Figure 4.8. Histogram and Polygon of A ЇBЇ
C. Data Analysis
Before analyzing the data using the two-way variance ANOVA to test hypothesis, the distribution of the sample must be normal and homogeneous. The
following are about the computation and the result of normality and homogeneity test applied to the gained data.
1. Normality Testing
Normality test is aimed to know whether a population is in a normal distribution or not. In this research, Lilliefors test is used to compute the
normality of the data. If L obtained is lower than
L table at the level of signific
ance α = 0.05 on Liliefors, then it can be concluded that the data are in a normal distribution. The summary of Normality test using Lilliefors can
be seen in table 4. 17. The complete computation is in Appendix 10. The
formula used in testing the normality is: where s =
or
or
Table 4.18 The summary of Normality test using Lilliefors
No Variables
Number of Data
α Status
1 Writing Score of the Students
Taught by Using Discovery Learning Method
36 0.134
0.148 0.05
Normal 2
Writing Score of the Students Taught by Using Direct
Instruction Method 36
0.056 0.148
0.05 Normal
3 Writing Score of the Students
having high creativity 36
0.146 0.148
0.05 Normal
4 Writing Score of the Students
having low creativity 36
0.108 0.148
0.05 Normal
5 Writing Score of the Students
having high creativity Taught by Using Discovery Learning
Method 18
0.179 0.200
0.05 Normal
6 Writing Score of the Students
having high creativity Taught by Using Direct Instruction
Method 18
0.083 0.200
0.05 Normal
7 Writing Score of the Students
having low creativity Taught by Using Discovery Learning
Method 18
0.122 0.200
0.05 Normal
8 Writing Score of the Students
having low creativity Taught by Using Direct Instruction
Method 18
0.109 0.200
0.05 Normal
The summary of normality test using Lilliefors in table 4. 17 shows that all of the values of
are lower that . Consequently, it can be concluded that
all of the samples are in normal distribution. 2.
Homogeneity Testing The homogeneity test is done to check whether the data are homogeneous
or not. This test is important as homogeneity of the data shows that the
population is well-formed. In this research, the homogeneity testing is conducted by using Bartlett formula.
The summary of homogeneity testing result is presented in the table 4.18 while for complete computation is provided in Appendix 11.
Table 4.19 The Summary of Homogeneity Test
Sample Df
1df s ²
log s1² dflog
s1² 1
17 0.059
19.87 1.30
22.07 2
17 0.059
30.25 1.48
25.17 3
17 0.059
32.24 1.51
25.64 4
17 0.059
22.47 1.35
22.98 ∑
95.86
Considering the result of the homogeneity test, it shows that the score of
. According to the table of Chi-Square distribution with the
significance level α = 0.05, the value of . Because of
is lower than or
1.36 7.81, it can be drawn the conclusion that the data are homogeneous.
D. Hypothesis Testing