DATA ANALYSIS

PART 1
1. To understand the processes followed by the researcher and the statistical
tools used by the statistician in the story, study Lessons 1 and2.

Lesson 1. Data Collection Procedure

Data collection or data gathering is the process of gathering and
measuring information on variables of interest in an established systematic
method that enables one to answer stated research questions, test hypotheses,
and evaluate outcomes.

In your research paper, this is the part where you give a narrative
description on the steps undertaken in data gathering. The discussion should
include the specifics of the procedures of the entire research process. The tense
of verbs to be used in research proposal is future tense. After the collection of
data, researchers change the tense of verbs to past tense.

In collecting data, there are ethics to consider. Here are the ethical
consideration:

1. Confidentiality and anonymity

Collecting data is often done under the assumption that information

provided is confidential and results will be anonymous. You should let your
respondents know when you will have to break confidentiality and whether
results will be anonymous or not.

Protect the data you collected from your respondents. Do not leave
anything with personal information in a place that can be easily accessed by
people who do not need to see the data. If possible, keep the information in a
secure or locked location.

2. Right to withdraw from the study

You have to explain to your target respondents that their involvement in

your study is voluntary. They are free to withdraw from any active data collection
at any point without pressure or fear of retaliation.

3. Extent of participation in the study

As a researcher, be neutral and unbiased. Don’t let your personal

preconceptions or opinions interfere with the data collection process.

4. Risks

2
 You have to avoid or minimize anything that will cause physical or
emotional harm to your respondents. Make your respondents aware of any
potential harms prior to their involvement in your study.

5. Plan for dissemination

After analyzing the data gathered, it is always good to share the result to
your respondents.

When you are done with your research paper, you are encouraged to
publish your work. You did not conduct a study to keep the results by yourself
but for people to learn. Your work might be useful for future researches.

In collecting data, there are process to follow. The common process for a
quantitative research is indicated below:

1. Seek approval from the head of the institution where you are going to
gather data. For schools, ask permission from the school head. For
any businesses, ask permission from the manager through a letter.
2. After the approval of the head of the institution, recruitment of
respondents follows. Your target respondents should be aware that they
are your respondents. It should be explained the risk of involvement in
the study, their right to withdraw, and know the results of your study.
3. Once your target respondents agreed, you may gather data using the
appropriate method in gathering data.
4. Analyze the data using the appropriate statistical tools.
5. Disseminate the results.

Sample of Data Gathering Procedure:

Data Gathering Procedure

The researchers sought the approval of the school president through a letter.
Upon his approval, the approved letter was presented to the instructors for the
scheduling of the administration of the questionnaire.
Before the questionnaires were distributed to the respondents, they were
informed of the risk in the involvement in the study, their right to withdraw, and
know the results of the study. They were requested to follow the given instructions
and ask for assistance if help is needed.
The questionnaires were retrieved on the same day. The responses of the
respondents were tabulated and analyzed.

3
Sample letter to the president:

Republic of the Philippines

Department of Education
Cordillera Administrative
Region Baguio City Division
BAGUIO CITY NATIONAL SCIENCE HIGH SCHOOL
Irisan, Baguio City, Philippines 2600

October 14, 2020

DR. LEO PETER N. DACUMOS

President
Saint Palasi University
Bonifacio St., Baguio City
2600

Sir:

We, the student researchers of Baguio City National Science High School, are
currently conducting a research entitled “Factors Influencing the Choice of Course
among College Students”. In line with this, we would like to ask permission from your
office to allow the administration of the attached questionnaires to the students of
Saint Palasi University.

Thank you for your favorable action on this request.

Very respectfully yours,

(Name of a researcher with signature) (Name of a researcher with

signature) Researcher
Researcher

Noted:

(Name of the teacher with signature)

4
Research Teacher

5
Sample letter to the respondents:

Republic of the Philippines

Department of Education
Cordillera Administrative
Region Baguio City Division
BAGUIO CITY NATIONAL SCIENCE HIGH SCHOOL
Irisan, Baguio City, Philippines 2600

October 14,

2020 Dear

respondent,

We, the student researchers of Baguio City National Science High School, are
currently conducting a research entitled “Factors Influencing the Choice of Course
among College Students”. In line with this, may we request you to accomplish the
attached questionnaire. Rest assured your answer in the questionnaire will be kept

Very respectfully yours,

(Name of a researcher with (Name of a researcher with

signature) signature)
Researcher
Noted:

(Name of the teacher with

signature)

Lesson 2. Data Analysis (Part 1)

Data analysis is the process of systematically applying statistical tool to

describe and illustrate, condense and recap, and evaluate data.

