HYPOTHESIS

TESTING
WHAT IS HYPOTHESIS?
- A statement of belief used in the
evaluation of population values

- Assumption, scientific or
educated guess, prediction
HYPOTHESIS TESTING

- A statistical procedure of drawing

conclusion that generally pertains
to the characteristic/s of
population through sample data.
Steps in hypothesis testing:
1. Formulation of Null and Alternative Hypotheses
2. Select the Level of Significance (α)
3. Determine the Test of Statistics to Use.
4. Define the Area of Rejection or the Critical Region
5. Compute for the value of the statistical test.
6. Decision
7. Interpretation or Conclusion
STEP 1: Formulation of Null and Alternative Hypotheses
TWO TYPES OF HYPOTHESIS
1. NULL HYPOPTHESIS
- Denoted by symbol 𝑯𝟎
- Refers to any claim or assertion pertaining to the parameter of
the population.
- The hypothesis that is intended to reject in conducting
hypothesis testing.
- It is known as the “NULL hypothesis” because of insufficient or
absence of statistical evidences or facts to warrant its truthfulness.
- Always a statement of equality
STEP 1: Formulation of Null and Alternative Hypotheses
TWO TYPES OF HYPOTHESIS
2. ALTERNATIVE HYPOPTHESIS
- Denoted by symbol 𝑯𝒂
- An assertion or claim that disagrees or contradicts the null
hypothesis
- Always a statement of inequality
z
The Symbol What It Means
Less than
< Fewer than
Lower than

Less than or equal to

≤ No more than
Does not exceed
At most

Greater than
> More than
Exceeds
Higher than

Greater than or equal to

≥ At least
No less than
STEP 1: Formulation of Null and Alternative Hypotheses
EXAMPLE: Formulate Ho and Ha using FORM I
z A Barangay Captain claims that the average
monthly income in a certain barangay is P15,000

Ho: The average monthly income in a certain

barangay is equal to P15,000.
µ = 15,000

Ha: The average monthly income in a certain

barangay is higher than P15,000.
µ > 15,000
 Mr. Ron P. Agaloos
z
Formulate Ho and Ha
A Statistics teacher wants to know whether the performance of
students in the morning session is better than those of afternoon
session. (Use Form I)

Ho:
The performance of students in the morning session is as good
as those in the afternoon session.

Ha:
The performance of students in the morning session is better
than those in the afternoon session.
EXAMPLE: Formulate Ho and Ha using FORM II
z The Dean of the College of Sciences claims that
the average IQ of his students is lower than 75%.

Ho: The average IQ of students is equal to 75%

µ = 75%

Ha: The average IQ of students is lower than 75%

µ < 75%
 Mr. Ron P. Agaloos
z
Formulate Ho and Ha
A medical practitioner would like to know whether traditional
medicine is less effective than modern medicine. (Use Form II)

Ho:
Traditional medicine is as effective as modern medicine.

Ha:
Traditional medicine is less effective as modern medicine.
EXAMPLE: Formulate Ho and Ha using FORM III
z The Dean of the College of Sciences claims that
the mean IQ of his students is lower than 75%.

Ho: The mean IQ of students is higher or equal to

75%.
µ ≥ 75%

Ha: The mean IQ of students is lower than 75%.

µ < 75%
 Mr. Ron P. Agaloos
z
Formulate Ho and Ha
The average monthly income of Nurses in NCR is higher than
₱25,000. (Use Form IV)

Ho:
The average monthly income of Nurses in NCR is not higher
than ₱25,000.
OR
The average monthly income of Nurses in NCR is less than or
equal to ₱25,000.

Ha:
The average monthly income of Nurses in NCR is
higher than ₱25,000.
EXAMPLE: Formulate Ho and Ha using FORM V
z The researcher claims that the average monthly
salary of employees of company A and B are the
same.
Ho: The average monthly salary of employees of
company A and B are the same.
µA = µ B

Ha: The average monthly salary of employees of

company A and B are not the same.
µA ≠ µ B
 Mr. Ron P. Agaloos
z
Formulate Ho and Ha
Suppose we want to know whether Face-to-face Learning
(physical) is as effective as the Online Learning (virtual) in
teaching high school Mathematics. (Use Form V)

Ho:
Face to face learning is as effective as Online Learning in
teaching HS Maths.

