Tuesday, April 07, 2009

Mistakes in JEE question papers

With this year's JEE just five days away, Manoj Mitta reminds us about the mistakes that crept into the math paper last year. While I did link to the story when it broke, I didn't realize the true magnitude of the errors: five questions worth 18 marks in all.

Given that each mark is worth some 70 ranks (for ranks in the range of 500 - 7500), these errors are horribly costly and unfair.

The high cost of the mistakes were compounded last year by the IITs' policy to not make the answers public immediately after the exam. As a consequence of this policy, they did not eliminate the effects of these mistakes during grading. Clearly, then, the 2008 rank list released by the IITs must have had huge flaws / holes.

Mitta's article alerts us that the IITs have not indicated when they'll make the answers public this year:

Though the Joint Admission Board of JEE 2009 discussed Joshi’s correspondence, as disclosed to TOI by its chairman Gautam Baruah, its information brochure gives no indication whether the model-answer sheet would be made public at least this time, immediately after the exam, so that any mistakes there could be corrected before the damage is done with the announcement of results.


  1. Anonymous said...

    This kind of half-hearted and incomplete reporting deeply irritates me. While Manoj has provided the question numbers of the doubtful questions, a lay reader has no direct access to either the questions, the model answers or KDJ's comments and "detailed calculations and drawings". I agree this would be difficult in a print edition, but why not provide appropriate links in the online edition. Also, nothing substantial discussed in each paragraph devoted to the individual errors, except fluff, maybe the questions could have been reproduced.

    Manoj has made no attempt to corroborate KDJ's claims or cross check with others, all of which is extremely easy, since in a nation of a billion people, there must be someone nearby who can solve 12th class maths problems. There is no mention whether the JEE Board 2008 discussed KDJ's correspondence or what their findings are. That JEE 2009 discussed his correspondence is different, did they agree with his claims? More importantly, did JEE 2008 board agree with his claims?

    Manoj would like us to conclude, based solely on KDJ's comments that indeed there were 5 errors in the JEE question paper. I would think this is very poor journalism. Of course, KDJ could be right, but one has to be convinced.

    In fact, I would go this far to say this (though I know I am being unfair), since you have been tracking JEE/IIT exams, maybe you could have provided us with links to JEE 2008, model answers etc. Or maybe, I should do that!


  2. Anonymous said...

    Ok, I wrote too soon. CHARU SUDAN KASTURI has provided one offending problem in his 2008 Telegraph story, but what about the other problems? There is some mention of JEE website, but which? But even in the 2008 article, there is no attempt to cross check. All the chairman says is that “I have received Professor Joshi’s complaint. But it is now too late to do anything. JEE 2008 is over,” Of course, the chairman may not accept the mistake on account of potential liability, but surely there are others who could check it. BTW, the one minute that I spent on the problem, I had no idea how to proceed!


  3. Anonymous said...

    KDJ has done much more than point out apparent mistakes. He has painstakingly analyzed JEE question papers for the past 5 years and written 90 page analyses.

    Your comment that "I didn't realize the true magnitude of the errors: five questions worth 18 marks in all."
    is unfair and wrong. KDJ's objections are more philosophical, pedagogical, technical and his own words moral. In fact, some of the so called mistakes are minor, as per his conlusions (page 86, line 2) and not earth shattering to cause a loss of 18 marks and loss of several hundred ranks.

    None of this is brought out in the either Manoj's articles or your posts.

    It took us under 5 minutes to locate KDJ's notes and realized that he has a much more nuanced position than what comes out in the newspaper articles or your posts.

    The only "wrong" answer (more serious error in KDJ's words) seems to be Q 7, paper 1. Even in that case, it might be a case of confusion between a strict inequality (<) and an inequality
    ( <=).

    Niket and Sridhar

  4. Abi said...

    Sridhar (I hope you are the same person in all the three comments!) and Niket: Thanks for the pointer to Prof. K.D. Joshi's analysis of JEE-2008 (as well as the previous JEEs). It's absolutely fascinating!

    After reading the relevant parts of KDJ, I'm not sure if we even read the same document: KDJ uses quite unambiguous phrases such as "disaster for the sincere candidate," "mockery of perseverence," and of course, "shocking indeed that such a mistake went unnoticed by the experts". His anguish at the "minor" errors (ambiguities such as the use of 'minimum' in place of 'local minimum', or the use of ">" in place of ">=") is very clear: he thinks that this is tantamount to sanctioning "wrong mathematics." I don't know how you can think of this indictment as less than serious.

    I am not sure whether Prof. Joshi himself would say the errors and ambiguities in the two papers would amount to 18 marks. But what I am sure about is that even a four-mark difference (the mistake in Q.7 in paper 2 is worth 4 marks; this mistake is not being disputed at all) would mean a difference of several hundreds (for ranks beyond 500).

    But those are details; let's return to the main point of Manoj Mitta's article. Clearly, Prof. Joshi would agree with it; here's what he says on p.87: "It will be highly desirable if the model answers prepared by the paper-setters and scrutinised by JEE experts are made open to public scrutiny before freezing." I agree with him, and I'm sure, so would you.

    I also think Manoj Mitta is right to point out that the JEE-2009 authorities have not said anything about it so far.