WHOLE BRAIN LEARNING SYSTEM

OUTCOME-BASED EDUCATION
Science, Technology, and Engineering (STE) Program
GRADE
BASIC STATISTICS
9

4
LEARNING QUARTER

MODULE WEEK 2-3

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 1

 MODULE IN
BASIC STATISTICS

QUARTER 4
WEEK 2-3

TESTS ON TWO MEANS

Development Teams

Writer: Gregorio P. Agatep, Jr.

Editors/Reviewers: Roque A. Domingo

George A. Alonzo

Illustrator: Gregorio P. Agatep, Jr.

Lay-out Artist: Gregorio P. Agatep, Jr.

Management Team:

Vilma D. Eda Arnel S. Bandiola

Lourdes B. Arucan Juanito V. Labao

Marlyn S. Ventura

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 2

 What I Need to Know

In the previous module, we dealt with hypothesis testing on a single mean. In this
module, we test hypotheses concerning two means using both the critical value and the p
value

Most Essential Learning Competencies:

The learners:

1. differentiate dependent and independent samples

2.. test hypotheses concerning two means
.
Objectives:

At the end of the module, the learners should be able to:

1. differentiate dependent and independent samples

2.. test hypotheses concerning two means

What I Know

Multiple Choice. Read the given questions carefully and answer each item correctly. Write
the letter of your choice on a separate sheet.

1. Which situation shows the samples are independent?

A. A researcher uses 10 mice, all of which have glucose levels higher than 220 mg/dL,
5 were treated with metformin, while the remaining 5 are given “serpentine” extract.
After 2 weeks of treatment, the glucose levels of the two groups were compared.
B. A researcher uses 10 mice, all of which have glucose levels higher than 220 mg/dL
Before administering solosa, the glucose levels of the mice were taken. Two hours
after the drug was administered, the glucose levels of the mice were again measured.
The glocuse levels before and after the administration of the drug were compared.
C. A researcher wanted to determine if there is any difference in the attitude of twins
toward a controversial subject.
D. Twelve lawyers in Laoag City and twelve lawyers in Batac City were asked to rate
the way the government is handling the issue regarding the West Philippine Sea. Is
there a significant difference in the way the lawyers from each city rated how the
government is handling the issue?

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 3

For items 2-5, refer to the situation below.

A manager wishes to compare the use of two different spreadsheets by having seven
employees work with each spreadsheet and then noting the number of errors that are
committed. Shown are the result of that comparison.
Number of errors
Employee Spreadsheet I Spreadsheet 2
Alex 3 4
Benny 4 2
Charlie 2 7
Don 7 3
Evans 5 2
Frank 3 1
Gio 2 5
Is there a difference in the mean number of errors committed? Use α =0.05

2. What is the most appropriate alternative hypothesis?

A. H 1 : μ 1> μ 2 B. H 1 : μ 1< μ 2 C. H 1 : μ 1 ≠ μ 2 D. H 1 : μ 1 ≤ μ 2

3. What statistical tool is used to test the hypothesis?

A. z -test B. t -test (pooled variances)
C. t -test (unequal variances) D. t -test (dependent data)

4. What is the critical value?

A. 2.447 B. 2.179 C. 1.943 D. 1.792

5. What is the mean used in finding the value of the test statistic?
A. 30.14 B. -0.14 C. 3.42 D. 3.71

Lesson

1 TESTS ON TWO MEANS

What’s In

Before taking on the lesson, do Activity 1 first.

Activity 1: Solve the problems

1. A fruit juice company claims that the average content of their tetra packs is 250 mL A
sample 100 yielded an average of 245 mL with a standard deviation of 7.5 mL Test the claim
of the company at the 0.05 level of significance.

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 4

2. A teacher wants to determine if listening to popular music affects the performance of
students. A class of 36 students were used in the study. The average score was 85 with a
standard deviation of 5. A previous study revealed that the mean was 82. Does this mean
that the scores have increased?

What’s New
 

Testing hypothesis need not be limited to only one sample. There are many cases
where two samples are involved. In dealing with two samples it is of utmost importance to
determine if the samples are dependent or independent because independence dictates the
test statistic to use.

