Archives

by Professor R. P. Gunawardane

It appears that there is a confusion with regard to the basis of university admissions this year. Various organizations, including University student bodies are making erroneous statements leading to further confusion in the minds of general public. This is mainly due to the fact that there is no proper public awareness programme launched to explain this new procedure.

In the circumstances, I have been compelled to write this article to present the background and basic facts regarding the Z Score scheme in ranking candidates for university admission. In this article I would be making an attempt to present Z Score method in a qualitative manner without using mathematical expressions.

At the outset it must be stated that the statistical procedure called Z Score method is used in ranking the students in each stream for the purpose of university admissions this year. This ranking will be done in respect of each of the four streams - Biological Science, Physical Science, Arts and Commerce all island as well as district-wise. The Department of Examinations have already released individual Z Scores and island-wide and district ranking for all candidates, who are qualified for University admission.

In short what has been done is to employ widely accepted Z Score instead of the aggregate of marks or average in ranking students in each stream. It must also be stated categorically that there is no change in the application of district quota system as outlined in the current university admission policy. As such there is absolutely no truth in the claim that district quota system is abolished with the introduction of the new scheme.

District quota system According to the present university admission policy, the admission to each course of study (e.g. Medicine, Engineering, Agriculture, etc.) is governed by the following quota system, except in the case of Arts where the admission is made on all island merit basis.

All island merit basis 40%

District basis 55%

Additional quota or educational Disadvantage districts 05%

This district quota will be allocated to the 25 administrative districts on the ratio of population of the district concerned to the total population in the country in respect of each course of study. Similarly, the 5% quota for the underprivileged districts will also be allocated in proportion to the population of each such district. There are 13 districts which have been declared as educationally disadvantaged districts by the University Grants Commission.

These districts are: Nuwara-eliya, Hambantota, Jaffna, Kilinochchi, Mannar, Mullaitivu, Vavuniya, Trincomalee, Ampara, Anuradhapura, Plonnaruwa, Badulla and Moneragala.

Available places in each course of study in the University system will be filled according to this national policy. This will be done by ranking candidates using their final Z Scores. After processing the university admission applications, the UGC will determine and announce the cut off Z Scores for each course of study in respect of each district, as practiced earlier.

In Sri Lanka aggregate of marks has been used to rank students in each stream until the year 2000 (G.C.E. A/L-1999). However, in the year 2001 (A/L-2000) the average has been used to rank the students since there were two groups of students: one group offering 3 (new scheme) and the other group offering 4 subjects (old scheme). Use of aggregate or average mark to determine merit order has serious defects as pointed out by many educationists.

All subjects are not equivalent. There are easy subjects as well as difficult subjects. There are high scoring subjects and low scoring subjects. Marking also can vary: there can be strict marking or liberal marking in different subjects. As such, aggregate of different subjects is not a fair criterion to give an overall rank to a student.

It is more unfair to give a common rank to students, when there are two or more groups of students who followed different subjects, different curricula, and different combinations etc. are involved. It is even worse, when one group of students has sat for four subjects, while the other group has sat for only three subjects.

Furthermore, different subjects have different spread or distribution with respect to the average. For instance, only about 3% of the students eligible for university entry obtain 'A' grade in Physics, while nearly 14% receive 'A' in Agriculture. Similarly, percentage of students who received A's in Botany was much higher than that of Biology in the year 2000. There are many examples in the Arts stream where striking differences are observed in the distribution of marks.

Although it is possible to rank the students in each subject using the raw marks, it is not proper to use the aggregate marks of different subjects to determine the overall ranking of a candidate.

Aggregate of marks of different subjects to determine the merit order in such situations is tantamount to the proverbial adding of apples and oranges.

If one examine the sets of marks in each subject, it is seen that the average mark and distribution of marks vary widely. Therefore, ranking many be done based on raw marks in one particular subject but it is not correct to add individual marks of subjects (having different averages and different distributions) to give an overall rank of the student on the aggregate.

As an example, we can take two subjects A and B and let us assume that A is easy or marking is liberal in A.

For example, in subject A average is 65, with a spread of 50-90, while in the subject B, average is 40, with spread from 20-60. It is very clear that the value of 50 marks in A is completely different from that of 50 marks in the subject B with regard to their ranking in the relevant subject. In the aggregate or average we assume that all subjects are equally difficult or easy and the marks have the same value irrespective of the subject.

This assumption is not valid.

Z Score method is widely used all over the world for this purpose. In the Z Score technique raw marks in each subject are converted into a standard scale - Z Scores, which are comparable across the subjects.

The average Z Score of 3 or 4 subjects can then be used to arrive at the final merit order. Extensive discussions over more than two years with a wide participation of educationists, university academics, school principals and others, and deliberation of several committees have led to the acceptance of Z Score method to rank the candidates for university admission. This proposal has been accepted by the National Education Commission and the University Grants Commission for implementation.

Z Score is a statistical concept which can be used to solve this problem satisfactory. Z Score takes into consideration the average mark as well as the range (measured by standard deviation) for the subject. The Z Score of the different subjects of a candidate can be averaged and comparisons made across the subjects/groups etc.

In the Z Score method the average of each subject is brought to 0 and the spread is usually from +3 through 0 to -3. Z Score gives the candidates the position in each subject with respect to the average mark scored by the candidates in that subject. Z Score therefore ranges from around +3 to -3 (e.g. +1.252, +2.012, -o.958 etc.). Then, the average Z Scores of the 3 or 4 subjects of the candidates are taken as the final Z Score, which will be used to rank the students in each stream, island-wide and district-wise.

It is clear that the Z Score not only corrects the disparity arising from aggregate raw marks of students taking different sets of subjects but also different number of subjects. Z Score is proved to be a much better index than the aggregate or the average, and such schemes are operative for national examinations in many countries.

This is accepted as a fair scheme and is much superior to the use of aggregate or average. If the subjects offered in one particular stream are the same (e.g. Medicine) it will not make much of a change from the previous ranking scheme.

If however, the subjects are different, like in the Arts stream or for admission to Law faculties, etc., the new scheme offers a much better determination of merit order leading to a near-perfect and extremely fair method of ranking.

Thus, ranking on this basis is more fair, highly reliable since the technique virtually ranks the students in each subject and the average ranking of the 3 or 4 subjects is taken as the final ranking of the student.

Introduction of Z Score is only one correction applied to the current admission scheme. There are other deficiencies in the existing admission policy which needs early review. The National Education Commission after much deliberations has already submitted a comprehensive proposal to the government to review the current university admission policy.

News | Business | Features | Editorial | Security
Politics | World | Letters | Sports | Obituaries |

Produced by Lake House
Copyright 2001 The Associated Newspapers of Ceylon Ltd.
Comments and suggestions to :Web Manager