Ha:
Face to face learning is as not effective as Online Learning in
teaching HS Maths.
STEP 2: Select the Level of Significance (α)
- It corresponds to the area in the critical region.
- It is also the degree of significance in which we reject or
recognize the 𝐻0 .
- When a test statistic fall or included with this area, the
result is referred to as significant level α.
- Measures the “risk” associated with making an error
STEP 2: Select the Level of Significance (α)

Level of Confidence (1-α) - degree of assurance (belief) that

a particular statistical statement is correct under
specified conditions. (1-α can be 99%, 95%, 90%)

Level of Significance (α) - is the degree of uncertainty (doubt) about

the statistical statement under the same conditions
used to determine the confidence level.
(α can be 1%, 5%, 10%)
EXAMPLE:
z

A Barangay Captain from a certain barangay in Valenzuela City

claims that the average monthly income of families with six members
from his vicinity is P12,000. The researcher is 97% confident in his claim.

What is the level of significance (α) used?

Solution:
Since 1- α = 97% level of confidence
Find α = 1 – 97% level of significance
α = 1 – 0. 97
Therefore, α = 0.03 or 3% level of significance
 Mr. Ron P. Agaloos
z
Example
The teacher claims that the average grade of
SHS students enrolled in Statistics class is
higher than 85 with a standard deviation of 7.8.
A random sample of 250 students are selected
and found out that their mean is 81. The
teacher is 95% confident in his claim. Find α

α = 0.05
STEP 3: Determine the Test of Statistics to use

● What type of test are we going to use to test our

hypothesis?
Consider the following:
Identify the given information. Statistical test
Determine the type of tailed test. Sample mean/s
How many mean/s involved? Tailed-test

Then compute.
STEP 3: Determine the Test of Statistics to use
 z

General Categories of Statistical Test

Parametric Test
Non-parametric Test
 z
PARAMETRIC TEST
- Parametric tests are test that require normal distribution.
- The levels of measurement of which are expressed in an interval or
ratio.
- It is more powerful than nonparametric test.
If the test concerns means, some parametric tests are :
z - test
t - test
paired t - test
Analysis of Variance (ANOVA)
 z
NON - PARAMETRIC TEST
Non-Parametric Test is being used based on the ff.
- The given are qualitative and usually measured in terms of
ordinal or nominal
- Researchers doubt the validity of the parametric test

If the test concerns means, some non-parametric tests are:

Sign Test
Wilcoxon signed-rank test
Wilcoxon rank-sum test
Kruskal-Wallis test
Chi-square test
 Test of Difference
z

Z-test and the T-test are two statistical tools which

can be used to compare or to study groups of data
through the values of their means.
Medicine - Has the quality of life improved for patients who took drug A as opposed to drug B?

Sociology - Are men more satisfied with their jobs than women ? Do they earn more?

Economics - Is the economic growth of developing nations larger than the economic growth of the
first world ?

Marketing - Does customer segment A spend more on groceries than a customer segment B?
 z Z - Test
- used for testing hypothesis when the population
standard deviation is known and the sample size is
at least 30.

Note:
In case σ is unknown; the value of sample SD, s, can be
used. The reason is that in z-test the sample size (n) is large
enough to represents the population.
 Mr. Ron P. Agaloos
Z - Test
z

z-test for One - sample Mean

𝑥ҧ − µ 𝑥ҧ − µ 𝒏
𝒛= σ or 𝒛=
σ
𝒏

𝑥ҧ − µ 𝑥ҧ − µ 𝒏
𝒛= 𝒔 or 𝒛=
𝒔
𝒏
 z Z - Test

z-test for Two-Independent Means

σ1 & σ2 known; n1 ≥ 30 & n2 ≥ 30

𝑥ഥ1 − 𝑥ഥ2
𝒛=
σ𝟐𝟏 σ𝟐𝟐
+
𝒏𝟏 𝒏𝟐
 z T - Test
- used for testing hypothesis when the sample
standard deviation, s, is known and the sample
size is less than 30.

