Experimental Class Control Class
No Student
Score No
Students Score
36 36
-5 36
36 -5
37 37
37 37
15 38
38 38
38 -5
39 39
39 39
-15
∑ 185
∑ 165
Mean 4.74
Mean 4.23
On Table 4.2, we can see that highest gain score from the experimental class was 35 and the lowest score was -20. Later on, from the control class the
highest gain score was 40 and the lowest score was -35. This calculation result indicates that some students, both in experimental class and control class,
performed better in their post-test than in their pre-test. But, in the other hand, some students performed worst in their post-test than in their pre-test.
In addition, based on the data on the Table 4.2, the average gain score for experimental class was 4.74, while the mode was 0, and the median was 0.
Meanwhile, the average gain score for experimental class was 4.23, while the mode was 5, and the median was 5.
4. Data Testing
In this research, the test of normality distribution and homogeneity variances taken from students’ pre-test and post-test scores was performed before
the calculation of t
test
value. It was done to determine if the data set was well- modeled by a normal distribution and to compute how likely it is for a random
variable underlying the data set to be normally distributed.
a. Normality of the Pre-test
The test of normality distribution of the pre-test data was analyzed by the use of Kolmogorov-Smirnov test in IBM SPSS Statistics 20. The result of the test
can be seen as follows:
Table 4.4 Normality of the Pre-test
One-Sample Kolmogorov-Smirnov Test
PreScoreExp PreScoreCont
N 39
39 Normal Parameters
a,b
Mean 74.36
69.62 Std. Deviation
9.402 9.893
Most Extreme Differences Absolute
.194 .157
Positive .129
.096 Negative
-.194 -.157
Kolmogorov-Smirnov Z 1.211
.978 Asymp. Sig. 2-tailed
.107 .295
a. Test distribution is Normal. b. Calculated from data.
The normality test above used One-Sample Kolmogorov-Smirnov Test. Table 4.4 shows that the absolute difference D of experimental class data was
0.194. It is smaller than D
table
with the closest Kolmogorov-Smirnov critical points of 40 with degree of significance = 0.05 which is 0.210, in other words, D
experiment
D
table
0.194 0.210. Meanwhile, the absolute difference D of control class data was 0.157, which is smaller than D
table
with the closest Kolmogorov-Smirnov critical points of 40 with degree of significance = 0.05 which is 0.210, in other
words, D
experiment
D
table
0.157 0.210. The table also shows that the Kolmogorov-Smirnov Z
experiment
was recorded at 1.211 which is bigger than Z
table
of 0.05 or p 0.05 and the Kolmogorov-Smirnov Z
control
was recorded at 0.978 which is bigger than Z
table
of 0.05 or p 0.05. It showed that there is no difference between the theoretical distribution and the empirical distribution which means
that the data from the experimental class and control class was normal.
b. Normality of the Post-test
The test of normality distribution of the post-test data was analyzed by the use of Kolmogorov-Smirnov test in IBM SPSS Statistics 20. The result of the test
can be seen as follows:
