Overall Student Performance Depends on His Performance in Each of the Unit Courses

Overall student performance was measured in the research by taking the average of a student’ s unit courses. It is helpful to find out in this investigation which unit course affects overall performance most. Table 3 shows the correlation coefficients, Pearson’ s r, between the unit courses and overall performance. The figures show that overall performance is most highly and positively correlated with BMAN10632(M) (10) - Fundamentals of Management Accounting, r = . 862, p < .01. This means that 86.2% of the variances in overall performance may be significantly explained by a student’ s BMAN10632 scores.

The other units that are highly correlated with overall performance are BMAN10522(M) (10) - Financial Decision Making (M), BMAN10732 (10) - Quantitative Methods for Bus & Mgmt 2, and BMAN10002 (10) - Economic Principles: Macroeconomics, with coefficients of r = . 833, p < .01, r = . 818, p< .01, and r = . 807, p < .01, respectively. On the other hand, the units that have the lowest correlation with overall performance are BMAN10780 (10) - Business & Management Skills and BMAN10791 (10) - People and Organisations, with coefficients of r = . 473, p < .01, and r = . 511, p < .01, respectively.

Units that have a low correlation with overall performance are those that do not particularly dictate a student’ s overall performance scores. It would seem that the unit courses that are highly correlated with overall performance are quantitative in nature. Thus, we will investigate later whether a previous math education, which may indicate an inclination towards higher quantitative familiarity, has any effect on a student’ s overall performance. It is also helpful to look into the relationships of unit courses with each other.

That is, which unit courses seem to be highly correlated with each other? Another importance of this portion of the investigation is that it allows us to identify and eventually eliminate courses that are highly correlated with each other when the data are subjected to regression analysis. This is because one requirement of effective regression analysis is that homoscedasticity must be avoided.

Close ✕