Abstract:
The quality of higher education is determined by the performance of the students who have the opportunity to receive higher education. At present, the raw aggregate marks of the G.C.E. (A/L) examination is used as the selection criteria of students for universities.
Whether raw aggregate marks can represent the performance of students at G.C.E. (A/L)examination is a question that is currently debated on.
This study was carried out to investigate the correlation among subject grades, to find out whether mean aggregate differences exist between (a) districts, (b) repeat and first attempt students and (c) male and female. It also investigates the distribution of grades in different subjects.
For this study G.C.E. (A/L) results (Science subjects) of 1995 and 1996 were used.
Univariate Analysis, Chi-square (contingency table analysis), ANOVA and Analysis of
Means procedure were used for analysis of data.
High degrees of correlation among subject grades were observed in the results. Most of the time mean aggregates of different districts were significantly different. No significant
difference was observed in repeat and first attempt students while male to female
difference was greater in Physical Science subjects. Distributions of grades in different
subjects are different to each other. in the same subject stream, some subjects showed left skewed distributions while other subjects showed right skewed distributions. In this context it can be concluded that adding of raw subject marks to form raw aggregate, which is the current practice, as a basis for admission to the universities is not a proper method of selection of students for university education. Some type of scaling for location dispersion of the distribution of marks is needed instead of using the aggregate marks as an indicator of performance.
