Search This Blog

Wednesday, November 25, 2015

Committee to Examine the JEE System

Yet another committee, and yet another report. However, for a change, here is one report with a difference. The only way to describe this report - it is music to my ears.

First, the background. The committee was set up in October, 2015, after a couple of committees had recommended that the board marks be not used for ranking even for admissions to NITs, IIITs, etc. The normalization was a big problem, and there were far too many administrative challenges in implementing it. Besides, a study by Joint Admissions Board had found that none of the expected benefits - more women in engineering, more rural folks in engineering, less coaching, etc., had been achieved in the three years that the system was in operation. It was decided that instead of just dropping the board marks from ranking, could we take this opportunity to clean things up a bit more, and hence a committee chaired by Prof. Ashok Misra, ex-Director of IIT Bombay, and Chairman of Joint Admissions Board (JAB) was set up.

Let us now look at the major recommendations and opine on them.

1. All Centrally Funded Technical Institutes (CFTIs) to admit students from the same exam, the JEE Advanced. This is very positive. Considering the board marks was a pain for all stake holders, and once that was gone, there was really no reason for some CFTIs to admit through JEE Mains and some through JEE Advanced. Of course, on philosophical grounds, I would prefer a system where each Institute can decide a different way of admitting students, for example, some may decide to admit based on just Physics and Maths marks in JEE. And the common counseling portal developed this year has the capability to allow such variations. But as long as the admission process is going to be dictated from the top, it is better that the top dictates a common exam than two separate exams.

2. The shortlisting for JEE Advanced to be done through an aptitude test. The committee has suggested that we should set up a National Testing Service which should conduct an aptitude test. This may take 1-2 years, and hence in the interim, we can continue with JEE Mains being the filtering test. This too is a very positive suggestion. This is certainly not going to be easy. To design an aptitude test which would be free of any cultural and other sorts of biases is not easy. On top of that, we want to offer that test in multiple languages, and we need to make sure that the test in each language is of similar difficulty level. Further that test will be offered several times in a year and hence we will need to have some normalization across different offerings. This may not be possible to do even by 2017, but a beginning has to be made sometime, and I am glad that this committee has recommended strongly that we don't wait any further to make that beginning. Of course, the committee has shown what a bunch of intelligent people can do - come up with solutions to serious problems. It is proposed that the aptitude test is used to filter 4-5 lakh students for JEE Advanced. This will make sure that those on the borderline or those who don't make it will not feel very disturbed and cheated, since they would most probably not have any hopes of getting into the top 35000 eventually anyway. So even if the normalization is less than perfect and some bias remains in the test, it won't materially impact the admissions at IITs, NITs, etc.

3. A suggestion to the government that the level of NITs and other CFTIs be raised and the gap between them and IITs be reduced. This is really an excellent suggestion, and I do hope MHRD will find ways to strengthen the tier 2. I have been repeatedly saying on this blog and on my facebook that the real reason for stress is that a few marks in JEE can take you from excellent institute to one you don't like as much. And unless this issue is resolved by reducing the gap between successive institutes, there is no hope of reducing stress. Of course, this would invariably mean giving more money (where is the money?) and giving more autonomy (which is very difficult for those who currently hold the levers of power, just compare the new NIT statutes with IIT statutes). But there is always hope.

4. Suggestion that somehow boards should improve. A motherhood statement really, but we must keep making such statements. Even if it encourages a few people somewhere, it can only have positive impact. In particular, they have mentioned the examination system of the boards be looked into. None of the boards in the country has a distribution of marks that you would expect from a large public exam in any other part of the world, and the results are completely inconsistent with the quality that we perceive of the schools around us.

5. IITs should create a large question bank and develop some system for mock JEE examinations. May be there can be lessons through MOOCs. I think IITs can really offer subject training in 12th class science subjects. Already, there are IIT professors like Harish Verma whose school level books are like bibles for 12th class students. We should be able to tap into such resources and come up with online courses in all three subjects which are available to anyone freely. If our school students have access to high quality courses to learn for JEE Advanced, the coaching culture will reduce anyway.

6. In the interim period (while the country plans an aptitude test), the JEE Mains will become a 6-hour exam and 2 lakh students to be filtered instead of 1.5 lakhs for the JEE Advanced. I have no comments on this, as I fail to see the benefits, but there is no harm either.

The only issue I have with the report (and all such discussions at the Ministry level) is that we are focusing too much on coaching. I think that if we ignore coaching and just do the right things - better admission strategies, better schools, better colleges, and so on, the coaching will either go away or will contribute to the educational efforts of the country.

