Lesson 7 – Hypothesis Testing

A statistical hypothesis is a statement about the numerical value of a population

parameter. It is a statement or tentative assertion which aims to explain facts about a
certain phenomenon. A hypothesis needs to be resolved whether it is true or not. Thus, it
must be subjected to statistical testing procedure known as test of hypothesis or
hypothesis testing. If the hypothesis is found to be true, it is accepted; if it is found to be
false, it is rejected.

TWO KINDS OF HYPOTHESIS

• Null hypothesis, denoted by Ho. is a statement that there is no difference between a
parameter and a specific value.
• Alternative hypothesis, denoted by Ha, is the opposite or negation of the null hypothesis.
It is statement that there exists a difference between a parameter and a specific value.
When we formulate the null and alternative hypothesis, we examine the claim or the
conjecture regarding the population parameter.
How do we formulate the null and alternative hypothesis for a given conjecture or claim?
Study the following examples:
Example 1
Claim: The average monthly income of Filipino families who belong to low
income bracket is Php 8,000.
Ho: The average monthly income of Filipino families who belong to low income
bracket is Php 8,000 (𝜇 = 8,000).
Ha: The average monthly income of Filipino families who belong to low income
bracket is not equal to Php 8,000 (𝜇 ≠ 8.000).
Notice that the null hypothesis is expressed through the use of the equal
symbol while the alternative hypothesis is expressed by the not equal symbol
because the claim or conjecture does not specify any direction.
Example 2
Claim: The average number of hours that it takes a ten-year-old child to
learn a certain task in a specific subject is less than 0.52 hour.
Ho: The average number of hours that it takes a ten-year-old child to
learn a certain task in a specific subject is equal to 0.52 hour (𝜇 =
0.52).
Ha: The average number of hours that it takes a ten-year-old child to
learn a certain task in a specific subject is less than 0.52 hour. (𝜇 <
0.52).
Observe that the alternative hypothesis is expressed with < symbol
because the claim or conjecture specifies to use the less than symbol.
Example 3
Claim: The average weight loss for a sample of people who exercise 30 minutes
per day for 6 weeks is greater than 3.7 kg.
Ho: The average weight loss for a sample of people who exercise 30
minutes per day for 6 weeks is equal to 3.7 kg (𝜇 = 3.7).

Ha: The average weight loss for a sample of people who exercise 30 minutes
per day for 6 weeks is greater than 3.7 kg (𝜇 > 3.7).
Observe that the alternative hypothesis is expressed with > symbol because the
claim or conjecture specifies to use the greater than symbol.
Based on the three examples, the null hypothesis is always expressed with the
“equal” symbol. It is always a statement that a parameter is equal to a specific
value. On the other hand, the alternative hypothesis is expressed with ≠, >, 𝑜𝑟 <
symbols, depending on the given conjecture or claim.
TYPES OF TESTS
A statistical test may either be directional (one-tailed) or nondirectional
(two-tailed). We can determine whether a test is directional or nondirectional by
looking at how the alternative hypothesis is expressed.
Directional Test
A test of any statistical hypothesis where the alternative hypothesis is
expressed, using less than (<) or greater than (>) is called directional test or one-
tailed test since the critical or rejection region lies entirely in one tail of the
sampling distribution.
Example 4
Claim: The average weekly allowance of college students is less than Php 1 500.
Ho: The average weekly allowance of college students is equal to Php 1 500 (𝜇 = 1 500).
H a: The average weekly allowance of college students is less than Php 1 500 (𝜇 < 1 500).
This is directional test or one-tailed. More specifically, this is a left-tailed test because
the “less than” symbol was used in expressing the alternative hypothesis. Thus, the critical
region or the rejection lies entirely in the left tail of the sampling distribution.
Example 5
Claim: The average weekly allowance of college students is greater than Php
1,500.
Ho: The average weekly allowance of college students is equal to Php 1,500
(𝜇 = 1,500).
Ha: The average weekly allowance of college students is greater than Php 1,500
(𝜇 > 1,500).
This is also a directional test or one-tailed test. More specifically, this is a
right-tailed test because the “greater than symbol was used in expressing the
alternative hypothesis. Thus, the critical region or the rejection region lies entirely
at the right tail of the sampling distribution.
Nondirectional Test
A test of any statistical hypothesis where the alternative where the
alternative hypothesis is written with a not equal sign (≠) is called a
nondirectional test or two-tailed test since there is no assertion made on the
direction of the difference. The rejection region is spilt into two equal parts, one
in each tail of the sampling distribution.
Example 6
Claim: The average weekly allowance of college students is Php 1 500.
Ho: The average weekly allowance of college students is equal to Php 1 500
(𝜇 = 1 500).
Ha: The average weekly allowance of college students is not equal to Php 1 500
(𝜇 ≠ 1 500).
Observed that the alternative hypothesis is expressed, using the “not
equal” symbol; the test is two-tailed.
Null hypothesis (Ho)
= is equal to, the same as, not changed from, is

