1 – 0.05/2) * (500) ) / (50) , 0)
2 2 2
n = Roundup( ( NORM.S.INV (0.975) * (500) ) / (50) , 0)
2 2 2
n = Roundup( ( (1.96) * (500) ) / (50) , 0)
n = Roundup( 384.1459, 0 )
n = 385
A minimum of 385 employees must be surveyed to be 95 percent certain that the average salary of the
sample is no more than $50 from the true average salary within the entire California DOT.
295
Binomial Distribution in Excel
Overview
The binomial distribution is the discrete probability distribution of the number of successes in n
successive independent binary trials. Each trial has the same probability p of a successful outcome of the
binary trial’s two possible outcomes.
p = probability of success in each and every binary trial
n = the total number of binary trials
k = a specific number of successful outcomes of n binary trials
X = the actual number of successful outcomes of n binary trials
The probability that X (the actual number of successful outcomes of n successive, independent binary
trials each having probability p of a successful outcome) equals a specific k is given as follows:
for k = 0, 1, 2, …, n,
where
is sometimes referred to as “n choose k.” This is called the binomial coefficient and the source of
the name of the binomial distribution. The binomial coefficient is equal to the number ways that k items
can be combined to create n elements. This is the number combinations that k items can be arranged to
create n total elements.
A different ordering of the same k items does not create a new combination. For example, the three
elements A, B, and C can only be arranged into one combination of three total elements. ABC is not a
different combination than CAB is. A different ordering of the same three elements does, however, create
a new permutation.
296
The Excel formula for the total possible number of combinations of k elements into n total elements is
COMBIN(n,k) is the Excel combination formula.

A combination should be differentiated from a permutation. The number of possible combinations is the
total number of ways that k elements can be arranged into n total elements when order does not matter.
The number of possible permutations is the total number of ways that k elements can be arranged into n
total elements when order does matter.
There are always more permutations of k objects than combinations because, for example, ABC and CAB
are two different permutations of the letters A, B, and C but not two different combinations of those three
letters. Re-arrangement of the same elements within a set creates a new permutation but does not create
a new combination.
The Excel formula for the total possible number of permutations of k elements into n total elements is
PERMUT(n,k) is the Excel permutation formula.
Note that the following is true:
Each Trial is a Bernoulli Trial

Each individual trial is called a Bernoulli trial. The outcome of a Bernoulli trial can be described by the
binomial distribution with n = 1. A Bernoulli trial follows the Bernoulli distribution, which is a special case
of the binomial distribution where n = 1.
Maximum Sample Size Rule

The binomial distribution should not be applied to a single random sample taken from a population unless
the population is at least 10 times larger than the sample size.
297
Binomial Distribution Has Two Parameters n and p
A binomial distribution is fully described by two parameters n (the total number of trials) and p (the
constant and unchanging probability of success on each trial). The binomial distribution is denoted as
follows:
B(n,p)
The binomial distribution is a discrete distribution because its probability equation Pr(X = k) is calculated
only for values of k, which can only assume integer values.
The binomial distribution describes the distribution of the number of successes, X, if the following four
conditions exists:
1) Each trial is a single Bernoulli trial, which is a binary event having only one of two outcomes: success
or failure.
2) The total number of trials = n.
3) Each trial is independent of all other trials.
4) The probability of success in each trial is p, which is constant in all trials. The probability of failure = q =
1 – p.
The binomial distribution requires that each sample taken is returned to the population before the next
sample is taken. Samples are always taken from the same population. This is called sampling with
replacement.
If samples taken are not returned to the population, the hypergeometric distribution is used in place of the
binomial distribution. This is called sampling without replacement. If sample size is much smaller than the
population from the sample was drawn, the binomial distribution provides a good approximation of the
hypergeometric distribution when sampling without replacement is performed. The population size should
be at least ten times as large of the sample size for this substitution to be valid.
Population Parameters of the Binomial Distribution

n = number of trial
p = probability of success on each trial
q = 1 – p = probability of failure on each trial
If X ~ B(n,p) (X is a binomially-distributed variable having n trials and the probability p of success on each
trial. X is a variable representing the number of successes given n and p), then the following is true:
Expected Value(X) = Mean(X) = μX = np
2
Variance(X) = σ X = npq
Standard Deviation(X) = σX =
Applying these formulas to a basic example illustrates the intuitive nature of these formulas. If a coin were
flipped 12 times (n = 12) and the coin was fair (p = 0.5 and q = 1 – p = 0.5), then the following are true
regarding X, the number of successes (heads) in n trials (coin flips) given the probability p of success
(heads) on each trial (coin flip):
Expected Value(X) = Mean(X) = μX = 12 * 0.5 = 6
2
Variance(X) = σ X = npq = 12 * 0.5 * 0.5 = 3
This makes sense because 6 heads is the number of heads that one would expect to occur with 12 flips
of a fair coin.
298
Binomial Distribution’s PDF and CDF
PDF = Probability Density Function
CDF = Cumulative Distribution Function
Binomial Distribution’s PDF - Probability Density Function

The binomial distribution’s PDF is given by the following:
for k = 0, 1, 2, …, n, where
f(k;n,p) = Pr(X = k) is the probability that X, the number of successes, equals k for n independent,
successive Bernoulli trials each having the probability p of success.
For example, for the following parameters:
k=4
n = 10
p = 0.5
Each unique binomial distribution is fully described by two parameters n and p. This binomial distribution
is the distribution X successes in n = 10 trials with p = 0.5 probability of a successful outcome in each and
every one of the 10 trials.
This binomial distribution’s PDF calculates that the probability that X (the actual number of successful
outcomes) equal k which is 4.
The Excel formula to calculate the binomial distribution’s PDF is the following:
f(k;n,p) = Pr(X = k) = BINOM.DIST(k, n, p , FALSE)
FALSE indicates that this Excel formula is calculating the binomial distribution’s PDF and not the CDF for
this k, n, and p. “False” answers the question of the calculation is cumulative (which is the case if
calculating the CDF – Cumulative Distribution Function) on not cumulative (which is the case if calculating
the PDF – Probability Density Function).
Excel 2010 and later also use the formula BINOM.DIST() which is equivalent to BINOMDIST() that is
used by earlier versions of Excel. It should be noted that many of the equivalent but upgraded formulas in
Excel 2010 are more accurate than the original versions and should be used when possible. Microsoft
recommends using the latest possible versions of any statistical formulas.
BINOM.DIST(4,10,0.5,FALSE) = 0.2051
This Excel binomial PDF calculation indicates that there is a 20.51 percent chance of exactly 4
successes in 10 independent, successive Bernoulli trials with 50 percent probability of success on each
trial.
299
A graph of the binomial distribution’s PDF for this unique binomial distribution (n= 10 and p = 0.5) shows
the probability that X = 4 is 0.2051.
The binomial distribution’s curve shifts from left to right as p increases. This is shown is the following
graphs of the binomial’s PDF for n = 10 when p = 0.2 and then as p = 0.8. Note that the graphs of p = 1/5
(0.2) and p = 4/5 (0.8) are mirror images of each other about the mean of 5.
300
301
Binomial Distribution’s CDF - Cumulative Distribution Function
The binomial distribution’s CDF is as follows:
F(k;n,p) = Pr(X ≤ k). This is the probability that X, the number of successes in n Bernoulli trials each
having the probability p of a successful outcome, equals up to k.
The binomial distribution’s CDF is given by the following:
is the “floor” under k, i.e., the greatest integer less than or equal to k.
The Excel formula to calculate the binomial distribution’s CDF is the following:
F(k;n,p) = Pr(X ≤ k) = BINOM.DIST(k, n, p , TRUE)
TRUE indicates that this Excel formula is calculating the binomial distribution’s CDF and not the PDF for
this k, n, and p.
BINOM.DIST(4,10,0.5, TRUE) = 0.3769
There is a 37.69 percent chance of up to 4 successes in 10 independent, successive Bernoulli trials with
50 percent probability of success on each trial. A graph of the binomial distribution’s CDF for this unique
binomial distribution (n= 10 and p = 0.5) shows the probability that X = 4 is 0.3769.
The CDF for any distribution always varies from a minimum value of 0 on the left to a maximum value of 1
on the right. Just as with the PDF, the binomial distribution’s CDF shifts from left to right as p increases.
This is shown is the following graphs of the binomial’s CDF for n = 10 when p = 0.2 and then as p = 0.8.
302
303
Problems Solved With the Binomial Distribution in Excel
Problem 1 – Solving With the Binomial’s PDF in Excel
Calculate the probability of getting exactly 4 heads in 6 flips of a fair coin.
n=6
k=4
p = 0.5
Pr(X=k) = BINOM.DIST(k, n, p, FALSE)
Pr(X=4) = BINOM.DIST(4, 6. 0.5, FALSE) = 0.234
There is a 23.4 percent chance of getting exactly 4 heads in 6 flips of a fair coin.
Problem 2 – Solving With the Binomial’s CDF in Excel

Calculate the probability of getting up to 4 heads in 6 flips of a fair coin.
n=6
k=4
p = 0.5
Pr(X≤k) = BINOM.DIST(k, n, p, TRUE)
Pr(X≤4) = BINOM.DIST(4, 6. 0.5, TRUE) = 0.891
There is a 89.1 percent chance of getting up to 4 heads in 6 flips of a fair coin.
Problem 3 – Calculating a Range of Binomial Probabilities in Excel