Note:
T-test is almost similar to Z-test, it is being used based on the
following assumptions:
the probability distribution of the random variable is
appropriately normal.
- n < 30
 z T - Test

t-test for One - sample Mean

Degrees of freedom: df = n - 1

𝑥ҧ − µ 𝑥ҧ − µ 𝒏
𝒕= 𝒔 or t= 𝒔
𝒏
 T - Test
z
Two-Independent Sample Means (Unpaired)
CASE 1: Unequal Variance: σ𝟐𝟏 ≠ σ𝟐𝟐 and unknown;
s1 & s2 given; n1 < 30 & n2 < 30
𝑥1 −𝑥2
t=
𝒔𝟐 𝒔
𝟏+ 𝟐
𝟐
𝒏𝟏 𝒏𝟐

*Degrees of freedom: df = n – 1
where n is the smaller sample size
 T - Test
z
Two-Independent Sample Means (Unpaired)

CASE 2: Equal Variance: σ𝟐𝟏 = σ𝟐𝟐 and unknown;

s1 & s2 given; n1 < 30 & n2 < 30
𝑥1 −𝑥2 Where:
t= 𝒏𝟏 −𝟏 𝒔𝟐𝟏 + 𝒏𝟐 −𝟏 𝒔𝟐𝟐
𝟏 𝟏 SP =
𝑺𝑷
+
𝒏𝟏 𝒏𝟐
(𝒏𝟏+𝒏𝟐) −𝟐

*SP – pooled variance

* Degrees of freedom: df = (n1 + n2) - 2
SELECTING THE CORRECT

Mr. Ron P. Agaloos

T-TEST
z

𝑙𝑎𝑟𝑔𝑒𝑟 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒
Note: To test equal or unequal variances use F test: F=
𝑠𝑚𝑎𝑙𝑙𝑒𝑟 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒
FSTAT < FCrit : Equal Variance FSTAT > FCrit : Unequal Variance
STEP 4: Define the Area of Rejection or Critical Region
● Rejection Region
- It is the region at far end of the distribution
- Also known as the Critical Region
- It is the area under the normal curve in which null
hypothesis is rejected based on the set condition (decision
rule) Critical value or Tabulated value (TV) -
- Denoted by α separates the area of rejection and the area
in which null hypothesis is not rejected
under the normal curve.
 Mr. Ron P. Agaloos
z

-2.4 -1.90 -1.45 0.35 2.75

B z = - 2.4 A
z = -1.90 A
t = - 1.45
t = 2.75
t = 0.35 B C

Identify the region where each given value falls?

STEP 5: Compute the Value of the Statistical Test
This step is for the computation of the
different tests that will be used to test the
hypothesis.
Computed value (CV) that will later compare with
the Critical / Tabulated value (TV)
STEP 6: Decision
● Reject Ho or Do not to reject Ho then cite
the level of significance used in the study.
Decision Rule:
REJECT Ho – if the test statistic falls in the critical region
DO NOT REJECT Ho – if the test statistic falls in the
non-rejection region
 Mr. Ron P. Agaloos
TYPES
z
of ERRORS in Hypothesis Testing

Type I Error – committed when the null hypothesis is true but

decided to go in contradiction. Denoted by α.

Type II Error – committed when the researcher decided to favor

a null hypothesis that is actually false. Denoted by β.

Decisions or making conclusion is

very crucial in statistical process.
Mr. Ron P. Agaloos
z
 Mr. Ron P. Agaloos
Two - tailed test (Ha is ≠)
z
 Mr. Ron P. Agaloos
One - tailed test (Ha is either < or >)
z

Non Rejection Region

Mr. Ron P. Agaloos
RULES TO FOLLOW:
z
STEP 7: Interpretation or Conclusion
● State the appropriate conclusion. In addition,
the conclusion should be affirmed in the context
of the problem and the level of significance.
Example:
Based on the computed value of _______ equals to ___
which is _____ the tabulated value of ____, using ___ level of
significance, hence ______ Ho. Therefore,
____________________________________________ . This
further implies that the claim is _______.
EXAMPLES
 Test the Hypothesis
z
Sample Scenario # 1
The teacher claims that the average grade of SHS students enrolled in
Statistics class is equal to 85 with a standard deviation of 7.8. A random
sample of 50 students were selected and found out that their mean is 81.
The teacher is 95% confident in his claim.

