15 views

Uploaded by Murali Dharan

h

- Evaluation of Shear Bond Strength
- Anova1 - Statistic
- Qt2week06 Anova
- Anova
- Anova
- T10_MixModels
- Science Teachers' Practices IJITL Pp 25-34
- Airline Data Prediction With Ln Feb 25
- Chapter 16
- Automation of Software Testing for Reusable Components
- Art [Vollmann et al, 2000] Spatial field variations in soybean (Glycine max [L.] Merr.) performance trials affect agronomic characters and seed composition.pdf
- Mockmidterm2015_1
- Analysis of Variance
- Experimental Design2
- TechAspects for RestructuringConcentratedPulp 07
- School Characteristics, Use of Project Method and Learner Achievement in Physics
- Assignment Thu
- Modified Rjpt_tween 20 Manuscript Dox
- foot over bridge.pdf
- Tw Anova Aranjat

You are on page 1of 8

The completely randomized design is the simplest form of

experimental designs. In a completely randomized design,

each treatment is applied to each experimental unit

completely by chance. Although the completely

randomized design is very simple, it has many

advantages:

(1) It is very flexible. Any number of treatments and any

number of experimental units can be used.

(2) The statistical analysis is easy even if the number of

experimental units in different treatments is different.

(3) The statistical analysis remains simple in the presence

of missing values. We can actually prove that the relative

loss of information due to missing data is smaller than any

other experimental designs. Completely randomized

design is extremely attractive in the case when

experimental units are homogeneous. For example, it is

the method of choice for many laboratory experiments,

e.g., in physics, chemistry, or cookery, where a quantity of

material, after through mixing, is divided into small

samples or batches to which the treatments are applied.

(ANOVA) for a Completely Randomized Design:

Step 1: To make sure the data come from a completely

randomized design. This means that each treatment is

assigned to an experimental unit completely by chance.

Step 2: Use a graphical procedure such as box-plots or

dot-plots to visualize the equal variance assumption. The

normality assumption is guaranteed if the data truly

comes from a completely randomized design. Suppose

that the equal variance assumption is not satisfied, you

can either use a nonparametric method, discussed in

Chapter 15, or find a statistical consultant to get help. The

method discussed in Chapter 15 is just one alternative

when the equal variance assumption is not satisfied. It is

not the "best" alternative.

Step 3: Create an ANOVA table that is similar to Table

10.1 below with any statistical package such as SAS (used

in this lecture) or SPSS. In Table 10.1, we assume that

there are p treatments and n experimental units.

Source

df

SS

MS

F

Treatment

SST

MST(a)

MST/MSE

(p 1)

Error

SSE

MSE(b)

(n p)

Total

SS(Total)

(n 1)

(b) MSE = SSE/(n p)

Note:

(1) The degrees of freedom for treatment is (p 1) in a

completely randomized design with p treatments.

(2) The degrees of freedom for error is (n p) in a

completely randomized design with p treatments and n

experimental units.

(3) SST + SSE = SS(Total)

(4) We can complete the ANOVA table if we know any

two of these three quantities: SST, SSE, or SS(Total).

(5) We can complete the ANOVA table if we know both

MSE and MST.

(6) Partially complete ANOVA table will always be

available in exams or practice problems.

perform the following test:

H0: 1 = 2 = . . . = p

Ha: At least two treatment means are different.

The output from SAS or SPSS always includes the pvalue of the above F-test. We can reject the null

hypothesis at significance level when "p-value ."

Step 5: Suppose the F-test suggests that one can reject the

null hypothesis, one can then use the multiple comparison

procedures discussed in Section 10.3 to find the mean

differences. However, all of those multiple comparison

procedures discussed are very conservative and will not

be discussed this semester.

Example 10.4:

A manufacturer of television sets is interested in the effect

on tube conductivity of four different types of coating for

color picture tubes. The following conductivity data are

obtained.

Coating Type

1

2

3

4

143

152

134

129

Conductivity

141

150

149

137

136

132

127

132

146

143

127

129

General Linear Models Procedure

Dependent Variable: RESP

Sum of

Source

DF

Squares

Model

(1)

844.68750

Error

(2)

(4)

Corrected Total(3)

1080.93750

Mean

Square

(5)

(6)

F Value

(7)

Pr > F

.00029

df(Error) = (n p), df(Total) = (n 1),

SSE = SS(Total) SST, MST = SST/(p 1),

MSE = SSE/(n p),

F = MST/MSE

Note: In the above SAS printout, Model = Treatment and

Pr > F = p-value.

