Can a Tier 2 institution have a culture of excellence

I have been to a large number of Tier-2 and Tier-3 institutions in the country, and at every place, I make it a point to argue that they can imbibe a culture of excellence, that they can do better than what they are currently doing. This is almost universally resisted. Typical argument on the other side is that a culture of excellence can only develop in a resource rich environment, where the quality of students and faculty is high, and all that. "We don't get IIT type students," "Our faculty is not like that in Stanford," "Our budget is only 10 percent of IIT Kanpur," it goes on and on.

Is it really true that we can only have a culture of excellence in a resource rich environment.

Let me change gear briefly and talk about one more experience. IIT Gandhinagar organizes a 2-day event every winter. On one day, they invite a select group of academicians from around the world to discuss excellence in academics at IITGN. On the second day, they invite people from academia, industry, government, alumni, and discuss excellence in non-academic issues at IITGN. It has been going on for a decade and I have been luck y to have been invited to each of these events (and I have attended all of them). I love those two days at IITGN as there is so much learning for me.

In one of the early years when they hadn't even graduated the first batch, they discussed excellence in one of these meetings. After a lot of discussion, it was opined that IITGN should try to be like a top class institution. You can not be the top ranked university in 10 years, but you can be like a top ranked university very soon. I wasn't sure if I understood the import of that statement fully. But after a few years, I started realizing how important that statement was.

Let us consider teaching. A tier-2 institution would argue that they don't have faculty of the same quality as Tier-1 institution and hence cannot offer teaching programs of the same quality. Even if this is assumed, can Tier-2 institution imbibe some of the cultural aspects. The difference is not just in the competence of the teacher (which in today's world does not account for much) but how other aspects of the course are handled. If you are teaching programming, for example, your students may not be capable of building a large project within that first year course (actually, one will be surprised, if one were to try, but let us leave it for now), but do you even give them a programming assignment every week, and do you even penalize when students copy. Those things do not require competence and rich infrastructure. Every engineering college today has a course on communication. Do you only do some grammar and have an in-class presentation, or do you ask your students to read up 4-5 books in the semester, may be even watch a few English movies and write a critique. (And the teacher can spend 5 minutes with each student to figure out whether that has indeed been done honestly or not.) Today, you don't need competence to suggest a few books/movies, and you don't need rich infrastructure for students to read/watch. One of the biggest problem with teaching/learning in Tier-2 institutions is that we don't engage students beyond the contact hours. And that requires neither very high level of competence nor resources.

Consider faculty recruitment. All IITs would go out of the way to attract faculty. They would all have a mailing list of people, typically department heads in top places, or alumni who are in academia, to whom they will keep writing. They would meet PhD students in conferences and tell them about faculty openings. They would invite PhD students closer to graduation to visit the department and give talks, etc. How many Tier-2 institutions do this. To create a mailing list of heads of 50 departments from where you expect most of your faculty to come will take a student in your department not more than a few hours. Sending an email to this list every 2-3 months will take 5 minutes of your time. Similarly, putting out this information on platforms like LinkedIn would take a few minutes, and all this is free. How do you treat an applicant. Do you think you are doing a favor by offering a job, or do you think that the process of recruitment is not just about "evaluation" but also about attracting that person to join you. Do these things require competence or resources. Of course, not.

When we do performance evaluation, clearly the criteria for recruitment/promotion, etc., would be different in MIT, IIT, and NIT. But what cannot be different is the need for performance evaluation. The need for performance evaluation and accountability is independent of competence and resources. If you start doing time bound promotions in academia, that is the end of excellence.

How do you treat your students. Are they kids to be "controlled" or are they grown up who can be trusted with many a decisions. You would find that in most Tier-1 institutions, students have a far greater freedom and a role in institute administration, while in Tier-2 institutions, there are all sorts of restrictions. Do you need a lot of resources to have students as members of various committees in the Institute.

There are so many other aspects of an academic institution where becoming "like" a better institution (or inculcating a culture of excellence) can improve  quality significantly, but these examples should suffice to make a point.

There is no doubt that greater competence, and greater resources can improve things a lot, and I would certainly like to see a larger funding of all academic institutions in the country. But one can be "like" a good institution even with lesser resources and we can have a higher quality of academics than what we see currently in our Tier-2 and Tier-3 institutions. What we need is a culture of excellence.