What is the probability that between 10 and 25 products out of 100 require service is 15 percent of all
products require service?
Pr( 10≤X≤25 )
This is equal to the intersection of the following two probabilities:
Pr(X≤25 ) AND Pr(X≥10)
which is
= Pr(X≤25 X≥10)
= Pr(X≤25 ) - Pr(X≤9)
Note that the Excel formula to calculate the binomial distribution’s CDF calculates Pr(X≤k). All cumulative
binomial probabilities must therefore be expressed as “up to” (less than or equal to) before they can be
converted to Excel.
= BINOM.DIST(25,100,0.15,TRUE) – BINOM.DIST(9,100,0.15,TRUE)
= 0.9970 – 0.0551
= 0.9419
There is a 94.19 percent probability that between 10 and 25 products out of 100 will need service if 15
percent of products need service.
304
Problem 4 – Sampling With and Without Replacement in Excel
A production line makes batches of 200 items. A random sample of 20 items is taken from every batch
and inspected. The entire batch is rejected of 3 or more of the sampled items are found to be defective.
The production line is known to have a 10 percent defect rate. If the sampling is done without
replacement, what is the probability that a batch will pass the inspection?
The hypergeometric distribution would be the correct distribution to solve this problem because sampling
is done without replacement. The binomial distribution can be substituted for the hypergeometric
distribution if the population size is at least 10 times as large as the sample size. This is the case here
since the population size (batch size) is 200 and the sample size is 20.
A batch passes inspection if 2 or fewer items are found to be defective in the inspection of 20 sampled
items. The probability of a batch passing inspection is equal to the probability that up to 2 items in each
sample are found defective. The CDF (cumulative distribution function) will be used to calculate this.
One reason to estimate the hypergeometric distribution’s CDF with the binomial distribution’s CDF is that
Excel only has a formula to calculate the hypergeometric distribution’s PDf but not its CDF. Calculating
the hypergeometric distribution’s CDF in Excel would require calculating the hypergeometric distribution’s
PDF at every point from 0 to k and then summing those numbers to obtain the hypergoemetric
distribution’s CDF at X=k. These calculations will be performed at the end of this problem’s solution and
compared with the binomial distribution’s CDF at X = k.
The hypergoemetric distribution’s CDF will be estimated using the binomial distribution’s CDF as follows:
A batch passes inspection if there are 2 or fewer defective items found in each sample of 20 items. The
defect rate is known to be 10 percent.
n = 20
k=2
p = 0.1
Pr(X≤k) = BINOM.DIST(k, n, p, TRUE)
Pr(X≤2) = BINOM.DIST(2, 20. 0.1, TRUE) = 0.6769
There is a 67.69 percent chance of finding up to 2 defects in each sample of 20 if the defect rate is 10
percent and the sample is done with replacement. The sampling is done without replacement so the exact
answer would require the use of the hypergeometric distribution. Since the population size is 10 times as
large as the sample size, the hypergeometric distribution can be approximated by using the binomial
distribution is as done here.
The exact answer using the hypergeometric distribution will be calculated and then compared to the
binomial distribution’s estimate as follows:
Excel only provides a formula to calculate the PDF of the hypergeometric distribution but not the CDF. All
versions of Excel calculate the hypergeometric distribution’s PDF with the formula HYPGEOMDIST().
Excel 2010 and later also calculate the hypergeometric distribution’s PDF with the formula
HYPGEOM.DIST(). Both produce equivalent results.
The hypergeometric distribution is a discrete distribution just like the binomial. This means that the
hypergeometric distribution only has values of its PDF and CDF at integer values of k. The distribution’s
CDF is calculated as follows:
Pr(X≤k)
Since the hypergeometric distribution is discrete and values only at integer values of k, the CDF can be
calculated by summing the PDF at each integer from 0 to k as follows:
Pr(X≤k) = Pr(X=0) + Pr(X=1) + Pr(X=2) + … + Pr(X=k)
This method will have to be used because Excel only has a formula to calculate the hypergeometric
distribution’s PDF but not its CDF.
305
The Excel formula to calculate the hypergeometric distribution’s PDF is as follows:
Pr(X=k) = HYPGEOM.DIST(k, n, K, N)
k = number of successes in each sample
n = sample size = 20
K = number of successes in the population = 20
= (Known defect rate) * (Population size)
= 0.10 * 200 = 20
N = population size =200
Calculating the hypergeometric distribution’s CDF would be done as follows in Excel (this uses the legacy
formula HYPGEOMDIST, which is equivalent to the current formula HYPGEOM.DIST):
An entire batch passes inspection if 2 or fewer items in the random sample of 20 items are found
defective. The probability of a batch passing inspection is therefore Pr(X≤2).
HYPGEOMDIST(0,20,20,200) = HYPGEOM.DIST(0,20,20,200,FALSE) = 0.1085

Pr(X≤2) = 0.6787
There is a 67.87 percent chance of finding up to 2 defects in each sample of 20 taken from a population
of 200 if the defect rate is 10 percent and the sample is done without replacement.
The binomial distribution estimate of Pr(X≤2) is 0.6769 which is fairly close to exact answer of 0.6787
found using the hypergeometric distribution.
306
Normal Approximation of the Binomial Distribution
For certain values of n and p, the binomial distribution can be closely approximated by the normal
distribution. This allows for quick and easy normal-distribution-based hypothesis testing and confidence
interval creation for sample data that are binomially distributed.
There are several different guidelines specifying the ranges of n and p that are appropriate for
approximation of the binomial distribution with the normal distribution. The normal approximation of the
binomial distribution works when n is large enough and p and q are not close to zero.
The most widely-applied guideline is the following: np > 5 and nq > 5. Other sources state that normal
approximation of the binomial distribution is appropriate only when np > 10 and nq > 10. this manual will
utilize the first rule-of-thumb mentioned here, i.e., np > 5 and nq > 5.
Each unique binomial distribution is completely described by its two parameters n (number of trials) and p
probability of success in each trial). Each unique normal distribution is completely described by its two
parameters μ (mean) and σ (standard deviation).
For the appropriate ranges of n and p, the normal distribution approximates the binomial distribution by
substituting np for μ and for σ.
The Excel formula to calculate the PDF for the normal distribution at X is given as follows:
NORM.DIST(X,μ,σ,FALSE)
The Excel formula to calculate the PDF for the binomial distribution at X is given as follows:
BINOM.DIST(X,n,p,FALSE)
This is closely approximated for the appropriate ranges of n and p with the normal distribution in Excel as
follows:
NORM.DIST(X,np,SQRT(npq),FALSE)
The CDF is calculated by substituting TRUE for FALSE in the Excel formula.
Following is a comparison of the normal approximation of the binomial distribution as n increases. Both
the PDF and the CDF for binomial distribution and the normal distribution’s approximation are calculated.
The PDF is very close even at low values of n. The CDF remains significantly different in this case until n
has reached 80.
307
Continuity Correction
The continuity correction is an addition of 0.5 to the normal CDF’s X value. This makes the normal CDF
more closely approximate the binomial CDF. The continuity correction of 0.5 is added to X to calculate the
normal CDF in the following Excel chart. It can be seen that the normal CDF is a significantly better
approximation of the binomial CDF with the continuity correction than without it.
A continuity correction factor of +0.5 is applied to the X value when using a continuous function (the
normal distribution) to approximate the CDF of a discrete function (the binomial distribution).
The binomial’s CDF is calculated in Excel by this formula:
BINOM.DIST(X,n,p,TRUE)
The normal approximation would now be calculated by the following formula with the continuity correction
of 0.5 added to X
NORM.DIST(X+0.5,np,SQRT(npq),TRUE)
The continuity correction is much less important than it used to be. Exact values of the binomial’s PDF
and CDF can be calculated with specific Excel formulas. The normal approximation of the binomial
distribution allows for quick and easy normal-distribution-based analysis tools such as hypothesis testing
and confidence intervals to be applied to binomially-distributed data. The continuity correction is not used
when performing these techniques.
308
Hypothesis Testing on Binomial Data
Overview
population.
Null Hypothesis
A hypothesis test is based upon a Null Hypothesis which states that the sample did come from a
hypothesized population. A hypothesis test compares a sample statistic such as a sample mean or
sample proportion to a population parameter such as the population’s mean or proportion. The amount of
difference between the sample statistic and the population parameter determines whether the Null
Hypothesis can be rejected or not.
different.
Null Hypothesis is Either Rejected or Not Rejected But Is Never

Accepted
309
following:
H1: x_bar ≠ 5
H1: x_bar > 5
or
H1: x_bar < 5
following:
H1: p_bar ≠ 0.3
H1: p_bar > 0.3
or
H1: p_bar < 0.3
One-Tailed Test vs. a Two-Tailed Test

Rejection.
310
Level of Certainty

following:
the normal distribution or the t distribution. The Test Statistic is derived directly from the sample statistic
such as the sample mean. If the Test Statistic is distributed according to the normal or t distribution, then
the sample statistic is also distributed according to normal or t distribution. This will be discussed is
sample statistic (the sample mean or proportion) to its own unique normal or t distribution. The sample
statistic is the distributed variable whose distribution is mapped according its own unique normal or t
distribution.
The Region of Acceptance is the percentage of area under this normal or t distribution curve that equals
variable’s mapped normal or t distribution curve.
Region of Rejection
The Region of Rejection is the percentage of area under this normal or t distribution curve that equals the
311
The Region of Acceptance includes 95 percent of the total area under the normal or t distribution curve
The Region of Rejection includes 5 percent of the total area under the normal or t distribution curve that
is not rejected.
Critical Value(s)
Test Statistic
mean.
Level of Certainty.
The Test Statistic is either the z Score or the t Value depending on whether a z Test or t Test is being

Each hypothesis test calculates Critical t or z Values. A Critical t Value is calculated for a t Test and a
Critical z Value is calculated for a z Test. A Critical t or z Value is the amount of difference expressed in
curve’s mean.
between the two outer tails.
The Test Statistic (the t Value or z Score) are compared with the Critical t or z Value on that side of the
mean.
312
If the Test Statistic is farther from the standardized mean of zero than the Critical t or z Value on that side,
the Null Hypothesis is rejected. The Test Statistic is the number of Standard Errors that the sample
statistic is from the curve’s mean. The Critical t or z Value on the same side is the number of Standard
Errors that the Critical Value (the boundary of the Region of Rejection) is from the mean. If the Test
Statistic is farther from the standardized mean of zero than the Critical t or z value, the sample statistic
lies in the Region of Rejection.
Relationship Between p Value and Alpha

The 3 Equivalent Reasons To Reject the Null Hypothesis

1) The sample statistic (the observed sample mean or proportion) is beyond the Critical Value. The
sample statistic would therefore lie in the Region of Rejection because the Critical Value is the boundary
of the Region of Rejection.
2) The Test Statistic (the t value or z Score) is farther from zero than the Critical t or z Value. The
Test Statistic is the number of Standard Errors that the sample statistic is from the curve’s mean. The
Critical t or z Value is the number of Standard Errors that the boundary of the Region of Rejection is from
the curve’s mean. If the Test Statistic is farther from farther from the standardized mean of 0 than the
Critical t or z Value, the sample statistic lies in the Region of Rejection.
3) The p value is smaller than α for a one-tailed test or α/2 for a two-tailed test. The p Value is the
curve area beyond the sample statistic. α and α/2 equal the curve areas contained by the Region of
Rejection on that side for a one-tailed test and a two-tailed test respectively. If the p value is smaller than
α for a one-tailed test or α/2 for a two-tailed test, the sample statistic lies in the Region of Rejection.