This is the part of your research paper where all the statistical tools used
have to be presented including how these were used and their formulas. Before
you decide on what statistical tool are you going to use, you have to consider the
type of variable, the distribution of data, and the research design.

For the distribution of data, let us discuss the parametric test and non-
parametric
test.

Parametric testsare those thatmake assumptions about parameters of the

population distribution from which the sample is drawn. This is often the
assumption that the population data are normally distributed. To assess whether
6
data are likely from a normal distribution,

7
there are statistical tests that can be used like the Kolmogorov-Smirnov Test, the
Anderson- Darling test and the Shapiro-Wilk test.

Non-parametric tests are called “distribution free” tests because they are based
on fewer assumptions (example: They do not assume that the outcome is
approximately normally distributed).

Parametric Tests Non-parametric Tests Equivalent

One-sample t -test One-sample Wilcoxon Signed Rank
Test
Paired-sample t -test Wilcoxon Matched Pair Signed-Rank
Two-sample t -test Mann-Whitney U Test
Pearson Correlation Spearman Correlation
One-way analysis of Kruskal-Wallis H Test
variance

For the research design and the type of variable, they were discussed in
Module 1 and Module 3, respectively.Let us discuss the type of research question
with the appropriate statistical tools to be used.

Types of Research Question

1. Descriptive Question – typically asks the question, “What is…?” with the
underlying purpose to describe the significance of a situation,
state, or existence of a specific phenomenon. This often associated
with revealing hidden or understudied issues.
- The commonly used statistical tool is mean. The formula is
Mean 
x
n
Where:
 is summation symbol
x is the responses of the respondents
n is the number of respondents

2. Inferential Question – asked to compare two or more variables or if you

want to find out the relationships of two or more variables.

a. Comparative Question – typically asks the question, “Is there a

significant

difference between two or more groups or treatments?”.

- This type of problem statement is used when the researcher compares
or contrasts two or more phenomena.This is a common approach to
defining a problem in the clinical social sciences or behavioral sciences.

The statistical tools that you may use for comparative question are
presented on the table below.

Parametric Tests Non-parametric Tests

One-sample t -test One-sample Wilcoxon Signed
Rank Test
Two-sample t -test Mann-Whitney U test

8
Paired-sample t -test Wilcoxon Matched Pair Signed
- Rank
k-sample ANOVA Kruskal-Wallis H Test
(using F -test)

9
 In this module, we will be focusing on the parametric tests. The discussion
of each test is presented below.

a.1. One-sample t-test - this is used to test a null hypothesis that the population
mean is equal to a specified value.
The formula is
xμ
t
s
n

Where: x is the sample mean

μ is the population mean
s is the sample standard deviation
n is the number of samples
Example:
A doctor in Baguio City claimed that the average pulse rate of women in
the city is 80 per minute. A nurse wanted to test the claim of the doctor so she
took pulse rate per minute of 20 adult women in the city. The data she recorded
were presented below.
Research Question: Is there a significant difference between the population mean and the
sample mean?
Null Hypothesis: The average pulse rate of adult women in Baguio City is 80 per minute.
Or you could say: There is no significant difference between the population mean and the
sample mean.

Respondent Blood Pressure ( x ) x  x x  x 2

1 73 -12.15 147.6225
2 78 -7.15 51.1225
3 82 -3.15 9.9225
4 67 -18.15 329.4225
5 90 4.85 23.5225
6 85 -0.15 0.0225
7 88 2.85 8.1225
8 92 6.85 46.9225
9 98 12.85 165.1225
10 64 -21.15 447.3225
11 74 -11.15 124.3225
12 90 4.85 23.5225
13 97 11.85 140.4225
14 83 -2.15 4.6225
15 79 -6.15 37.8225
16 83 -2.15 4.6225
17 86 0.85 0.7225

10
 18 95 9.85 97.0225
19 99 13.85 191.8225
20 100 14.85 220.5225

  x  x
2
 x  1,703  2,074.55

Steps in finding the computed t -value:

1. Complete the table above.
The second column in the table is the set of data gathered by the nurse.
The third column was computed by deducting the mean from each blood
pressure then squared the difference for the fourth column. Sample computation
is illustrated below.

