Thursday, October 20, 2011

Quality of Students at the IITs

The BusinessStandard has published two responses to a recent allegation by N.R. Narayana Murthy of Infosys that IIT students are not as good as their seniors, and his identification of the cramschool culture as the main reason behind this deterioration in quality. The responses are from big hitters from the IIT system: Prof. Gautam Barua (Director, IIT-G) and Prof. S. Prasad (former Director, IIT-D). NRN gets results!

Here's Barua:

So how do we improve the “quality” of IIT graduates? Based on the points above, the obvious answers are to increase the numbers of those who are really interested in a career in engineering or science, and to reduce the cases of mental fatigue. As far as the latter is concerned, the IIT Council has been discussing this issue and it has been decided in principle to do away with the Joint Entrance Exam (JEE) and instead use school results and the results of an aptitude test to decide admission. The wide variety in school board exams is sought to be handled by using the percentile rank of a student as the absolute marks of the school result. This will mean that the marks a student obtains will depend on her rank in her Board and on the size of the Board in which she is appearing.

And here's Prasad:

It is nobody’s case that the admission processes of the IIT system are perfect. Having reduced the question paper to a multiple-choice, objective test, imperfections have crept in, which the coaching institutions have exploited. Therefore, it is impossible to guarantee that everyone who has cracked the Joint Entrance Examination is brilliant. There is no doubt that there is great scope for improving our admission processes and factoring in more information about the candidate than performance in a single test. Perhaps factoring in school results, as is being considered, will help. Perhaps we need to include a component of subjective testing, as used to be done in the past. There are many dimensions for bringing in such improvements.

In other news, the IITs announced their most serious attempt to address the issue of student quality: students now need to score 10% in each subject (and 35 % in the aggregate) in JEE-2012.


  1. Anonymous said...

    Barua said:

    "In my assessment, about 50 per cent of the students in a batch are not interested in a career in engineering after graduation. Another 30 per cent are not sure what they want, or are struggling through their programme owing to handicaps they have brought with them, or because they are burnt out. So, only 20 per cent of the students are “good”. N R Narayana Murthy has come up with a similar figure."

    Multiple-choice reaction (pick any one):
    (a) nice!
    (b) ITYS!

    Prasad said:

    "Contrary to popular perception, there has always been significant variation in the calibre of students joining the IITs, almost from day one."

    Multiple-choice reaction (pick any one):
    (a) nice!
    (b) ITYS!


  2. Ungrateful Alive said...

    Best to let the big guys work out the big picture while we little folks worry about useless things like how to make the next lecture more interesting, and why the sysads could not get the wireless network up and running in the seminar hall in four months...

  3. Dilip D'Souza said...

    Am I missing something? 10% in each subject and 35% in the aggregate? What kind of standards are these?

  4. gautam barua said...

    You are not missing anything. These are "cut-offs" and they have been chosen based on the results of the last three years. What is surprising here? When 10,000 students have to be admitted, a 35% cutoff gives us only 312 marks (65* 4.8 -> total marks are 480) to discriminate and rank these 10,000 students (and please note that the higher reaches are pretty sparse). The papers are set so that we can get "enough" spread.
    I am surprised that none of the newspapers have reported the main reason for the change in cut-offs this year. A judgement of the Delhi High Court which was upheld by the Supreme Court on appeal, states that the method used by JNU to decide the cut-off of OBC candidates ( being .9 of the marks of the mark obtained by the last general student admitted) is not correct. Cut-offs must be declared a priori and must not depend on the performance of general category students. Since IITs have been following a system similar to JNU's, this change was required.
    Gautam Barua, Director IIT Guwahati