313
Power of a Test
following formula:
Power = 1 – β
Effect Size
Effect size in a t-Test or z Test is a convention of expressing how large the difference between two
groups is without taking into account the sample size and whether that difference is significant.
values.
Hypothesis Tests of Mean – Basic Definition
Hypothesis Tests of mean are covered in detail separate sections on t Tests and z Tests.
t Tests are also summarized at the end of the section on the t distribution.
z Tests are also summarized at the end of the section on the normal distribution.
Hypothesis Tests of Proportion – Basic Definition
Data observations in the sample taken for a Hypothesis Test of Proportion are required to be distributed
according to the binomial distribution. Data that are binomially distributed are independent of each other,
314
binary (can assume only one of two states), and all have the same probability of assuming the positive
state.
The binomial distribution can be approximated by the normal distribution under the following two
conditions:
1) p (the probability of a positive outcome on each trial) and q (q = 1 – p) are not too close to 0 or 1.
2) np > 5 and nq > 5
A z Test can be performed on binomially-distributed data if the above conditions are met. Hypothesis
Test of Proportion only use z Tests and not t Tests because the binomial distribution is approximated by
the normal distribution, not the t distribution.
The Test Statistic for a z Test is a z Score.
A Hypothesis Test of Proportion is performed in a very similar manner to a Hypothesis Test of Mean. A
general description of the major steps is as follows:
1) A sample of binary data is taken. The sample proportion is calculated. Examples of a sample
proportion is the proportion of sampled people who are of one gender or the proportion of sampled
production units that are defective.
normal graph that maps the distributed variable.
5) Critical z Values, the Test Statistic (z Score) and p Value are calculated.
6) The Null Hypothesis is rejected if any of the following equivalent conditions are shown to exist:
a) The observed sample proportion, p_bar, is beyond the Critical Value.
b) The z Value (the Test Statistic) is farther from zero than the Critical z Value.
The Null Hypothesis is not rejected in the output of the following Hypothesis Test of Proportion because
none of the above equivalent conditions exist. This is evidenced in the following graph:
315
This z-Test was a two-tailed test as evidenced by the yellow Region of Rejection split between the both
The mean of this non-standardized normal distribution curve is 0.30. This indicates that the Null
H0: p_bar = 0.30
H1: p_bar ≠ 0.30
This one-sample z-Test is evaluating whether the population from which the sample was taken has a
population proportion that is not equal to 0.30. This is a non-directional z-Test and is therefore two-tailed.
The sample statistic is the observed sample proportion of this single sample taken for this test. This
observed sample proportion is calculated to be 0.42.
The boundaries of the Region of Rejection occur at 0.17 and 0.43. Everything beyond these two points is
in the Region of Rejection. Everything inside of these two points is in the Region of Acceptance. These
two Critical Values are 1.95 Standard Errors from the standardized mean of 0. This indicates that the
Critical t Values are ±1.96.
The graph shows that the sample statistic (the sample proportion of 0.42) falls inside the right Critical
Value of 0.43 and is therefore in the Region of Acceptance.
The sample statistic is 1.85 Standard Errors from the standardized mean of 0. This is closer to the
larger than α/2 which is 2.5 percent of the total curve area because alpha was set to 0.05.
The Null Hypothesis is not rejected. As the graph shows, none of the three equivalent conditions have
been met to reject the Null Hypothesis. It cannot be stated with at least 95 percent certainty that the
proportion of the population from which the sample was taken does not equal the hypothesized
population proportion of 0.30.
It should be noted that failure to reject the Null Hypothesis is not equivalent to accepting the Null
Hypothesis. A hypothesis test can only reject or fail to reject the Null Hypothesis.
316
Uses of Hypothesis Tests of Proportion
1) Comparing the proportion of a sample taken from one population with the another population’s
proportion to determine if both populations have the different proportions. An example of this would be to
compare the proportion of monthly purchases returned in a sample of retail stores from one region to the
national mean monthly return rate to determine if the monthly proportion of sales returned in all stores in
the one region is different than the national monthly return rate.
2) Comparing the proportion of a sample taken from one population to a fixed proportion to
determine if that population’s proportion is different than the fixed proportion. An example of this might be
to compare the proportion of a specified chemical measured in a sample of a number of units of a product
to the company’s claims about that product specification to determine if the actual proportion of the
chemical in all units of that company’s product is different than what the company claims it is.
3) Comparing the proportion of a sample from one population with the proportion of a sample
from another population to determine if the two populations have different proportions. An example of
this would be to compare the proportion of defective units of a sample of production runs by one crew
with the proportion of defective units of a sample of production runs by another crew to determine if the
two crews have consistently different proportions of defective units in all of their runs.
anything has changed between measurements. An example of this would be to evaluate whether there is
difference in the proportion of the same people passing a standardized test before and after a training
program to determine if the training program makes a difference in the proportion of all people who take
the standardized test before and after undergoing the training
be to evaluate whether the proportion of total household income brought in by the husband and the wife is
different in a sample of married couples to determine if there is a difference in the proportions of total
household income brought in by husbands and wives in all married couples.
It is important to note that a hypothesis test is used to determine if two populations are different, The
Types of Hypothesis Tests of Proportion

The 3 types of Hypothesis tests of Proportion discussed here are the following:
One-sample Hypothesis Test of Proportion
Two-Independent-Sample, Pooled Hypothesis Test of Proportion
Two-Independent-Sample, Unpooled Hypothesis Test of Proportion
A description of each of these three hypothesis tests of proportion is as follows:
317
1) One-Sample Hypothesis Test of Proportion in Excel
Overview
This hypothesis test analyzes a single sample to determine if the population from which the sample was
taken has a proportion that is equal to a constant, p. In many cases, a one-sample Hypothesis test of
Proportion is used to determine if one population has the same proportion as the known proportion of
another population, p.
This hypothesis test analyzes a single sample to determine if the population from which the sample was
taken has a proportion that is equal to a constant, p. In many cases, a one-sample Hypothesis test of
Proportion is used to determine if one population has the same proportion as the known proportion of
another population, p.
p_bar = observed sample proportion
p_bar = X/n = (Number of successes in the sample)/(Number of trials in the sample)
q_bar = 1 – p_bar
p = Constant (is the hypothesized population proportion, which is often a known population proportion)
q=1–p
SE is calculated using population parameters p and q, not sample statistics p_bar and q_bar.
The z Value for this test, which is the Test Statistic, is calculated as follows:
Null Hypothesis H0: p_bar = Constant = p

1) The observed p_bar is beyond the Critical Value.
2) The z Value (the Test Statistic) is farther from zero than the Critical z Value.
318
Example of a One-Sample, Two-Tailed Hyp. Test of
Proportion in Excel
Following is a one-sample hypothesis test of proportion that is two-tailed:
Over the course of one entire year, 30 percent of all units produced by one production line had at least
one defect. During the next year the first 21 out of 50 units produced by the production line had a defect.
Determined with 95 percent certainty whether production line’s performance has changed.
Note that the question asks only whether there has been any change, not whether there has been a
specific change such as whether there has been a worsening of performance. This means that the
hypothesis test will be a two-tailed test and not a one-tailed test.

p = Population proportion defective for last year = 0.30
The example evaluates whether the population proportion for this year equals p (last year’s proportion
defectives) based on a sample taken from this year’s production.
q = Population proportion not defective for last year
q = 1 – p = 0.70
X = number of defects detected in the sample taken from this year’s production = 21
n = Sample size = 50
p_bar = Observed sample proportion = X/n = 21/50 = 0.42
Required Level of Certainty = 95% = 0.95
Null and Alternative Hypotheses

This Null and Alternative Hypotheses are the following:
H0: p_bar = Constant = p
H0: p_bar = p = 0.3
H1: p_bar ≠ Constant, which equals 0.3
Therefore:
H1: p_bar ≠ 0.3
The “not equals” sign in the Alternative Hypothesis indicates that this is a two-tailed hypothesis test.
Standard Error
SE = SQRT[ (0.3*0.7)/50 ]
SE = 0.0648
319
Note that SE is calculated using p (0.3) from the Null Hypothesis and not p_bar (0.42). q is derived from p
and q_bar is derived from p_bar.
Critical Values
Critical Values = p ± (Number of Standard Errors from Mean to Region of Rejection) * SE
Critical Values = p ± NORM.S.INV(1-α/2) * SE
Critical Values = 0.3 ± NORM.S.INV(1- 0.05/2 ) * 0.06428
Critical Values = 0.3 ± NORM.S.INV(0.975) * 0.06428
Critical Values = 0.30 ± 0.13
Critical Values = 0.43 and 0.17
It should be noted that the p in the Critical Values formula refers to is the p hypothesized by the Null
Hypothesis to be the curve’s mean.
Region of Rejection
The Region of Rejection is therefore everything that is to the right of 0.43 and everything to the left of
0.17.
z Value
z Value = Test Statistic
z Value = (0.42 – 0.3)/0.0648

z Value = 1.85
This means that the observed p_bar (0.42) is 1.85 Standard Errors from the curve’s mean.
Critical z Values
Critical z Valuesα=0.05,two-tailed = ±NORM.S.INV(1-α/2)
Critical z Valuesα=0.05,two-tailed = ±NORM.S.INV(0.975) = ±1.96
This means that the left and right boundaries of the Regions of Rejection are 1.96 Standard Errors from
the curve’s mean.
p Value
p Value = MIN(NORM.S.DIST(z Value,TRUE),1-NORM.S.DIST(z Value,TRUE))
p Value = MIN(NORM.S.DIST(1.85,TRUE),1-NORM.S.DIST(1.85,TRUE))
p Value = 0.032
This means that curve area beyond (to the right of) the observed p_bar (0.42) equals 3.2 percent of the
total area under the normal curve.
320
Determine Whether to Reject Null Hypothesis
Perform the Critical Value Test, the p Value Test, or the Test Statistic Test
1) Compare p_bar with Critical Value

If the observed value of p_bar (0.43) falls into the Region of Acceptance (the blue region under the
curve), the Null Hypothesis is not rejected. If the observed value of p_bar falls into the Regions of
Rejection (either of the two yellow outer regions), the Null Hypothesis is rejected.
The observed p_bar (0.42) is closer to the curve’s mean (0.3) than the Critical Value (0.43) and falls in
the blue Region of Acceptance. We therefore do not reject the Null Hypothesis. We cannot state with 95
percent certainty that there is a real difference between the overall defect rates of this year and last year
based upon the defect rate of the sample taken from this year’s production.
2) Compare z Value with Critical z Value

The z Value is the number of Standard Errors that the observed p_bar is from the standardized mean of
zero. The Critical z Value is the number of Standard Errors that the Critical Value is from the mean.
If the z Value is closer to the standardized mean of zero than the Critical z Value, the Null Hypothesis is
not rejected. If the z Value is farther from the standardized mean of zero than the Critical z Value, the Null
The z Value (1.85) is closer to the curve’s standardized mean of zero than the Critical z Value (1.96) so
the Null Hypothesis is not rejected.
3) Compare the p Value to Alpha

The p Value is the percent of the curve that is beyond the observed p_bar (0.42). If the p Value is smaller
than Alpha/2 (since this is a two-tailed test), the Null Hypothesis is rejected. If the p Value is larger than
Alpha/2, the Null Hypothesis is not rejected.
The p Value (0.032) is larger than Alpha/2 (0.025) and we therefore do not reject the Null Hypothesis.
The following Excel-generated graph shows that the red p Value (the curve area beyond p_bar) is larger
than the yellow Alpha/2 Region of Rejection in the outer right tail.
321
2) Two-Sample, Pooled Hypothesis Test of Proportion
This hypothesis test analyzes two independent samples to determine if the populations from which the
samples were taken have equal proportions. This test is often used to determine whether two sample
proportions are likely the same. The test is a called pooled test because the Null Hypothesis states that
the two sample proportions are the same. The formula for Standard Error uses a pooled proportion that
combines the proportions of both samples. This formula is shown in the following set of formulas.
p_bar1 = observed sample 1 proportion
p_bar1 = X1/n1
= (Number of successes in sample 1)/(Number of trials in sample 1)
p_bar2 = X2/n2
The z Value for this test is the Test Statistic and is calculated as follows:
Null Hypothesis H0: p_bar2 - p_bar1 = Constant = 0