Computation of Mean:
x
 x  1703  85.15
n 20

Sample computation for the third column:

x  x
Sample Computation for the fourth column:

 x  x 2
 12.15 147.6225

b. Compute the sample standard deviation s.

 x 
The formula s 
is x  2
n

 x  x 
2
Where: is the value computed in the table above
n is the number of sample

Let us substitute the given the simplify.

2074.5500
s 20 1 Substitution.

2074.5500
=
19 Subtract the denominator.

= 109.1868
Divide.
=10.4493 Get the square root.

c. Compute the t -value using the formula given.

xμ

11
 Formula
t
s
n

12
 85.15 
t80 Substitution
10.4493
20
5.15
t Subtract the numerator and get the square root of 20

10.4493
4.4721
Divide the denominator.
5.15
t  2.3366 Divide.
t  2.2041

The computed t -value is 2.2041. Using the t -distribution table on page 30

with 5%
level of significance and the degrees of freedom of n 1 , the tabular t - value is
2.0930.
Since the absolute value of the computed t -value which is 2.2041 is
greater than the absolute value of the tabular t -value which is 2.2041, reject
the null hypothesis. The claim of the doctor is rejected. This indicates that the
average pulse rate of adult women in Baguio city is not 80 per minute.

a.2. Two-sample t -test – it is known as independent sample t -test.

This is used to compare means of two sets of data taken from different
group of respondents.

Examples:
1. Comparing the pre-test scores taken from two sections of grade 12
students of Baguio City National Science High School.
2. Comparing before and after training test scores of two different groups

of students. The formula for two-sample t -test is

x1  x2
t .

2 1 1
s   
 n1 n2 

Where: x1 is the mean of the first group

x2 is the mean of the second group
n1 is the number of the first group
n2 is the number of the second group
s 2 is the pooled sample variance

Before you can use the formula above, the value of pooled sample variance
2
(s )

13
,should be computed first. The formula for pooled sample variance is
2

 x  x   
2
s 
2
1 2  x2
1
 x n 1 
n2  2

14
 Where:  is summation symbol
x1 refers to the data of the first group
x2 refers to the data of the second group
x1 is the mean of the first group
x2 is the mean of the second group
n1 is the number of the first group
n2 is the number of the second group

Note: Get the absolute value of the computed t - value.

Sample Testing of Hypothesis:

A teacher wants to assess her students by comparing the pre-test scores
ofmales and females from He 11 students of BCNSHS. The data she gathered are
presented below.
Research Question: Is there a significant difference between the pre-test sores of HE 11
students of BCNSHS when they are grouped according to sex?
Null Hypothesis:There is no significant difference between the mean of pre-test scores of
HE 11 students of BCNSHS when they are grouped according to sex.
Respondent Male Female
x  x  2
x x  x x  2
( x1  x1 ) 2 2
Score Score 1 1 2 2

( x1 ) ( x2 )
1 3 20 - 4.20 17.64 1 1
2 3 13 - 4.20 17.64 -6 36
3 3 13 - 4.20 17.64 -6 36
4 12 20 4.80 23.04 1 1
5 15 29 7.80 60.84 10 100
Total 36 95 136.80 174

The steps in finding the computed t -value are the following:

a. Find the mean scores

Mean of Pre-test Scores for Male Students:
36
x   7.20
1 5
Mean of Pre-test Scores for Female Students:
95
x   19
2 5
b. Compute the pooled sample variance using the formula given.

n2 2
x
s
2
1 
 x1 
2
2
 2x2 
 x n 1 
15
Formula

16
 2 136.80
s 
174 Substitute the corresponding values found in
552
s  38.85
2
the table. Simplify.

c. Compute the t -value using the given formula above.

x1  x2
t Formula
1 1
s2   
 n1 n2 
7.20  19
t Substitute the corresponding values.