Industry Interaction with Engineering Education

It is now a cliche that engineering education in India, barring a few honorable exceptions, is of very poor quality. I will take examples from CS and while I do recognize that there are differences between education/training in different disciplines, I believe the larger picture remains the same.

We have been talking about poor quality of our graduates for a few decades. But things seem to be going from bad to worse. Why?

Is quality education too expensive and a poor country like ours cannot put enough resources to it. To a small extent, yes. To be in the same league as top 500 in the world is expensive, but to make the graduates employable is not so expensive.

Is the issue of quality education too complex and there are no easy answers. Well, again, it is complex if we want to be the best in the world, but making the graduates employable is not so complex.

So, what is it. The real issue is that we haven't identified the real problem, or perhaps we don't wish to. When we talk about our graduates, we keep talking about them being unemployed because they don't learn the latest that industry is using, that the curriculum is outdated, that we don't get industry folks to come and teach, or tell the faculty what to teach. That they don't learn soft skills. And to justify this line of argument, one can show examples where a college started teaching something which was immediately needed in the industry and had a few soft skill workshops before the placement season, and the number of students getting jobs improved.

No one asks a question that if indeed those few skills were the main problem in employment, then how come those things help only a few thousands and several lakhs are still unemployable.

To just pick up one skill (programming) as an example, the problem is not that our graduates do not know the LATEST programming language (the one which is more heavily used in industry today), but that our graduates do not know ANY programming language. Most of our graduates cannot qualify GATE without dedicated coaching for months. Most of our graduates cannot write even a pseudo code for any simple algorithm (say, just sorting), or for any simple data structure (say, insert in a binary search tree).

If you believe that the problem is only the lack of knowledge of the latest industry stuff, you will do what we have been doing for decades, and be happy that you have helped a few students. But if you believe that the problem is that our graduates know nothing, you can actually do things that would improve the quality of education significantly.

I have talked to a lot of technical folks in industry who would tell me that if their problem is not that the recruits do not know the latest, but that they can't learn the latest in a small period of time. That the training periods are too long and even after that, they haven't picked up as many skills as industry people would desire. If they knew, for example, a couple of programming languages well, they can pick up the next programming language very soon. So indeed, even the enlightened folks in industry would agree that the problem is not lack of latest knowledge, but lack of any knowledge.

In terms of pedagogy, the problem is easy to solve (at least to the extent of bringing the students to the level of being employable), but only if there is a will to solve. Why do computer science graduates do not have programming skills when they are supposed to have had two full courses doing just programming, and at least 10 other courses in which they would have done programming projects. Because, no one insisted that they do those things and gave them marks anyway in all courses in all 4 years. And the students were not self motivated to learn on their own.

At least in computer science, you don't need great teachers to reach the level of being employable. (Of course, great teachers can motivate you to do better than just being employable.) There are enough online resources to learn. If the teacher can just give assignments (any book would have a number of them) and make sure that students do them honestly, submit them, and they are graded honestly, the problem is solved. It is really that trivial. (Again, I am only looking at them becoming employable in the current market scenario, and not competing with the best in the world, for which a lot more will need to be done.)

But trivial things are sometimes hard to do. Our affiliated colleges have no incentive to do this (and most of our higher education happens in affiliated colleges). If one college starts getting strict with internal marks, their students will have poor ranks in the university results since other colleges will continue to give liberal marks. And there will be student protests and they will attract fewer students next year.

Even universities do not have an incentive to do these things. In a government university, you can't penalize an instructor who does not want to work. In a private university, if the failure rates go up as they will in the beginning, there will be market pressure on them. In fact, our accreditation bodies will see this as negative that a number of students are having backlogs.

So our regulatory bodies take an easy way out. Keep blaming the university for poor curriculum. Keep blaming industry for not offering enough internships. Keep blaming schools for not preparing them with adequate soft skills. Sympathize with colleges because there is overall shortage of faculty and resources. But never admit that the problem is within our colleges and universities and never take action against them for poor quality education.

Frankly, if the quality is poor for whatever reason, shouldn't it be reflected in denial of accreditation to a large number of institutions/programs every year.

All this is not to say that there should not be any industry interaction with engineering education. But the real benefits come from internships, both summer internships and semester internships, or part-time jobs. Outside India, it is common for students to do internships/jobs. Would Indian students be interested in graduating in 5 years. And, by the way, if I were to encourage my students to do industry jobs during the BTech, both accreditation bodies and ranking bodies would not like it. They only want industry to control my curriculum and teach my students the latest stuff.

