Saturday, September 11, 2010

IITs and the JEE

HRD Minister Kapil Sibal chaired yesterday's IIT Council meeting.

One of the issues for this September meeting of the august body was how to reconcile IITs' exceptionalist stand on entrance exams with MHRD's keenness on running just one nation-wide test, and on giving a fairly big weight to the students' performance in Board exams.

IITs are too fond of JEE to let go of it, and they insist on the use of Board exams and NAT as filters to get a hundred thousand students who will then be subjected to a JEE.

So, what happened in the meeting? The issue will be examined again by a committee headed by DST Secretary, Dr. T. Ramasami.

In a post-meeting interview, IIT-Roorkee director said this:

JEE is a good format and this is helping in filtering students well. Till a suitable system evolves, it cannot be done away with.

There's nothing objectionable about the second statement. But what about the first statement? Wasn't there a Committee of IIT Directors -- including the Director of IIT-Roorkee! -- whose mandate was to establish whether the JEE is indeed "a good format" and if "this is helping in filtering students well"?

Here, then, is the ringing endorsement from the Committee (this excerpt is from Section 6 of the Committee's draft report):

Analysis of JEEs and Suggestions for Change

An analysis of the performance of the relatively few students admitted to the IITs over the last decade in the IIT-JEE and subsequently in the IITs ( 2 tier JEE was conducted between 2000 and 2005 and a single objective-type examination has been conducted since 2006) leads the following broad and somewhat expected conclusions:

  • There is a strong correlation between the Standard X and Standard XII marks and CGPA including the final performance in IIT.

  • Both AIR and percentage marks at Standard XII are better correlated to the CGPA only upto the end of the first year.

  • There is poor correlation between AIR and the CGPA of GE and OBC candidates from 2nd year onwards.

  • Percentage of marks at XII level better explains group performance in later years.

  • Students with high AIR (less than 1000) have higher score at XII level while aberrations are more prominent at lower AIRs.

  • An analysis of the performance of students in the screening and main tests of IIT JEE between 2000 and 2005 showed a considerable overlap between the sets of top 5000 students although their ranks within the sets showed little correlation. Hence it would be expedient to settle for a completely objective single examination.


  1. kk said...

    I have no idea how the calculations were done. Did it took into account that most of the students opt for branch based on their AIR and subsequent CGPA(after first year) is branch specific? In first year, all the students do same courses and their CGPA can be correlated to AIR but from second year onwards, except 1 or 2 elective courses(which may be department elective), all courses are done in one's own department. Therefore someone with AIR 800 may be in branch A cannot be compared to someone with AIR 1200 and in branch B from second year onwards. Do you have any idea how the correlation was established?

  2. Anonymous said...

    kk: true...

    "Students with high AIR (less than 1000) have higher score at XII level while aberrations are more prominent at lower AIRs."

    -So, the JEE is not missing out on students who sincere abt their board exams.... Then what's the point in giving separate weightage to XII board?

  3. Ninad said...

    I agree with kk ....
    Since most of the students choose branch AIR wise correlation is likely to be very low second year onwards .. also for smaller branches/specializations sometimes range of AIR's(difference between opening and closing AIR)is 300 - 400 .. translating into hardly 5-6 marks i.e max 2 question in JEE (especially true for 4 digit AIR's).. so correlation should not be expected as such ..