Search This Blog

Wednesday, June 8, 2011

Should placement be a criteria for choosing your program of study

In the last one month, I would have received no less than a few hundred emails asking for all sorts of advice, sometimes complimenting me, and sometimes strongly disagreeing with my views. The strongest disagreement has been on the issue of placement.

I have always advised that students and parents should not look at the placement records of different colleges or programs before deciding their choices. And I get asked many times, don't I like money. Why do I want everyone to be a professor. Apparently, professors are satisfied with less money.

The truth is that I love money too. When 6th pay commission almost doubled my salary, I did not protest. And if 7th pay commission doubles it again, I will not be protesting at that time.

And I don't ever say you should not worry about money. I only say that you should not be bothered about placement information.

But this question has been raised so many times, that I thought I will write my views on it in some detail, though you will find parts of it in some of my earlier blog articles.

First of all, placement information is so unreliable, that in most cases it cannot be relied upon. Some colleges will tell you the top salary, which only one student would have got. If you know a bit of statistics, you would know that a single sample cannot be used to deduce a general principle (of ordering programs in this context). The most reliable barometer of the current employability and the economic value of a program, is the median salary of that program. But only if the program has large enough number of students, say at least 40-50, who are interested in taking up a job, and indeed get one. Hardly any institute will give you median salaries. And a large number of programs have much fewer students graduating, or having an interest in placement. If colleges give you median income, they will not tell you how many people did not even get a job. Sometimes, they will remove weak students from the list, and then give you placement statistics. Sometimes, they will tell you the number of jobs and number of students, but will count two jobs to a student as two - so you can't figure out how many people don't have jobs. In top places, sometimes students would not go for placement but for higher studies. That is not captured by the placement statistics. Then there is a huge fudge factor that both companies and colleges introduce in terms of what constitutes salary - is it take home income, is it pre-tax salary, is it the cost-to-company, etc.

Second, the placement is only an indication of how a group was valued by those wanted to recruit that sort of talent. The individual differences are lost in the statistics. May be the guys who got good jobs are those who also had good soft skills, something that the program did nothing to add value to. I have no doubt that there is a difference in the market value of different academic programs. But that difference cannot be judged by placement figures, since they get biased by the presence or absence of certain categories of students. For example, if we assume that CS education adds more market value than Chemistry education, the degree of the difference cannot be judged by placement figures alone. The difference in median value of salaries offered will be partly due to CS versus Chemistry, but also partly due to the fact that CS program attracted much higher ranked individuals to begin with. So some difference is due to the set of individuals who are joining the program. (If you believe that 100 percent of the difference is because of the discipline, then essentially you are admitting that you bring no value as an individual to your recruiter. Your recruiters should be careful with that attitude.)

Third, placement data is for last year's batch. You will be graduating 4-5 years from now. What will be the market valuation of your discipline/program, we don't know. Things may change in the interim period. Need to take that into account as well.

Fourth, you are looking at last year's placement data because you want to be rich. Fair enough. But there is no study done looking at the correlation between the first month salary, and the career earnings. If you want to be rich (and why not), you should be looking at a decent living immediately after the graduation, but most importantly, a hefty salary after 30-40 years. Your work life is going to be 50 years. Too bad, but you will be working till the age of 75. (Even today most people retire between 65 and 70 years of age, and with the rapid advances in medical science and technologies, it would be at least 75 for you.) Today's placement data gives you absolutely no indication of what your salary will be 30-40 years from now. In fact, I just did a small informal check up on my batchmates. We entered IIT Kanpur about 30 years ago. I guessed the incomes of people by their designation/company/location, etc. For some people, I knew. And based on this very crude, unreliable, statistically insignificant experiment, I can still confidently say that the median income of Computer Science graduates of our batch is nothing to write home about compared to many other programs.