In summary, the problem in Indian academia is NOT lack of industry interaction. The problem is that there appears to be no incentive for any honest evaluation of any learning outcome. Lack of quality faculty ensures that students don't get motivated to learn. So students will learn only if it improves grades, but that will not happen because there is no honest evaluation.

Subsidy on MTech programs to be reduced

Last week, IIT Council has taken a series of very important decisions. One of the most important decision was to reduce the subsidy on MTech programs in IITs. It was decided that over the next three years, the tuition fee will be increased to make it same as that for BTech program, which is currently, Rs. 2 lakhs per annum. The current tuition fee is Rs. 50,000 per annum. However, students who are getting an assistantship (which is almost all the students), pay only Rs. 10,000 per annum. So, for all practical purposes, the tuition fee is being hiked 20 times over the next three years. The reduction of subsidy does not stop there. It is also proposed that the financial assistance of Rs. 12,400 per month that is given to all those students who have qualified in GATE (which means pretty much all the students in IITs) will also be stopped. Of course, IITs can provide support to some students by recruiting them as Teaching Assistants, Research Project staff, etc.

When I saw the news, I thought it was a step in the right direction with some minor caveats about a proper implementation strategy. However, over the next couple of days, media quoted three IIT Directors regarding their reasons for reduction of subsidy (all three had identical reason), and I was aghast, to say the least.

They argued that a lot of people join MTech programs and then leave when they get a job. That leaves a large number of vacancies in their MTech programs which could have been used to train other interested students. And that this increase in tuition and removal of assistantship will discourage those students who are not serious about the MTech program.

Really! This is only to discourage non-serious students. Would this not discourage serious students.

Let us look at the issue of students dropping out. Most of the students drop out in the first couple of months, because of PSU jobs being offered in the months of August/September timeframe. What is the extra cost for non-serious students: Rs. 95,000 extra in tuition, and Rs. 24,800 in not getting assistantship for two months. That is roughly Rs. 1.2 lakhs. Now, let us look at serious students. They pay Rs. 95,000 extra for all four semesters, and do not get Rs. 12,400 for all 24 months. Their penalty is about Rs. 6.8 lakhs. So in order to discourage non-serious students, you put a penalty of Rs. 1.2 lakhs on non-serious students and a penalty of Rs. 6.8 lakhs on serious students. How wonderful their logic is.

Could there be other ways to handle the non-serious students. If Rs. 1.2 lakhs is all it takes to dissuade a non-serious MTech student, what about taking a deposit of Rs. 1.2 lakhs from all students, and returning the money to those who graduate in time. Make it Rs. 1.5 lakhs, if your data shows that people leave after 3-4 months.

There are other possible solutions. If you want to admit 50 students, and historically you are seeing 40% attrition, and only 30 remain, why not admit 70-80 students. Yes, some years in some departments you might end up with more than 50 students, but in a large system some small perturbations can be handled easily. This is something that I have suggested dozens of times in different meetings, social media, in my blogs, over the last decade.

There is yet another solution. Shift GATE exam. Currently, the result comes out in the second half of March. PSUs despite being requested repeatedly take their own time to recruit. What if the result was out in January or December. PSUs would have enough time to complete their recruitment before July. In most universities, the core part of the program is completed by 3rd year, and 4th year is mainly electives and projects. So holding GATE at the end of 7th semester or even in the middle of 7th semester or even at the end of 6th semester is fine. In fact, holding GATE early would help a lot of things to be done well in time. It will help PMRF recruitment, your MTech/PhD recruitment, and also many students who are confused between industry and higher studies will move towards higher studies if they get a good GATE score before they get a job.

I wonder if any of these options were considered by IIT Directors.

A reader of this blog is bound to ask me this question. If the logic of IIT Directors is so flawed, then why did I welcome this reduction of subsidy on the first day. So here are my reasons for the same (some of them were informally discussed in a meeting of Deans of Academic Affairs of various IITs more than 5 years ago):

Long time (a couple of decades) ago, the PhD programs in India were very small, particularly in Engineering. MTech students were the primary research manpower we had access to, and creating new knowledge, solving problems is important to any society. Therefore, it was important to attract the best graduates to MTech programs. The best students would not want to be dependent on their parents after completing a college degree. Hence they needed large subsidies, both in terms of reduced or no tuition, and supporting them with a stipend. Today, the times have changed. Most engineering departments have a fair number of PhD students and PhD applicants every semester.