Step 1: Formulate the Hypotheses

Ho: µ = 85
The average grade of SHS students enrolled in Statistics
class is 85.

Ha: µ ≠ 85
The average grade of SHS students enrolled in Statistics
class is not equal to 85.
 Test the Hypothesis
z
Sample Scenario # 1
The teacher claims that the average grade of Grade 11 students enrolled
in Statistics class is equal to 85 with a standard deviation of 7.8. A random
sample of 50 students are selected and found out that their mean is 81.
The teacher is 95% confident in his claim.

Step 2: Level of Significance

Level of confidence: 1 – α = 95%

α = 1- 0.95

α = 0.05
 Test the Hypothesis
z
Sample Scenario # 1
The teacher claims that the average grade of Grade 11 students enrolled
in Statistics class is equal to 85 with a standard deviation of 7.8. A random
sample of 50 students are selected and found out that their mean is 81.
The teacher is 95% confident in his claim.

Test: z-test
Mean/s: 1-sample mean
Tailed-test: Two-tailed test (since Ha is ≠)
 Test the Hypothesis
z
Sample Scenario # 1
The teacher claims that the average grade of SHS students enrolled in
Statistics class is equal to 85 with a standard deviation of 7.8. A random
sample of 50 students were selected and found out that their mean is 81.
The teacher is 95% confident in his claim.

Step 1: Formulate the Hypotheses

Ho: µ = 85
The average grade of SHS students enrolled in Statistics
class is 85.

Ha: µ ≠ 85
The average grade of SHS students enrolled in Statistics
class is not equal to 85.
 Test the Hypothesis
z
Sample Scenario # 1
The teacher claims that the average grade of Grade 11 students enrolled
in Statistics class is equal to 85 with a standard deviation of 7.8. A random
sample of 50 students are selected and found out that their mean is 81.
The teacher is 95% confident in his claim.

Test: z-test
Mean/s: 1-sample mean
Tailed-test: Two-tailed test (since Ha is ≠)
 Test the Hypothesis
z

Step 4: Area of Rejection

Tabulated Value: ±1.96
 Test the Hypothesis
z

Step 4: Area of Rejection

Tabulated Value: ±1.96
 Test the Hypothesis
z

Step 5: Computed Value

 Test the Hypothesis
z

Step 6: Decision
RULE:
REJECT Ho – if the test statistic falls in the critical region
DO NOT REJECT Ho – if the test statistic falls in the non-rejection region

Computed Value: - 3.63

Tabulated Value: ±1.96
 Test the Hypothesis
z
Step 6: Decision: REJECT Ho, since CV < TV
-3.63 < - 1.96
RULE:
REJECT Ho – if the test statistic falls in the critical region
DO NOT REJECT Ho – if the test statistic falls in the non-rejection region

CV TV
 Test the Hypothesis

\
z
Step 7: Interpretation and Conclusion
Based on the computed z-value of -3.63 which is
less than the tabulated value of -1.96, using 5% level of
significance, hence reject Ho. Therefore, the average grade
of Grade 11 students enrolled in Statistics class is not equal
to 85%.
This further implies that the claim is not true.
 Test the Hypothesis
z
Sample Scenario # 2
A Barangay Captain from a certain barangay in Valenzuela City claims that the average
monthly income of families with five members from his vicinity is higher than P12,000. But
when the City Statistics Office (CSO) conducted a survey to 100 families with five
members, to his barangay randomly they found out that they only have an average
monthly income of 10,800 with a standard deviation of 1,500. With this information the
CSO assert that the claim is not true. Using 1% level of significance, test the claim of the
Brgy. Captain.

Step 1: Formulate the Hypotheses

Ho: µ ≤ 12,000
The average monthly income of families with five members in his vicinity
is not higher than P12,000.

Ha: µ > 12,000

The average monthly income of families with five members in his vicinity
is higher than P12,000.
 Test the Hypothesis
z
Sample Scenario # 2
A Barangay Captain from a certain barangay in Valenzuela City claims that the average
monthly income of families with five members from his vicinity is higher than P12,000. But
when the City Statistics Office (CSO) conducted a survey to 100 families with five
members, to his barangay randomly they found out that they only have an average
monthly income of 10,800 with a standard deviation of 1,500. With this information the
CSO assert that the claim is not true. Using 1% level of significance, test the claim of the
Brgy. Captain.