Fifth, the emails that I have been getting, they only ask for "scope" of two programs whose perceived economic value is very similar. Nobody has ever asked me for an opinion between English literature in Delhi University versus Computer Science at IIT Kanpur. The perceived economic value of the two programs are so vastly different, after all. And if you consider money in such a situation, you are being normal. Almost everyone will prefer 10 lakhs over 2 lakhs, even if one has to give up one's interest. But in India, we are worrying about very small amounts of money. I never get asked to compare Computer Science with Mining. I only get asked comparison between EE and CS, or between Mechanical and Chemical, etc. And the reason for all the confusion, all the stress is that between these two programs, the difference in the median salary last year would not be more than 5 percent, and people are really trying to figure out whether this difference of 5 percent will stay for 4-5 more years, and indeed 40-50 more years. When the difference is 50 percent, then people automatically assume that the difference may come down or go up, but it will never be erased. So they are sure in such circumstances which program they want to put ahead of the other.

Now, if the economic value of a degree program differs from the other one by only 5 percent, you should know that your interest and passion in a program can at least cover up that 5 percent difference. In fact, the interest and passion can often compensate very significantly in a job/career, much more than the 5 percent.

Across the length and breadth of this country, parents and students are not even attempting to know their interests, their passions, their aspirations, in the hope of getting that 5 percent extra money. Worse, many people will directly, shamelessly, tell me, that they are interested in one discipline, but will like to get admission in the other, because of a possibility (not even a guarantee) that they might get slightly better placement. And that bothers me.

And what I feel is that if so many of our youth at an age where they should be most idealistic are willing to give up their interest and passion for 5 percent, when they grow up and become less idealistic (and more "practical"), would they not be willing to give up a principle or two for 10 percent. Wouldn't such students in an academic institution be willing to do plagiarism, copy an assignment, and cheat in the exam.

And what would they be willing to do for 100 percent, just the imagination of it fills me with horror.

I don't think Anna Hazare will be successful.


ദിനൂസ് |dinoos said...

This is an interesting observation. Probably the link you've high-lighted is the reason for the corrupt society around us.

gautam said...

Dheeraj, I could not agree with you more. Placement statistics can be very misleading.
I would like to add another dimension to this decision making process. Let us assume that higher the JEE rank, the brighter the student. So you have a rank of 100 in GE and you are delighted that you will get into CSE at IITB. But hang on! You are likely to be no. 40 in ranks among those who will get into IITB CSE. So what is the chance that you will graduate with 9 points? Very small. More likely, you will get 7.5 to 8.0 with a lot of hard work. Now come placement time, companies are not going to ask your JEE rank. They are going to compare you with the other IITB CSE students. Now, suppose you join IIT Patna. Starting rank last year was 2925. So you will a star at IITP! With normal effort you will be a 9 pointer. You will get one of the best jobs that will be available to IITP CSE students. That job is not likely to be much different than the top job at IITB (moneywise; check the max salary at all the IITs that have graduated students. IITP hasn't yet). So why the hell are you going to join IITB? Because money deos not matter. Or does it? You decide.

Rathod said...

very relevant article indeed and an interesting observation to conclude with. I could not help but think abt the movie 3 idiots where the stress again is on doing what is your passion. The courage and freedom to follow one's dreams shud be ingrained right frm childhood. A profession shud not be glorified by the success others have achieved in it.

Dheeraj Sanghi said...

Gautam, you are right. I wouldn't want to be the last person in the department. However, the way parents look at it is that the "median" or "maximum" salary for IITB is likely to be higher than IITP. And, of course, my son has been the topper all through. He just had a bad time on JEE day. So he got only a rank of 100. He will obviously be topper 4 years from now. And, he will obviously bag that top job. And, therefore, select IITB over IITP.

Anonymous said...

Excellent and very very relevant analysis, Sir.

However, I think parents, or students themselves, don't consider placement alone; it is always thought of in conjunction with the perceived rank of the department. It just so happens that this perceived rank, which is precisely the order in which dept seats fill in, often goes hand-in-hand with the placement ranks. Which is the cause and which is the effect, it is difficult to tell.