38. 85 1  1 
 5 5 
 11.80
t 1
Convert to decimal then add.
5

38.850.40
 11.80
t Multiply 38.85 by 0.40.
15.54
 11.80
t  3.94 Get the square root of 15.54.

t  2.99
Divide.

Get the absolute value of the computed t -value. Therefore, the

computed t - value is 2.99.
Now, look for the tabular t - value on the t -distribution table on page 30
using 5%
level of significance for a two-tailed test and (n  2) degrees of freedom. You will
notice that the tabular t -value is 2.306.

Since the computed t - value which is 2.99 is greater than the tabular t -
valuewhich is 2.306, reject the null hypothesis. This shows that there is a
significant difference between the mean of the pre-test scores of male and
female He 11 students of BCNSHS.

a.3. Paired-sample t-test – it is known as dependent sample t-test.

- This is used to compare means of two sets of data taken from the same
group of respondents.

Examples:
1. Comparing before and after training test scoresof the same group of
students.
2. The difference of two blood pressure measurements of the same group
of senior citizen using different equipment.
17
3. Comparing the pre-test and post-test scores taken from the same
section of grade 12.

The formula for paired sample t -test is

18
   x  y
n
t   x  y   2 
 x  y   
2

 n 
n 1n
Where: n is the number of respondents

is summation symbol
x is the group of first data
y is the group of second data
Note: Get the absolute value of the computed t - value.

Sample Testing of Hypothesis:

A teacher from Philippine Science High School wants to assess the HE 11
male students from BCNSHS using the pre-test and post-test scores in Earth and
Life Science subject. She then selected 5 male students of HE 11 from BCNSHS.
The scores were presented below.
Research Question:Is there a significant difference between the pre-test scores and post-
test scores of HE 11 male students from BCNSHS?
Null Hypothesis:There is no significant difference between the pre-test scores and the post-
test scores of HE 11 male students from BCNSHS.

Respondent Pre-test Score ( x Post-test Score ( y ) ( x  y) (x  y)2

)
1 3 20 - 17 289
2 3 13 - 10 100
3 3 13 - 10 100
4 12 20 -8 64
5 15 29 - 14 196
SUM - 59 749

The y -values(post-test scores) were subtracted from the x -values(pre-test

score)
then placed the values on the 4 th column. The values under the 4 th column were
squared then placed under the 5th column. Let us now use the formula above to
compute the t -value. That is,
 59
5 11.8 11.8
t 
749 
749    59 
2
   3481 
749   696.2
 5   5 
45 20
5 15

11.8 11.8

  7.26
52.8 2.64
20 19
 Get the absolute value of the computed t -value. Therefore, the computed t -
value
is 7.26.

Now, look for the tabular t - value on the t -distribution table on page 30
using 5%
level of significance for a two-tailed test and degrees of freedom of (n 1) . You
will see that the tabular t -value is 2.776.

Since the absolute value of the computed t - value which is 7.26 is

greater than the absolute value of the tabular t - value which is 2.776, reject the
null hypothesis. This indicates that there is a significant difference between the
mean of the pre-test scores and the mean of the post test scores of HE 11
students from BCNSHS.

a.4. ANOVA (using F-test)

ANOVA stands for Analysis of Variance. As it was mentioned before, this is
used for comparing means of two or more groups. It is either one-way ANOVA or
two-way ANOVA. Let us look at each.

a.4.1. One-way ANOVA – is used when you are going to test one independent
variable with two or more groups.

Examples:
1. Comparing the pre-test scores of the learners from three sections of
STEM 11 of BCNSHS.
2. Comparing before and after training test scores of three different
groups of students.

The formulas to be used in computing the variance ratio ( F ) are:

x  x
2

MSSW 
nk

MSSB   x  xG 
MSSB
F
MSSW
df B  k 1

dfW  n  k

Where:
MSSW is the within groups mean sum of squares
MSSB is the between groups mean sum of squares
x refers to the raw data
x is the mean of each group
xG is the mean of the means or the grand mean
df is the degrees of freedom

20
k is the number of groups
n is the number of respondents

21
 F is the variance ratio

Note: Get the absolute value of the computed variance ratio ( F ).

Example:
A teacher from Regional Science High School wants to assess the 3
sections of STEM 11 from BCNSHS based on their General Mathematics pre-test
scores.