Step 2: Level of Significance

α = 0.01
 Test the Hypothesis
z
Sample Scenario # 2
A Barangay Captain from a certain barangay in Valenzuela City claims that the average monthly
income of families with five members from his vicinity is higher than P12,000. But when the City
Statistics Office (CSO) conducted a survey to 100 families with five members, to his barangay
randomly they found out that they only have an average monthly income of 10,800 with a standard
deviation of 1,500. With this information the CSO assert that the claim is not true. Using 1% level of
significance, test the claim of the Brgy. Captain.

𝜇 = 12,000 𝑛 = 100
𝑥ҧ = 10,800 𝑠 = 1,500

Test: z-test
Mean/s: 1-sample mean
Tailed-test:
 Test the Hypothesis
z
Sample Scenario # 2
A Barangay Captain from a certain barangay in Valenzuela City claims that the average
monthly income of families with five members from his vicinity is higher than P12,000. But
when the City Statistics Office (CSO) conducted a survey to 100 families with five
members, to his barangay randomly they found out that they only have an average
monthly income of 10,800 with a standard deviation of 1,500. With this information the
CSO assert that the claim is not true. Using 1% level of significance, test the claim of the
Brgy. Captain.

Step 1: Formulate the Hypotheses

Ho: µ ≤ 12,000
The average monthly income of families with five members in his vicinity
is not higher than P12,000.

Ha: µ > 12,000

The average monthly income of families with five members in his vicinity
is higher than P12,000.
 Test the Hypothesis
z
Sample Scenario # 2
A Barangay Captain from a certain barangay in Valenzuela City claims that the average monthly
income of families with five members from his vicinity is higher than P12,000. But when the City
Statistics Office (CSO) conducted a survey to 100 families with five members, to his barangay
randomly they found out that they only have an average monthly income of 10,800 with a standard
deviation of 1,500. With this information the CSO assert that the claim is not true. Using 1% level of
significance, test the claim of the Brgy. Captain.

𝜇 = 12,000 𝑛 = 100
𝑥ҧ = 10,800 𝑠 = 1,500

Test: z-test
Mean/s: 1-sample mean
Tailed-test: one-tailed test (since Ha is >)
 Test the Hypothesis
z

Step 4: Area of Rejection

Tabulated Value: +2.326

Critical
value

Rejection
Region
Non-Rejection Region
 Test the Hypothesis
z

Step 4: Area of Rejection

Tabulated Value: +2.326

Critical
value

Rejection
Region
Non-Rejection Region
 Test the Hypothesis

\
z

Step 5: Computed Value

 Test the Hypothesis
z

Step 6: Decision
RULE:
REJECT Ho – if the test statistic falls in the critical region
DO NOT REJECT Ho – if the test statistic falls in the non-rejection region

Computed Value: - 8
Tabulated Value: +2.326
 Test the Hypothesis
z

Step 6: Decision: DO NOT REJECT Ho, since CV < TV

- 8 < +2.33

Critical
value

Rejection
Region
Non-Rejection Region
 Test the Hypothesis
z

Step 7: Interpretation and Conclusion

Based on the computed z-value of -8 which is less
than the tabulated value of +2.33, using 1% level of
significance, hence do not reject Ho. Therefore, the
average monthly income of families with five members in
his vicinity is not higher than P12,000.
This further implies that the claim is not true.
 \
z

Two-Independent
Sample Means
(for z and t tests)
 Test the Hypothesis
z
Example # 1
A Physics Professor claims that there is no significant difference between
the mean scores obtained by students in the afternoon and morning
sessions in an examination in Physics 1. Date are shown below. If the
professor is 95% confident with his claim, perform the hypothesis testing.

Morning Afternoon
Mean 85 83
SD 15 10
n 40 40
 Test the Hypothesis
z
A Physics Professor claims that there is no significant difference between the mean
scores obtained by students in the afternoon and morning sessions in an examination in
Physics 1. Date are shown below. If the professor is 95% confident with his claim,
perform the hypothesis testing.
 Test the Hypothesis
z
A Physics Professor claims that there is no significant difference between the mean
scores obtained by students in the afternoon and morning sessions in an examination in
Physics 1. Date are shown below. If the professor is 95% confident with his claim,
perform the hypothesis testing.