This Null Hypothesis that states that p_bar1 = p_bar2 indicates that this is a pooled test.
1) The observed p_bar2 - p_bar1 is beyond the Critical Value.
2) The z Value (the Test Statistic) is farther from zero than the Critical z Value.
322
Example of a Two-Sample, Two-Tailed Pooled Hyp. Test of
Proportion in Excel
Following is result of a two-independent-sample pooled hypothesis test of proportion that is two-tailed:
A new 401K plan is being proposed to the employees of a large company as a replacement for the
existing plan.
90 out of 200 randomly sampled male employees prefer the proposed plan over the existing plan.
59 out of 100 randomly sampled female employees prefer the proposed plan over the existing plan.
Determine with 95 percent certainty whether there is a difference between the proportions of male and
female employees who prefer the proposed plan.
Note that this will be a pooled z Test because this objective of this hypothesis test is determine whether
there is a difference between p_bar1 and p_bar2, i.e. p_bar1 – p_bar2 = 0 or p_bar2 – p_bar1 = 0.
Two-independent-sample Hypothesis Tests of Proportion remain slightly more intuitive and allow for
consistent use of the variable name p_bar2–p_bar1 if the larger sample proportion is always designated
as p_bar2.and the smaller sample proportion is designated as p_bar1.

First Sample (Male Employees)
X1 = number of positive outcomes in n1 trials = 90
n1 = number of trials (sample size) = 200
p_bar1 = X1/n1 = 90/200 = 0.45
Second Sample (Female Employees)

X2 = number of positive outcomes in n2 trials = 59
n2 = number of trial (sample size) = 100
p_bar2 = X2/n2 = 59/100 = 0.59
p_bar2 – p_bar1 = 0.59 – 0.45 = 0.14

The Null and Alternative Hypotheses are the following:
H0: p_bar2–p_bar1 = Constant = 0
H1: p_bar2–p_bar1 ≠ Constant, which equals 0
Therefore:
H1: p_bar2–p_bar1 ≠ 0
323
(The “not equal” sign indicates a two-tailed hypothesis test)
Alpha (α) = 1 – Level of certainty = 1 – 0.95 = 0.05
ppooled
ppooled = (X1 + X2)/(n1 + n2) = (90 + 59)/(200 + 100) = 0.50

qpooled = 1 - ppooled = 1 – 0.50 = 0.50
Standard Error
SEDiff = SQRT[ 0.50 * 0.50 * (1/200 + 1/100) ] = 0.06
Critical Values
Critical Values = Mean ± NORM.S.INV(1-α/2) * SEDiff
Critical Values = 0 ± NORM.S.INV(1- 0.05/2 ) * 0.06
Critical Values = 0 ± NORM.S.INV(0.975) * 0.06
It should be noted that the Mean in the Critical Values formula refers to is the p_bar2–p_bar1
hypothesized by the Null Hypothesis to be the curve’s mean. This is hypothesized to be 0.
Region of Rejection
This means that the boundaries of the Region of Rejection in the right and left tails are located at -0.12
and +0.12.
z Value
z Value (the Test Statistic) = (p_bar2–p_bar1 - Constant) / SEDiff

z Value (the Test Statistic) = (0.59 – 0.45 – 0)/0.06
324
z Value (the Test Statistic) = 2.29
This means that the observed p_bar2–p_bar1 is 2.29 Standard Errors from the mean.
Critical z Values
Critical z Valuesα=0.05,two-tailed,right tail = ±NORM.S.INV(1-α/2)
Critical z Valuesα=0.05,two-tailed,right tail = ±NORM.S.INV(0.975) = 1.96
This means that the boundaries of the Region of Rejection in the right and left tails are located at -1.96
and +1.96 Standard Errors from the mean.
p Value
p Value = 0.0111
This means that curve area beyond (to the right of) the observed p_bar 2–p_bar1 (0.14) equals 1.11
percent of the total area under the normal curve.

1) Compare p_ba2-p_bar1 with Critical Value

If the observed value of p_bar2–p_bar1 (0.14) falls into the Region of Acceptance (the blue region under
the curve), the Null Hypothesis is not rejected. If the observed value of p_bar2–p_bar1 falls into the
Regions of Rejection (either of the two yellow outer regions), the Null Hypothesis is rejected.
The observed p_bar2–p_bar1 (0.14) is farther to the curve’s mean (0) than the Critical Value on the right
side (+0.118) and falls in the yellow Region of Rejection. We therefore reject the Null Hypothesis. We can
state with 95 percent certainty that there is a real difference between the overall proportions of male and
female employees who prefer the proposed plan based upon the small samples taken each of the two
populations.

The Test Statistic for this Hypothesis Test is called the z Value. The z Value is the number of Standard
Errors that the observed p_bar2–p_bar1 is from the mean of zero. The Critical z Value is the number of
Standard Errors that the Critical z Value is from the mean.
The z Value (2.29) is farther to the curve’s standardized mean of zero than the Critical z Value (1.96) so
the Null Hypothesis is rejected.
3) Compare the p Value to Alpha.
The p Value is the percent of the curve that is beyond the observed p_bar2–p_bar1 (0.14). If the p Value is
smaller than Alpha/2 (since this is a two-tailed test), the Null Hypothesis is rejected. If the p Value is larger
than Alpha/2, the Null Hypothesis is not rejected.
The p Value (0.0111) is smaller than Alpha/2 (0.025) and we therefore reject the Null Hypothesis.
The following Excel-generated graph shows that the red p Value (the curve area beyond p_bar2–p_bar1)
is smaller than the yellow Alpha/2 Region of rejection in the outer right tail.
325
326
3) Two-Sample, Unpooled Hypothesis Test of
Proportion
Overview
This hypothesis test analyzes two independent samples to determine if the populations from which the
samples were taken have equal proportions. This test is often used to determine whether two sample
proportions are likely different by some specific proportion.
The test is called an unpooled test because the Null Hypothesis states that the two sample proportions
are not the same. The formula for Standard Error uses a unpooled proportion that does not combine the
proportions of both samples into a single, pooled proportion as a pooled test does. This formula is shown
in this following set of formulas.
p_bar1 = X1/n1
q_bar1 = 1 - p_bar1
p_bar2 = X2/n2
q_bar2 = 1 - p_bar2
The z Value for this test is the Test Statistic and is calculated as follows:
H0: p_bar2–p_bar1 = Constant = a non-zero proportion
1) The observed p_bar2 - p_bar1 is beyond the Critical Value.
2) The z Value (the Test Statistic) is farther from zero the Critical z Value.
327
Example of a Two-Sample, One-Tailed Unpooled Hyp. Test of
Proportion in Excel
Following is result of a two-independent-sample, unpooled Hypothesis Test of Proportion that is one-
tailed:
It is believed that Production Line B produces 5 percent more defects than Production Line A.
Both production lines manufacture the same products and have the same types of machines that are all
in similar condition. Both production lines operate approximately the same number of hours. The only
difference between the production lines is the experience of the crews. The crews that operate Production
Line A have more experience the crews on Production Line B.
Completed units from both production lines were sampled and evaluated over the same period of time as
follows.
12 out of 200 randomly sampled units produced on Production Line A were nonconforming. The
proportion of sample units from Production Line A that were nonconforming was 0.06 (6 percent).
39 out of 300 randomly sampled units produced on Production Line B were nonconforming. The
proportion of sample units from Production Line B that were nonconforming was 0.13 (13 percent).
Determine with 95 percent certainty whether production Line B’s overall proportion nonconforming
exceeds that of Production Line A by more than 5 percent. In other words, determine whether difference
between Production Line B’s overall percent defective and Production Line A’s overall percent defective is
greater than 5 percent.
Note that this will be a unpooled z Test because the proportions of the two populations are assumed to be
different. The Null Hypothesis will state that the difference between the proportion of defectives of the two
populations from which the samples are taken is equal to 5 percent. The Alternative Hypothesis states
that this difference is greater than 5 percent.
Sample results for the two samples cannot be pooled if they are known to be different, as stated by the
Null Hypothesis. Sample results can only be pooled if the Null Hypothesis states that the proportions of
the two samples are the same.
Two-independent-sample Hypothesis Tests of Proportion remain slightly more intuitive and allow for
consistent use of the variable name p_bar2–p_bar1 if the larger sample proportion is always designated
as p_bar2.and the smaller sample proportion is designated as p_bar1.

First sample – Production Line A (the experienced crews)
X1 = number of positive outcomes (defects) in n1 trials = 12
n1 = number of trials (random units inspected) = 200
p_bar1 = X1/n1 = 12/200 = 0.06
q_bar1 = 1 - p_bar1 = 0.94
Second sample – Production Line B (the inexperienced crews)
X2 = number of positive outcomes (defects) in n2 trials = 39
n2 = number of trials (random units inspected) = 300
p_bar2 = X2/n2 = 39/300 = 0.13
q_bar2 = 1 - p_bar2 = 0.87
p_bar2 – p_bar1 = 0.13 – 0.06 = 0.07
328
Alpha (α) = 1 – Level of certainty = 1 – 0.95 = 0.05

The Null and Alternative Hypotheses are the following:
H0: p_bar2–p_bar1 = Constant = non-zero proportion = 0.05
H1: p_bar2–p_bar1 > Constant, which equals 0.05
Therefore:
H1: p_bar2–p_bar1 > 0.05
The “greater than” sign in the Alternative Hypothesis indicates that this is a one-tailed hypothesis test in
the right tail.
Standard Error
SEDiff = SQRT[ (p_bar1*q_bar1/n1) + (p_bar2*q_bar2/n2) ]

SEDiff = SQRT[ (0.06*0.94/200) + (0.13*0.87/300) ]
SEDiff =0.026
Critical Values
Critical Value = Mean + NORM.S.INV(1-α) * SEDiff
Critical Value = 0.05 + NORM.S.INV(1- 0.05 ) * 0.026
Critical Value = 0.05 + NORM.S.INV(0.95) * 0.026
Critical Value = 0.05 + 0.042
Critical Value = 0.092
It should be noted that the Mean in the Critical Value formula refers to is the p_bar2–p_bar1 hypothesized
by the Null Hypothesis to be the curve’s mean.
Region of Rejection
This means that the Region of Rejection for this one-tailed test includes therefore everything under the
curve that is to the right of 0.092. If alpha equals 0.05, then 5 percent of the total curve area lies beyond
0.092.
329
z Value
z Value (the Test Statistic) = (p_bar2–p_bar1 - Constant) / SEDiff

z Value (the Test Statistic) = (0.13 – 0.06 – 0.05)/0.026
z Value (the Test Statistic) = 0.78
This means that the observed p_bar2–p_bar1 (0.07) is 0.78
Standard Errors from the curve’s mean (0.05).
Critical z Value
Critical z Valueα=0.05,two-tailed = NORM.S.INV(1-α)
Critical z Valuesα=0.05,two-tailed = NORM.S.INV(0.95) = 1.65
This means that the boundary of the Region of Rejection is 1.65 Standard Errors to the right of the
curve’s mean.
p Value
p Value = 0.2180
This means that curve area beyond (to the right of) the observed p_bar 2–p_bar1 (0.07) equals 21.80
percent of the total area under the normal curve.