Research Question: Is there a significant difference between the scores of STEM 11

students from BCNSHS when they are grouped according to section?

Null Hypothesis:There is no significant difference between the scores of the STEM 11 from
BCNSHS when they are grouped according to section.

STEM A STEM B STEM C ( (x 1  x 1)2 (x 2  x 2)2 (x 3  x 3)2

( x1 ) ( x2 ) x3 )
7 5 1 0.1600 0.0237 4.9765
7 6 3 0.1600 0.7261 0.0533
6 3 4 1.9600 4.6389 0.5917
9 5 3 2.5600 0.0237 0.0533
8 4 1 0.3600 1.3313 4.9765
7 6 1 0.1600 0.7161 4.9765
6 5 2 1.9600 0.0237 1.5149
7 4 6 0.1600 1.3313 7.6685
8 5 5 2.5600 0.0237 3.1301
9 5 4 2.5600 0.0237 0.5917
6 3 0.7161 0.0533
7 4 3.4085 0.5917
6 5 0.7161 3.1301
SUM=74 SUM=67 SUM=42 SUM=10.40 SUM=13.6923 SUM=32.3077

Compute the mean score of each group. The mean score of STEM A was
used to complete the 4th column, the mean score of STEM B was used to
complete the 5th column, and the mean score of STEM C was used to complete
the 6th column.

Mean score of STEM A

7  7  6  9  8  7  6  7  8  9 74
x1  10 10
  7.4000
Sample computation for the 4th column
(x1  x 1)2  7  7.4000  (0.4000)2  0.1600

Mean score of STEM B

5  6  3  5  4  6  5  4  5  5  6  7  6 67
x 2 13

13
 5.1538
Sample computation for the 5th column
(x2  x2 )2  5  5.1538  (0.1538)2  0.0237

22
 Mean score of STEM C

1 3  4  3 11 2  6  5  4  3  4  5 42
x 3 13

13
 3.2308
Sample computation for the 6th column
(x 3  x 3)2  1 3.2308  (2.2308)2  4.9765
Let us identify or solve the needed information in computing the variance
ratio then solve for the variance ratio.
Given:
* n  36
*k3
* df B  k 1  3 1  2
* dfW  n  k  36  3  33

* xG 
 x  7  7  6  9  8  7  6  7  8  9  5  ...  3  4  5  183  5.0833
n 10 13 13 36
x  10.4000 13.6923  
*
MSSW 
x
2 
32.3077
36  3

56.4000
33 1.7091
nk

MSSB

nx  x G

10(7.40  5.0833)2 13(5.1538  5.0833)2 13(3.2308  5.0833)2
3 1

2

k 1
10(5.3671) 13(0.0050) 13(3.4218) 53.6710  0.0650  44.6134
 2  2
98.3494
  49.1747
2
Solution:

F  49.1747
MSS B   28.7723
MSSW 1.7091

The computed F -value is 28.7723. For the tabular F -value, you will find it at
the
F -distribution table on page 31 with 5% level of significance, the numerator
degrees of
freedom of df B  3 1  2 , and the denominator degrees of freedomof dfW  36  3
 33
You can see that the tabular F -value is 3.285.
.

Since the absolute value of the computed F -value which is 28.7723 is

greater than the absolute value of the tabular F -value which is 3.285, the
decision is to reject the null hypothesis. This shows that there is no significant
difference between the scores of STEM 11 students from BCNSHS when they are
grouped according to section.

23
a.4.2. Two-way ANOVA – compares multiple groups of two factors. It considers
the effect of two factors and the effect of two categorical factors on each other.
It meets the three principles of design of experiments which are replication,
randomization, and local control. In this case, there are 3 null hypotheses that
are being tested.

24
Example: A teacher wants to assess her students based on their scores in General
Mathematics considering their age and sex. The data were presented below.

Research Questions:
1. Is there a significant difference between the scores of students when they are grouped
according to their age group and sex?
2. Is there an interaction between age group and sex?

Null Hypotheses:
1. There is no significant difference between the scores of students when they are
grouped according to sex.
2. There is no significant difference between the scores of students when they are
grouped according to age.
3. There is no interaction (relationship) between age and sex.

