Table 4.9 above showed that the total number of post-test data in control class was 43. The highest score of post-test in control class was 88 and the lowest score
was 56. The total score was 3168. The mean score of post-test score in control class was 73.67. The standard deviation is 7.419.
According to the table above, it can be formed a table of frequency distribution which us presented as follows:
Table 4.10 Frequency Distribution of Post-test Result of Control Class
Frequency Valid
56 1
60 3
64 1
68 7
72 10
76 10
80 5
84 4
88 2
Total 43
From Table 4.10, it can be known that in post-test result in control class the student who got 56 was 1 student, the students who got 60 were 3 students, the
student who got 64 was 1 students, and so on until the total number of frequency were 43 data.
Based on the table of frequency distribution above, it also can be presented in a diagram as follows:
Diagram 4.4 Post-test Result of Control Class
From the diagram of post-test in control class above, it can be seen the most frequent score is 72 and 76 which was got by 10 students. The second frequent
score is 68 which were got by 7 students. The less frequent score is 56 and 64 which was got only by 1 student. The highest score is 88 which were got by 2
students. From result of pre-test and post-test score in control class above showed that there is difference on students’ reading comprehension achievement between
pre-test and post-test achievement although as not significant as post-test in experimental class.
B. Analysis of the Data
1. Normality of the Data
Before analyzing the hypothesis. the writer had to analyze the normality of the data. This analysis is to measure that the data got in the research was normally
distributed or not. The writer used SPSS v.22 for windows with criteria α 0.05.
The result of normality can be seen by somparing the value of T
max
to T
table
. The criteria of hypothesis is:
H
1
: T T
table
H
O
: T ≥ T
table
2 4
6 8
10 12
56 60
64 68
72 76
80 84
88
a. Normality of Pre-test 1. Normality of Pre-test in Experimental Class
Hypothesis: H
: Data of X is normally distributed H
1
: Data of X is not normally distributed.
Table 4.11 Normality Pre-test Results of Experiment Class
Shapiro-Wilk Statistic
Df Sig.
Pre-test Experimental Class
,966 43
,225
From Table 4.11 above, it can be seen that the significance of pre-test score in experimental class based on Shapiro-Wilk was 0.225. If the data is
higher in a significance α = 0.05 it means that data was normal distributed
hence it can be concluded that the data is normally distributed because 0.225 is higher than 0.05 0.2250.05.
2. Normality of Pre-test in Control Class Hypothesis:
H : Data of Y is normally distributed
H
1
: Data of Y is not normally distributed
Table 4.12 Normality Pre-test Results of Control Class
Shapiro-Wilk Statistic
Df Sig.
Pre-test Control Class
,967 43
,253 a. Lilliefors Significance Correction
From Table 4.12, it can be seen that the significance of pre-test score in control class based on Shapiro-Wilk was 0.253. It can be
concluded that the data is normally distributed because 0.253 0.05 or 0.253 is higher than 0.05.
b. Normality of Post-test 1. Normality of Post-test in Experimental Class
Hypothesis: H
: Data of X is normally distributed H
1
: Data of X is not normally distributed.
Table 4.13 Normality Post-test Results of Experiment Class
Shapiro-Wilk Statistic
Df Sig.
Post-test Experimental Class
,965 43
,213
From Table 4.13, it can be seen that the significance of post-test score in experimental class based on Shapiro-Wilk in Liliefors
Significance Correction was 0.213. It can be concluded that the data is normally distributed because 0.213 0.05.
2. Normality of Post-test in Control Class Hypothesis:
H : Data of Y is normally distributed
H
1
: Data of Y is not normally distributed
Table 4.14 Normality Post-test Results of Control Class
Shapiro-Wilk Statistic
df Sig.
Post-test Control Class
,962 43
,160
From Table 4.14, it can be seen that the significance of post-test score in control class based on Shapiro-Wilk was 0.160. It can be
concluded that the data is normally distributed because 0.160 0.05.
2. Homogeneity of the Data
a. Pre-test Homogeneity Test Based on the calculation of normality, the writer got the result that all
data in pre-test and post-test of both experiment class and control class have been normally distributed. The next step of the calculation was finding the
homogeneity of the data. The purpose of this calculation was to see whether the data in both classes were homogenous or heterogeneous. The writer used
SPSS v.22 to find the homogeneity of the data by looking at the significant of the data. If it is higher than 0.05 it means that the data is homogeneous.
Table 4.15 Homogeneity of Pre-test Results between Experimental and Control Class
Levene Statistic df1
df2 Sig.
,085 1
84 ,772
Table 4.15 showed that the significance of pre-test score between experimental class and control class 0.772. Therefore, it can be inferred that
the pre-test data of both classes were homogenous since 0.772 is higher than 0.05 or 0.772 0.05.
b. Post-test Homogeneity Test After analyzing the homogeneity of pre-test class of experimental class
and control class, then, the writer looked for the homogeneity of post-test class of experimental class and control class by using SPSS v.22. The result
of post-test homogeneity test was described in a table as follows:
Table 4.16 Homogeneity of Post-test Results between Experimental and Control Class
Levene Statistic df1
df2 Sig.
,597 1
84 ,442