Alternative hypothesis (Ha)

≠ not equal, different from, changed from, not the same as

> greater than, above, higher than, at least

< less than, below, lower than, at most

Claim: The average age of jeepney drivers in Metro Manila is 35.5
years old.
Ho: The average age of jeepney drivers in Metro Manila is equal to
35.5 years old. (𝜇 = 35.5)
Ha: The average age of jeepney drivers in Metro Manila is not equal
to 35.5 years old. (𝜇 ≠ 35.5)
Claim: The average number of calories of a low-calorie meal is at
most 250.
Ho: The average number of calories of a low-calorie meal is 250. (𝜇 =
250)
Ha: The average number of calories of a low-calorie meal is at most
250. (𝜇 < 250)
In 2017, it was recorded that around 35% of the population were not
married. A researcher surveyed a random sample of 500 couples. He
found out that 18% were living together but unmarried. Test at 5%
significant level if the current percentage of unmarried couples is
different from 35%.

Ho: The current percentage of unmarried couples is 35%. (𝜇 = 35%)

Ha: The current percentage of unmarried couples is different from
35%. (𝜇 ≠ 35%)
A medical trial is conducted to test whether or not a new medicine
reduces uric acid by 50%.
Ho: The new medicine can reduce uric acid by 50%. (𝜇 = 50%)
Ha: The new medicine cannot reduce uric acid by 50%. (𝜇 ≠ 50%)
Activity 6.
State the null and alternative hypotheses for the given statements.
1. The school record claims that the mean score of Grade 7 students
in Math is 80.3
2. The body mass index of pupils under the feeding program is at
least 19.2 kg
3. We want to test whether the mean height of Grade 11 students is
at most 170 centimeters.
4. We want to test whether the general average of students in
Chemistry is less than 82.
5. The average waiting time for the food to be served at McDonalds
is below 4 minutes.
Steps in Testing the Hypothesis
Step 1. Identify the claim and formulate the null (Ho) and alternative (Ha) hypothesis.
Step 2. Set the level of significance and determine whether the test is one-tailed or two-
tailed by looking at how the alternative hypothesis is expressed. Decide on the test
statistic to be used and find the critical value for the test. Draw or illustrate the rejection
region.
Step 3. Compute the test value, using the test statistic or formula for the test.
Step 4. Make a decision whether to accept or reject the null hypothesis.
Step 5. Formulate a conclusion by answering the research question.

Level of Significance
Type of Test 𝛼 = 0.01 𝛼 = 0.05
One-tailed ±2.33 ±1.65
Two-tailed ± 2.58 ±1.96
Accepting or Rejecting the Null Hypothesis
1. Determine the critical value, using appropriate statistical tables.
2. Draw the rejection region and the critical value.
3. If the test value or the computed value falls in the rejection region, then reject the null hypothesis;
otherwise, accept the null hypothesis.
Comparing the Sample Mean and the Population Mean in a Large Sample Size
In many instances when we are testing hypotheses, we want to find out if the mean difference is
statistically significant or not. The z-test can be used for this purpose. Generally, the z-test is used when the
following conditions are satisfied.
The population standard deviation is known or given.
The population standard is unknown but the sample size is sufficiently large, (i.e. greater than or equal to thirty. n
≥ 30). In this case, we use the sample standard deviation (s) to replace the population standard deviation (𝜎).

