Sunday, October 25, 2009

Extreme Coaching and Entrance Exams

Today's DNA has two stories where lots of different people offer their views on coaching schools. While I have heard lots of complaints from faculty (in private conversations and through anonymous comments), it is good to hear them echoed by an IITian:

Says Karthik Shashidhar who graduated from IIT, Madras in 2004 and now works as an investment banker in Bangalore, "At IIT, we refer to the coaching centres as factories. I went to the biggest factory - BASE. But it was low intensity compared to most others. We have seen that more intense the preparation, worse they do at IIT. I have seen fairly sad cases of people dropping out, taking 6 years to complete the course and so on. Those who overstretch themselves to clear the JEE, tend to take it easy once they get into IIT. They just give up in life later. It could also be that people who don't have the aptitude somehow manage to scrape through with intense coaching, and then can't cope at IIT."

* * *

Many people have pointed out that the culture of coaching schools cannot be eliminated, given the huge demand for IIT seats. Some people have even argued that coaching may actually be good, because it teaches kids vital skills that their regular schools fail to impart. All this is true; but these arguments misinterpret the concern about coaching schools.

The main complaint is primarily against Extreme Coaching exemplified by the Kota-style residential centres, which encourage students to develop and internalize a disdain for 'regular' studies -- a disdain that many students take to the IITs!

I think it's fair to say that it's worth pursuing ways of minimizing -- if not eliminating -- the pernicious influence of Extreme Coaching.

* * *

One of them is to use multiple rounds of contests, like the math and physics Olympiads do.

* * *

A simpler version -- involving a two-step process -- could also work. In fact, IITs did use (for a brief period that ended in 2005) a screening test that selected a small subset of JEE applicants for the final exam.

However, this didn't do anything to reduce the need for Extreme Coaching primarily because the screening test itself was as brutal as ever.

I think the trick is to make the screening test really simple. Something that a bright, diligent student should be able to ace without needing serious coaching. Ideally, the screening test should also be done about a year in advance, so that a huge majority of students can get on with their lives and make other plans.

With this scheme, Extreme Coaching loses its sting: where the numbers are huge (the screening test) coaching is not needed, and where coaching might be needed, the numbers are small.

* * *

An even better option is to give some weight to broader measures of students' achievement, including performance in X and XII board exams (BTW, this is not as outlandish as it sounds -- IIM-B uses them). Use of X and XII results would also reward consistent performance -- something that was found in a study to correlate well with the performance of IIT-M students.

Given their stature and prestige, IITs will not have any trouble at all in getting the boards to declare percentile scores along with marks in Class X and Class XII. The percentile scores can also be used as an input into the ranking exercise in AIEEE, Pre-Medical Test, and other such tests.


  1. Pratik Ray said...

    I am not so sure that making the screening test less brutal will do anything for less emphasis on coaching. Coaching centers, and students themselves prepare at 2 levels - at the level of screening test, as well as at the level of the main exam, since they have to assume/back themselves to clear the screening.

    Given the brutality of the main JEE, if someone postpones the prep for the main exam until after the screening, he is quite likely to flounder. Take it from me - first hand experience - first batch of students who appeared the JEE when the 2-stage process was re-started in 2000 ;)

    I do think though, that the 2-stage process has some merit. I would prefer to see a situation where the school syllabii is a lot more balanced. Back then at least, it was as if you are given 2 yrs to learn this much of stuff. So finish mechanics and thermodynamics in class XI (school exam syllabii) and then finish electromagnetism and optics in Class XII. If this can be changed so that the students spend class XI on the basics and build on them in Class XII that would be so much better.

    An indirect pay-off would be that the screening process could be conducted at the end of Class XI, so that those who dont qualify have the ENTIRE year to prepare for board exams and other entrance exams.

    The downside of course, is that a number of students, if they dont qualify in the first shot, take a second shot at the JEE - these folks will probably re-take the screening again at the end of Class XII if they dont qualify in class XI.

