Homegeneity Test Data Analysis of Testing Requirement

was obtained that significance level or probability value of F 0.180 was higher than si gnificance degree α = 0.05. This result meant that variance of data in both the experimental and the control class was homogenous. Because of obtaining homogenous variance data, the researcher had to use t value t = -.729 and df = 50. Based on the result of independent sample t-test for pretest above, the result was gained from the calculation that t count is -.729 whereas t table 0.025 is 2.009 gained based on df = 50 and ⁄ α = 0.025 . Because of obtaining Sig. 2-tailed from the result of independent sample t-test, the degree of significance α = 0.05 had to be divided by 2. It meant that t count -.729 was lower than t table 2.009. Similarly, the value of Sig. 2- tailed was 0.469. It meant that the significance level 0.469 or probability value p was higher than the degree of significance ⁄ α = 0.025. The result of Independent-Sample T Test on pretest p = 0.469. To interpret the result of the independent sample t test of pretest explained above, the resercher formulated the criteria of hypothesis testing which stated that there was no extreme difference in narrative writing skill between the experimental and the control class before the treatment was given; if t count t table or Sig 2-tailed ⁄ α . Meanwhile, if Sig 2-tailed ⁄ α or t count t table, it could be assumed that there was difference in narrative writing skill between the experimental and the control class before the treatment was given. Hence, the result of the independent sample t test of the experimental and control class’ pretest could be concluded that there was no extreme difference in both the experimental and the control class’ narrative writing skill before the treatment was given. In other words, both the classes had equal ability in writing narrative text before the treatment was given. Furthermore, after obtaining the result of Independent-Sample T Test in both the the experimental and the control class’ pretest which proved that both of the classes had no extreme difference in narrative writing skill before the treatment was performed, the last step is to test the hypothesis. The following was the result of the experimental and the control class’ posttest based on the Independent Sample T Test calculation: Table 4.9 The Mean, the Standard Deviation, and Standard Error Mean of the Experimental and Control Class’ Posttest Score Group Statistics Class N Mean Std. Deviation Std. Error Mean Posttest Score Experimental 26 60.865 14.5248 2.8486 Control 26 51.827 11.9699 2.3475 The table 4.9 above showed that the mean score of posttest in the experiment class was 60.87 whereas in the control one was 51.83. It indicated that the mean score of the experimental class’ posttest was higher than the control one after the treatment was given. Table 4.10 The Result of Independent Sample t-Test of the Experimental and Control Posttest Independent Samples Test Levenes Test for Equality of Variances t-test for Equality of Means F Sig. t Df Sig. 2- tailed Mean Difference Std. Error Difference 95 Confidence Interval of the Difference Lower Upper Posttest Score Equal variances assumed .914 .344 2.449 50 .018 9.0385 3.6912 1.6245 16.4525 Equal variances not assumed 2.449 48.239 .018 9.0385 3.6912 1.6177 16.4592