Step 2:
α = 0.05

Step 3: Test Statistics

Test: z-test
Means: two-sample means
Tailed-test: two-tailed

Step 4: Critical or Tabulated Value: z = ± 1.96

 Test the Hypothesis
z
A Physics Professor claims that there is no significant difference between the mean
scores obtained by students in the afternoon and morning sessions in an examination in
Physics 1. Date are shown below. If the professor is 95% confident with his claim,
perform the hypothesis testing.

Step 2:
α = 0.05

Step 3: Test Statistics

Test: z-test
Means: two-sample means
Tailed-test: two-tailed

Step 4: Critical or Tabulated Value: z = ± 1.96

 Test the Hypothesis
z
Step 5: Computed Value
Morning Afternoon

Mean 85 83
SD 15 10
n 40 40
 Test the Hypothesis
z

Step 6: Decision: DO NOT REJECT Ho, since CV < TV

+0.702 < +1.96
Computed Value: + 0.702
Tabulated Value: ± 1.96

TV CV TV
 Test the Hypothesis
z

Step 7: Interpretation and Conclusion

Based on the computed z-value of +0.702 which is

less than the tabulated value of +1.96, using 5% level of
significance, hence do not reject Ho. Therefore, there is a
no significant difference between the average scores
obtained by morning and afternoon session.
This further implies that the claim of the Professor is
true.
 Test the Hypothesis
z
Example # 2
A group of nursing students selected two brands of pain reliever and test
the average time of each to take effect. For each brand the following
were obtained.
Brand A Brand B
Mean 5.2 min 4.7 min
SD 1 min 1.6 min
n 15 trials 14 trials

Assume difference between the average time of unequal variance and

unknown. Test the hypothesis that there is no significant difference
between brand A and B to take effect, using 0.01 as alpha.
 Test the Hypothesis
z
A group of nursing students selected two brands of pain reliever and test the average
time of each to take effect. For each brand the following were obtained.
 Test the Hypothesis
z
Assume difference between the average time of unequal variance and unknown. Test
the hypothesis that there is no significant difference between brand A and B to take
effect, using 0.01 as alpha.

Step 2: α = 0.01

Step 3: Test Statistics

Test: t-test
Means: two-sample means
Tailed-test: two-tailed

Step 4: Critical or Tabulated Value: t = ± 3.012

df = (n2 – 1) : 13
(smaller n)
 Test the Hypothesis
z
Assume difference between the average time of unequal variance and unknown. Test
the hypothesis that there is no significant difference between brand A and B to take
effect, using 0.01 as alpha.

Step 2: α = 0.01

Step 3: Test Statistics

Test: t-test
Means: two-sample means
Tailed-test: two-tailed

Step 4: Critical or Tabulated Value: t = ± 3.012

df = (n2 – 1) : 13
(smaller n)
 Test the Hypothesis
z
Step 5: Computed Value
Brand A Brand B
Mean 5.2 min 4.7 min
SD 1 min 1.6 min
n 15 trials 14 trials
 Test the Hypothesis
z

Step 6: Decision: DO NOT REJECT Ho, since CV < TV

+1.00095 < +3.012
Computed Value: + 1.00095
Tabulated Value: ± 3.012

TV CV TV
 Test the Hypothesis
z

Step 7: Interpretation and Conclusion

Based on the computed t-value of +1.00095 which

is less than the tabulated value of +3.012, using 1% level of
significance, hence do not reject Ho. Therefore, there is no
significant difference between the average time of brand A
and B to take effect.
This further implies that the claim of group of
nursing students is true.
z

ASSIGNMENT # 2
 Test the Hypothesis
z

* Formulate the Hypotheses (use form V)

Sample Scenario # 4
The average IQ of the adult population is 100. A researcher believes the average IQ of
adults is lower than 100. A random sample of 5 adults are tested with a sample mean of
89, and sample standard deviation is 15.81. Is there enough evidence to suggest the
average IQ is lower? The researcher is 99% confident in his claim.

* Formulate the Hypotheses (use form III)