What is It

Dependent and independent Samples

Two samples are dependent if one group provides information about the other.
Repeated measurements involving only one group fall into this category. Paired or matched
observations are also included here. Dependent samples are sometimes called correlated
samples.

Illustrations:

1. Students were given a pretest before an intervention and a posttest after the.
- Results of the pretest and posttest are dependent samples because one group
provide information about the other group

2. Husbands and their wives were surveyed on their level of knowledge how to combat
COVID-19.
- Here, the level of knowledge of the husband is paired with the level of knowledge of
his wife.

Two samples are independent if one group does not provide any information about
the other group. The groups are not related and there is no meaningful way to pair them.

Illustrations:

1. In an experiment one group was given a placebo while the other group was given a
treatment
- There are two different results taken from two different groups.

2. Two groups of cyclist were used in an experiment. One group was given tubeless tires
and the other group was given radial tires. Both groups were required to travel a distance of
25 km.

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 5

 - The data gathered from the cyclists are independent since the groups were never
related.

Testing Two Dependent Sample Means

When two samples are dependent, we use t−test for two dependent samples. The
formula is given by
d́
t=
sd , df =n−1
√n
sd =standard deviation of the difference df = degrees of freedom

Example 1: Companies train their employees to make them more productive. The output
level of 10 employees before training and after an extensive training program. The results
are measured in units per hour. Was the training effective in increasing output? Set α =0.05 .
Employee Output
Before Training After training
1 31 40
2 35 42
3 42 53
4 49 47
5 37 45
6 32 45
7 41 57
8 31 42
9 30 47
10 31 49

Solution:

I. Statement of the null and alternative hypotheses

H 0 :μ 1=μ2
H 1 : μ 2> μ 1 (“increase” signifies “greater than” – the direction of the test)

II. Level of Significance

α =0.05
t -test (dependent data
degrees of freedom = 9
critical value = 1.833
decision rule: If t c (computed) is greater than the critical value, reject H 0, otherwise
accept H 0

III. Computation
Steps: 1. Construct the column d (after minus before)
2. Find d́ (mean of d ), sd (standard deviation of d )
3. Solve for t

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 6

 d́
t=
sd , df =n−1
√n
Employee Output d
Before Training After training
1 31 40 9
2 35 42 7
3 42 53 11
4 49 47 -2
5 37 45 8
6 32 45 13
7 41 57 16
8 31 42 11
9 30 47 17
10 31 49 16

d́=
∑ d = 106 =10.6 sd =5.64
n 10
d́ 10.6
t= = =5.94
sd 5.64 , df =n−1=10−1=9
√ n √10
IV. Decision: Since If t c (5.94) is greater than the critical value (1.833), reject H 0.

V. Conclusion: The training effectively increases output.

Example 2: A group of subjects was given an attitude test on a

controversial subject. Then they were shown a film against the x1 x2
subject and the attitude test was readministered. Is there a change 24 16
in the attitude toward the subject? Use α =0.01 20 18
24 20
Solution: 28 24
30 24
20 22
I. Statement of the null and alternative hypotheses
24 20
H 0 :μ 1=μ2 (The first set is the attitude test before) 22 18
H 1 : μ 2< μ 1 18 10
18 8
II. Level of Significance 24 20
α =0.01
t -test (dependent data)
degrees of freedom = 10
critical value = -2.764
decision rule: If t c (computed) is beyond the critical value, reject H 0, otherwise
accept H 0

III. Computation

Steps: 1. Construct the column d (after minus before)

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 7

 2. Find d́ (mean of d ), sd (standard deviation of d )

3. Solve for t
d́
t=
sd , df =n−1
√n

x1 x2 d
24 16 -8
20 18 -2
24 20 -4
28 24 -4
30 24 -6
20 22 2
24 20 -4
22 18 -4
18 10 -8
18 8 - -10
24 20 -4

d́=
∑ d = −52 =−4.73 sd =3.26
n 11
d́ −4.73
t= = =−4.81
sd 3.26 , df =n−1=11−1=10
√ n √ 11
IV. Decision: Since If t c (-4.81) is beyond the critical value (-2.764), reject H 0.