Friday, November 20, 2015

Fee Control on Engineering and Management Education

It has been said a million times that the quality of education that our institutions offer is rather poor. We keep hearing that 75 percent of our engineering graduates are not just unemployed, but unemployable. That is, the quality of their learning is so poor that even a finishing school or a company training for several months would not make them productive engineer. Lots of experts keep giving various reasons for this poor quality, lack of autonomy, lack of faculty, greed on part of private colleges, poor preparation in the school education, and so on. But I disagree.

The poor quality education is a result of our higher education policy. Our policy over the last 3-4 decades have considered cost, access, and equity as much more important parameters than quality. We have worked really hard over this period to ensure that the cost of higher education (primarily in engineering and management) remains among the lowest in the world. It has not mattered to policy makers that the result is a poor quality education. And higher education policy is one of the shining examples of policy success, and not an example of policy failure. We consciously decided that everyone should be able to get an engineering degree at low cost near his/her home, even if that degree will not get one a job, and we have succeeded in it beyond anyone's expectation. A low-cost low-quality degree is strongly preferred over higher-cost, higher-quality degree. Today, engineering education is a buyer's market. Anyone can get admission to an engineering course.

So it does not come as a surprise that yet another committee has decided to control tuition at levels which can only provide poor quality education. Here is the news item. If this report is accepted, the highest fee that any engineering institution can charge for BTech program is Rs. 1.58 lakhs per year if the institution is located in a Tier 1 city. If it is a high quality institution (as evidenced by an accreditation), then an extra 20% is allowed. And another 1% can be charged for miscellaneous services, a total of about Rs. 1.92 lakhs per annum for the best institutions in the country.

I did a quick search of tuition cost of private institutions who are trusted by society as provider of high quality education and believed to be not a profit making entity. Both BITS and IIIT Hyderabad had higher tuition in 2015 than the maximum allowed by this committee. And both will have to increase their tuition substantially to take care of not just annual inflation, but also the higher salary costs imposed by the 7th pay commission report. In fact, other popular, well-known private institutions, including Thapar, Manipal, LNMIIT, JIIT, and many more also had 2015 fee higher than the maximum allowed by this committee report.

There is another way to look at these numbers. What is the budget per student per year at IITs. The budget is about 3 times these numbers. This can only mean two things. One, IITs are a den of corruption and waste. That is why they are spending so much money, when good quality education can be provided for a fraction of the cost. Two, the costs of good quality education are indeed higher, and this committee has, in its wisdom, decided that private sector can not be allowed to provide quality education. I suspect that it is the latter. The committee is really saying that we must follow our national policy on higher education, which requires low-quality low-cost education to be dished out to our students, and we can not really allow private sector to violate that policy by attempting to provide higher quality education.

What is even more interesting is that the committee allows MCA programs to charge 10 percent more than BTech programs. This is absolutely ridiculous. What is the expense that MCA programs have that BTech students do not have. In fact, MCA students need only computer labs, while BTech students need many more labs, which are far more expensive to maintain. But in our socialistic mindset, the price is not related to cost. First degree students should pay less, and second degree students should pay more. This is the reason a large number of colleges continue to have MCA programs in the country - they have lower cost and higher price than under-graduate programs. Otherwise, there is no rationale for this program to exist in so many places.

And if one looks at MBA programs, even the government institutes charge more than the upper limit proposed by this committee.

The government policy for several decades is also ensuring that a large number of our students are going abroad for under-graduate studies. By focusing on quality at home, we could have kept these students within India, and also attracted foreign students to our campuses. We could teach the world and earn a lot of money through it.  While it would be impossible to close down poor quality institutions, and may be there is something positive about a poor farmer selling everything he has to see his son having a worthless engineering degree with no jobs, but the least we can attempt is to have high quality institutions to co-exist with poor quality institutions.

Sunday, November 15, 2015

Railways increase cancellation charges

Indian Railways have changed the rules for refund when reserved tickets are cancelled. The cancellation charges have been increased rather steeply and they now have to be cancelled much more in advance than earlier. Since the cancellation charges were enhanced not too long ago, this really came as a surprise.

It has been reported that this decision has been taken to discourage the touts from booking a large number of tickets. If the cancellation charges are small, then inability to sell those tickets in black market only results in small losses, but if the cancellation charges are large, then touts will be discouraged to book lots of tickets since the potential losses could be larger. And more genuine passengers will be able to get confirmed reservation.

