Sie sind auf Seite 1von 48

Tests of

Hypothesis
Identifying the Appropriate
Rejection Region for a
given Level of Significance
Objectives
 the population variance is assumed to be
known
 the population variance is assumed to be
unknown
 the Central Limit Theorem is to be used
A preschool teacher wanted to know the
mean age of kindergarten students in her
town. She took a random sample of 40
students and the mean age was five years.
The population standard deviation was
known to be 1.5 years. Suppose the level of
significance was set to ∝ = 𝟎. 𝟎𝟓 What is the
rejection region?
In hypothesis testing, a normal distribution of
data is needed. But what will happen for a
non-normal distribution? The Central Limit
Theorem allows it with certain conditions. It
states that any data set that is randomly
sampled from a population, regardless of
the type of distribution, will be
approximately normal if the sample size
taken is large, that is, 𝑛 ≥ 30.
When the population standard deviation 𝜎 is
known or when 𝜎 is not known but (𝑛 ≥ 30)
by Central Limit Theorem, z–test can be
used to find the rejection region. The table
below shows which tail test to perform with
its corresponding rejection region indicated
by the alternative hypothesis.
The rejection region is an interval of values for which
the null hypothesis is rejected. If the test statistic falls
within the rejection region, the null hypothesis is
rejected and the alternative hypothesis is accepted.
When is 𝜎 unknown and the sample size is
small (𝑛 < 30) use the t-test for rejection
region. The same method can be used as
mentioned above except that it is t-test
instead of z-test.
A preschool teacher wanted to know the
mean age of kindergarten students in her
town. She took a random sample of 40
students and the mean age was five years.
The population standard deviation was
known to be 1.5 years. Suppose the level of
significance was set to ∝ = 𝟎. 𝟎𝟓 What is the
rejection region?
Step 1: List the known facts of the problem.
𝑥ҧ = 5 years
n = 40
𝜎 = 1.5 years
𝛼 = 0.05
Step 2: Identify the null and alternative hypotheses.
𝐻0 = 5 years old
𝐻𝐴 ≠ 5 years old
Step 3: Identify the appropriate test statistic and
determine what tail test to perform.
The sample size is large, 𝜎 is known, and the
alternative hypothesis contains "≠" symbol. Thus, two-
tailed z-test can be used.
Step 4: Draw the graph, and identify the
rejection region.
Since it is a two-tailed test, divide 𝛼 = 0.05 by
∝ 0.05
two. = = 0.025
2 2
The area of the rejected region is 0.025 in each
end tail of the distribution, and the area of the
acceptance region is 0.95.
Step 5: Identify the critical value/s.
Divide 0.95 by two. The area from each critical value
to the center is 0.475. Using the z-table, find the z-
value that corresponds to 0.475. The equivalent z-
value for the area 0.475 is 1.96. Thus, the critical values
are −𝑧∝ = −1.96 and 𝑧∝ = 1.96
2 2

The rejection area is 2.5% in each end tail of the


distribution. The rejection regions are values above
1.96 and values below -1.96. Reject the null hypothesis
if the test statistic falls within the rejection region. Then,
accept the alternative hypothesis.
Try it!
The mean age of graduating high school
students in Paranaque City is at most 18
years old. A random sample of 25
graduating students is taken, and the
sample mean is 19 years with a standard
deviation of 1.5 years. Determine the critical
value and the rejection region at 0.10 level
of significance.
Key Points
 The level of significance is the risk of rejecting
the null hypothesis when it is true.
 The hypothesis is a two-tailed test if the
alternative hypothesis contains “≠” symbol.
 The hypothesis test is a left-tailed test if the
alternative hypothesis contains “<” symbol.
 The hypothesis test is a right-tailed test if the
alternative hypothesis contains “>” symbol.
 A right-tailed or left-tailed test has only one
rejection region while a two-tailed test has
two rejection regions.
Tests of
Hypothesis
Test of Hypothesis on the
Population Mean
Objectives
 compute for the test-statistic value
(population mean)
 draw conclusion about the population
mean based on the test-statistic value
and the rejection region
 solve problems involving test of hypothesis
on the population mean
A high school basketball coach reported
that all high school basketball players are
doing well in Math. On average, they are
getting at least a grade of 77. The
supervisor randomly sampled 20 players
and found that their mean Math grade
was 75 with a standard deviation of 3.
Does it appear that the mean Math grade
of the basketball players is less than 77?
Using 5% level of significance, is the
sample mean significantly different from
the basketball coach’s claim to decide
that his claim is false?
The following are the steps in conducting a
test of hypothesis:
 Step 1: Identify the given values and the
level of significance to be used.
 Step 2: Formulate the null and alternative
hypotheses.
 Step 3: Choose the appropriate test
statistic and determine the critical value
of the test statistic.
 Step 4: Find the value of the test statistic
using the sample data.
Step 5: Make a decision.

