Testing a great institution

Entrance tests have several drawbacks. One, can a written test measure the complex qualities

expected from students — social abilities for instance? Two, does it discriminate well enough? For instance, the permitted enrolment in engineering is over 360,000. With four papers of a 100 marks each, there will be on an average nearly a thousand students with the same aggregate marks, possibly several thousand at some levels. How do we choose when several thousand students have the same marks? Three, how reliable are the tests? How far do expensive coaching classes mask real ability, and impose a penalty on those who cannot afford them? Four, what happens when a student is indisposed on the day of the test? Five, these are the days when people swear by the utility of competition. Then, why should we have one monopoly test for the whole country? Also, mega tests inevitably attract mega crooks.

However scientifically designed, every test is liable to suffer from two types of statistical error. Type I error is said to occur when deserving candidates are rejected; Type II error occurs when an undeserving candidate is admitted. It is impossible to eliminate completely either kind of error. Wisdom demands that we accept this as unavoidable and try only to avoid them. For instance, as in the case of selection for a national sports team, for IIMs and IITs too, it is desirable that no undeserving candidate is admitted even if some deserving ones are left out. In a small private college, the policy could be the exact opposite — no deserving candidate should be left out, even if several undeserving ones creep in. At the same time, for shortlisting, even IITs and IIMs would like that no deserving candidate is left out even if several undeserving ones are included. Thus, whether Type I or Type II error is more acceptable depends on the circumstances. For that reason, IIMs and IITs may, at best, use entrance tests for shortlisting. It is not wise to treat them as final arbiters.

For years now, many of us in the IITs have been unhappy with the JEE system, and the way coaching classes are distorting the assessment of the real ability of a candidate. Yet, we have acquiesced because we have not been able to see any other way of protecting the IITs from political interference. Hopefully, soon the country will become a developed nation and IITs will be able to scrap the JEE system safely to admit students the way Harvard and Oxford do.

In those world-renowned institutions, the institution the candidate comes from is the primary criterion. A student from an IIT is given careful consideration but another from Budhuram Adarsh Vidyalaya from Nakalpur may not be considered at all even if the marks are the same in both cases. Apart from the grades obtained in the IIT, the comments made by the teachers who act as the student’s referees are given much importance. The candidate may have written some common tests like GRE, GMAT, TOEFL and the like. Those tests are used only as a cross check. In effect, great universities primarily shortlist institutions, and then only shortlist the applicants. Similarly, the IITs and the IIMs can prepare their own shortlist of worthy schools and colleges. That can be done quite objectively. For instance, the number of candidates shortlisted from any school may be set at the total number admitted from that school in the previous three to four years. Then, the shortlist is barely three to four times the numbers to be admitted, and manageable enough to be tested systematically, rather than mechanically by an entrance test. The candidates may even go through a test like GRE or GMAT. Those scores could be taken as a cross check, not as a decider.

This admission process accords high respect to the teachers and the institutions where the candidates had studied earlier. Mechanical entrance tests destroy all such respect given either to feeding institutions or to their teachers. Stanford and Harvard identify and promote good quality schools; IITs do not. The sufferers are the IITs themselves and the nation. Our socialistic culture will find this form of elitism repugnant. People could ask, with some truth, why a student from Budhuram Adarsh Vidyalaya of Nakalpur should be denied admission. The answer is there is no perfect solution. Even then, the doors may not be totally closed on Budhuram Adarsh Vidyalaya. Such students may have an alternate entry through entrance tests — conducted districtwise and not nationwide. As in the case of select schools, each district may be given a shortlist quota equal to the numbers that were admitted from there in the previous three-four years. In this multi-stage selection process, if the schools and districts become corrupt, and send sub-standard students, less of their number will be finally selected, and their quota will dwindle rapidly. Hence, those who are careless or dishonest will be penalised promptly. Thereby, a healthy competition is set among schools and districts. Such competition is a better safeguard against mischief than any amount of policing. It is not a perfect solution; no solution is.

World renowned institutions have recognised that it is better to trust feeding institutions than conduct entrance tests directly. That is the lesson we should learn. Further, it is a fundamental principle that multi-stage tests are better than any single-stage process one may devise.

The writer is a former director of IIT, Madras