V. Conclusion: The film effectively changed the attitude toward the controversial subject..

Testing Two Independent Sample Means

When testing two independent sample means, if the population variances are
known, use the z -test
x́ 1−x́ 2
z=
σ 21 σ 22
√ +
n1 n 2
where x́ 1 = mean of the first group, x́ 2 = mean of the second group
σ 12 = population variance of the first group n1= sample size of the first group
σ 22 = population variance of the second group n2 = sample size of the second group

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 8

 If the population variances are unknown and the samples are both more than
30, then the sample variances approximate the population variances, use the z-test
x́ 1−x́ 2
z=
s 21 s22
√
where x́ 1 = mean of the first group,
+
n1 n2
x́ 2 = mean of the second group
s21 = variance of the first group n1= sample size of the first group
s22 = variance of the second group n2 = sample size of the second group

The more prevalent situations concerning tests on two means, however, are those in
which variances are unknown. Although the variances are unknown, it can be tested if the
variances are equal or unequal. Unfortunately, such topic is beyond the scope of this
module. As a rule of thumb though, if the variances of the samples are somewhat close to
each other, we may assume that the variances are equal. But if the variances are somewhat
far apart, then we can assume that the variances are unequal.
If the variances are unknown but equal and at least one sample is less than 30, then
we use the two-sample t -test, also known as pooled t -test
x́ 1− x́ 2
t= ( n1 −1 ) s12+ ( n2−1 ) s 22 , df =n +n −2
sp
√ 1 1 ,
+
n1 n2
s p=
√ n1+ n2−2
1 2

where x́ 1 = mean of the first group, x́ 2 = mean of the second group

2
s = variance of the first group
1
n1= sample size of the first group
2
s = variance of the second group
2
n2 = sample size of the second group

Example 3: A sample of 5 nonsmokers and a sample of 7 smokers joined a study whose

lung destructive index scores are shown in the table below. A larger score indicates greater
lung damage. Is there enough evidence that smokers have greater lung damage compared
to nonsmokers using the destructive lung index? Set α =0.05 . Assume the two distributions
are normally distributed.

Lung Destructive Index

Nonsmokers 11.2 8.5 14.2 15.3 16.1
Smokers 14.8 12 18.6 20.4 18.8 21.4 20.4

Solution:
I. Statement of the null and alternative hypotheses
H 0 :μ 1=μ2 (The first group are the nonsmokers)
H 1 : μ 1< μ 2

II. Level of significance

α =0.05
Pooled t – test (independent samples)
Degrees of freedom: df =n1 +n2−2=5+7−2=10
Critical value: -1.812
Decision rule: If t c (computed) is beyond the critical value, reject H 0, otherwise

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 9

 accept H 0

III. Computations
x́ 1=13.06 s21=9.953 x́ 2=18.06 s21=11.72 n1 =5 n2 =7
( n1 −1 ) s12+ ( n2−1 ) s 22
s p=
√ n1+ n2−2
=

x́ 1− x́ 2
√ 4 ( 9.953 ) +6(11.72)
10
13.06−18.06
√
=

−5
110.132
10
=3.319

t= = = =−2.57
1 1 1 1 3.319(0.586)
sp
√ +
n1 n2 √
3.319 +
5 7

IV. Decision: Since t c (-2.57) is beyond the critical value (-1.812), reject H 0.

V. Conclusion: There is enough evidence that the lung damage of smokers is greater than
nonsmokers

Example 4. To find out if a new serum will treat leukemia, 9 mice, all with an advanced
stage of the disease, are selected. Five mice receive the treatment and 4 do not. Survival
times, in years from the time the experiment started are as follows:
Treatment 2.1 5.3 1.4 4.6 0.9
No treatment 1.9 0.5 2.8 3.1
At the 0.05 level of significance can the serum be said to be effective. Assume the two
distributions to be normally distributed.
Solution

