1

αJust select the tools that are used in your study

α
α
Statistical Treatment of the Data
The tools that were used by the researcher in analyzing the data were the
following:

Percent. This was used to show how a part is related to a whole. It was used
in presenting the profile and Mathematics performance of the students.
Formula:
part
Percentage= ×100
whole
Where:
part = frequency of students in a particular category
whole = total number of students
Weighted Mean This was used to get the extent of attitude of the students

towards Mathematics.

Formula:

w x̄ =
∑ fx
n
where w x̄ = weighted mean/average
x = rating
n = total number of students
f = frequency/number of students who responded in a
particular category
∑fx = sum of all the products of frequency and the rating
The following scale was also applied:

5-Point Likert’s Scale

Range Verbal Attitude Explanation

 2

Description
4.21 – 5.00 Strongly Very Positive The attitude is 81% - 100% manifested
Agree by the student.
3.41 – 4.20 Agree Positive The attitude is 61% - 80% manifested by
the student.
2.61 – 3.40 Moderately Neutral The attitude is 41% - 60% manifested by
Agree the student.
1.81 – 2.60 Disagree Negative The attitude is 21% - 40% manifested by
the student.
1.00 – 1.80 Strongly Very Negative The attitude is 1% - 20% manifested by
Disagree the student.
4 Point Scale
Legend: 3.25-4.00 - Very Serious

2.50-3.24 - Serious

1.75-2.49 - Partially Serious

1.00-1.74 - Not a Problem

Mean . This will was used in getting the extent of performance of the
students.
Formula:

x̄=
∑x
n
where x̄ = mean/average
x = scores/rating
n = number of students

The proficiency level or academic performance at which the students were

performing was based on the following criteria (DepEd Order No. 8, s 2015).

Rating Verbal *Explanation s

Equivalent
90% and Outstanding The student at this level exceeds the core
requirements in terms of knowledge, skills and
 3

above understanding, and can transfer them automatically

and flexibly through authentic performance tasks.

85% - 89% Very The student at this level has developed the
Satisfactory
fundamental knowledge and skills and core
understandings, and can transfer them
independently through authentic performance tasks.

80% - 84% Satisfactory The student at this level has developed the
fundamental knowledge and skills and core
understandings, and with little guidance from the
teacher and/or with some assistance from peers, and
can transfer these understandings through authentic
performance tasks.

75% - 79% Fairly The student at this level possesses the minimum
Satisfactory
knowledge and skills and core understandings, but
needs help throughout the performance of authentic
tasks.

74% down Did Not Meet The student at this level struggles with his/her
Expectations
understanding; prerequisite and fundamental
knowledge and/or skills have not been acquired or
developed adequately to aid understanding.

*The explanation is based on DepEd Order No. 73, s. 2012

Mann Whitney U Test. This was used to test the significant difference in the
Perceptions of the Grade III Teachers and the School Heads on the Extent of
Implementation of the Learning Action Cell (LAC) sessions
 4

Formula:

Where :
U1 = Mann Whitney U value of sample 1
U2 = Mann Whitney U value of sample 2
n1 = sample size for sample 1
n2 = sample size for sample 2
R1 = sum of the ranks in sample 1
R2 = sum of the ranks in sample 2

Pearson Product Moment Coefficient of Correlation This is used in identifying

the relationship between the readiness of the pupils and their performance.

N ( ∑ xy )−(∑ x)(∑ y )
r=
√¿ ¿ √ ¿ ¿

where x is the extent of readiness of the pupils

y is the academic performance of the pupils
r is the coefficient of correlation

Spearman Rank Correlation Coefficient (rs). This was used to determine

the relationship between the two variables (a. extent of attitude of the students and
 5

their performance in Mathematics; b. profile (year level) of the students and their

attitude towards Mathematics).

Formula:

where r s = Spearman rho

d2 = sum of the squared differences between ranks

n = total number of students

Legend: Value of r Strength of Relationship (Statistical Correlation, 2009)
Between ± 0.50 to ± 1.00 ± strong relationship
Between ± 0.30 to ± 0.49 ± moderate relationship
Between ± 0.10 to ± 0.29 ± weak relationship
Between ± 0.01 to ± 0.09 ± very weak relationship

Source:
Statistical Correlation (2009). https://explorable.com/statistical-correlation
Explorable.com (May 2, 2009). Statistical Correlation. Retrieved Oct 25, 2016 from
Explorable.com: https://explorable.com/statistical-correlation

Point-Biserial Correlation (rpbi).This will be used to measure the

relationship between two variables when one or both of the variables are

dichotomous. A dichotomous variable is one for which there are exactly two

categories (Gravetter and Wallnay 252). The calculation of the relationship precedes

as follows:

1. Assign numerical value to the two categories of the dichotomous variable. In this

research, male students were assigned a value of 1 and female students were

assigned a value of 2.

