From the table and the diagram above, it presented some information. The numbers of the class interval K were 6 and the length of the interval I was 4.
Then, the lowest score was 65 and the highest one was 88. Its diagram is taken from 17 students‟ scores.
The most frequency appearance on distribution was the score between 69- 72 with the numbers of absolute frequency 5 or 29.41 in relative frequency.
While, the least appearance were interval 81-84 and 85-88, the frequencies were 1 and 1 with relative frequency 5.88.
5. Normality Test
In gaining normaility test, the writer used SPSS software. The result of normality test was presented in table 4.7 and 4.8. Some steps in gaining normaility
test by using SPSS as follows
1
: a
Open the data b
Select analyze c
Select descriptive statistic d
Choose the variable will be analyzed and put into variable colomn e
Select option f
Check list skewness and kurtosis g
Select continue h
Select Ok a.
Normality Test of Pre-test Table 4.7. The Result of Normality Test of the Pre-test
Descriptive Statistics
Class N
Skewness Kurtosis
Statistic Statistic Std. Error Statistic Std. Error Experimental
17 -.140
.550 .953
1.063 Controlled
17 -.581
.550 -.829
1.063
1
Budi Susetyo, Statistika Untuk Analisis Data Penelitian, Bandung: Refika Aditama, 2010, p.272.
Valid N listwise
17
After analyzing the data, the result of normality calculation showed by skewness that the highest L
value
of pre-test of the experimental class was -0.140. While the highest L
value
of pre-test of the controlled class was -0.581. The test with high normality distribution if the score between 0-1. Hence, based on the
output of SPSS it was showed that both of the pre-test of the experimental class and pre-
test of controlled class‟s score not more than 1, it meant that the samples of both classes came from normal distributed population.
b. Normality Test of Post-Test
Table 4.8. The Result of Normality Test of the Post-Test Descriptive Statistics
N Skewness
Kurtosis Statistic Statistic Std. Error Statistic Std. Error
Experimental 17
.768 .550
1.041 1.063
Controlled 17
.357 .550
-.615 1.063
Valid N listwise
17
From the table we knew that the samples of both classes came from normal distributed population. The statistic score of skewness in experimental
class was 0.768 and the statistic score of skewness in controlled class was 0.357. Both of them are not more than 1, we can conclude they have normal distribution
of population.
6. Homogeneity Test
An experimental research involves two classes, the experimental and the controlled class. Before the treatment is given, both classes must have equal
starting points. If such condition is fulfilled, the two classes can be said homogeneous and the experiment can be valid.
To measure if the two classes are homogeneous, the approach commonly used is by comparing the variances of both classes. The statistic hypothesis as
follows: a
H :
1 2
=
2 2
= There is no difference between the variance of the experiment and control class.
b H
i
:
1 2
≠
2 2
= There is any difference between the variance of the experiment and control class.
Homogeneity was tested by using SPSS, some steps by using SPSS as follows
2
: a
Select Analyze b
Select Compare Means c
Select One-Way ANOVA d
Fulfill the independent and factor colomn with the score of pre-test and post-test in experimental and controlled class
Table 4.9 The Result of Homogeneity Test of the Pre-Test ANOVA
Pre Test Sum of
Squares df
Mean Square F
Sig.
Between Groups 393.275
10 39.327
1.010 .519
Within Groups 233.667
6 38.944
Total 626.941
16 From the table, it showed that the degree of freedom was 16, and F
value
was 1.010. By determining the significance level and the quantifiers with significance
level α 5, moreover, the F
table
was 2.110. To achieve the conclusion whether
2
Ibid., p. 296.