Respondent Sex Age Group Score Mean

1 1 1 4 468
2 1 1 6 x 3 6
3 1 1 8
4 1 2 6 669
5 1 2 6 x 3 
6 1 2 9 7
7 1 3 8 8  9 13
8 1 3 9 x 3 
9 1 3 13 10
10 2 1 4 489
11 2 1 8 x 3 
12 2 1 9 7
13 2 2 7 7 10 13
14 2 2 10 x 3  10
15 2 2 13
16 2 3 12 12 14 16
17 2 3 14 x 3 
18 2 3 16 14

Codes for Sex:

Male - 1
Female - 2

Codes for Age Group:

15 years old - 1
16 years old - 2
17 years old - 3

Computation of Mean for Male ( xM )

Add the scores of all males divided by the number of males.

4  6  8  6  6  9  8  9 13 69
xM  9 9
  7.6667

Computation of Mean for Female ( x F )

Add the scores of all females divided by the number of females.

25
 4  8  9  7 10 13 12 14 16 93
xF  9 9  10.3333

Computation of the Grand Mean ( xG )

Add the scores of all students divided by the number of students.

4  6  8  6  6  9  8  9 13  4  8  9  7 10 13 12 14 16 162

xG   9
18 18
Computation of Sum of Squares for Sex ( SSg )

The formula
is SSg
n
 x  n
2
 xG 
x  G
x
1 M 2 F
Where:
n1 is the number of males
is the number of females
xM is the mean score of
males xF is the mean score
of female xG is the grand
mean

Solution: SSg  97.6667  9  910.3333  9

2 2

= 15.9992 15.9992

= 31.9984

Computation of Sum of Squares for Age ( SSa )

The formula
is SSa  n (x  x )  n
2
x x  2
n x x  2

a a G b b G c c G
Where:
na is the number of 15 years old
students nb is the number of 16 years
old students nc is the number of 17
years old students
xa is the mean score of 15 years old
students xb is the mean score of 16
years old students xc is the mean score
26
 of 17 years old students

Compute first the mean scores according to age:

4  6  8  4  8  9 39
xa    6.5
6 6

27
 6  6  9  7 10 13 51
xb    8.5
6 6
8  9 13 12 14 16 72
xc  6

6
 12
Solution:
SSa  6(6.5  9) 2  68.5  9  612  9
2 2

= 37.5 1.5  54  93

Computation Sum of Squares Within

(Error) / SSW

The formula
  x  x
2
is SS 
w

Where:  is summation symbol

x refers to the raw scores
x is the mean by age considering the sex (refer to the table above)

Solution:

Sex Age Score Mean x  x  2

1 1 4 6 4  62  4
1 1 6 6  62  0
1 1 8 8  62  4
1 2 6 7 6  72  1
1 2 6 6  72  1
1 2 9 9  72  4
1 3 8 10 8 102  4
1 3 9 9 102  1
1 3 13 13 102  9
2 1 4 7 4  72  9
2 1 8 8  72  1
2 1 9 9  72  4
2 2 7 10 7 102  9
2 2 10 10 102  0
2 2 13 13 102  9
2 3 12 14 12 142  4
2 3 14 14 142  0
2 3 16 16 142  4
 x  x
2
 68

28
   x  x
2
Thus, SS   68
w

Let us now compute the F - values considering the hypotheses.

a. For the hypothesis, There is no significant difference between the scores of
students when they are grouped according to sex, let us use the sum of squares for
sex.

First, compute the mean square for gender using the formula:

MSg = SS
g
df
g

Where: SSg is the sum of squares for sex

dfg is the degrees of freedom for age/ n 1
MSg is the mean square for sex

Since there are 2 identified sex which are male and female then n  2 .
31.9984
Solution: MS    31.9984
SSg g g 1
df

Second, compute the mean square for sum of squares within (error). The
formula is
SSw
MS w 
df
w

Where: SSw is the sum of squares within

dfw is the degrees of freedom

Since there are 3 males that are 15 years old, 3 males that are 16 years old,
3 males that are 17 years old, 3 females that are 15 years old, 3 females that
are 16 years old, 3
females that are 17 years old, the degrees of freedom is 12 using the formula
n 1 . Subtract 1 from each group considering the age and sex. Thus,
68
MSw   5.6667
12

Last, compute the F -value for sex using the formula,

g
MSg
F  MS
Wh ere:
29
w
MSw
MSg is the mean square for sex
is the mean square for within
MSg 31.9984
F    5.6467
g
MS w 5.6667

30
Solution:

The computed F -value by hand is 5.6467. Using a software application,

the computed F -value is 5.6471. There’s a little difference due to rounding off
numbers.