1) Compare p_bar with Critical Value

If the observed value of p_bar2–p_bar1 (0.07) falls into the Region of Acceptance (the blue region under
the curve), the Null Hypothesis is not rejected. If the observed value of p_bar2–p_bar1 falls into the
Regions of Rejection (either of the two yellow outer regions), the Null Hypothesis is rejected.
The observed p_bar2–p_bar1 (0.07) is closer to the curve’s mean (0.05) than the Critical Value on the
right side (0.092) and falls in the blue Region of Acceptance. We therefore reject the Null Hypothesis. We
cannot state with 95 percent certainty that the difference between the proportions defective of Production
Line A and Production Line B is greater 0.05.

The Test Statistic for this Hypothesis Test is called the z Value. The z Value is the number of Standard
Errors that the observed p_bar2–p_bar1 is from the mean of 0.05. The Critical z Value is the number of
Standard Errors that the Critical z Value is from the mean.
330
The z Value (0.67) is closer to the curve’s standardized mean of zero than the Critical z Value (1.65) so
the Null Hypothesis is not rejected.
3) Compare the p Value to Alpha

p Value = 0.2180
The p Value (0.2180) is much larger than Alpha (0.05) and we therefore cannot reject the Null
Hypothesis.
The following Excel-generated graph shows that the red p Value (the curve area beyond p_bar2–p_bar1)
is larger than the yellow Alpha Region of Rejection in the outer right tail.
As mentioned, another entire section is dedicated to explaining hypothesis tests of proportion in much
greater detail than the brief summary just provided in this section.
331
Confidence Intervals of a Population Proportion
Overview
Confidence intervals covered in this manual will either be Confidence Intervals of a Population Mean or
Confidence Intervals of a Population Proportion. A data point of a sample taken for a confidence
interval of a population mean can have a range of values. A data point of a sample taken for a confidence
interval of a population proportion is binary; it can take only one of two values.
Data observations in the sample taken for a confidence interval of a population proportion are required to
be distributed according to the binomial distribution. Data that are binomially distributed are independent
of each other, binary (can assume only one of two states), and all have the same probability of assuming
the positive state.
A basic example of a confidence interval of a population proportion would be to create a 95-percent
confidence interval of the overall proportion of defective units produced by one production line based
upon a random sample of completed units taken from that production line. A sampled unit is either
defective or it is not. The 95-percent confidence interval is range of values that has a 95-percent certainty
of containing the proportion defective (the defect rate) of all of the production from that production line
based on a random sample taken from the production line.
The data sample used to create a confidence interval of a population proportion must be distributed
according to the binomial distribution. The confidence interval is created by using the normal distribution
to approximate the binomial distribution. The normal approximation of the binomial distribution allows for
the convenient application of the widely-understood z-based confidence interval to be applied to
binomially-distributed data.
The binomial distribution can be approximated by the normal distribution under the following two
conditions:
1) p (the probability of a positive outcome on each trial) and q (q = 1 – p) are not too close to 0 or 1.
2) np > 5 and nq > 5
The Standard Error and half the width of a confidence interval of proportion are calculated as follows:
Margin of Error = Half Width of C.I. = z Valueα, 2-tailed * Standard Error

Margin of Error = Half Width of C.I. = NORM.S.INV(1 – α/2) * SQRT[ (p_bar * q_bar) / n]
332
Example of a Confidence Interval of a Population Proportion
in Excel
In this example a 95-percent confidence interval of a population proportion is created around a sample
proportion using the normal distribution to approximate the binomial distribution.
This example evaluates a group of shoppers who either prefer to pay by credit or by cash. A random
sample of 1,000 shoppers was taken. 70 percent of the sampled shoppers preferred to pay with a credit
card. The remaining 30 percent of the sampled shoppers preferred to pay with cash.
Determine the 95-percent confidence interval for the proportion of the general population that prefers to
pay with a credit card. In other words, determine the endpoints of the interval that is 95 percent certain to
contain the true proportion of the total shopping population that prefers to pay by credit card.

p_bar = sample proportion = 0.70
q_bar = 1 – p_bar = 1 – 0.70 = 0.30
p = population proportion = Unknown (This is what the confidence interval will contain.)
n = sample size = 1,000
α = Alpha = 1 – Level of Certainty = 1 – 0.95 = 0.05
SE = Standard Error = SQRT[ (p_bar * q_bar) / n]

SE = SQRT[ (0.70 * 0.30) / 1000] = 0.014491
As when creating all confidence intervals of proportion, we must satisfactorily answer these two questions
and then proceed to the two-step method of creating the confidence interval of proportion.

What Type of Confidence Interval Should Be Created?
The Two-Step Method For Creating Confidence Intervals of Mean Are the Following:
333
The Initial Two Questions That Need To Be Answered Before Creating a Confidence Interval of the Mean
or Proportion Are as Follows:

This is a confidence interval of a population proportion because sampled data observations are binary:
they can take only one of two possible values. A shopper sampled either prefers to pay with a credit card
or prefers to pay with cash.
The data sample is distributed according to the binomial distribution because each observation has only
two possible outcomes, the probability of a positive outcome is the same for all sampled data
observations, and each data observation is independent from all others.
Sampled data points used to create a confidence interval of a population mean can take multiple values
or values within a range. This is not the case here because sampled data observations can have only two
possible outcomes: a sampled shopper either prefers to pay with credit card or with cash.

A confidence interval of proportion is always created using the normal distribution. The binomial
distribution of binary sample data is closely approximated by the normal distribution in certain conditions.
The next step in this example will evaluate whether the correct conditions are in place that permit the
approximation of the binomial distribution by the normal distribution.
It should be noted that the sample size (n) equals 1,000. At that sample size, the t distribution is nearly
identical to the normal distribution. Using the t distribution to create this confidence interval would produce
exactly the same result as the normal distribution produces.
This confidence interval will be a confidence cnterval of a population proportion and will be
created using the normal distribution to approximate the binomial distribution of the sample data.

a) Normal distribution can be used to approximate the binomial distribution
The most important requirement of a confidence interval of a population proportion is the validity of
approximating the binomial distribution (that the sampled objects follow because they are binary) with the
normal distribution.
The binomial distribution can be approximated by the normal distribution sample size, n, is large enough
and p is not too close to 0 or 1. This can be summed up with the following rule:
The binomial distribution can be approximated by the normal distribution if np > 5 and nq >5. In this case,
the following are true:
n = 1,000
p = 0.70 (p is approximated by p_bar)
q = 0.30 (q is approximated by q_bar)
np = 700 and nq = 300
It is therefore valid to approximate the binomial distribution with the normal distribution.
The binomial distribution has the following parameters:
Mean = np
Variance = npq
334
Each unique normal distribution can be completely described by two parameters: its mean and its
standard deviation. As long as np > 5 and nq > 5, the following substitution can be made:
Normal (mean, standard deviation) approximates Binomial (n,p)
If np is substituted for the normal distribution’s mean and npq is substituted for the normal distribution’s
standard deviation as follows:
Normal (mean, standard deviation)
becomes
Normal (np, npq), which approximates Binomial (n,p)
This can be demonstrated with Excel using data from this problem.
n = 1000
n = the number of trials in one sample
p = 0.7 (p is approximated by p_bar)
p = the probability of obtaining a positive result in a single trial

q = 0.7 (q is approximated by q_bar)
q=1–p
np = 700
npq = 210
At arbitrary point X = 700

(X equals the number positive outcomes in n trials)
BINOM.DIST(X, n, p, FALSE) = BINOM.DIST(700, 1000, 0.7, FALSE) = 0.0275
The Excel formula to calculate the PDF (Probability Density Function) of the normal distribution at point X
is the following:
NORM.DIST(X, Mean, Stan. Dev, FALSE)
The binomial distribution can now be approximated by the normal distribution in Excel by the following
substitutions:
BINOM.DIST(X, n, p, FALSE) ≈ NORM.DIST(X, np, npq, FALSE)
NORM.DIST(X, np, npq, FALSE) = NORM.DIST(700,700,210,FALSE) = 0.0019
BINOM.DIST(X, n, p, FALSE) = BINOM.DIST(700, 1000, 0.7, FALSE) = 0.0275
The difference is less than 0.03 and is reasonable close. Note that this approximation only works for the
PDF (Probability Density Function) and not the CDF (Cumulative Distribution Function – Replacing
FALSE with TRUE in the above formulas would calculate the CDF instead of the PDF).
335
We now proceed to the two-step method for creating all confidence intervals of a population proportion.

Half the Width of the confidence interval is sometimes referred to the Margin of Error. The Margin of Error
will always be measured in the same type of units as the sample proportion is measured in, which is
percentage. Calculating the Half Width of the Confidence Interval using the t distribution would be done
as follows in Excel:
Margin of Error = Half Width of C.I. = z Valueα, 2-tailed * Standard Error
Margin of Error = Half Width of C.I. = NORM.S.INV(1 – α/2) * SQRT[ (p_bar * q_bar) / n]
Margin of Error = Half Width of C.I. = NORM.S.INV(0.975) * SQRT[ (0.7 * 0.3) / 1000]
Margin of Error = Half Width of C.I. = 1.95996 * 0.014491
Margin of Error = Half Width of C.I. = 0.0284, which equals 2.84 percent