It does not sound believable. What fraction of tickets are being bought by touts. Remember, the ticket has a name, gender and age, and each reserved passenger is expected to carry an identity card which should verify name, gender and age. Yes, TTEs are not very careful at times, and some can carry fake identity cards. The problem of touts today is different from what it used to be. Today, I can get a Tatkal ticket through a tout who has a setting with the reservation office so that my ticket will be booked before anyone else standing in the queue. With Advanced Reservation Period (ARP) of 120 days, touts do not block a whole lot of money 4 months in advance to do block booking in the hope that they will be able to sell all of those at a good profit closer to the travel date. I don't think increasing cancellation charges would have any impact on touts.

It seems to me that the real reason for this increase is to generate more revenue. Given that the Railway finances have been allowed to deteriorate for so many years, I am all for Railways trying to generate more revenue. My friends in Railways tell me that the expectation is that the new rules would add more than Rs. 1,000 crores to their revenue in a full financial year, certainly not a small change. However, they could have collected the same amount of additional revenue just by increasing the fare by 2-3 percent.

Of course, all transport providers charge those who do not travel to ensure that the cost of travel for those who do end up traveling is lower. And Railways should follow the industry practice. However, there are some problems with the specified rules.

The minimum cancellation charges are very steep. This causes some problems. One, what if I book a wait listed ticket, which does not get confirmed for several days. I decide to cancel it and go by alternate method (air, bus). I would expect that if the transport provider is unable to provide me a confirmed booking, it would charge me a bare minimum amount to cover its costs of booking and cancellation. Two, in some cases of short distance travel, the minimum cancellation charges are almost comparable with the total cost of the ticket. It means that there is absolutely no incentive for the passenger to cancel the ticket. Now, short distance bookings (~200 KM) are not the ones that touts go for, since the passengers invariably would prefer a bus instead of paying a significant premium to touts. From a revenue maximization goal, it would have been better for the Railways to get some cancellation charges and re-sell that seat to another passenger. So revenue goal is also not being satisfied by such high cancellation charges. The only one happy with such a system would be a corrupt TTE who can sell that seat to someone in the coach, which will only cause the Railways image to be tarnished.

The other serious concern that I have is on no refund rule on waitlisted tickets if they are not presented at least 30 minutes prior to scheduled departure of the train. Invariably, a wait-listed passenger will wait till the chart preparation time, which is roughly 4 hours prior to the scheduled departure of the train, and if the ticket is still wait-listed, it would first try to arrange for an alternate method of travel and not focus on ticket cancellation. Earlier, it was possible to cancel the ticket till several hours after the train has departed. Recently, it was brought down to 2 hours after the train has departed. Now, it is reduced further. This one really hurts. Railways should sympathize with their customers whom they have not been able to provide accommodation, and not use them to get maximum revenue out of them. (Of course, people like me who book tickets only online will get automatic cancellation of wait-listed tickets. It is only the PRS tickets that will have to be cancelled in that small time duration.)

In fact, even with confirmed tickets, the four hours prior to scheduled departure of train is an issue. Earlier, if the train was a couple of hours late, I could cancel the ticket and at least get 50 percent refund. Now, I can get (full) refund only if the train is 3 hours late, and that too if I cancel my ticket before the train departs.

Some increase in cancellation charges was due (despite an increase not too long ago), but playing with the time duration has brought in an element of inconvenience.

Of course, there is a huge positive news associated with all these changes in refund rules. Railways is promising to make all the cancelled seats available for current bookings (after clearing the waiting list, of course), and the current booking can also be done online. I think in most popular trains on most days, this will not be a big advantage, since waiting list is typically large, but this will be of help to people on many trains.

I think most of the problems that Railways face are due to the fact that passenger fares are subsidized even in AC classes. People are willing to pay a much higher fare and that gap is exploited in various ways. Railways keep coming up with some mechanism or the other to make the system more fair, and get them some extra revenue, but it does not work. I am told (and media has speculated about it too) that Railways is thinking of introducing dynamic fares (something that I have been advocating for years) in all trains in all classes. I strongly believe that having dynamic fares in all AC classes is an absolute must. Of course, the dynamic fare algorithm needs to be much more complex than the current algorithm used in Suvidha trains. One can not just change fares based on percentage of tickets sold, but it also has to consider the rate of selling those tickets, and how much time is left before the train departure date. Also, they may want to introduce several "classes" of tickets - refundable and non-refundable, cheap tickets which only give you middle berths, etc.