The table below are guidelines to make an


appropriate decision.

Step 6: Draw a conclusion.


A high school basketball coach reported
that all high school basketball players are
doing well in Math. On average, they are
getting at least a grade of 77. The
supervisor randomly sampled 20 players
and found that their mean Math grade
was 75 with a standard deviation of 3.
Does it appear that the mean Math grade
of the basketball players is less than 77?
Using 5% level of significance, is the
sample mean significantly different from
the basketball coach’s claim to decide
that his claim is false?
Step 1: Identify the given values and the level of
significance to be used.
𝜇 = 77
𝑥ҧ = 75
n = 20
𝑠=3
𝛼 = 0.05
Step 2: Formulate the null and alternative hypotheses.
𝐻0 𝜇 ≥ 77
𝐻𝐴 𝜇 < 77
 Step 3: Choose the appropriate test
statistic and determine the critical value
of the test statistic.
 Since the alternative hypothesis contains
“less than” symbol, the hypothesis test is a
left-tailed test. The sample size is less than
30 and the population standard deviation
is unknown. Thus, the appropriate test
statistic is the t-test statistic.
 Since it is a one-tailed test with 𝛼 = 0.05
the rejection area is 0.05 in left end tail of
the distribution.
Step 4: Find the value of the test statistic
using the sample data.
𝑥ҧ − 𝜇
𝑡= 𝑠
𝑛
77 − 75
𝑡=
3
20
𝑡 = -2.98
 Step 5: Make a decision.
Since 𝑡 = -2.98 is located to the left of the
critical value -1.729 reject the null
hypothesis. Thus, accept the alternative
hypothesis.

 Step 6: Draw a conclusion.


The sample data do not support the claim.
There is no sufficient data to conclude that
the average Math grade of high school
basketball players is at least 77 at 5% level
of significance.
Try it!
According to a local newspaper, the average
age of private cars in Quezon City is 5 years. A
large car rental company in Quezon City
selects a sample of 40 cars. The average age of
the cars in the sample is 6 years and the
standard deviation is 2.6 years. At 1% level of
significance, is the sample data sufficient to say
that the average age of the cars in the car
rental is different from the private cars in
Quezon City?
Key Points
 Applying the formula for test statistic will
get the corresponding standardized
value.
 The decision is either to reject or not to
reject the null hypothesis.
 If the null hypothesis is rejected, the
alternative hypothesis is accepted.
 Draw a conclusion about the population
mean in the context of the original claim.
Tests of
Hypothesis
Test of Hypothesis and the
Central Limit Theorem
Objectives
 formulate the appropriate null and
alternative hypotheses on a population
proportion
 identify the appropriate form of the test
statistic when the Central Limit Theorem is
to be used
 identify the appropriate rejection region
for a given level of significance when the
Central Limit Theorem is to be used
 Remember that in formulating a
hypothesis, the null hypothesis 𝐻0 should
contain a statement of equality while the
alternative hypothesis 𝐻𝐴 should contain a
statement of inequality as indicated by
the symbols <, >, or ≠.
 Also remember that the type of tail test is
indicated by the alternative hypothesis.
 If 𝐻𝐴 contains >, you do a right-tailed test.
If 𝐻𝐴 contains <, you do a left-tailed test.
If 𝐻𝐴 contains ≠, you do a two-tailed test.
 In some instances, one might be
interested in making inferences about the
proportion of observations in the
population with a certain property. When
the variable is categorical, a hypothesis
test is used to test the claim about a
population proportion.
 The Central Limit Theorem for proportion
states that if a random variable 𝑝Ƹ is a
sample proportion with a random sample
size 𝑛 then 𝑝Ƹ is approximately normally
𝑝𝑞
distributed with mean and standard
𝑛
 The test statistic used for hypothesis testing for
a population proportion is the z-test for
proportion if 𝑛𝑝 ≥ 5 and 𝑛𝑞 ≥ 5
 Below is the formula which is used for z-test for
population proportion.
𝑝Ƹ − 𝑝
𝑧=
𝑝𝑞
𝑛
where:
z = z-value
𝑝Ƹ = sample proportion
p = hypothesized population proportion
q = complement of p
n = sample size
The table shows the rejection regions for
common values of ∝
Example
 A doctor claims that 70% of Filipino mothers
prefer Brand A powdered milk for their kids. It
is decided to test this claim and ask a random
sample of 100 Filipino mothers whether they
prefer Brand A powdered milk over other
brands. Seventy-two prefer Brand A
powdered milk. Using 0.05 level of
significance, find the following:
1. null and alternative hypotheses
2. appropriate form of test statistic
3. rejection region/s
1. To formulate the hypotheses, state the
claim, which is “70% of Filipino mothers
prefer brand A milk powder”, then translate
it to mathematical sentence p = 0.70 . This
mathematical sentence is the 𝐻0 . The
complement is which is p ≠ 0.70, which is the
𝐻𝐴

Thus, the hypotheses are:


𝐻0 : p = 0.70
𝐻𝐴 : p ≠ 0.70
2. Before using the formula, verify first if the
sampling distribution of the sample distribution is
suitable for the approximation of normal
distribution.