Using the F -distribution table on page 31 with 1 degree of freedom for

numerator and 12 degrees of freedom for denominator at 5% level of
significance, the tabular F -value is 4.747.

Since the absolute value of the computed F -value which is 5.6467 is

greater than the tabular F -value which is 4.747, reject the null hypothesis.This
indicates that there is a significant difference between the scores of students
when they are grouped according to sex.
b. For the hypothesis, There is no significant difference between the scores of
students when they are grouped according to age, let us use the sum of squares for
age.

First, compute the mean square for age using the formula:

MSa = SSa
df
a

Where: SSa is the sum of squares for age

dfa is the degrees of freedom for age / n 1
MSa is the mean square for age
Since there are 3 identified ages, n  3 which are 15, 16, and 17 years
old.

Solution: MSa 93
=
 46.5
2
Last, compute the F -value for age using the formula,
MS
Fa  MS a
w

Where:
MSa is the mean square for
MSw age is the mean square
for within

MSa
Solution: F  46.5  8.2058

a
MS w 5.6667

The computed F -value by hand is 8.2058. Using the software application,

the computed F -value is 8.2059.
31
 Using the F -distribution table on page 31 with 2 degrees of freedom for
numeratorand 12 degrees of freedom for denominatorat 5% level of significance,
the tabular F -value is 3.885.

Since the absolute value of the computed F -value which is 8.2058 is

greater than the tabular F -value which is 3.885, reject the null hypothesis.
This implies that there is a

32
significant difference between the scores of students when they are grouped
according to age.
c. For the hypothesis, There is no interaction(relationship) between age and sex, we
have several steps to do.
First, compute the total sum of squares by deducting the grand
mean from each score then square the difference. That is,

SSt
 x  x  4  9  6  9  8  9  6  9  6  9  9  9  8  9 
2 2 2 2 2 2 2 2

9  92  13  92  4  92  8  92  9  92  7  92  10  92  13  92 
12  92  14  92  16  92
= 25  9 1 9  9  0 1 0 16  25 1 0  4 116  9  25  49

= 200

Second, compute the sum of squares for both factors (age and
gender). The formula is,
SS  SS  SS  SS  SS 
b t g a w

Where:
SSb is the sum of squares for both factors
SSt is the total sum of
squares SSa is the sum of
squares for age SSw is the
sum of squares within
SSg is the sum of squares for gender

Solution: SSb  200  32  93  68   7

Third, compute the mean square for both factors.The formula is:

MSb = SSb
df
b

Where:
SSb is the sum of squares for both factors
dfb is the degrees of freedom for both factors
MSb is the mean square for both factors

For the degrees of freedom ( dfb ), multiply df g . So,

df a and
dfb  df a (df g )  2(1)  2 .
b
b df
Solution: MS 
SSb

33
 7 2
  3.5

Last, compute the F -value for both factors using the formula,

34
 MSb
F 
b MS w

Wher MSb is the mean square for both

e: factors is the mean square for
MSw
Solution: within

35
 MSb 3.5
F    0.6176
b
MS 5.6667
The computed Fw-value is 0.6176.

Using the F -distribution table on page 31 with 1 degree of freedom

for numerator and 2 degrees of freedom for denominator at 5% level of
significance, the tabular F -value is 3.885.

Since the absolute value of the computed F -value which is 0.6176 is

lesser than the tabular F -value which is 3.885, we failed to reject the null
hypothesis. This means that there is no interaction (relationship) between the
age and the sex.
The summary is presented on the table below.

SS df MS F -computed F -critical Decision

SSg 31.9984 1 31.9984 5.6471 4.745 Reject Ho

SSa 93 2 46.5 8.2058 3.885 Reject Ho

SSw 68 12 5.6667

SSb 7 2 3.5 0.6176 3.885 Retain Ho

SSt 200 17

t-Distribution Table (Two-tailed Test)

36
F -Distribution Table, ALPHA = 0.05

37
38