The Descriptive Statistic of Pre-test Result

48 experimental group was quite varied. For the complete scores of experimental group’s post-test, please refer to APPENDIX 11. Figure 4.4. The Distribution of Post-test Result of Experimental Group The graphic shows positively skewed distribution of the post-test result, as the distribution was concentrated on the left side in the figure. As seen above, the highest frequency was 70. As the control group functioned as comparison, no treatment was given. Table 4.5. and figure 4.5. show the descriptive statistic of pre-tests of control group. 49 Table 4.5. The Distribution Table of Post-test Result of Control Group Posttest Frequency Valid 47.00 2 50.00 4 53.00 6 56.00 7 59.00 2 63.00 5 65.00 1 66.00 7 69.00 2 72.00 3 75.00 1 78.00 4 81.00 1 84.00 2 88.00 1 91.00 2 94.00 1 97.00 1 Total 52 The information in table 4.5. presents the distribution of the post test results of the control group. There was bimodal : 56 and 66. The mean score for the post-test was 65.67, and the standard deviation was 13.08. The difference from the highest to lowest score was 50 points, almost five times standard deviation, which ranges from 47 to 97. This reflects that the score distribution of the post test of the control group was quite varied. For the complete score of control group’s post-test, please refer to APPENDIX 12. Statistics Po sttest N Valid 52 Missing Mean 65.6731 Std. Deviat ion 13.08602 50 Figure 4.5. The Distribution of Post-test Result of Control Group Figure 4.5. shows positively skewed distribution of the post-test results, as the distribution was concentrated on the left side in the figure. As seen above, the highest frequency belonged to the score of 50 and 70. The difference found in the mean score of the pre-tests and post-tests of the experimental and control group did not automatically call for significant difference. Therefore, to ensure that the difference was significant, the researcher performed a t-test, that will be discussed in the next topic. 51

4.2 Data Analysis

There was sufficient evidence to reject the null hypothesis which says that the mean score of post test was lower than the mean score of pre test. Therefore, the alternative hypothesis was used for this research. The mean score of post-test was higher than the mean score of pre-test. The difference between two mean score was significant t25 = 3.02, p .05. This finding became the evidence to answer the first research question at the beginning of this research; GAIL significantly improves SMA students’ writing skills. The results of both pre-test and post-test scores from experimental and control group were calculated by using SPSS ver.20. The aim was to observe the significance of the mean gain difference by applying an independent t-test Balnaves Caputi, 2001. Below is the result of the t-test of experimental group: Table 4.6. The T-test Result of Experimental Group One-Sample Statistics N Mean Std. Deviation Std. Error Mean change 53 5,1698 12,44949 1,71007 One-Sample Test Test Value = 0 T Df Sig. 1-tailed Mean Difference 95 Confidence Interval of the Difference Lower Upper Change 3,023 52 ,002 5,16981 1,7383 8,6013 The t-test results of the pre-test and the post-test of the experimental group showed M= 1.71, SD=12.44, t 52 = 3.02, p 0.05. There was a significant difference between the pre-test and post-test of the experimental group.