General Linear Models Procedure

Dependent Variable: RESP

Sum of

Source

DF

Squares

Model

3

844.68750

Error

12

236.25000

Corrected Total 15

1080.93750

Mean

Square

281.56250

19.68750

F Value

14.30

Pr > F

.00029

the alternative that at least two of the means differ. Use

= 0.05.

Solution:

H0: 1 = 2 = 3 = 4

Ha: At least two means differ

Test Statistic: Fc = 14.30

p-value: p-value = 0.00029

Since the p-value = 0.00029 < = 0.05, one can reject

the null hypothesis.

Example 10.5:

A manufacturer suspects that the batches of raw material

furnished by her supplier differ significantly in calcium

content. There is a large number of batches currently in

the warehouse. Five of these are randomly selected for

study. A chemist makes five determinations on each batch

and obtains the following data.

Batch 1

23.46

23.48

23.56

23.39

23.40

Batch 2

23.59

23.46

23.42

23.49

23.50

Batch 3

23.51

23.64

23.46

23.52

23.49

Batch 4

23.28

23.40

23.37

23.46

23.39

Batch 5

23.29

23.46

23.37

23.32

23.38

General Linear Models Procedure

Dependent Variable: RESP

Sum of

Source

DF

Squares

Model

(1)

(4)

Error

(2)

(5)

Corrected Total(3)

(6)

Mean

Square

0.0242440

0.0043800

F Value

(7)

Pr > F

.00363

SSE = (MSE)(n p), SS(Total) = SST + SSE

Dependent Variable: RESP

Sum of

Source

DF

Squares

Model

4

0.0969760

Error

20

0.0876000

Corrected Total 24

0.1845760

Mean

Square

0.0242440

0.0043800

F Value

5.54

Pr > F

.00363

batch to batch? Use

= 0.05.

Solution:

H0: 1 = 2 = 3 = 4 = 5

Ha: At least two means differ

Test Statistic: Fc = 5.54

p-value: p-value = 0.00363

Since the p-value = 0.00363 < = 0.05, one can reject

the null hypothesis that all means are equal. Thus, there

to batch, for = 0.05.

- Evaluation of Shear Bond StrengthUploaded byChirag Patil
- Anova1 - StatisticUploaded byLusius
- Qt2week06 AnovaUploaded byYc Ong
- AnovaUploaded byAnonymous JONRi7VVs
- AnovaUploaded bysiti
- T10_MixModelsUploaded bymaleticj
- Science Teachers' Practices IJITL Pp 25-34Uploaded byDr. Muhammad Naqeeb ul Khalil Shaheen
- Airline Data Prediction With Ln Feb 25Uploaded byKalana Kariyawasam
- Chapter 16Uploaded byNdomadu
- Automation of Software Testing for Reusable ComponentsUploaded bySEP-Publisher
- Art [Vollmann et al, 2000] Spatial field variations in soybean (Glycine max [L.] Merr.) performance trials affect agronomic characters and seed composition.pdfUploaded byIsmael Neu
- Mockmidterm2015_1Uploaded byPETER
- Analysis of VarianceUploaded byvada_so
- Experimental Design2Uploaded byDanna Claire
- TechAspects for RestructuringConcentratedPulp 07Uploaded bymisabu
- School Characteristics, Use of Project Method and Learner Achievement in PhysicsUploaded byAlexander Decker
- Assignment ThuUploaded byVu Thi Thu
- Modified Rjpt_tween 20 Manuscript DoxUploaded bySantoshgada Gada
- foot over bridge.pdfUploaded byAnonymous xlTHYk1ODE
- Tw Anova AranjatUploaded byCodruta Dura
- A Study of Agricultural Land Use Change and Water Quality AnalysisUploaded byTaylor Fleet
- Intestinal Hystology and Enterocytes Height Variaton in Rainbow Trout (Oncorhyncus Mykiss) Grown in Cages Effects of Enviormental Conditions - N. Savić, B. Rašković, Z. Marković, V. PoleksićUploaded byCk_psih
- IIUploaded byAinun Irvanto
- dae.pdfUploaded bytantiba
- The Aligned Rank Transform for Nonparametric Factorial Analyses Using Only A NOVA ProceduresUploaded byAsura Nephilim
- BRM final reportUploaded bysufyanbutt007
- Blender PerformanceUploaded byJulia Turpo Suarez
- Life Cycle SMEsUploaded bymoheebgul
- 1-s2.0-S1877042813033818-main-1.pdfUploaded byMarc Leonen
- 939-2917-1-PBUploaded byCharith Koggala Liyanage