2. Use the regular correlation formula to calculate the relationship.

 6

Chi-square. To determine the relationship between the profile of the

respondents and their extent of readiness of the pupils.

Formula:

2 ( fo−fe ) 2
x =∑
fe

Where:

x2 = chi- square

fo = observed frequency

fe= expected frequency

Pearson’s Contingency Coefficient (c). This was used in identifying the degree

or strength of association between the two variables.

Formula:

where: x2 = chi- square value

N = grand total of observations

Describing Strength of Association

Legend: Scale Strength of Association

> 0.5        High Association
0.3 - 0.5         Moderate Association
 7

0.1 < 0.3        Low Association

0.0 < 0.1        very Low
Source: http://www.acastat.com/Statbook/chisqassoc.htm)

Legend: Value of c Strength of Association

> 0.5         High Association

0.30 - 0.50          Moderate Association

0.10 - 0.29        Low Association

0.00 - 0.09        Very Low

Source: http://www.acastat.com/statbook/chisqassoc.htm

To identify the strength of association between the variables, the researcher applied the

following description from Schubert and Leimstoll (2007):

Contingency Coefficient (CC) Strength of Association

0.3< CC - Strong Association

0.2< CC < 0.3 - Medium Association

0.1< CC < 0.2 - Weak Association

0.0 < CC < 0.1 - Not appreciable Association

To be places sa References:

“Coefficient of Measuring Association”. 2015.

http://www.acastat.com/statbook/chisqassoc.htm

4-Point Likert’s Scale

Scal Range Verbal Interpretatio Explanation

e Description n
4 3.25-4.00 Always High Exposed to media
everyday
3 2.50-3.24 Sometimes Moderate Exposed to media 4-6 days
 8

a week
2 1.75-2.49 Seldom Low Exposed to media 1-3 days
a week
1 1.00-1.74 Never Very Low Not exposed to media at
all..
Analysis of variance. This will be used in identifying the significant

difference among the variables. This will be utilized since the data are in ratio scale.

F = MSSb/MSSw

Note:

TSS = total sum of squares

SSb = between-column sum of squares

SSw = within-column sum of squares

MSSb = mean sum of squares between- column

MSSw = mean sum of squares within-column

N = total number of items or entries

F = F test

Fisher's exact test of

independence
Summary
Use the Fisher's exact test of independence when you have two
nominal variables and you want to see whether the proportions of
one variable are different depending on the value of the other
variable. Use it when the sample size is small.

When to use it
Use Fisher's exact test when you have two nominal variables.
You want to know whether the proportions for one variable are
 9

different among values of the other variable. For example, van

Nood et al. (2013) studied patients with Clostridium
difficile infections, which cause persistent diarrhea. One nominal
variable was the treatment: some patients were given the antibiotic
vancomycin, and some patients were given a fecal transplant. The
other nominal variable was outcome: each patient was either cured
or not cured. The percentage of people who received one fecal
transplant and were cured (13 out of 16, or 81%) is higher than the
percentage of people who received vancomycin and were cured (4
out of 13, or 31%), which seems promising, but the sample sizes
seem kind of small. Fisher's exact test will tell you whether this
difference between 81 and 31% is statistically significant.
A data set like this is often called an "R×C table," where R is the
number of rows and C is the number of columns. The fecal-
transplant vs. vancomycin data I'm using as an example is a 2×2
table. van Nood et al. (2013) actually had a third treatment, 13
people given vancomycin plus a bowel lavage, making the total data
set a 2×3 table (or a 3×2 table; it doesn't matter which variable you
call the rows and which the columns). The most common use of
Fisher's exact test is for 2×2 tables, so that's mostly what I'll
describe here.
Fisher's exact test is more accurate than the chi-square
test or G–test of independence when the expected numbers are
small. I recommend you use Fisher's exact test when the total
sample size is less than 1000, and use the chi-square or G–test for
larger sample sizes. See the web page on small sample sizes for
further discussion of what it means to be "small".

Null hypothesis
The null hypothesis is that the relative proportions of one
variable are independent of the second variable; in other words, the
proportions at one variable are the same for different values of the
second variable. In the C. difficile example, the null hypothesis is
that the probability of getting cured is the same whether you
receive a fecal transplant or vancomycin.