By the way, I am totally impressed with the news coming out of Rail Bhavan these days. Quietly, a revolution is taking place. I may disagree with the specifics of cancellation charges, but overall I am very excited about the changes that are taking place, and I will hopefully write another blog soon about them.

Wednesday, November 4, 2015

Gender Bias (Women in IITs)

Yesterday, I made a post on my facebook page. This was a link to the article which gave information on how many women were admitted to different IITs last year. The numbers are abysmally low, less than 10 percent. And I suggested that this indicates that there is something wrong with the admission process of IITs.

Questions started coming in.

Could it be that women are not interested in engineering. How many women gave JEE.
I pointed out that 25% women gave JEE while only 10% succeeded. I also pointed out that in colleges which take admission through other routes, the percentage of women is significantly higher. I also pointed out that BITS Pilani saw a significant drop in women admission when it moved from admissions based on 12th class marks to admission based on an entrance test. All this does not seem to indicate that women are not interested in engineering.

May be women don't have merit. How do I know that women deserve to be in IITs in larger numbers.
Well, if we look at board performance across pretty much any board, in the top ranks, women perform better than men. You would find more than 50 women in the top 100, while the percentage of women giving the exam is significantly less than 50 percent. And the same is true not just overall, but specifically in science stream, where students have taken Physics, Chemistry and Mathematics, the same subjects that JEE tests on.

But aren't 12th class board exams testing something different than JEE. And may be women are not inherently good at what JEE tests.
Let us assume that they are testing different things. Let us also assume that there is something genetic about women not doing well in JEE. Should not then one ask what is more important for IIT education. I pointed out that in 2011, when there were lots of debate on changing IIT admission process and including 12th class performance in some way, it was pointed out that there have been three studies done in three different IITs in terms of what is a better predictor of success in IITs - 12th class marks, or JEE performance, and all three studies said that the correlation between 12th class marks and IIT performance is higher than correlation between JEE performance and IIT performance. While I consider all three studies as too small a sample size, and not scientifically rigorous, they do raise a doubt on whether JEE Advanced (which is causing only 10% women getting admitted) is bringing in the best talent to IITs. And since this is serious enough matter, we should at least do more research into it, and do a bigger, more rigorous study.

(By the way, I would still not support incorporating 12th class marks in the admission process, since the foul play in boards will increase, and many other reasons which I had articulated at that time.)

Do you have any hypothesis that could perhaps be tested by a research team.
Yes, I do. I believe that today coaching has become absolutely necessary for good performance in JEE. And most families are reluctant to send their girl child for coaching, particularly out of town places like Kota. And that is the reason for their poor performance in JEE.

Should IITs be bothered about societal biases. Is it not good enough that the same question paper is given to all candidates, no cheating is allowed, and an objective merit list is prepared.
If IITs are interested in attracting the best talent, they should be bothered about societal biases and any other limitations that their admission process may face. They need to think whether the current admission process gets them the best students or is there any other way to get even better students.

Conclusion: My own conclusion is that less than 10% women in the incoming class is far too less and has not been adequately explained by any study so far. There is enough reason to suspect that certain societal biases could be limiting chances of women in the current admission process of IITs. And hence, IITs must share all their admission related data for research or conduct research themselves to find out whether there is a problem or not, and if there is no problem why women are less than 10% in IITs.

Interestingly, the responses on my facebook post yesterday (and indeed, I have raised this issue earlier a few times as well, and the responses have been exactly on the similar lines) were essentially saying that there is no reason to suspect anything and there is no reason to do any research. Note that I am not recommending any reservation, or affirmative action, or any reduction of seats for men as of now. I am only asking for more research to be done. And a majority of male respondents don't even support research to be done into this issue.

I checked the profile of these men. Most are students/alumni of IITs.

My own take is that most of those respondents have following problems with the idea of doing research in to this issue. One, they realize that a research is likely to show that indeed there is a bias against women. And, more generally, JEE rank is not the best indicator of merit. And most of IIT alumni immediately fear that this may mean that they themselves perhaps did not deserve to be in an IIT. So they would oppose any research or any major change in the admission process. Two, the male elite does not want to give up any of their territory without a fight. The issue is not of fairness, but why give up our entrenched benefits.

And we always blame the poor and uneducated for gender bias.

EDITED: Nov 5, 2015
After I published this, I was sent the following link about women not being represented fairly in business schools also.

Yo Chanda Kochhar, if there aren't enough women in B-schools it isn't because they can't do the hard stuff