An interesting comparison, from my times at IITK, would be Aerospace/MME - perhaps the only tie that didn't need breaking from placement/perception angle. Out of the two, people always chose in terms of claimed interests/passions or on account of what they thought was cooler. All other branches were anyway hierarchized. So if someone chooses Chemical over Mechanical, the reasons for perceived superiority of the latter over former are so well established that one needs to explain one's reasons for having done so repeatedly, never the other way around.

The point is that even though placements hold a huge currency in parental vocabulary, often it is one of those questions that finally contribute nothing to the decision. It is as if decisions are already made for everyone. You only need to have a logic for making an alternative decision. With the new IITs the matrix must have become more complicated but I am sure on a general basis, the choices are still made on the outside than in any other terms.

Biswajit said...

Regarding having to choose between CS and Mining, I have been on both sides of the fence. Worldwide numbers suggest that median Mining/Petroleum grads make more than median CS grads today. The situation is so bad that schools are getting few graduate students and are in danger of losing the faculty pool.

More importantly, there is a false dichotomy between disciplines that is engendered by super-specialization. The interconnected of disciplines is obvious for anyone who has worked as an engineer. Perhaps CS grads can afford to know nothing about other fields, but mining engineers definitely cannot.

vishv said...

Hi Dheeraj,
I love your postings and try to read when I can. This time I thought I should start adding my thoughts too. The core of your advice is in the last two paragraphs (or three if I count last one line as a para). Rest is introduction -- needed but distracting to your main message!

I call it distracting as it pushes us into a wrong message -- money is not the end goal. But even if it is, those who are chasing last 5% advantage in the start salary in 3 to 5 years time may be losing big money by going into a discipline that may not be their dream discipline.
I would say that students must chose the discipline that they have passion for. For this is where they will excel and this is where they will find satisfaction and success lot easier than in a discipline that they may chose because of social prestige or peer pressure.
To steal idea from your last para, I have my opinions about Baba Ramdev but I do not think any IIT graduate beats him in success. Baba is in a discipline he loves and has passion for. That alignment propels him to the height he has attained.
So I suggest to everyone, to follow your own dream not your parents or advisers. But first watch that movie "3 Idiots".

Koganti Vamshi said...

Sir, i don't agree with you are telling.. i don't think people with real passion chooses the branch like that...if a student completes his 12th class and does not know much about disciplines,then they will choose the branch which fetches him better salary..I know some people who choose a less avg. salary branch Aeronautical Engg. ,when he get a very good rank in JEE...In general people wont have much passion to do pursue a particular discipline at the completion of 12 so parents come in and search the high paying branch for him.... you can contact me on hope to see a reply from you...

vishv said...

Dear Koganti Vamshi, I see what you are saying; and you are not wrong from the experience you are speaking. But, I feel sad that the culture and education system that we have does not prepare us, even after 12 years of schooling, to see the world without the aid of others. We still rely on others to tell us what is good for us!

Worst is that it is the JEE rank that determines your choice of discipline. If you are rank 100 you are good computer hack; if it is 100 you are great Aero enggy. Do you see the joke that this poker machine is playing?

By the time we complete year 12, we are (nearly) legally adult -- ready to choose the nation's government; but still not ready to say what we like!

Around the world, the notion of schooling is to provide basic skills that every citizen should have. Education beyond schooling is option of taste and ambition.

Unfortunately, IIT admission treadmill takes the youth at the prime of their life -- when they must see and experience world -- and forces them into years of blinkered life preparing for JEE. These bright kids when they win the admission lack wisdom to know there own strengths and desires. They are Hanumans who do not know that they can fly!

Where are Anna Hazare and Baba Ramdev of our education system? Fortunately, not every bright kid in the country is dazzled by IIT glamour.

Dheeraj Sanghi said...

