the participants who were taught writing through integrating it with reading comprehension and those who received treatment as integrating speaking and writing. As it is presented in Table 4.1.both groups involved 30 participants. The mean score of the participants in reading-writing integration class is more than those in speaking-writing integration class. For comparison of the findings presented by both groups for the same test, the two independent samples were subjected to t-test. Nonetheless, since this technique works with a normally distributed sample feature, it was crucial to undertake a study to determine whether or not the distributions were normal.
If the distributions were found not to be normal, then, nonparametric tests would have been used to compare the samples (Majzub & Abu, 2010). However, the analysis indicated the viability of applying t-test technique in interpreting the test that was administered to both groups. Observations ought to be independent, implying a lack of relationship between the groups or observations. This is an issue of study design more than it could be tested for. Nonetheless, it remains a crucial assumption with regards to independent samples t-test.
No significant outliers should exist. Outliers refer to single data points in the data which defy the usual pattern. Outliers present the challenge of negatively affecting the independent t-test thus jeopardising the accuracy of the results. From Table 4.2, it would be observed that the two samples have an arithmetic average of values of n, referred to as the mean, of 76.63 and 61.13 respectively. Higher values are posted by the group involved in writing integrated with reading as compared to the other group.
Appreciating the requirement for group normality for t-tests with regards to independent samples, both groups underwent the Kolmogrov-Smirnov normality test to determine whether or not the sample was normal. Table 4.3
...Download file to see next pages Read More