The z-test of One-Sample Mean

The z-test of one sample mean is used when we want to test if the sample mean x differs significantly
from a population mean or hypothesized mean . To illustrate, suppose that the mean of population is known to be
𝜇o. if we take a random sample size n from this population and obtain a sample x which is somewhat different
from 𝜇o., so there is a reason to believe that there is a significant difference between 𝜇o and x.
The z test of one-sample mean and population mean. We use the z-score formula for sample means
because we are dealing with sample means. Thus, by the Central Limit Theorem we have
𝑥− 𝜇 𝜎 Where
z= , where 𝜎 = X = mean of the sample
𝜎 𝑛
𝑥
𝜇 = mean of the population
The formula can be written as n = size of the sample
(𝑥− 𝜇) 𝑛 𝜎 = standard deviation of the population
z= ,
𝜎
Identify if the statement is one-tailed or two-tailed.
1. The school record claims that the mean score of Grade 7 students
in Math is 80.3
Answer: two-tailed
2. The body mass index of pupils under the feeding program is at
least 19.2 kg
Answer: one-tailed
3. We want to test whether the mean height of Grade 11 students is
at most 170 centimeters.
Answer: one-tailed
4. We want to test whether the general average of students in
Chemistry is less than 82.
Answer: one-tailed
Afterwards, we compare the test value of z against the critical value of z by
drawing the rejection in the sampling distribution. The table 5.1 shows the critical
values of z for the given type of test and level of significance.
Level of Significance
Type of Test 𝛼 = 0.01 𝛼 = 0.05
One-tailed ±2.33 ±1.65
Two-tailed ± 2.58 ±1.96
Example 1
A new drug on the market is claimed by its manufacturers to reduce overweight women by 4.55 kg per month with
a standard deviation of 0.91 kg. Ten women chosen at random have reported losing an average of 4.05 kg within a month.
Does this data support the claim of the manufacturer at 0.05 level of significance?
Solution:
Follow the five-step procedure in testing hypothesis.
Step 1: Ho: The average weight loss per month using a new drug is equal to 4.55 kg (𝜇 =4.55).
Ha: The average weight loss per month using a new drug is not equal to 4.55 kg (𝜇 ≠ 4.55).
Step 2: Type of test: Two-Tailed or nonconditional test.
Critical Value: With the use of table 5.1, 𝛼 = 0.05, two-tailed test, the critical value is z = ±1.96.
Rejection Region
Step 3: Compute the test value, using the formula or test statistic.
Given:
X = 4.05 (𝑥− 𝜇) 𝑛 (4.05−4.55) 10
z= = = -1.74
𝜇 = 4.55 𝜎 0.91

n = 10 The test value or computed value is z = - 1.74

𝜎 = 0.91

Rejection Region Rejection Region

Step 4: Decision: Since the computed or test value does not fall within the rejection region, we
accept the null hypothesis.
Step 5: Conclusion: There is no significant difference between the sample mean and population
mean. Thus, the manufacturer is correct in claiming that the new drug in the market can reduce
overweight women by 4.55 kg per month.
Example 2
A sociologist believes that if costs more than Php 90 000 to raise a child from birth to age one. A random
sample of 49 families, each with a child is selected to see if this figure is correct. The average expenses for these
families reveal a mean of Php 92 000 with a standard deviation o Php 4 500. Based on these sample data, can it be
concluded that the sociologist is correct in his claim? Use 0.05 level of significance.
Solution
Step 1: Ho: The average cost to raise a child from birth to age one is equal to Php 90 000 (𝜇 = 90 000).
Ha: The average cost to raise a child from birth to age one is more than to Php 90 000 (𝜇 > 90 000).
Step 2: Type of test: one-tailed or directional test (right-tailed).
Critical value: With the use of table 5.1, 𝛼 = 0.05, one-tailed test, the critical value is z = ± 1.65.
Rejection region
Step 3: Compute the test value, using the formula or test statistic.
Given:
X = 92 000 (𝑥− 𝜇) 𝑛 (92 000−90 000)( 49)
z= = = 3.11
𝜇 = 90 000 𝜎 4 500
n = 49 The test value or computed value is z = 3.11
𝜎 = 4500

Step 4: Decision: Since the computed or test value falls within the rejection region, we reject the null and accept the
alternative hypothesis.
Step 5: Conclusion: There is significant difference between the sample mean and population mean. Thus, the sociologist is
correct in claiming that the cost to raise a child from birth to age one is more than Php 90 000.