Confidence Interval = Sample Proportion ± (Half Width of Confidence Interval)
Confidence Interval = p_bar ± 0.0284
Confidence Interval = 0.70 ± 0.0284
Confidence Interval = [ 0.6716, 0.7284 ], which equals 67.16 percent to 72.84 percent
We now have 95 percent certainty that the true proportion of all shoppers who prefer to pay with a credit
card is between 67.16 percent and 72.84 percent.
A graphical representation of this confidence interval is shown as follows:
336
337
This minimum sample size, n, can be derived by the following equation:
(Half-width of C.I.) = z Valueα, 2-tailed * SQRT[ (pest * qest) / n]
Estimates of population parameters p and q must be used in this equation because sample statistics
p_bar and q_bar are not available since a sample has not been taken.
(Half-width of C.I.) = z Valueα, 2-tailed * SQRT[ (pest * qest) / n]
(Half-width of C.I.) = NORM.S.INV(1 – α/2) * SQRT[ (pest * qest) / n]
2 2
(Half-width of C.I.) = NORM.S.INV (1 – α/2) * pest * qest / n
Further algebraic manipulation provides the following:
2 2
n = [NORM.S.INV (1 – α/2) * pest * qest] / (Half-width of C.I.)
or, equivalently
2 2
n = [NORM.S.INV (1 – α/2) * pest * qest] / (Margin of Error)
2 2
n = Roundup( ( NORM.S.INV (1 – α/2) * pest * qest ) / (Half-width of C.I.) ,0)
pest and qest are estimates of the actual population parameters p and q. The most conservative estimate of
the minimum sample size would use pest = 0.50.
If pest = 0.05, then qest = 1 – p = 0.50
The product pest * qest has its maximum value of 0.25 when pest = 0.50. This maximum value of pest * qest
produces the highest and therefore most conservative value of the minimum sample size, n.
If p is fairly close to 0.5, then pest should be set at 0.5. If p is estimated to be significantly different than
0.5, pest should be set to its estimated value.
Example 1 in Excel – Min Number of Voters to Limit Poll Margin of Error

Two candidates are running against each other in a national election. This election is considered fairly
even. What is the minimum number of voters who should be randomly surveyed to obtain a survey result
that has 95 percent certainty of being within 2 percent of the nationwide preference for either one of the
candidates?
pest should be set at 0.5 since the election is considered even.
pest = 0.5
qest = 1 – pest = 0.5
338
Half-width of the confidence interval = Margin of Error = 2 percent = 0.02
2 2
n = Roundup( ( NORM.S.INV (1 – α/2) * pest * qest ) / (Half-width of C.I.) ,0)
2 2
n = Roundup( ( NORM.S.INV (1 – 0.05/2) * 0.50 * 0.50 ) / (0.02) , 0 )
2 2
n = Roundup( ( NORM.S.INV ( 0.975) * 0.250 ) / (0.02) ), 0 )
n = Roundup(2400.912, 0 )
n = 2401
The preferences of at least 2,401 voters would have to be randomly surveyed to obtain a sample
proportion that has 95 percent certainty of being within 2 percent of the national voter preference for one
of the candidates.
Example 2 in Excel – Min Number of Production Samples to Limit Defect Rate Margin
A production line is estimated to have a defect rate of approximately 15 percent of all units produced on
the line. What would be the minimum number of completed production units that should be randomly
sampled for defects to obtain a sample proportion of defective units that has 95 percent certainty to being
within 1 percent of the real defect rate of all unites produced on that production line?
pest should be set more conservatively than its estimate. The more conservative that p is, the higher will
be the minimum sample size required. The most conservative setting for p est would 0.5. pest should be set
between its estimate of 0.15 and 0.5. A reasonable setting for pest would be 0.25.
pest = 0.25
qest = 1 – pest = 0.75
Half-width of the confidence interval = Margin of Error = 1 percent = 0.01
2 2
n = Roundup( ( NORM.S.INV (1 – α/2) * pest * qest ) / (Half-width of C.I.) ,0 )
2 2
n = Roundup( ( NORM.S.INV (1 – 0.05/2) * 0.25 * 0.75 ) / (0.01) ,0 )
2 2
n = Roundup( ( NORM.S.INV ( 0.975) * 0.1875 ) / (0.01) ,0 )
n = Roundup(7202.735, 0 )
n = 7203
At least 7,203 completed units should be randomly sampled from the production line to obtain a sample
proportion defective that has 95 percent certainty of being within 1 percent of the actual proportion
defective of all units produced on that production line.
If pest were set at 0.15 instead of the more conservative 0.25, the minimum sample size would have been
reduced to 4,898.
339
Family of Closely-Related Distributions
Binomial Distribution – Calculates the probability of k successes in n Bernoulli trials with replacement.
Sampling with replacement means that each sample is placed back into the population before another
sample is taken.
Hypergeometric Distribution - Calculates the probability of k successes in n Bernoulli trials without
replacement. Sampling without replacement means that each sample taken is not placed back into the
population. This is not a binomial experiment because a binomial experiment requires that the probability
of success be constant on every trial.
Multinomial distribution – This is a generalization of the binomial distribution. Each multinomial trial has
two or more possible outcomes. A binomial trial is a multinomial trial with only two possible outcomes.
The multinomial distribution calculates the probability of a specific combination of outcomes occurring in n
multinomial trials. Sampling is done with replacement during a multinomial experiment.
Negative Binomial Distribution – Calculates probability of a specific number of successes occurring
before a specified number of failures. Each trial is a Bernoulli trial. Sampling is done with replacement
during a negative binomial experiment. Like the binomial distribution, the probability of success on every
trial is the same and the trials are independent, i.e., the outcome of one trial does not affect the outcome
of any other trial.
Geometric Distribution – Calculates the probability that a success will occur for the first time on the nth
Bernoulli trial. This is a special case of the negative binomial distribution in which the number of
successes that will occur before a specified number of failures is equal to one. Sampling is done with
replacement during a geometric experiment. Like the binomial distribution, the probability of success on
every trial is the same and the trials are independent, i.e., the outcome of one trial does not affect the
outcome of any other trial.
Following will be presented more information about each of the distributions just mentioned. Information
about each distribution will include its PDF, CDF, Excel formulas, and a brief problem solved with the PDF
of that distribution in Excel.
340
Binomial Distribution
Calculates the probability of k successes in n Bernoulli trials with replacement. Sampling with
replacement means that each sample is placed back into the population before another sample is taken.
The binomial distribution’s PDF is given by the following:
for k = 0, 1, 2, …, n,
where
is the binomial coefficient and is sometimes referred to as “n choose k.”

f(k;n,p) = Pr(X = k) is the probability that X, the number of successes, equals k for n independent,
successive Bernoulli trials each having the probability p of success.
The Excel formula to calculate the binomial distribution’s PDF is the following:
FALSE indicates that this Excel formula is calculating the binomial distribution’s PDF and not the CDF for
this k, n, and p. “False” answers the question of the calculation is cumulative (which is the case if
calculating the CDF – Cumulative Distribution Function) on not cumulative (which is the case if calculating
the PDF – Probability Density Function).
Excel 2010 and later also use the formula BINOM.DIST which is equivalent to BINOMDIST used by
earlier versions of Excel.
The binomial distribution’s CDF is as follows:
F(k;n,p) = Pr(X ≤ k). This is the probability that X, the number of successes in n Bernoulli trials each
having the probability p of a successful outcome, equals up to k.
The binomial distribution’s CDF is given by the following:
341
is the “floor” under k, i.e., the greatest integer less than or equal to k.
The Excel formula to calculate the binomial distribution’s CDF is the following:
F(k;n,p) = Pr(X ≤ k) = BINOM.DIST(k, n, p , TRUE)
A basic problem solved by the binomial distribution’s PDF in Excel is as follows:
What is the probability of getting exactly 4 “fives” if a single die is rolled 7 times?
The probability of getting a five on any roll of a die is one sixth.
p = 1/6 = 0.1667
n = 7 trials (rolls of the die)
k=4
This requires the PDF to solve for Pr(X = k)
f(4;7,0.1667) = BINOM.DIST(4,7,0.1667,FALSE) = 0.0156
There is a 1.56 percent of getting exactly 4 “fives” if a single die is rolled 7 times.
342
Hypergeometric Distribution
Calculates the probability of k successes in n Bernoulli trials without replacement. Sampling without
replacement means that each sample taken is not placed back into the population. This is not a binomial
experiment because a binomial experiment requires that the probability of success be constant on every
trial.
The hypergeometric’s distribution’s PDF is given by the following:
k = targeted number of successes in n trials
K = total number of successes in the population
n = number of trials
N = population size
where

The Excel formula to calculate the hypergeometric distribution’s PDF is the following:
Pr(X = k) = HYPGEOM.DIST(k, n, K, N, FALSE)
Excel 2010 and later also use the formula HYPGEOM.DIST which is similar but not exactly equivalent to
HYPGEOMDIST used by earlier versions of Excel. The earlier version of the formula, HYPGEOMDIST(),
calculates only the PDF.
HYPGEOMDIST(k, n, K, N) = HYPGEOM.DIST(k, n, K, N, FALSE)
The hypergeometric distribution’s CDF is as follows:
Until Excel 2010 there is no Excel formula to calculate the hypergeometric distribution’s CDF. Prior to
2010 the hypergeometric distribution’s CDF has to be calculated in Excel by summing successive PDF
calculations as follows:
343
The hypergeometric distribution is a discrete distribution whose CDF is the sum of sum of the specific
point values that make up the PDF of the hypergeometric distribution because it is a discrete distribution.
In Excel 2010 and beyond, the CDF of hypergeometric distribution can be calculated as follows:
Pr(X≤k) = HYPGEOM.DIST(k, n, K, N, TRUE)
A basic problem solved by the hypergeometric distribution’s PDF in Excel is as follows:
What is the probability of getting two kings in ten successive draws from a deck of 52 cards. There are
four kings in a deck of cards. Each card drawn is not replaced back into the deck?
k=2
K=4
n = 10
N = 52
Pr(X = k) = HYPGEOM.DIST(k, n, K, N , FALSE)
Pr(X = 2) = HYPGEOM.DIST(2, 10, 4, 52 , FALSE) = 0.1431
There is a 14.31 percent probability of getting two kings in 10 successive draws from a 52-card that initial
contains four kings if each card is discarded after being drawn.
344
Multinomial Distribution
This is a generalization of the binomial distribution. Each multinomial trial has two or more possible
outcomes. A binomial trial is a multinomial trial with only two possible outcomes. The multinomial
distribution calculates the probability of a specific combination of outcomes occurring in n multinomial
trials. Sampling is done with replacement during a multinomial experiment.
The multinomial’s distribution’s PDF is given by the following:
n = number of trials
m = number of possible outcomes in each trial
i as i goes from 1 to m = ith possible outcome of the 10 total possibilities. For example, if the three colors
st
of marbles that can be drawn are red, blue, and green, then red is the 1 possible outcome (ired = 1), blue
is the second possible outcome (iblue = 2), etc.
ki = the targeted number of successes for the ith possible outcome. For example, if the targeted number
of blue marbles is 2 then k2 = 2
pi = the probability of each trial producing a success for the ith possible outcome. For example, if 5 out of
20 total marbles are blue, then p3 = 5/20 = 0.25. These probabilities remain the same because each
sample is replaced before the next trial.
There is no Excel formula to calculate multinomial probabilities.

