47
From the diagram above, there is an improvement of in the students’ achievement in writing Hortatory Exposition text which is shown by their improvement in each
aspect of writing. Each group had different achievement. The achievement on experimental group is higher than the control group’s.
4.4 Test of Significance
To know the significant difference of the experiment, I used t-test formula. It was done by getting the t-value, I consulted to the critical value on the table column to
check whether the difference was significant or not.
Figure 4.6 The Result of the Control Group’s Mean Scores in Each Writing Aspect
3,25 3,37
3,52 3,51
3,37 3,28
3,5 3,54
3,78 3,67
2,9 3
3,1 3,2
3,3 3,4
3,5 3,6
3,7 3,8
3,9
Grammar Vocabulary
Spelling Content
Fluency Pre test
Post test
48
Best 1981: 271 suggested that for the subject which requires computation such as mathematics or physics the 1 level of significance can be used,
whereas for the psychological and education circle is 5 level of significance can be used as a standard for rejection of a null hypothesis. Since the study is
education consideration, the level of significance is 5. The number of the subject in this study is 63 students of both experiment
and control groups. And the degree of freedom df is 61 which was obtained from the computation formula Nx+Ny-2 = 32+31-2 = 61. At the alpha 5 level of
significant. Since there was no definite critical value in the table, it was necessary to find out the definite value using interpolation.
t-table for 60 = 2.00 120 = 1.98
61 = ...?
1.99 For reading, the t-value of the calculation is 2.02. It is higher than the t-
table that is 1.99. So it can be concluded that the differences is significant. The following was the computation:
0.51 1.46
49
45.75
192
So the t-test computation:
+ + , + + ,
2.02 The mean score of the experiment group is 1.46 and the control group is
0.51 and the difference between the two means is 0.95. The t-test scores showed that it is 2.02. For the t t
table
, for the t= , for =5 with df= 61, the t
table
= 1.99 be obtained.
Since t t
table
it is found that there is difference in the mean score increase between the experimental and control group, where the mean score of the
experimental group increase is higher than that of the control group. The computation showed that t =
1.99.
50
For writing, the t-value of the calculation is 3.66. It is higher than the t- table that is 1.99. So it can be concluded that the differences is significant. The
following was the computation: +
+ +
- .
+
++
So the t-test computation:
+ +
+ + , + + ,
+
51
The mean score of the experiment group is 2.63 and the control group is 0.67 and the difference between the two means is 1.96. The t-test scores showed
that it is 3.66. For the t t
table
, for the t= + , for =5 with df= 61, the t
table
= 1.99
be obtained. Since t t
table
it is found that there is difference in the mean score increase between the experimental and control group, where the mean score of the
experimental group increase is higher than that of the control group. The computation showed that t =
+ 1.99.
4.5 Discussion of the Result