Also, for admission to PhD programs in Engineering, most universities required an MTech degree and hence keeping that pipeline full of best students was important. No longer. Most good places today would admit students after their under-graduate degrees to the PhD program.

Whether we like it or not, MTech programs in India are not considered "higher" than BTech programs. That is, it is hugely unlikely that a BTech from an IIT will join MTech program in the same (or similar era) IIT. You mostly find BTechs from Tier 2 institutions join MTech programs in Tier 1 institutions since they want to become equal to BTechs of Tier 1 institutions. Or BTechs from Tier 3 institutions will join MTech programs in Tier 2 institutions for the same reason. This has huge value addition for them which is reflected in their placement in the job market. Since it is only a 2-year program and as a result, the costs are lower, and there is a better predictability in terms of value addition, it is easy to take an educational loan and repay it, which is not so in case of under-graduate programs. Please note that educational loans were not all that prevalent 20-30 years ago, but they are now more easily available.

Further, as of now, since all MTech students are paid a stipend, most faculty members in IITs recruit non-students as their project staff who may leave as soon as they get a job. In fact, that uncertainty is far more detrimental to our research efforts. Now, I would hope that most faculty members would like to recruit MTech students as project employees. So anyone who does not want to take a loan or be dependent on their parents and have genuine interest and capability to do research would get recruited as a project staff (except that he would have to pay all the first semester charges anyway). So even if one wanted to consider MTech as a source of research manpower, their support should have depended on the quality of their research and not be without any performance criteria. This decision will improve research projects output in IITs.

If we do not consider MTech as primary research manpower, then the obvious question is whether under-graduate education is a more important public good, or post-graduate education is more important public good. I am sure there is no doubt that first degree is more of a public good than 2nd degree and hence the subsidy should be higher for 1st degree than 2nd degree.

Of course, any service provided by a public institution should be accessible to people on the basis of their merit, and hence appropriate steps need to be taken to support those few who may come from financially weak background and are unable to afford the new tuition. How will that structure work is yet to be seen, but I am confident that IITs will come up with some model.

So I am in favor of the decision but not for the reason that IITs seem to have done it.

Added on 9/10/19

Received an anonymous email giving an interesting reason why subsidy at the current high levels should continue for MTech programs. 

Only IIT under-graduate students get really high level of subsidy. Till now, for someone who got a few marks less in JEE, there was a second chance to get similar high quality of education with high level of subsidy. Now that chance is being taken away. Should India only support those students who do well in JEE, or should others get a second chance is an important question.

My reply:

It is an important question, and certainly students in India either should have access to quality education in very large numbers, or should get multiple chances to get access to quality education if the quality is going to remain limited. But we need to look at the details of how individual IITs would implement this. In any case, the maximum tuition will be same as UG tuition. So the same level of subsidy (in fact, higher, since MTech courses may require more specialized labs) will be granted by having same tuition. The way it is likely to pan out is that one may have to bear the first semester costs without financial assistance (except that tuition waivers for SC/ST/PD and BPL families will be there, as for UG students) but from second semester onward, most students will get either TAship or RAship, which may include some stipend and tuition amount. I recall that limiting the budget drastically for financial assistance was tried in IIT Kanpur (perhaps all IITs) around late 90s. I was responsible in CSE department for raising financial support for MTech students from external sources. We raised enough sponsored projects from industry that there was not a reduction of even one student in CSE department. So I believe that students doing well in Tier 2 institutions will continue to get a reasonable chance at a low enough cost to study in Tier 1 institutions even after this.

Compulsory Attendance at Indian Universities

The college life in India is considered as an extension of school life, not by students themselves, but by elders, including parents and teachers. One of the ways in which this manifests itself is the attendance policy in Indian universities. A large number of universities would insist that attendance be marked of every student coming to a class in every class of every course, including labs and tutorials. There is a certain minimum requirement of attendance, mostly 75 percent.