Koganti Vamshi, To add to what Vishv has to say, the theory that 12th class student does not know enough has only limited validity. To an extent, it is obviously true. But the point is whether s/he tries to gain that knowledge.

Does anyone ask a question like, "I am good at Mechanics, I like algebra, but not calculus, I like to play with electronic gadgets, I hate Chemistry. From these can you suggest what my aptitude may be for." Or "Are there sites on the Internet where I can answer a few questions, and find out my aptitude." No. Never.

When people are willing to do huge amount of research to find out the placement statistics, why are they not willing to do something about finding their own interests and aptitude.

Second, I get enough emails of the type, "I have an interest in X, but I want to do Y for the sake of placements." How do you explain these.

Third, let us look at the students who are into the system. Clearly, in one year, one would have been exposed to some additional information about various disciplines, not necessarily as courses but just talking to seniors, etc. What happens during the branch change process. Out of 92 CS students, is there even one who says I don't find CS exciting. Out of 100+ EE students, is there even one who says I don't find EE exciting, except for joining CS. In IITK, we also have a course called, "Introduction to Profession" in the first year, which each department tells their students what to expect in that discipline. So there is a lot of exposure, and we still see the placement rules the choices.

So the theory that 12th class student does not know enough has only limited validity.

This is, of course, not to say, that every single person does this. There are exceptions.

Prashb said...

@ Dr. Sanghi

I agree with the problem you have identified. But I would also like to point out that the issue, in part, has been created by the JEE itself.

At this point of time, India actually has some good creative, technical talent. Open source developers, robotics enthusiasts, aero-modelling enthusiasts with creativity, students who represent India in Robocon, Microsoft Imagine Cup and hey, the IIT Techfests itself which for IITians themselves are opportunities to showcase their "organizational and management" talent alone.

If JEE is unable to identify this talent it is because it is a standardized test in Physics, Chemistry, Mathematics which is being used to allocate to seats for everything from Computer Science to BioTech to MSc Economics. I do not downplay the importance of a standardized test but if that is the only admission criteria then it is only reasonable that IIT should expect standardized products.

The JEE does often correctly identify good theorists ( Karmakar, Vinod D, AKS Primality, etc. ). Moving forward, it should also have some way to identify people who have practical talent as well. It should be able to give us a Zuckerberg or Blake Ross(Firefox visionary) as well.

I personally feel that before and after a student joins IIT, there absolutely needs to be focus on some liberal arts/humanities component to help students develop creativity, structured and independent thinking, apart from language and communication skills.
Caltech has something like 25% Humanities courses in their curriculum. IIT needs to have at least six to eight such courses - maybe a small one every semester.

Please take a look at this video of an interview with this year's JEE topper.

The kid is young, probably nervous in front of the camera and English might not be his language- so we can excuse him to some extent. He is obviously a genius and no one is denying that. But I find it unfortunate that he is going to IIT only because "IIT was a dream" and he has no specific interest etc. to talk about. The lack of a goal beyond getting into IIT is a problem most IITians have and this interview shows it. I have seen many such cases get burnt out at IIT despite top notch AIRs at JEE.
Also, a brilliant student like him should not be so handicapped at articulating himself. This is where language skills and liberal education will fix a lot of issues.

Basically the JEE currently involves social suicide so you should not be surprised at the attitude to life, interests, values etc in JEE-qualifiers.

Dheeraj Sanghi said...

@Prashb, I fully agree with you that JEE has serious limitations. The problem is that IITs have never articulated what kind of student they are looking for. Are they looking for students who would after a training in an IIT would be successful in certain set of careers, or are they looking for students who would just do well in the studies at IITs, and it does not matter what they do and what they achieve later on. This is just one example of the decision that the admission process has to take.