    So, may be, try an early screening exam after class XI AND make it easy enough to clear without extremely intense prep (although I am not sure how easy it would be to design an exam that is simple enough so as not to need lot of extra prep beyond Class XI, and yet be hard enough to screen out a couple of lakhs of candidates

  2. Vishu said...

    Wouldn't two or multiple rounds of tests put more stress on students appearing for them? Also, what makes coaching centers hand adapt to such a JEE pattern? Olympiads don't earn jobs or degress, so I am not entirely sure of comparison between them.

  3. Vishu said...

    Correction: "what makes coaching centers not adapt to such a JEE pattern?"

  4. Ankur Kulkarni said...

    The ministry's goal may be to eliminate coaching, though I believe that is not entirely possible. But the goal of the IITs needn't be this. IITs only need to ensure that they select the same quality of students as they would have if coaching was absent. The JEE itself needs to be changed in order to do this. For this purpose, apart from putting multiple sieves (say by having a multilayered JEE, or having multiple criteria - like XII marks, or other attributes etc) one of the things the JEE needs to do is bring in randomness, and bring in qualitative judgement.

    For eg, the JEE could randomize it's pattern, so that it is not easy to for coaching centres to adapt students to the JEE and crack the system. This, they are already doing in some measure (at least they did it at my time). Secondly, apart from problem solving, have deeper questions with large weightage that test the depth of the student's knowledge and have examiner(s) evaluate the student on the basis of their qualitative judgement of the student's understanding. For eg, student could be asked to explain a certain concept in physics or chemistry. Or may be explain and clarify some "paradox". In mathematics they could be presented with puzzle type questions which belong to no particular branch of the curriculum and only require thinking. Or they could presented a set of calculations which lead to an absurd answer and asked to find out where the conceptual mistake (say division by 0, or differentiation where right and left derivatives don't match etc) is.

  5. Anonymous said...

    Pay for IIT coaching faculty

  6. Anonymous said...

    "The highest salary that a teacher can earn at our institute is Rs 2 crore per annum. This includes their performance-related bonus. And this is all in white money," says Vivek Khanna, vice president, Resonance, a Kota-based coaching centre which recently entered the Mumbai market.

  7. ansumali said...

    It is no surprise that IIM-B is a place where mostly upper-middle class people from urban background studies. Any institute which will put good schooling as a criteria will end up discriminating rural and/or poor people.

  8. Anonymous said...

    I like Abi's idea of a spaced out 2-stage process. If you can choose 30,000 via round 1, then you have reduced the coaching centres' collective customer base by 90%. For this to succeed, we need to get the stage 1 process really simple so that the coaching centres have lesser role here. But that wont happen. And here's why: In India, we tend to break our heads in choosing the "best". Can JEE select the best 10,000 of 3,00,000 applicants? Given the definition of "best" is at best fuzzy, lets assume that JEE while attempting to select the "best" 10,000, actually succeeds in selecting "some" 10,000 of the best 20,000 amongst the 3,00,000 applicants. Then, it may be possible to design a simpler exam which also succeeds in selecting "some" 10,000 of the best 20,000. (The 20,000 number was arbitrary, but you get the point).

  9. Anonymous said...

    Sooner or later, the fraction of GDP in circulation for various sectors will decide how people spend their lives. Given the amount of coaching fees paid far exceeds the total budget of IITs and IISc combined, small wonder we are drifting toward multi-round "solutions" where many many more people work on "getting into IITs" than "working when in IIT". a JEE coach can make 10x the pay of an IIT professor. Folks, this is a litmus test of a terminally sick society. There is no fix.

    Either IITs will have to become ETS, or a separate Indian incarnation of ETS has to crop up. Warts and all, the JEE has managed to maintain a decent level of integrity and avoided subversions common in regional exam systems. In a country where top national milk brands are routinely adulterated, it remains to be seen if an ETS outside the direct control of IIT professors can remain untainted. I won't bet my (meager) salary on it.

  10. Mainak said...

    An interesting idea. Nonetheless, it boils down to - Coaching centres are making money and you shouldn't let them make money.

    What role does a coaching centre have on a student who fails to clear his course in 4 years. Who was in charge? IIT or the coaching centre?

    Has the quality of intake gone down? The numbers have gone up - but the best - have nowhere to go. They still come to IITs.

    I think we need to stop putting the blame on the coaching centres and focus on training the students we have in IITs