- 6.3.2.1. Shewhart X-bar and R and S Control ChartsUploaded byMurali Dharan
- 2019 CalendarUploaded byVadivel
- 9783319160610-c2Uploaded byMurali Dharan
- Statistical Process Control MC QuestionsUploaded byMurali Dharan
- Different Types of Reliability.pdfUploaded byMurali Dharan
- Nerd's Notes_ How We Did the ClinicalTrials.gov Data AnalysisUploaded byMurali Dharan
- Output and Discussion for the Certification ProjectUploaded byMurali Dharan
- 6.1.6. What is Process CapabilityUploaded byMurali Dharan
- https.docxUploaded byMurali Dharan
- zxzxUploaded byMurali Dharan
- Bootstrap Sample_ Definition, Example - Statistics How ToUploaded byMurali Dharan
- Confidence Intervals for Effect Sizes_ Applying Bootstrap Resampling - Practical Assessment, Research & EvaluationUploaded byMurali Dharan
- 7. List in R_ Create, Select Elements With Exampled1p7Uploaded byMurali Dharan
- Worksheets ( Downloads ) - WestgardUploaded byMurali Dharan
- AdvisorUploaded byMurali Dharan
- Chapter IV Acceptance SamplingUploaded byMurali Dharan
- Course OutlineUploaded byMurali Dharan
- Factor Analysis in Social ScienceUploaded byMurali Dharan
- Forest plot - Wikipedia.pdfUploaded byMurali Dharan
- lecturenotes1.pdfUploaded byMurali Dharan
- Course Outline SQCUploaded byMurali Dharan
- Monitoring the Performance of Bayesian EWMA Control Chart Using Loss FunctionsUploaded byMurali Dharan
- Analysis of Longitudinal and Survival Data Joint MUploaded byMurali Dharan
- Course outline.docxUploaded byMurali Dharan
- Montgomery Ch09 Sec02 03Uploaded byMurali Dharan
- Chapter I Introduction.pdfUploaded byMurali Dharan
- Decision Flow Chart for Double Sampling PlanUploaded byMurali Dharan
- a154b413-793d-4b47-9b43-54c58edc270d (1) (1)Uploaded byMurali Dharan
- TQM Review Lecture 2010Uploaded byVicky de Chavez
- Different Types of ReliabilityUploaded byMurali Dharan

- Caliornia Geology Magazine Mar-Apr 1992Uploaded bybornite
- Introduction to Survey Weights Pri VersionUploaded bykhanny96
- Geological and Geotechnical ModelUploaded byhugoq21hugo
- Reduction of Inpatient Hospital Length of Stay in Lumbar Fusion Patients With Implementation of an Evidence Based Clinical Care PathwayUploaded byShweh Fern Loo
- The Effect of Age, Gender and Attitudes on Self-regulation in DrivingUploaded byGheorghe Andreea
- Mock Aime IIUploaded byLeon Fone
- Monitoring and Evaluation of Community Systems StrengtheningUploaded byKataisee Richardson
- SAES-A-100Uploaded bypandiangv
- Marketing Vol 44 No 3Uploaded byJelena Ljubojevic
- Ritz Carlton Case Study-2Uploaded byABDALLAH
- FOR-NONUploaded byf
- c12Uploaded byRadian Sigit
- Recruiting and HiringUploaded byBrian_Bateman_4845
- A FMEA-based Approach to Prioritize Waste Reduction in Lean ImplementationUploaded byLi Near
- unit planUploaded byapi-318391076
- joavi1Uploaded bybigdaddy223
- Res 122667Uploaded byTheodoreSmith
- Stata Help Hausman Test.pdfUploaded byDr-Habib Ahmad
- Barnes FarrellMatthewsAgeandWorkAttitudesChapter2007Uploaded byAnonymous Gcgf2DYBJD
- Model- vs. design-based sampling and variance estimationUploaded byFanny Sylvia C.
- WLF2 2011 0373 Marchi RevisedUploaded bybene5997
- Using Importance-Performance Analysis to Evaluate E-Business StrategiesUploaded byHamid Usman
- Della 2006 Herbs of Paphos Larnaca in Cyprus.pdfUploaded byCiera Corca
- Exercise 14.11Uploaded byLeonard Gonzalo Saavedra Astopilco
- Chap 4 - Detection-ClassificationUploaded byFiras Zak
- OpenIntro StatisticsUploaded byblazingarun
- Chap17-Program Design for Resistance TrainingUploaded byPaulo Camelo
- 4.1. Uncertainty.pptxUploaded byanima1982
- Angela Logomasini - Consumer Guide to Chemical RiskUploaded byCompetitive Enterprise Institute
- Introducing Geophysical InversionUploaded byvelkus2013