How the test works

 10

Unlike most statistical tests, Fisher's exact test does not use a
mathematical function that estimates the probability of a value of a
test statistic; instead, you calculate the probability of getting the
observed data, and all data sets with more extreme deviations,
under the null hypothesis that the proportions are the same. For
the C. difficile experiment, there are 3 sick and 13 cured fecal-
transplant patients, and 9 sick and 4 cured vancomycin patients.
Given that there are 16 total fecal-transplant patients, 13 total
vancomycin patients, and 12 total sick patients, you can use the
"hypergeometric distribution" (please don't ask me to explain it) to
calculate the probability of getting these numbers:
fecal vancomycin

sick 3 9
cured 13 4
P of these exact numbers: 0.00772

Next you calculate the probability of more extreme ways of

distributing the 12 sick people:
fecal vancomycin

sick 2 10
cured 14 3
P of these exact numbers: 0.000661

fecal vancomycin

sick 1 11
cured 15 2
P of these exact numbers: 0.0000240

fecal vancomycin

sick 0 12
cured 16 1
P of these exact numbers: 0.000000251
 11

To calculate the probability of 3, 2, 1, or 0 sick people in the fecal-

transplant group, you add the four probabilities together to
get P=0.00840. This is the one-tailed P value, which is hardly ever
what you want. In our example experiment, you would use a one-
tailed test only if you decided, before doing the experiment, that
you were only interested in a result that had fecal transplants being
better than vancomycin, not if fecal transplants were worse; in
other words, you decided ahead of time that your null hypothesis
was that the proportion of sick fecal transplant people was the
same as, or greater than, sick vancomycin people. Ruxton and
Neuhauser (2010) surveyed articles in the journal Behavioral
Ecology and Sociobiology and found several that reported the
results of one-tailed Fisher's exact tests, even though two-tailed
would have been more appropriate. Apparently some statistics
textbooks and programs perpetuate confusion about one-tailed vs.
two-tailed Fisher's tests. You should almost always use a two-tailed
test, unless you have a very good reason.
For the usual two-tailed test, you also calculate the probability of
getting deviations as extreme as the observed, but in the opposite
direction. This raises the issue of how to measure "extremeness."
There are several different techniques, but the most common is to
add together the probabilities of all combinations that have lower
probabilities than that of the observed data. Martín Andrés and
Herranz Tejedor (1995) did some computer simulations that show
that this is the best technique, and it's the technique used by SAS
and most of the web pages I've seen. For our fecal example, the
extreme deviations in the opposite direction are those
with P<0.00772, which are the tables with 0 or 1 sick vancomycin
people. These tables have P=0.000035 and P=0.00109,
respectively. Adding these to the one-tailed P value (P=0.00840)
gives you the two-tailed P value, P=0.00953.

Post-hoc tests
When analyzing a table with more than two rows or columns, a
significant result will tell you that there is something interesting
going on, but you will probably want to test the data in more detail.
For example, Fredericks (2012) wanted to know whether checking
termite monitoring stations frequently would scare termites away
 12

and make it harder to detect termites. He checked the stations

(small bits of wood in plastic tubes, placed in the ground near
termite colonies) either every day, every week, every month, or just
once at the end of the three-month study, and recorded how many
had termite damage by the end of the study:

Percent
Termite termite
damage No termites damage

Daily 1 24 4%

Weekly 5 20 20%

Monthly 14 11 56%

Quarterly 11 14 44%

The overall P value for this is P=0.00012, so it is highly significant;

the frequency of disturbance is affecting the presence of termites.
That's nice to know, but you'd probably want to ask additional
questions, such as whether the difference between daily and weekly
was significant, or the difference between weekly and monthly. You
could do a 2×2 Fisher's exact test for each of these pairwise
comparisons, but there are 6 possible pairs, so you need to correct
for the multiple comparisons. One way to do this is with a
modification of the Bonferroni-corrected pairwise technique
suggested by MacDonald and Gardner (2000), substituting Fisher's
exact test for the chi-square test they used. You do a Fisher's exact
test on each of the 6 possible pairwise comparisons (daily vs.
weekly, daily vs. monthly, etc.), then apply the Bonferroni
correction for multiple tests. With 6 pairwise comparisons,
the P value must be less than 0.05/6, or 0.008, to be significant at
 13

the P<0.05 level. Two comparisons (daily vs. monthly and daily vs.
quarterly) are therefore significant

P value

Daily vs. weekly 0.189

Daily vs. monthly 0.00010

Daily vs. quarterly 0.0019

Weekly vs. monthly 0.019

Weekly vs. quarterly 0.128

Monthly vs. quarterly 0.57

You could have decided, before doing the experiment, that

testing all possible pairs would make it too hard to find a
significant difference, so instead you would just test each treatment
vs. quarterly. This would mean there were only 3 possible pairs, so
each pairwise P value would have to be less than 0.05/3, or 0.017,
to be significant. That would give you more power, but it would also
mean that you couldn't change your mind after you saw the data
and decide to compare daily vs. monthly.

UNDERPINING THEOREIS
1- Any background work that has been done in the field that will support your research and
thesis.
2- During your frame work, you need to support every relationship between each IV and DV
with the help of a theory.
 14

3- One theory at a time can be used to support more than one relationship
4- Three is enough and four is maximum to support the theoretical background.
For example for the relationship of leverage and firm performance, we can use Pecking order
theory, trade off theory, capital structure theory which are related to the relationship of
leverage and firm performance. While we should always remember the opposing theory also,
it may be asked in viva during presentation, as in the above example the Miller & Modi
theory is the opposing theory state that there is no such relationship.