I. Statement of the null and alternative hypotheses

H 0 :μ 1=μ2 (The first group receive the treatment)
H 1 : μ 1> μ 2

II. Level of significance

α =0.05
Pooled t – test (independent samples)
Degrees of freedom: df =n1 +n2−2=5+ 4−2=7
Critical value: 1.895
Decision rule: If t c (computed) is greater than the critical value, reject H 0, otherwise
accept H 0

III. Computations
x́ 1=2.86 s21=3.883 x́ 2=2.08 s21=1.363 n1 =5 n2 =4
( n1 −1 ) s12+ ( n2−1 ) s 22
s p=
√ n1+ n2−2
=

x́ 1− x́ 2
√ 4 ( 3.883 ) +3(1.363)
7
2.86−2.08
√
=
19.621

0.78
7
=1.674

t= = = =0.694
1 1 1 1 1.674 (0.671)
sp
√ +
n1 n2
1.674
√ +
5 4

IV. Decision: Since t c (0.694) is less than the critical value (1.895), do not reject H 0.

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 10

V. Conclusion: There is not enough evidence that the new serum is effective.

What’s More
Activity 2: More Tests!

Solve the problems.

1. A spelling test administered to two groups of grade 9 students are shown below. If the
scores are normally distributed, test if there is a significant difference in the result of the
spelling test at α =0.05 .
Group A 16 14 12 12 10 8 6 4
Group B 14 8 7 6 4 4 12
2. A researcher wanted to determine if a new drug effectively lowers glucose levels in the
blood of 15 diabetic patients. He took the glucose levels of the patients then administers the
drug and after two hours took again the glucose levels of the patients. The results are shown
below.
Before 240 180 160 195 203 220 178 158 224 176 226 190 187 165
After 225 187 160 190 180 188 185 172 190 170 190 176 195 165
Is there enough evidence to conclude that the new drug is effective n lowering blood glucose
levels?

What I Have Learned

1. Two samples are dependent if one group provides information about the other. Repeated
measurements involving only one group fall into this category. Paired or matched
observations are also included here. Dependent samples are sometimes called correlated
samples.
2. Two samples are independent if one group does not provide any information about the
other group. The groups are not related and there is no meaningful way to pair them.
3. When two samples are dependent, we use t−test for two dependent samples. The
formula is given by
d́
t=
sd , df =n−1
√n
sd =standard deviation of the difference df = degrees of freedom
4. When testing two independent sample means, if the population variances are known,
use the z -test
x́ 1−x́ 2
z=
σ 21 σ 22
√ +
n1 n 2

where x́ 1 = mean of the first group, x́ 2 = mean of the second group

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 11

 σ 12 = population variance of the first group n1= sample size of the first group
σ 22 = population variance of the second group n2 = sample size of the second group

5. If the population variances are unknown and the samples are both more than 30,
then the sample variances approximate the population variances, use the z-test
x́ 1−x́ 2
z=
s 21 s22
√ +
n1 n2

where x́ 1 = mean of the first group, x́ 2 = mean of the second group

2
s = variance of the first group
1 n1= sample size of the first group
2
s = variance of the second group
2
n2 = sample size of the second group

6. If the variances are unknown but equal and at least one sample is less than 30, then we
use the two-sample t -test, also known as pooled t -test
x́ 1− x́ 2
t= ( n1 −1 ) s12+ ( n2−1 ) s 22 , df =n +n −2
sp
√ 1 1 ,
+
n1 n2
s p=
√ n1+ n2−2
1 2

where x́ 1 = mean of the first group, x́ 2 = mean of the second group

2
s = variance of the first group
1
n1= sample size of the first group
2
s = variance of the second group
2
n2 = sample size of the second group

What I Can Do
WRITTEN WORK
Snow white buys her seven dwarfs new shovels. The amounts that each dwarf could
dig in their mine with their old shovels and their new shovels are shown below. At the 5%
level of significance, did Snow White’s gift improve performance?
Dwarf Daily Output In Tons
Old Shovel New Shovel
Doc 1.8 1.9
Happy 1.5 1.5
Grumpy 2.2 2.1
Bashful 1.9 2.0
Sleepy 2.1 2.2
Dopey 1.3 1.4
Sneezy 1.9 1.8

