Abstract:
The common credit and qualification framework for Sri Lankan University system proposed by the Quality Assurance
and Accreditation Council is to maintain the consistency and comparability of university level qualifications,
and to promote student mobility by creating more flexible arrangements for students learning. This
method is based on Cumulative Grade Point Average (CGPA) of marks obtained by students. To minimize drastic
variations, all degree programmes were categorized into six and a separate framework has been developed
for each. Considering the complexity associated with handling data, this study only focuses the framework developed
for Commerce and Management degree programmes. The negative impact of the CGPA method on performance
evaluation, especially in determining class or pass of the degree, is investigated and analysed through
this research. This study also makes important suggestions to overcome the drawbacks of the new method. This
study is based on statistically generated data due to practical impossibility of obtaining actual results data which
are highly confidential. Therefore, marks of students are generated from a multivariate normal distribution to
accommodate the correlation and normality of marks of a single student. Marks of thousand students for fourty
subjects are generated using MINITAB 14 and converted into grade points. Final results (class/pass) intended
under both CGPA (new) and final average (old) methods are then cross-tabulated and compared. The research
discovers that the proposed method of evaluation is inefficient, as it always underestimates performance of students,
relative to the existing method. Through an in-depth analysis of intended results under both methods, cutoff
marks for the CGPA method were redefined for Commerce and Management degree programmes, so that it
will not make any significant deviation from the existing method of determining class/pass of students.
