Hypothesis

Testing

ECON 22:
Introduction to Mathematical Statistics
Hypothesis
• A hypothesis (plural: hypotheses), is a testable statement about
the relationship between two or more variables or a proposed
explanation for some observed phenomenon.

Hypothesis Testing
• It is an essential tool in inferential statistics that enables
researchers to make informed decisions about the population
parameters based on sample statistics.
 Step Procedures of Hypothesis Testing

State the Hypothesis

Determine the Test Statistic

Calculate the Test Statistic

Decision
Do not reject Ho Rule Reject Ho

Conclude H0 may be Conclude Ha is true.

true.
 Two Types of Statistical Hypothesis
1. Null Hypothesis (H0)
- This hypothesis states that there is no difference between
groups or no relationship between variables. It is a presumption of
no change (“=“).
2. Alternative Hypothesis (Ha)
- This hypothesis is a contradictory to H0 and shows that there is
difference between the two variables (“≠“, “<”, “>”).
Example:
H0: There is no difference in the salary of factory workers based on gender.
Ha: Male factory workers have a higher salary than female factory workers.
 One-Tailed Test vs Two-Tailed Test
- The alternative hypothesis will tell us if a test is one-tailed or two-tailed.
• Two-Tailed Test – it is used if the alternative hypothesis is non-
directional. Non-directional asserts that one value is different from
another (Not equal to).
Example:
H0: There is no significant difference in the scores between the two groups of students.
Ha: There is significant difference in the scores between the two groups of students.

• One-Tailed Test – it is used if the alternative hypothesis is directional.

Directional asserts that one measure is less than (or greater than)
another measure of similar nature.
Example:
H0: The discussion method is as effective as the lecture method.
Ha: Discussion method is more effective than the lecture method.
 One-Tailed Test vs Two-Tailed Test
(area of rejection)

• One-Tailed Test – a test • Two-Tailed Test – a test where the

where the area of rejection is areas of rejection are at both sides of
at either side. the distribution.
 Determine the Test Statistic
• Test statistic describes how closely the distribution of your
data matches the distribution predicted under the null
hypothesis of the statistical test you are using.
• The use of a particular test statistic depends upon the type
of data, purpose of study, and sometimes the number of
observations.
 The Z-test and T-test
• These test statistics are used for quantitative data. It used to compare
sample mean with population mean or are used to compare between
two sample means.
• Use the diagram below to determine what test will be used between
z-test and t-test :
Is population standard deviation known or given?
Use z-test
𝑥̄−μ
Use z-test No Yes 𝑧 = 𝑠
Yes
𝑛
𝑥̄−μ
𝑧= 𝜎 Is n ≥ 30? Use t-test
𝑥̄−μ
𝑛 No
𝑡= 𝑠
𝑛
3 2 1
Example Problem 1:
A manufacturer of ball pens claims that a certain pen he manufactures has
a mean writing life of 400 pages with a standard deviation of 20 pages. A
purchasing agent selects a sample of 100 pens and puts them for test. The
mean writing life for the sample was 390 pages. Should the purchasing
agent reject the manufactures claim at 1% level?
Given: 1. State the hypothesis.
Sample size: 100 pens H0: There is no significant difference
Sample mean: 390 pages between the sample mean and the
Population mean: 400 pages population mean of writing life of pen
he manufactures.
Population SD: 20 pages
Significance level: 1% or 0.01 Ha: There is significant difference
between the sample mean and the
population mean of writing life of pen
he manufactures.
2. Determine the test statistic.
𝑥̄−μ
Z-Test: 𝑧= 𝜎
𝑛

3. Calculate the test statistic.

Given:
𝑥̄−μ 390−400 −10 −10
𝑧= 𝜎 𝑧= 𝑧 =
20 20 𝑧 = 2
Sample size: 100 pens 10
𝑛 100
Sample mean: 390 pages
Population mean: 400 pages 𝑧 = −5
Population SD: 20 pages
Significance level: 1% or 0.01
4. Decision Rule

z-value = -5
critical value = ±2.575
z-value (absolute value) > critical value
5 > 2.575, reject H0