Claim Evidence Reasoning

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 12

 PERFORMANCE TASK

Performance Standard: Conduct tests on hypotheses

Learning Competency: Test hypotheses concerning two means

Ask 10 couples (husband and wife) to rate the way the government is
handling the pandemic (COVID-19) in a scale of 1 – 20, with 1 being the lowest and
20 being the highest. In recording your data, you may adopt the table below
Husband Wife

Using α =0.05 , test the hypothesis that there is no significant difference in the ratings of the
couple. What conclusions can be drawn from the decision?
Your work will be evaluated in terms of the accuracy of solution and correctness of
conclusions.
PERFORMANCE TASK RUBRIC
Criteria 4 3 2 1 0
Outstanding Satisfactory Developing Beginning Missing
Accuracy The solution is The solution Some parts of Most parts No
accurate and is accurate the solution of the solution is
precise are incorrect solution are presented
incorrect
Correctness The conclusion The The conclusion The No
is correct. conclusion is is partially conclusion conclusion
Additional correct correct is incorrect is
explanation are provided
added to
facilitate
understanding

Assessment

Multiple Choice. Read the given questions carefully and answer each item correctly.
Write the letter of your choice on a separate sheet.

1. Which situation shows the samples are dependent?

A. Fifteen doctors in Laoag City and fifteen doctors in Batac City were asked to rate
the way the government is handling the COVID -19 pandemic. Is there a significant
difference in the way the doctors from each city rated how the government is handling
the pandemic?

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 13

 B. A researcher uses 20 mice, all of which have glucose levels higher than 170 mg/dL,
10 were treated with glucophage, while the remaining 10 are given “banaba” extract.
After 4 weeks of treatment, the glucose levels of the two groups were compared.
C. A researcher uses 20 mice, all of which have glucose levels higher than 220 mg/dL
Before administering solosa, the glucose levels of the mice were taken. Two hours
after the drug was administered, the glucose levels of the mice were again measured.
The glocuse levels before and after the administration of the drug were compared.
D. A researcher wanted to determine if there is any difference in the attitude of parent
and child toward a controversial subject.

For items 2-5, refer to the situation below.

A manager wishes to compare the use of two different machines by having seven
employees work with each machine and then noting the number of errors that are
committed. Shown are the result of that comparison.
Number of errors
Employee Machine I Machine 2
Zeny 5 2
Yolly 1 3
Xylene 2 5
Winnie 3 7
Vicky 7 2
Udora 2 4
Tess 4 3
Is there a difference in the mean number of errors committed? Use α =0.05

2. What is the most appropriate alternative hypothesis?

A. H 1 : μ 1> μ 2 B. H 1 : μ 1< μ 2 C. H 1 : μ 1 ≠ μ 2 D. H 1 : μ 1 ≤ μ 2

3. What statistical tool is used to test the hypothesis?

A. z -test B. t -test (pooled variances)
C. t -test (unequal variances) D. t -test (dependent data)

4. What is the critical value?

A. 2.447 B. 2.179 C. 1.943 D. 1.792

5. What is the mean used in finding the value of the test statistic?
A. 30.14 B. -0.14 C. 3.42 D. 3.71

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 14

 Answer Key

What I Know (Page 3)

1. A 2. C 3. D 4. A 5. B

Assessment
1. C 2. C 3. D 4. A 5. B

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 15

References
.
Downie, N.M. and R.W. Heath. 1984. Basic Statistical Methods. Harper and Row Publishers,
Inc.

Hanneman, R.A., A.J. Kposowa, and M.D. Riddle. 2013. Basic Statistics for Social
Research. John Wiley and Sons, Inc.

Walpole, R.E., R.H. Myers, and S.H. Myers. 2000. Probability and Statistics for Engineers
and Scientists. Prentice Hall, Inc.

Webster, A. 1992. Applied Statistics for Business and Economics. Richard D. Irwin, Inc.

Ymas, S.E. Jr., B.C. Dayrit, and A.G. Turqueza. 2007. College Statistics. Ymas Publishing
House.

WBLS-OBEMELC-AlignedSelf-Learning Module Basic Statistics 16