how many numbers true and how many number false. Then, based on that numbers, the writer gives the score in range 0-100 point.
2. Data analysis in pilot test
The pilot test aims to check validity and reliability of the instrument. It was conducted before doing pretest and post-test. If the respondents were able
to understand the given instruction it was concluded that instrument can be used as pretest and post-test.
3. Data analysis on pretest and post-test
Pretest and post-test were given to both experimental and controlled groups in the same procedures. A hypothesis was started with the alpha level
at 0.05. The data gathered through pretest and post-test computed one by one by Using SPSS 20.0 for windows. three steps were accomplished covering
normality test, homogeneity test, and independent t-test
1
. The details of statistical procedures as follows:
a. Normal Distribution Test
Normal distribution test was calculated before the t-test. It aims to investigate whether or not the distribution of pretest and post-test both of two
groups were normally distributed. The statistical calculation of normality test
used Kolmogorov-Smirnov by following three steps below:
1 Setting the level of significance p at 0.05 and establishing the hypothesis
as follows: H
: the variances of experimental and controlled group are normally distributed.
H
1
: the variances of experimental and controlled group are not normally distributed.
2 Analyzing the normality distribution with Kolmogorov-Smirnov test.
Comparing the asymp.sig 0.05, the null hypothesis is not rejected and alternative hypothesis is rejected, and the distribution of data is normal.
Hence, if the asymp.sig 0.05, the null hypothesis is rejected and alternative hypothesis is not rejected, and it means the data is not normally
distributed.
Table 3.1 Normality test of pre-test
Kolmogorov-Smirnov Shapiro-Wilk
Statistic Df
Sig. Statistic
Df Sig.
EXPERIMENTAL GROUP CONTROL GROUP
.148 .119
24 24
.162 .200
.949 .949
24 24
.239 .234
The normality test above used Kolmogorov-Smirnov critical points table in determining the ttable. According to the consideration of n=45 and significance
level 0.05, the ttable of this normality test was 0.198. Furthermore, table showed that the t value of experimental class was 0.115 and the tvalue of controlled. Class
was 0.128. The values were smaller than ttable, in other words, tvalue ttable 0.1150.198 for experimental class, and 0.1280.198 for controlled class.
Therefore, Ho was accepted and Hi was rejected. It meant that the distribution of the data was normal. Also, the significance of experimental class was 0.184 and
the significance of controlled class was 0.070. The result shown that the significance of both classes are above 0.05. Therefore, the distribution of posttest
score was normal.
Table 3.2 Normality test of post-test
Kolmogorov-Smirnov Shapiro-Wilk
Statistic Df
Sig. Statistic
df Sig.
VIIIA VIIIB
.159 .149
24 24
.102 .160
.936 .936
24 24
.122 .117