Regarding curriculum too, I am in full agreement with you. In the 60s, IIT Kanpur decided to have 20% of the courses to be humanities, when typical US universities had 25%. 20 percent itself was too radical for the times, but since then it has been a steady decline, and today we have less than 10 percent. Last year, I had written three articles in EDU magazine on curriculum, emphasizing the need for more broad-based education, and taking HSS component back to at least 20%. IIT Gandhinagar has decided to have close to 20% component to be non-science, non-technical.

Dheeraj Sanghi said...

@Prashb, somehow I did not get the link to the video. But thanks for mentioning that. I could search for it, and I thought readers of this blog would be interested in the video. So here is the link to the Interview of JEE topper.

gautam said...

The central issue that faces the IITs and the govt is: how do we ensure a level playing field for admissions? Any scheme that selects students based on their special skills is subject to objection through the legal system as being unfair. So we have ended up with a single exam with Physics, Chemistry and Maths (why these three? Because they are taught in all schools in India - equal opportunity!) which we all agree has severe shortcomings, many of which have been pointed out in this blog. But will any of the alternatives be acceptable to civil society and to the legal system? That is the problem. It is not the attitude of the IITs, their faculty, etc., that is the issue. The IIT Council is in agreement that the JEE system needs change. The Chairman of the IIT Council (Minister Kapil Sibal) wants the Board exam results to play a major role in deciding entry into higher educational institutes, including IITs. All of us in the Council are in agreement. But how to implement this? Two years have gone by, and no solution is in sight. No solution is in sight because the issue is complex. There are 30+ Boards in the country with different syllabi (some progress has been made in getting an uniform set of syllabi), different examination norms, and, most importantly, different levels of trustworthiness. If anyone can suggest a practical, implementable solution, everyone will be happy. Please remember Education is in the "concurrent list". This means that both the State Govts and the Central Govt can handle education. The centre therefore has little power over decisions made by State Assemblies.
But please look into reactions to any scheme and our ability to handle these reactions (eg migration of students to "easier Boards through dubious means", inflation of marks by Boards).

To pu ti another way: any scheme to identify the best talent for a particular course is likely to be accused of being subjective, and hence discriminatory. When demand so outstrips supply, this is a major stumbling block.

Dheeraj Sanghi said...

Gautam, I fully appreciate the complexity. In fact, I am convinced that there can be no scheme of equating marks of different boards that can be acceptable to all stake holders. So, can we solve the problem partially. The idea of looking at school marks has been that it represents longer term learning, it includes language, it encourages a different (read, better) pedagogy than that of coaching classes, etc.

If we can't consider the 12th class directly for ranking the students, and if we can't keep a very high cutoff, since again the same issue of comparison will come in, can we at least make the cutoff higher than what it is right now.

I would like to see JEE do some research. How many candidates are denied admission because they get less than 60% (let us talk of unreserved only). My understanding is that the number is about 0.1 percent. What would be this number if the cutoff was 61, or 62, and so on. We should also see what boards do these students belong to who despite being successful in JEE will not get IIT admission.

On top of that, one needs a bit of long term research looking at correlation between school marks and performance at an IIT.

Such research will give insights to all members of JIC and JAB, and I think they will be take better decisions. For example, if you notice that at 65 percent, you only bar a couple of percent of students that most of these students are from CBSE and not from a "tough" board (like UP Board), then it stands to reason that the cutoff should be increased to 65. (And if you can show a correlation between 12th class marks and CPI, your decision becomes unassailable.) And if you deny admission to 1-2 percent students, immediately you will increase the seriousness with which students take schooling. (It will also increase repeaters in JEE, but that is a small price to pay.)

And if 60 percent is legally defensible, 65 percent will also be legally defensible. (If BITS can defend 80 percent across all boards without any normalization, IITs should be able to defend 65 percent, but someone needs to do the research.)

The point is that while we may not be able to solve all problems, can we attempt moving in a certain desirable direction. A bit ad hocish, may be, but better than no movement at all.