-2.575 +2.575

The computed z-value of -5 falls under the area of

rejection at 1 percent level of significance, thus, reject the
null hypothesis. This indicates that there is significant
difference between the sample mean and the population
mean of writing life of pen he manufactured.
Example Problem 2:
Records show that it takes an average of 45 minutes for a student to enroll. The
registrar of a certain school introduced a new enrollment system and increased the
number of advisers. After the introduction of the new enrollment system, the
registrar wants to find out whether there had been an improvement in the flow of
enrollment. A random sample of 75 students was taken and the sample revealed a
mean of 41 minutes with a standard deviation of 6.3 minutes. Assuming the
enrollment time takes a normal distribution, test whether the enrollment time is
reduced. Use the 0.05 significance level.
Given: 1. State the hypothesis.
Sample size: 75 students H0: The amount of time for students to
Sample mean: 41 minutes enroll is equal to 45 minutes.
Population mean: 45 minutes Ha: The amount of time for students to
Population SD: unknown enroll is less than 45 minutes.
Sample SD: 6.3 minutes
Significance level: 5% or 0.05
2. Determine the test statistic.
𝑥̄−μ
Z-Test: 𝑧= 𝑠
𝑛

3. Calculate the test statistic.

Sample size: 75 students 𝑥̄−μ 41−45 −4 −4
𝑧= 𝑠 𝑧= 𝑧 = 𝑧=
Sample mean: 41 minutes 6.3 6.3 0.7274
Population mean: 45 minutes 75 8.6602
𝑛
Sample SD: 6.3 minutes
𝑧 = −5.50
Significance level: 5% or 0.05
4. Decision Rule

z-value = -5.50
critical value = -1.645
z-value (absolute value) > critical value
5.50 > 1.645, reject H0

-1.645

The computed z-value of -5.5 falls under the area of

rejection at 5 percent level of significance, thus, reject the
null hypothesis. This means to say that the amount of
time for students to enroll is less than 45 minutes and the
new enrollment system had improved the length of
enrolment time by reducing the amount of time spent.
Example Problem 3:
In a particular barangay, it was found out that the average monthly electricity bills
were P3,992.60. A seminar was conducted to educate residents for the proper use of
appliances and other equipment that requires electricity for its use and operation.
Several months after the seminar, a resurvey was conducted and a sample of 25
residents was taken. Results revealed a sample mean of P3,062.00 and a standard
deviation of P1,685.00. At the 0.01 significance level, can we conclude that there is
reduction in the monthly electricity bill?

Given: 1. State the hypothesis.

Sample size: 25 residents H0: The amount of monthly electricity
Sample mean: P3,062.00 bill is equal to P3,992.60.
Population mean: P3,992.60 Ha: The amount of monthly electricity
Population SD: unknown bill is less than P3,992.60.
Sample SD: P1,685.00
Significance level: 1% or 0.01
2. Determine the test statistic.
𝑥̄−μ
T-Test: 𝑡= 𝑠
𝑛

3. Calculate the test statistic.

Sample size: 25 residents
𝑥̄−μ 3,062 − 3,992.60 −930.6
𝑡= 𝑠 𝑡=
1,685
𝑡=
1,685
Sample mean: P3,062.00
25 5
Population mean: P3,992.60 𝑛
Sample SD: P1,685.00 −930.6
Significance level: 1% or 0.01
𝑡=
337
𝑡 = −2.76
4. Decision Rule

z-value = -2.76
critical value = -2.492

-2.492

The computed z-value of -2.76 falls under the area

of rejection at 1 percent level of significance with a
degree of freedom of 24, thus, reject the null
hypothesis. This means to say that the amount of
monthly electricity bill is less than P3,992.60.
Example Problem 4:
A manufacturer of car batteries claims that it takes an average of 14 months before a
battery runs out of life. To prove the claim of the manufacturer, a customer
conducted a study of 40 cars and found out that the mean life of batteries used by
these cars was 12.5 months with a standard deviation of 4 months. Assuming that
the distribution for the life of the batteries is normal, is the mean life of batteries
different from 14 months? Use the 0.01 significance level.

Example Problem 5:
A manufacturer claims that the average content of one bottle of cooking oil is 237ml.
Retails stores claim otherwise. To determine whose claim is true, sample of 20 bottles
of cooking was tested and revealed a sample mean of 235ml and with a standard
deviation of 2.5ml. Is there a difference between the sample mean and the
population mean. Use the 0.05 significance level.
End of Presentation.
 References:
• Bonamente, Max (2018). Statistics And Analysis of
Scientific Data.
• Newbold, P. Carlson, W., and Thorne B. (2007).
Statistics for Business and Economics. 8th Edition.
USA: Pearson
• Pagala, R. (2011). Statistics. Manila: Mind shapers
Co. Inc.
• Rumsey, Deborah J (2019). Statistics Essentials For
Dummies.
• Santos, David (2011). Probability: An Introduction.
Prepared by:
Cherlen J. Cayetano
Instructor
Department of Economics, CEMDS
Cavite State University