Verify if 𝑛𝑝 ≥ 5
(100)(0.70) ≥ 5
70≥ 5
Verify if 𝑛𝑞 ≥ 5
(100)(0.30) ≥ 5
30≥ 5
Since both conditions were satisfied, the z-test
statistic for population proportion can be used.
3. The alternative hypothesis contains ≠
symbol. Thus, the hypothesis test is a two-
tailed test. There are two rejection regions,
one in each end tail of the distribution. For a
0.05 level of significance, the rejection area
is 0.025 in each end tail of the distribution.
The critical values are 1.96 and –1.96. The
rejection regions are values that are greater
than 1.96 and values that are less than –
1.96.
Try it!
 A computer technician claims that at least
60% of Filipino adults in Metro Manila have
internet access on their mobile phones. In a
random sample of 80 Filipino adults in Metro
Manila, 65% have internet access on their
mobile phones. Find the following using 0.1
level of significance:
1. null and alternative hypotheses
2. appropriate form of test statistic
3. rejection region/s
Key Points
 Hypothesis test on the population proportion is
used when dealing with categorical variables.
The sample proportion is tested against the
hypothesized population proportion.
 The test statistic used for hypothesis testing for
a population proportion is the 𝑝z-test
ො −𝑝
statistic
for population proportion 𝑧 = 𝑝𝑞 or It can
𝑛
only be used if 𝑛𝑝 ≥ 5 and 𝑛𝑞 ≥ 5. This is to
verify that the sampling distribution of the
sample proportion is approximately normal.
Tests of
Hypothesis
Test of Hypothesis on the
Population Proportion
Objectives
 compute for the test statistic value
(population proportion)
 draw conclusion about the population
proportion based on the test statistic
value and the rejection region
 solve problems involving test of hypothesis
on the population proportion
 Remember that a sample proportion is
computed by taking the ratio of the
outcome of interest of the sample to the
sample size.
𝑥
𝑝Ƹ =
𝑛
where:
𝑝Ƹ = sample proportion
x = is the outcome of interest
n = sample size
The complement of 𝑝Ƹ is 𝑞ො
ො 1 - 𝑝Ƹ
𝑞=
A study showed that at least 60% of the
employees in the company prefer to go to
Boracay during summer holidays. The
manager took a random sample of 50
employees and found out that 56% prefer to
spend their summer holidays in Boracay. Is
there a significant evidence to suggest that
the proportion of employees in the
company who prefer to go to Boracay has
decreased? Use 5% level of significance.
 Step 1: Identify the given values and the
level of significance to be used.
 Step 2: Formulate the null and alternative
hypotheses.
 Step 3: Choose the appropriate test
statistic and determine the critical value
of the test statistic.
 The critical value and rejection region
depend on the type of test to be
performed (left-tailed, right-tailed, or two-
tailed) and the level of significance.
 Step 4: Find the value of the test statistic using
the sample data.
 Before using the z-test statistic formula for
population proportion, verify if 𝑛𝑝 ≥ 5 and
𝑛𝑞 ≥ 5
𝑝ො −𝑝
 𝑧= 𝑝𝑞
𝑛
where:
where:
z = z-value
𝑝Ƹ = sample proportion
p = hypothesized population proportion
q = complement of p
n = sample size
 Step 5: Make a decision.
 The table below shows how to make the
appropriate decision.

 Step 6: Draw a conclusion.


Step 1: Identify the given values and the
level of significance to be used.
Try it!
A nutritionist said that 40% of Filipino teens
eat breakfast everyday. In a random
sample of 200 Filipino teens, 68 say that they
eat breakfast everyday. Using 10% level of
significance, is there enough evidence to
reject the nutritionist’s claim?
Key Points
 The
hypothesis test for population
proportion is an inference about the
proportion of a population.
𝑝ො −𝑝
 Use the formula 𝑧 = for population
𝑝𝑞
𝑛
proportion if 𝑛𝑝 ≥ 5 and 𝑛𝑞 ≥ 5
 Draw a conclusion about the population
proportion in the context of the original
claim.

Das könnte Ihnen auch gefallen