STestMAP as an alternative for significance test selection / Tang Howe Eng... [et al.]

One major problem faced by many researchers and students when confronted with the hypothesis testing procedure is to select the relevant statistical test for the purpose of calculating the test statistics. This study aims to determine the effectiveness of STestMAP as a tool to help in the selection of the significance test. It also aims at investigating the flexibility afforded by the STestMAP procedure in catering to the needs of both linear and non-linear users. The research design divides the study into two phases: Phase I where quantitative analysis dominates while Phase II involves qualitative analysis. Four instruments, which consisted of prior knowledge test, pre-test, post-test and interview, were adopted and administered. Data collection was carried out through paper-and-pencil tests and interviews. Both quantitative (descriptive and inferential analyses) and qualitative data analysis techniques were utilized. In Malaysia, a random sample of 49 respondents from two public universities in Sarawak, Malaysia was chosen out of which 6 respondents were purposively selected and interviewed. The Wilcoxon Signed Rank Test showed that there was a statistically significant difference in the pre-test and the post-test for the control group, Z=-3.978, p=0.001 as well as for the experimental group, Z=-4.275, p=0.001. In comparing the gains between the experimental and the control groups, however, the test indicated insignificant difference in the gain scores, Z=-0.142, p=0.887. In Singapore, 29 respondents were selected on a voluntary basis from one of the public universities. Eight respondents were purposively selected and interviewed. Paired t- Test showed that there was significant gain in the pre and post-tests (t=7.678,p=0.001). For both Malaysia and Singapore, a more detailed analysis with programme of study and Cumulative Grade Point Average (CGPA) as independent variables.