The PTF report (Page 2-3 in the Higher Ed report) describes the reasons for the delays in university admissions - the UGC has to wait for A/L recorrections, validation of results, they work peaks at a certain time resulting in extreme workloads, they have to reprocess vacancies arising from non-registration, etc.
They recommend that only the latter (vacancies arising from non-registration) be delegated to universities.
In my view, as the number of universities and programs increase, centralised processing of admissions will not only require more work and more time, but the gap between what each program of study requires and what is assessed by A/Ls will increase too.
My recommendation is that
-
The entire work of admissions be delegated to each university.
-
The admission should be based on assessments unique to each program of study (e.g. there will be an admission test for Engineering that is different to the admission test for Physical Science). Each university is responsible for their admission test but probably will collaborate with others (e.g. all Engineering Faculties can have a common admission test). This will allow for a better “fit” in determining if the student is capable of the requirements of the course.
-
A/L scores / z-scores should not be considered at all - all that is required is that they have passed the necessary subject in the A/L. In case they have failed, but have applied for re-scrutiny, they can still sit the university’s admission test; if they pass on the re-scrutiny they can enrol, if not they will have to repeat A/L. This will eliminate the delay arising from re-scrutiny of A/L, and also reduce the hyper-competitiveness of the A/L exams. Of course, it can be possible that candidates want their admission test results rescrutinized, but as these are smaller batches it will take less time. It is noteworthy that the “Kannangara Report” was inclined to have separate school graduation and university entrance exams, but decided to combine them as they felt it was not viable to have two exams at around the same time. But that was at a time when we had 1 university with a few campuses; in the current environment the benefits of separating them outweigh the benefits
-
If there is a need to correct for some schools being more disadvantaged than others (which should then be based on the level of each school rather than the district, as there is considerable variation within a district), it can be modelled as an adjustment to the passmark for A/L (e.g. a poor district/school might have a handicap score of 10 (added to the subject marks) whereas an elite district/school might have a handicap of 0), and/or to the score obtained in the university’s assessment test.
-
With this model, the year the student took the A/L is irrelevant. i.e. if they do not score well enough in the university’s admission test to gain a place, they can try again for the next intake (or even try for other programs / other universities) using their previous A/L results.
-
This will also eliminate the need to standardise across multiple A/L syllabi (a problem that occurs whenever there is a syllabus change and a university intake has to consider students who sat the exam under new syllabus and repeated under old syllabus); all that matters is that they passed the exam in whatever syllabus they sat it in.
-
Another benefit of this approach is that there will be much less incentive for leaking A/L papers, as the “high stake” role shifts to each university’s exam - as this occurs at smaller scales, it will be easier to prevent leaking, and it will be much harder for exam-prep tutors to “guess” questions (this industry will nevertheless adapt and try to predict questions for in-demand courses like engineering and medicine)
-
While there are benefits in holding these examinations at approximately the same time, this is not strictly required - and as students could strive for different programs and sit different tests for them, some degree of spreading would be desirable. Some of them can even be held before A/L results are released.