This formula probably appears more onerous that it actually is. The following example will hopefully
simplify it.
A bag contains 21 marbles whose colors are distributed as follows:
6 marbles are red
5 marbles are blue
10 marbles are green
9 marbles are individually drawn, evaluated, and then placed back into the bag before the next marble is
drawn. Calculate the probability that the 9 trials will produce exactly the following results:
4 red marbles
2 blue marbles
3 green marbles
n = 9 trials
m = 3 possible outcomes (red, blue, green)
p1 = 6/21 = probability of getting outcome 1 (red) in each trial
p2 = 5/21 = probability of getting outcome 2 (blue) in each trial
p3 = 10/21 = probability of getting outcome 3 (green) in each trial
k1 = 4 = the targeted number of successes of outcome 1 (red) in n trials
k2 = 2 = the targeted number of successes of outcome 2 (blue) in n trials
345
k3 = 3 = the targeted number of successes of outcome 3 (green) in n trials
4 2 3
Pr(X1=4, X2=2, X3=3)= 9! / (4!*2!*3!) * (6/21) * (5/21) * (10/21)
Pr(X1=4, X2=2, X3=3)= 0.0514
There is a 5.14 percent probability of obtaining 4 red marbles, 2 blue marbles, and 3 green marbles in 9
successive random samples taken with replacement from a set of 6 red marbles, 5 blue marbles, and 10
green marbles.
The CDF of the multinomial distribution can be calculated but its formula quickly becomes unmanageably
large as the number of outcomes and targeted number of each outcome grows.
346
Negative Binomial Distribution
Calculates probability that a specific number of failures will occur before a specified number of successes
will. Each trial is a Bernoulli trial with the same probability of a successful outcome. Sampling is done with
replacement during a negative binomial experiment. Like the binomial distribution, the probability of
success on every trial is the same and the trials are independent, i.e., the outcome of one trial does not
affect the outcome of any other trial.
The negative binomial distribution calculates the probability of a specific number of trials to obtain a fixed
number of successes. The number of trials is varied and the number of successes is fixed.
The binomial distribution calculates the probability of a specific number of successes in a fixed number of
trials. The number of successes is varied and the number of trials is fixed.
The negative binomial distribution’s PDF is given by the following:
x = targeted number of failures occurring before the rth success
r = total number of successes that occur. The rth success occurs immediately following the xth failure.
p = the probability of success on each trial
where

The Excel formula to calculate the hypergeometric distribution’s PDF is the following:
Prior to Excel 2010: Pr(X = x) = NEGBINOMDIST(x, r, p)
or
Excel 2010 and beyond: Pr(X = x) = NEGBINOM.DIST(x, r, p, FALSE)
NEGBINOM.DIST can calculate both the PDF and the CDF. The earlier version of this formula prior to
Excel 2010, NEGBINOMDIST, will only calculate the PDF.
The negative binomial distribution’s CDF is as follows:
347
Only Excel 2010 has a formula to calculate the negative binomial distribution’s CDF. The CDF can be
calculated in Excel 2010 or later as follows:
Pr(X ≤ k) = NEGBINOM.DIST(x, r, p, TRUE)
Calculating the negative binomial distribution’s CDF in versions of Excel earlier than 2010 would require
summing the PDF as i goes from 0 to x as follows:
A basic problem solved by the negative binomial distribution’s PDF in Excel is as follows:
th
A fair coin is continuously flipped. What is the probability that the 4 head will occur immediately following
th
the 10 tail? A “head” will be counted as a success.
x = 10 = targeted number of failures (tails) occurring before the rth success (heads)
r = 4 = total number of successes (heads) that occur. The rth success occurs immediately following the
xth failure (tails).
p = 0.5 = the probability of success on each trial
Pr(X = x) = NEGBINOM.DIST(x, r, p, FALSE)
Pr(X = 4) = NEGBINOM.DIST(10, 4, 0.5, FALSE) = 0.0174
th th
There is a 1.74 percent probability the coin flip following the 10 tail will produce the 4 head.
348
Geometric Distribution
Calculates the probability that a success will occur for the first time on the nth Bernoulli trial. This is a
special case of the negative binomial distribution in which the number of successes that will occur before
a specified number of failures is equal to one. Sampling is done with replacement during a geometric
experiment. Like the binomial distribution, the probability of success on every trial is the same and the
trials are independent, i.e., the outcome of one trial does not affect the outcome of any other trial.
As mentioned, the geometric distribution is equivalent to the negative binomial distribution with r (the
number of successes) set to one.
A basic problem solved by the geometric distribution’s PDF in Excel is as follows:
A fair coin is continuously flipped. What is the probability that the 1st head will occur immediately
following the 3rd tail? A “head” will be counted as a success.
x = 3 = targeted number of failures (tails) occurring before the rth success (heads)
r = 1 = total number of successes (heads) that occur. The rth success occurs immediately following the
xth failure (tails).
p = 0.5 = the probability of success on each trial
Pr(X = x) = NEGBINOM.DIST(x, r, p, FALSE)
Pr(X = 1) = NEGBINOM.DIST(3, 1, 0.5, FALSE) = 0.0625
rd st
There is a 6.25 percent probability the coin flip following the 3 tail will produce the 1 head.
349
Check Out the Latest Book in the Excel Master Series!
Click Here To Download This 200+ Page Excel Solver Optimization Manual Right Now for $19.95
http://37.solvermark.pay.clickbank.net/
For anyone who wants to be performing optimization at a high level with the Excel Solver quickly, Step-
By-Step Optimization With Excel Solver is the e-manual for you. This is a hands-on, step-by-step,
complete guidebook for both beginner and advanced Excel Solver users. This book is perfect for the
many students who are now required to be proficient in optimization in so many majors as well as industry
professionals who have an immediate need to become up-to-speed with advanced optimization in a short
time frame.
Step-By-Step Optimization With Excel Solver is 200+ pages .pdf e-manual of simple yet thorough
explanations on how to use the Excel Solver to solve today’s most widely known optimization problems.
Loaded with screen shots that are coupled with easy-to-follow instructions, this .pdf e-manual will simplify
many difficult optimization problems and make you a master of the Excel Solver almost immediately.
The author of Step-By-Step Optimization With Excel Solver, Mark Harmon, was the Internet marketing
manager for several years for the company that created the Excel Solver and currently develops it for
Microsoft Excel today. He shares his deep knowledge of and experience with optimization using the Excel
Solver in this book.
Here are just some of the Solver optimization problems that are solved completely with simple-to-
understand instructions and screen shots in this book
● The famous “Traveling Salesman” problem using Solver’s Alldifferent constraint and the Solver’s
Evolutionary method to find the shortest path to reach all customers. This also provides an advanced use
of the Excel INDEX function.
350
● The well-known “Knapsack Problem” which shows how optimize the use of limited space while
satisfying numerous other criteria.
● How to perform nonlinear regression and curve-fitting on the Solver using the Solver’s GRG Nonlinear
solving method
● How to solve the “Cutting Stock Problem” faced by many manufacturing companies who are trying to
determine the optimal way to cut sheets of material to minimize waste while satisfying customer orders.
● Portfolio optimization to maximize return or minimize risk.
● Venture capital investment selection using the Solver’s Binary constraint to maximize Net Present Value
of selected cash flows at year 0. Clever use of the If-Then-Else statements makes this a simple problem.
● How use Solver to minimize the total cost of purchasing and shipping goods from multiple suppliers to
multiple locations.
● How to optimize the selection of different production machine to minimize cost while fulfilling an order.
● How to optimally allocate a marketing budget to generate the greatest reach and frequency or number
of inbound leads at the lowest cost.
Step-By-Step Optimization With Excel Solver has complete instructions and numerous tips on every
aspect of operating the Excel Solver. You’ll fully understand the reports and know exactly how to tweek all
of the Solver’s settings for total custom use. The book also provides lots of inside advice and guidance on
setting up the model in Excel so that it will be as simple and intuitive as possible to work with All of the
optimization problems in this book are solved step-by-step using a 6-step process that works every time.
In addition to detailed screen shots and easy-to-follow explanations on how to solve every optimization
problem in this e-manual, a link is provided to download an Excel workbook that has all problems
completed exactly as they are in this e-manual.
Step-By-Step Optimization With Excel Solver is exactly the e-manual you need if you want to be
optimizing at an advanced level with the Excel Solver quickly.
Reader Testimonials
"Step-By-Step Optimization With Excel Solver is the "Missing Manual" for the Excel Solver. It is pretty
difficult to find good documentation anywhere on solving optimization problems with the Excel Solver.
This book came through like a champ!
Optimization with the Solver is definitely not intuitive, but this book is. I found it very easy to work through
every single one of the examples. The screen shots are clear and the steps are presented logically. The
downloadable Excel spreadsheet with all example completed was quite helpful as well.
Once again, it really amazing how little understandable documentation there is on doing real-life
optimization problems with Solver.
For example, just try to find anything anywhere about the well-known Traveling Salesman Problem (a
salesman needs to find the shortest route to visit all customers once).
Step-By-Step Optimization With Excel Solver is the "Missing Manual" for the Excel Solver. It is pretty
difficult to find good documentation anywhere on solving optimization problems with the Excel Solver.
This book came through like a champ!
Optimization with the Solver is definitely not intuitive, but this book is. I found it very easy to work through
every single one of the examples. The screen shots are clear and the steps are presented logically. The
downloadable Excel spreadsheet with all example completed was quite helpful as well.
351
Once again, it really amazing how little understandable documentation there is on doing real-life
optimization problems with Solver.
For example, just try to find anything anywhere about the well-known Traveling Salesman Problem (a
salesman needs to find the shortest route to visit all customers once)
It is a tricky problem for sure, but this book showed a quick and easy way to get it done. I'm not sure I
would have ever figured that problem out, or some the other problems in the book, without this manual.
I can say that this is the book for anyone who wants or needs to get up to speed on an advanced level
quickly with the Excel Solver. It appears that every single aspect of using the Solver seems to be covered
thoroughly and yet simply. The author presents a lot of tricks in how to set the correct Solver settings to
get it to do exactly what you want.
The book flows logically. It's an easy read. Step-By-Step Optimization With Excel Solver got me up to
speed on the Solver quickly and without to much mental strain at all. I can definitely recommend this
book."
Pam Copus
Sonic Media Inc
“As Graduate student of the Graduate Program in International Studies (GPIS) at Old Dominium
University, I'm required to have a thorough knowledge of Excel in order to use it as a tool for interpreting
data, conducting research and analysis.
I've always found the Excel Solver to be one of the more difficult Excel tools to totally master. Not any
more. This book was so clearly written that I was able to do almost every one of advanced optimization
examples in the book as soon as I read through it once.
I can tell that the author really made an effort to make this manual as intuitive as possible. The screen
shots were totally clear and logically presented.
Some of the examples that were very advanced, such as the venture capital investment example, had
screen shot after screen shot to ensure clarity of the difficult Excel spreadsheet and Solver dialogue
boxes.
It definitely was "Step-By-Step" just like the title says. I must say that I did have to cheat a little bit and
look at the Excel spreadsheet with all of the book's example that is downloadable from the book. The
spreadsheet was also a great help.
Step-By-Step Optimization With Excel Solver is not only totally easy to understand and follow, but it is
also very complete. I feel like I'm a master of the Solver. I have purchased a couple of other books in the
Excel MaSter Series (the Excel Statistical Master and the Advanced Regression in Excel book) and they
have all been excellent.
I am lucky to have come across this book because the graduate program that I am in has a number of
optimization assignments using the Solver. Thanks Mark for such an easy-to-follow and complete book
on using the Solver. It really saved me a lot of time in figuring this stuff out."
Federico Catapano
Graduate Student
International Studies Major
Old Dominion University
Norfolk, Virginia
"I'm finished with school (Financial Economics major) and currently work for a fortune 400 company as a
business analyst. I find that the statistics and optimization manuals are indispensable reference tools
throughout the day.
352
I keep both eManuals loaded on my ipad at all times just in case I have to recall a concept I don't use all
the time. Its easier to recall the concepts from the eManuals rather then trying to sift through the
convoluted banter in a text book, and for that I applaud the author!
In a business world where I need on demand answers now this optimization eManual is the perfect tool.
I just recently used the bond investment optimization problem to build a model in excel and help my VP
understand that a certain process we're doing wasn't maximizing our resources.
That's the great thing about this manual, you can use any practice problem (with a little outside thinking)
to mold it into your own real life problem and come up with answers that matter in the work place.!"
Sean Ralston
Sr. Financial Analyst
Enogex LLC
Oklahoma City, Oklahoma
"Excel Solver is a tool that most folks never use. I was one of those people. I was working on a project,
and was told that solver might be helpful. I did some research online, and was more confused than ever. I
started looking for a book that might help me. I got this book, and was not sure what to expect.
It surpassed my expectations! The book explains the concepts behind the solver, the best way to set up
the "problem", and how to use the tool effectively. It also gives many examples including the files. The
files are stored online, and you can download them so you can see everything in excel.
The author does a fantastic job on this book. While I'm not a solver "expert", I am definitely much smarter
about it than I was before. Trust me, if you need to understand the solver tool, this book will get you
there."
Scott Kinsey
Missouri
“The author, Mark, has a writing style that is easy to follow, simple, understandable, with clear examples
that are easy to follow. This book is no exception.
Mark explains how solver works, the different types of solutions that can be obtained and when to use
one or another, explains the content and meaning of the reports available. Then he presents several
examples, goes about defining each problem, setting it up in excel and in solver and interpreting the
solution.
It is a really good book that teaches you how to apply solver (linear programming) to a problem.”
Luis R. Heimpel
El Paso, Texas
353
Click Here To Download This 200+ Page Excel Solver Optimization Manual Right Now for $19.95
http://37.solvermark.pay.clickbank.net/
354
Meet the Author
Mark Harmon is a university statistics instructor and statistical/optimization consultant. He was the
Internet marketing manager for several years for the company that created the Excel Solver and currently
develops that add-in for Excel. He has made contributions to the development of Excel over a long period
of time dating all the way back to 1992 when he was one of the beta users of Excel 4 creating the sales
force deployment plan for the introduction of the anti-depressant drug Paxel into the North American
market.
Mark Harmon is a natural teacher. As an adjunct professor, he spent five years teaching more than thirty
semester-long courses in marketing and finance at the Anglo-American College in Prague, Czech
Republic and the International University in Vienna, Austria. During that five-year time period, he also
worked as an independent marketing consultant in the Czech Republic and performed long-term
assignments for more than one hundred clients. His years of teaching and consulting have honed his
ability to present difficult subject matter in an easy-to-understand way.
This manual got its start when Mark Harmon began conducting statistical analysis to increase the
effectiveness of various types of Internet marketing that he was performing during the first decade of the
2000s. Mark initially formulated the practical, statistical guidelines for his own use but eventually realized
that others would also greatly benefit by this step-by-step collection of statistical instructions that really did
not seem to be available elsewhere. Over the course of a number of years and several editions, this
instruction manual blossomed into the Excel Master Series of graduate-level, step-by-step, complete,
practical, and clear set of guidebooks that it is today.
Mark Harmon received a degree in electrical engineering from Villanova University and MBA in marketing
from the Wharton School.
Mark is an avid fan of the beach life and can nearly always be found by a warm and sunny beach.
355
Advanced Statistical & Solver Optimization Project Consulting
Availability
If you have an advanced statistical or optimization problem or project that you would like finished quickly,
explained clearly (just like the book you are reading), and performed completely in Excel so that you can
interact with the data and results, I might be able to provide just what you need.
What sets me apart from other statistical consultants is my ability to perform advanced statistical work in
Excel and provide crystal-clear, simple, and easy-to-follow explanations of the statistics and Excel
procedures. I make every effort to design all Excel work so that you can make changes to any of the data
or input variables and conditions and then immediately observe the final result of your changes. Excel
provides the only platform that makes all of this possible for you without extensive training and/or
expensive statistical software package purchase. You already have and know Excel.
You will have total ownership of the result because you will be able to open up and manipulate the
finalized data in Excel on your computer and then fully understand how the statistical analysis was
performed in Excel and what the final result means for you.
Statistical consulting services that I can provide for you using Excel are as follows:
1) Selection of the correct statistical techniques to solve the problem.
2) Configuring the Excel workbook for data input, data processing, and clearly-understandable output. I
try to confine all Excel work to formulas so that changes to input data are immediately reflected in output
figures, charts, and graphs in the Excel workbook.
3) Predictive analytics such as binary logistic regression, nonlinear and linear regression, and dummy-
variable regression combined with techniques such Conjoint Analysis.
4) Data mining and manipulation with Excel’s tools to locate patterns and develop segments.
5) Sample and population comparison with nearly all types of ANOVA, ANCOVA, z-Tests, and t-Tests
along with ALL nonparametric alternatives when the data does not meet parametric requirements of
normality and all alternative tests when sample group variances are not similar such as Welch’s ANOVA
and the Brown-Forsythe F test.
6) Major nonparametric tests including the Mann-Whitney U test, the Wilcoxon Signed-Rank test, the
Kruskal-Wallis test, the Sign test, the Scheirer-Ray-Hare test (a two-factor ANOVA substitute), and many
others.
7) Major normality tests including the Shapiro-Wilk test, the Kolmogorov-Smirnov test, the Anderson-
Darling test, and the Chi-Square Goodness-of-Fit test.
8) Major variance comparison tests such as Levene’s test, the Brown-Forsyth test, and the Chi-Square
Population Variance test.
9) Confidence intervals of a population mean or proportion along with calculation of the minimum sample
needed to limit the width of the margin of error to a specific size.
10) Factor independence testing using as the Chi-Square Independence test.
11) Post hoc, Power Analysis, and Effect Size of most tests.
12) Degree of relationship between variables using the Pearson and Spearman correlations along with
calculation of the r Critical and the p value that determine whether the calculated correlation is significant
or not.
13) Detailed analysis of binomial data (data that can assume only two outcomes) and Poisson data (data
that are counted and have an average frequency of occurrence across a specific interval or time or
space).
356
13) Any optimization problem using the Excel Solver. Optimization means the calculating the correct
allocation of resources that will minimize or maximize an output variable. I wrote one of the most widely-
sold books on this topic.
14) Statistical process control and analysis problems.
If you are faced with any of the above statistical issues and would like easy-to-follow results in Excel
quickly, feel free to contact me at the following email address:
mark@ExcelMasterSeries.com
My fees are negotiable but are typically around $75 per hour.
357

