Thursday, January 11, 2007

CATalysis

Yes I know the title I have used is a very tired cliche. Lets just attribute it to my lack of imagination and move on. Of course, I am about to pour gyaan on the entrance test that has caught the fancy of everyone who is anywhere in the vicinity of any media whatsoever. Since most reactions to the various changes in pattern, marking etc. tend to be either disgustingly predictable 'Expect the unexpected in CAT' or knee-jerk and inflammatory ' CAT this year was a mess, the IIMs didnt select me, I seriously dont know whats wrong with them they'll all burn in hell and die etc etc', a slightly informed critique is rather necessary.

First of all the basics for those who are as yet uninformed - 75 questions 300 marks...equally distributed over the three sections, very easy quantitative section, very ambiguous verbal section. Assuming that the shift to easy math and inference based english was not a result of accident, it is worthwhile to examine why the shift was effected in the first place. There is but one obvious answer - Indian students in general, and techies and managers in particular, are renowned for their quantitative skills and the general impression, not entirely untrue, is that the communication skills of even the best among them are not of equivalent level. Standardized tests administered by the US agencies, for eg SAT, GRE and GMAT tend to have much simpler math sections and conversely, have verbal sections that are either equivalent to or tougher than their Indian counterparts. It is not uncommon to see Indian engineers with GRE scores of 1300 with a break-up of Math:800, Verbal:500. Inspite of this, I have always believed that the American tests cannot be considered ideal in their balance either. To this day, I vividly remember the absolute shock I got when I realised that the first question in Math in the SAT I exam was basically asking me to subtract 6 from 11. I mean, was this what 12th graders were supposed to be asked? Ofcourse, at the same time I was also expected to know the meaning of the word 'marionette' which I didn't know then, and I've seen used only once in all of my habitual readings of newspapers etc in the 6 years that have passed by since that exam. So while a move towards giving equivalent weight to quantitaive and verbal skills is most necessary for CAT, I don't think a sudden dumbing down of the math level and a corresponding exponential leap in the verbal section was the answer.

Secondly, and more importantly, even if the basic plan was perhpas spot-on, the implementation was far from it. It is easy to see that if the relative toughness of a section is either very high or very low, people who are naturally more skilled in that department start with a huge advantage over those who are only moderately good in that section but more balanced. For eg, a very tough math section would mean that there would be a lot of questions that only people with strong mathematical backgrounds will be able to attempt. Also, if the math section is much easier than average, those who are good with numerical speed have a huge advantage over the others. To give most people a fair shot in quant. and to ensure that it doesn't become make or break, it is absolutely imperative that the section is of average difficulty.

As far as the verbal section goes, the huge amount of ambiguity introduced by the fact that most questions were inferential in nature rendered the concept of the 'educated guess' equivalent or even inferior as a tactic to wild guessing. The only safe option was to attempt a fewer number of questions, to basically be more risk-averse, which is not necessarily a quality that managers in the current Indian context should possess. More pertinently, atleast some questions had answers according to the official IIM key that could be debated for hours on end. There was a paragraph completion question, where the paragraph was taken from a paper by Noam Chomsky and the key obviously showed the answer as the sentence that Chomsky had himself used in his paper. Now Chomsky may have used those words because they seemed to him as most appropriate to carry forward the general idea he was propagating through the paper(as opposed to that one particular paragraph), yet there was nothing in the given four lines to suggest that the sentence given in the official answer key was more appropriate to round off the paragraph than atleast two other options that were given. This year's verbal section can be described as a test of a particular way of thinking, rather than that of a general proficiency in communication and language, and there's no reason to presume that people with that way of thinking would necesarily make better students/managers.

And then ofcourse there was the case of the cut-offs decided by IIM-A. This institute had very initially declared that sectional cut-offs will not be less than 25% marks, and it stuck to its word, with the Verbal cut-ff being the lowest at exactly 25 marks, which gave a percentile of 95.33 in that section. Now, this percentile was strangely extended to the other two sections in an extremely rigid algorithmic fashion, ensuring that those with a score of 45% in DI/LR(95.22 percentile) and 50% in math(95.27 percentile) lost out on sectional-cutoffs by extremely narrow margins, despite some of them having brilliant overall scores. Ahmedabad may have had its reasons, the number of students they wanted to shortlist for the interviews seems like a highly possible one, but I genuinely feel they have lost out on a significant number of very brilliant students who will probably grace one of the other IIMs.

And now for the facts ....

1) The topper had 204 marks
2) 100th percentile was achieved at 195 marks
3) The highest percentage in the three sections was, respectively, 86(DI/LR), 61(VA) and 100(QA).
4) A total of 123+ with a sectional distribution of 46+, 25+ and 51+ would ensure all six IIM calls.
5) The total marks obtained by the students seemed to fit beautifully in a log-normal curve, with a mean of around 36 and a standard deviation of around 41 on the positive side.

All factual analysis is courtesy the information gleaned from the discussion forum at http://www.pagalguy.com

4 comments:

shivchakra said...

again...brilliant analysis

although i believe that the shift of emphasis from quant to verbal ability has done away with the unfair competitive advantage in favour of mechanical problem solving. Simple yet intellectually stimulating formulae-less questions are sufficient to test quantitative ability.

at the same time care should be taken that the inherent risk of subjectivity creeping into the verbal ability section (as has happened this time) is avoided.

increasingly...empathising, envisioning, selling dreams, etc. are becoming more dominant in the management arena than pure number crunching skills.

vibhor said...

am still shocked.. did i read it rightly? "no iim call"????
kya hua???????????(infinite question marks)

Ritwik said...

@vibhor : ohh dude...that particular statement was within single quotes...i was just phrasing something written by the people who didnt fare too well which i read at a lot of places....and opposed as i am to revealing personal info on this blog....i got all six...

also, welcome to the blog...i didnt know that u read it...leave a few comments whenever u feel like..tangible readership is always appreciated!!

Sanjana Govindan said...

hhaha i love * infinite question marks... good going vibhor.
also
i am not going to say brilliant analysis or anything grand , cause everyone has already stole my thunder..
all i am going to say is " hmmmmm interseting .. . "