There are other things that JEE can consider. Is it really necessary to have equal number of questions in PCM. Can we do some sort of correlation study with marks in individual subjects and their performance in IITs, and decide that some subjects will have a greater focus.

When I gave JEE, there was an English test also. Today, it is perhaps not politically possible to have an English test, but what about a language test to be given in top 5-7 languages.

It may not be possible, but I am only giving examples of things that JEE can do. The impression one gets (and as you know, I would have talked to at least 10 JEE Chairmen and Vice Chairs) is that most people want "their" JEE to run smoothly. Experiments can wait till next year, unless forced in some way.

vishv said...

Well, mates, today I have decided to hurt everyone. Without exception. But, first let me mention that there were several unintended typos in my previous posting to this article. Second 100 was intended to be 1000. A use of "there" should be corrected to "their". And, many more.

Now let me spray bullets randomly. I arrive at Kanpur central late in night and I have two 10 rupee notes and SBI card. My friend has arranged me to be picked from station to IIT but taxi is late. A youngish girl with obviously malnourished child comes along and I give her a tenner. A minute later I am surrounded by a large group each more deserving of help than the other. What do I do? I ask them to leave me alone. I ask them to work. ... Finally, I choose the most meek-looking, small, and relatively clean kid and I hand my second tenner and walk away, hoping that world is a sane supporting place again.

Now let me press the trigger. Take JEE. Replace English with Hindi. Replace urban arrogance with humbleness and meekness. Let the keenness and hope remain unaltered. Am I not back to my previous para. Half the IITians do not shower everyday anyway:-)

The problems are same. Solutions are same. The country can not throw unbounded amount of resources to provide placement for every capable applicant that comes out of JEE. Criteria for reward will remain arbitrary and will never please everyone. But, still IITs and JEE loves the situation as it preserves and nurtures the sense of elitism in their product.

But the real solution is once again in providing awareness of alternatives. Providing training and resources to other opportunities, inside IITs and outside IITs, for youth to pursue. Why have we ended up with a system that values IITs to the exclusion of other avenues.

Reason is simple. Their is shortage of real opportunities in the real world. When an employer needs to fill a position, they find it assuring to slot an IIT graduate preferrably one that is ranked higher. (Skills for most jobs can easily be learned relatively quickly on the job. Four years in a discipline are really not needed!) In turn IITs provide this assurance by only training those that are at the top of available year 12 passouts.

So solution one must look for is not in the reform of IITs or JEE but the wider system. Solution is in diversity rather than in stratification of education systems.

gautam said...

Last year (or was it two years ago?), Kapil Sibal had suggested that the Board cut-off should be increased and made around 75-80% (I forget the exact number he suggested). Unka Putla Patna aur Lucknow mein phook diya gaya (His effigy was burnt in Lucknow and Patna). There were protests that bright students studying in poorly equipped schools, especially in rural areas are not able to do well in Board exams because of the poor quality of teaching. The JEE exam gives them a chance. Reference was made to the "Super 30" of Bihar during those protests.
Since then, the question of raising the cut-off is not being discussed.
Some of us have been pushing for reforms in exam paper setting. One of the suggestions that has been made is that there should be a school teacher in each paper setting board. After all, IIT faculty are not the best judge of what students are taught at school. But, I agree there is resistance to change in this area. Confidentiality is used as an argument against bringing in "outsiders". I agree the syndrome of "is that most people want "their" JEE to run smoothly. Experiments can wait till next year" is also there.

rahul said...

I have also read in a newspaper that he wants to become an IAS officer. Then why is he joining IIT!!!!! I think the IIT system is a complete failure except for branches like CS and EE.And why everyone is talking about leadership, why not research ? What about science branches? Does technology means CS and EE only. MIT,Caltech (both institutes of technology have very good science departments too ) IITs should introduce a seminar course in the first year where a faculty member from each department will give presentation on research scope. Students should be encouraged to publish their research work and such students should be given scholarships. Others who are in MBA mentality should be charged heavily.