Graduate Statistics in Excel Manual 1 S

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Graduate Statistics in Excel Manual 1 S

Hochgeladen von

Copyright:

Verfügbare Formate

Manual 1 of 4

Practical and Clear Graduate Statistics

The Excel Statistical Master

Histograms and Basic Graphing in Excel

Scatterplot Graph in Excel ..................................................................... 44

Bar Chart in Excel ................................................................................... 58

Combinations and Permutations in Excel............................................. 74

Normal Distribution in Excel .................................................................. 82

The Central Limit Theory in Excel ....................................................... 109

Hypothesis Tests Using the Normal Distribution in Excel ................. 122

2) Two-Independent-Sample, Unpooled z-Test in Excel .................... 145

3) Paired (Two-Sample Dependent) z-Test in Excel ........................... 157

z-Based Confidence Intervals of a Population Mean in Excel ........... 168

t-Distribution in Excel ........................................................................... 175

1) One-Sample t-Test in Excel .............................................................. 206

2) Two-Independent-Sample, Pooled t-Test in Excel ......................... 220

3) Two-Independent-Sample, Unpooled t-Test in Excel ..................... 242

4) Paired (Two-Sample Dependent) t-Test in Excel ............................ 264

t-Based Confidence Intervals of a Population Mean .......................... 284

Binomial Distribution in Excel ............................................................. 296

Hypothesis Testing on Binomial Data ................................................. 309

1) One-Sample Hypothesis Test of Proportion in Excel..................... 318

2) Two-Sample, Pooled Hypothesis Test of Proportion ..................... 322

3) Two-Sample, Unpooled Hypothesis Test of Proportion ................ 327

Confidence Intervals of a Population Proportion ............................... 332

Family of Closely-Related Distributions.............................................. 340

Meet the Author .................................................................................... 355

Advanced Statistical & Solver Optimization Project Consulting

This brings up a blank Axis Labels dialogue box as follows:

This produces the following labels for the X-Axis:

Step 4 – Delete the Legend

The following basic scatterplot chart will appear:

This brings up a blank Axis Labels dialogue box as follows:

This produces the following labels for the X-Axis:

Step 4 – Delete the Legend

Excel Functions For Combinations

PERMUT ( n, k ) = Number of Permutations of n Objects Taken k at a time Sequentially

Problem 1: Combinations of Investment Proposals

Problem 2: Combinations of Newly-Opened Offices

Problem 3: Multiple Combinations of Newly-Opened Offices

Total number of combinations =

Problem 5: Combinations of Sub-Groups

Problem 6: Permutations of Delivery Routes

Problem 7: Permutations of Seating Arrangements

Problem 8: Permutations of Executive Groups

Problem 10: Permutations of Letter Groups

Uses of the Normal Distribution

Useful Normal Approximations of Other Distributions

Normal Approximation of the Binomial Distribution

Normal Approximation of the Poisson Distribution

Statistical Tests That Require Normality of Data or Residuals

History of the Normal Distribution

Probability that potatoes weigh less than 22 pounds = NORM.DIST(22,20,1,TRUE)

f(X = 2, µ = 0, σ = 1) = NORM.DIST(2,0,1,FALSE) = 0.05399

Φ(0) = 0.5 = 50%

Φ(X) = 1 - Φ(-X) and therefore Φ(X) + Φ(-X) = 100%

The t-Distribution’s Convergence to the Standard Normal Distribution

For Normally-Distributed Data, Use the Empirical Rule

Empirical Rule Problem in Excel

For Data of Unknown Distribution, Use Chevbyshev’s Theorem

Chebyshev’s Theorem Problem in Excel

Why the Central Limit Theorem Is So Important to Statistics

Population (Sample size n = 1)

Total number of samples = 2,500