If a student attends less classes than that, there is a severe consequence, usually not allowed to take the final exam and given a fail grade in the course. In many universities, a fail grade would amount to a student being marked as a "backlogger" and no company coming for placement would touch him by a 10 feet pole. So the student not only has to repeat the course, which sometimes may mean spending a semester extra in the system, paying extra tuition, delaying earnings from a job, but it may also mean that there is no job through campus placement. Basically, the punishment for missing a few boring classes is very harsh with no consequence whatsoever to the teacher who delivered those boring lectures. After all, all teachers hired through proper selection procedures must be assumed to be great, and all students missing a class must be bad students.

Over the last 25 years of my academic career, I have had discussions on attendance in several forums and with a large number of colleagues in different institutes, and frankly, I still haven't figured out the genesis of capital punishment for missing classes. Here are the reasons that I have heard so far.

If students attend classes. they learn better. Let us assume this to be true. Well, if they don't attend and consequently don't learn, shouldn't the grades or marks reflect their learning. As a teacher, I want my students to learn, and if they don't attend my class and are not performing well, I can counsel them, and if they still don't learn, I must assign them the grade that reflects their learning. How many faculty members would take the pains of counseling students. How many faculty members would give a Fail grade to someone who has not learnt. If you don't want to do either of these two things, then forcing attendance is not for helping students, but for helping yourself.

Also note that many faculty members will also argue that if someone was ill, or if someone had a family member die during the semester or had other "genuine reasons", then we could be lenient with them. These people are not realizing (or perhaps they are realizing and still believe that it is the right thing to do) that they are asking for grades to be based on sympathy or "genuine reasons" and not academic learning. Consider two students. Both have attended 70% classes. Both have identical marks in all exams, quizzes, projects, etc. One had his father die during the semester and submits death certificate. The other had his father ill and submits medical certificates of his father. What would we do. We can't question the death certificate. (I am deliberately taking the most extreme reason to make a point.) But all medical certificates, particularly from a private doctor, are assumed to be fake. So, one students is barred from taking the final exam and is awarded a Fail grade. The other student is allowed to take the final exam, and passes the course. What have we done. Between two identical students, we have given fail grade to one, and pass grade to the other, simply because we had sympathy with one and not the other.

There are aspects of learning that happens by attending classes which cannot be evaluated. This could be true for some courses and may not be true for all courses. Can we have attendance requirements in some courses and not the others. And even in courses where some learning happens in classroom which cannot be evaluated, may be such learning can be quantified in terms of fraction of the grade. So just like we have in our mind various learning outcomes and we evaluate learning of those learning outcomes through exams, quizzes, projects, presentations, assignments, and so on, and assign some weight to learning of each of those outcomes, we could similarly assign some weight to learning of those outcomes which cannot be evaluated through traditional means. So if that weight is 10% or 20%, then absence can be penalized in proportion of those weights. Why award a capital punishment when a small deduction of marks will take care of matching learning and grades.

What is very interesting is that lately the regulatory bodies are almost forcing the universities to give students credit for online courses offered through Swayam portal. In such courses, the student studies online from wherever s/he wants, whenever s/he wants. That is, there are no classes. In some instances, there may be some discussion sessions at best. So one can learn well without attending a single class in 20% of the courses. But in the other 80% of the courses, one can only learn if one attends at least 75% of the classes. At the very least, this is an acceptance of the principle that in some courses, attendance is not required for learning. Once we accept this, shouldn't we then consider each course carefully to decide whether attendance is necessary for learning in that course.

There are non-academic learning which are important for careers. For example, you learn to discuss, communicate, dress up, pay respect to your elders (teachers) and what not. Let us assume that all these are indeed important for career growth, and we want to encourage them to learn these during their college days. Why not just put a small monetary fine. Students from poorer background will find it difficult to bunk classes. And rich kids, you don't have to worry about their careers. Their rich parents will take care of that. Let them pay fine and get away with it. May be the fine can be exponentially increasing with every course in which there is lack of attendance.

It is the discipline, stupid! The most commonly heard complaint is that if attendance is not forced then campus romance will flourish. After all, what will they do with all the free time. And that is somehow bad. And that, of course, will rise to indiscipline. Empty mind is devil's workshop or something like that. Again, let us assume that this is a genuine concern. But can we avoid this indiscipline by having a smaller penalty. Should capital punishment be the only penalty for missing some classes (and causing indiscipline in campus).

I hope someone can come up with a rational argument in favor of compulsory attendance, one which explains why online courses are fine, why missing classes due to some reasons is fine, and why a smaller punishment than failing the course and barring the student from campus placement will not work.