Difference between Top educational institutions and the next

There is one question that has bothered me for long. The academic preparation or the quality in general of students entering IITs (and other top institutions) and Tier 2 institutions is only marginally different. But the quality of graduates from Tier 1 and Tier 2 is vastly different. From Tier 2 institutions, the best students do extremely well, even better than an average Tier 1 graduate, but if we compare the average Tier 1 and Tier 2 graduates, there is a significant gap.

(Is this gap only my biased perception? I do see a lot more IITians in the top rung of most tech companies. Is this bias and not truth. If truth, is this because of Old Boys networks of IITs. If there is no such gap, are all students and parents wrong in choosing Tier 1 institutions over Tier 2 institutions. One can have all these questions. But for the purpose of this blog, I will assume that there is a significant gap.)

Why a minor gap at admission time expands to a significant gap at the graduation time.

The obvious answer will be that the quality of inputs at Tier 1 institutions are far better than the quality of inputs at Tier 2 institutions. The faculty quality is better. The labs and infrastructure are better. They have access to more people who can help them in better curriculum and all that. Greater amount of resources can lead to better outcomes.

But I am not satisfied with this obvious answer. There must be other factors which are not based on resources alone. Are there cultural factors, for example. And the reason I am trying to find non-resource factors is that if indeed there are such factors, we can bring them to Tier 2 institutions and improve their outcomes without additional resources.

I believe I now understand some of those factors and sharing them in this blog.

Student motivation and aspirations: I think the biggest factor is that when a student joins a Tier 2 institution, the whole world tells him that he has failed in securing admission to Tier 1. This demotivates the student and his aspirations become lower. He stops working as hard as he used to do in school. Only a few remain motivated and hard working. And these students do extremely well throughout their careers.

Mentorship and Networking: In a Tier 1 institution, there are so many visitors, seminars, workshops, lectures. Industry leaders, successful alumni, top academicians, all are visiting the campus frequently. If a student attends even a very small percentage of these events, or interacts with very few of these visitors, that has significant positive impact. One usually finds seniors, recent graduates if not successful old alumni, who can mentor and advice. In Tier 2 institutions, one has lesser number of events due of resource crunch, but more importantly, when they do invite people, they are shy of inviting the top notch people. When a top person does visit a Tier 2 institution, the institution is too bothered about his/her comfort, keep a lot of time in the itinerary for "rest" and also not spend a lot of time interacting with students because the institution does not have confidence in how the students will behave. So the students don't get proper mentorship and advice during their academic program.

Culture: Excellence is a matter of culture. To give a simple example, if there is no attendance requirement, what percentage of students will attend a class in Tier 1 institution, and what percentage of students will attend a class in Tier 2 institution. How many students will try to get an internship on their own six months before the start date. How much is the role of students in taking decisions, giving them a stake in marching towards excellence.

Having recognized some of these non-resource factors, we have tried to make a difference in JK Lakshmipat University.

When students join us in the first year, we repeatedly tell them that not being in top 0.1% of the country on a given day does not define them. In any case, the country is not going to be built by 0.1% people. Many of the students when they realize that what they were considering a failure was only inability to be in the top few on a given day, they do gain some confidence and motivation. There are many sessions in our induction program from experts which makes them believe in themselves.

We have initiated a formal mentorship program in the campus. Every student gets a faculty mentor who meets them regularly and there is also a possibility of getting an external mentor to those who seek one, typically aligning with the career goals of a student. For example, those who are thinking of entrepreneurship, can get a mentor who has done a startup.

We have lots of visitors on campus from both industry and academia and we leverage these visits quite a bit. The visitors are not here to rest in the guest house. There will be people who will interact with them during their travel from airport/station to campus and back and once on campus, there is going to be a series of interactions.

We have a strong Student Mobility Program under which lots of our students spend a semester in IITs and other top institutions to get exposure to the kind of culture we want to develop here.

In our experience, these non-academic interventions really help in improving learning by our students and almost all of these don't require major financial resources to implement.

Software Industry Trainee Layoffs

 A few weeks ago, there were about 400 trainees in Infosys who were separated from the company. This led to two narratives.

Infosys narrative was essentially this: We recruit a large number of graduates without them being ready for the job because the Indian academia is poor quality (obviously, not their words). We look for students who are eager to learn. We make them go through a very high quality training. We invest a lot in them including paying them salary during the training period. Most of these trainees convert into our regular employees. But some have difficulty. We hand hold them, allow them three chances to prove the minimal learning. But every year, a few are not able to learn sufficiently well, and we have to let them go. This is the process we have consistently followed for the last 20 years, and the same process was followed this year too.

The narrative of the trainees was this: This wasn't a simple case of a few trainees not able to learn what they were taught. This was mass layoffs. About 400 trainees were laid off. The company suddenly increased the syllabus, reduced the training time, made the question paper tougher, and increased the pass percentage substantially. All this was designed to deliberately show this as trainees' inability to learn but it is essentially layoffs.

There is also an additional issue of how the activity was conducted with bouncers, and not allowing them to stay on even for just overnight, and all that. If true, I think Infosys could have been a bit more sensitive around the process but frankly, the main issue is separation from the company. So let us focus on that.

The trainees were asked to learn what they should have learnt in their four years. Every Computer Science program in the country would have a curriculum that has a significant overlap with what Infosys wanted trainees to learn. If they knew what they were supposed to have learnt in the college, the training and the testing would be a breeze. But it was difficult only because they didn't learn sufficiently well in the 4 years of college. No one has claimed so far that the syllabus of training was not related to a computer science degree program.

Students not learning much in 4 years of the college is not just a problem, it is an epidemic. But if you are enjoying your college life to the extent of ignoring academics completely, you are taking a risk. The risk is that you may not get a job at all. And if you get a job, you may not be able to learn new things quickly and get fired. And if you took this risk as an adult, you have to now face the consequences. You can't now point to others who also took the same risk, and were luckier, and demand that you be treated same as them. Remember, there were lakhs of students who took the same risk, and were even less lucky than you were, and they didn't even get a job.

Many trainees claimed that they were given a joining date 2.5 years after the job offer. This is certainly not helping your cause. Did you not apply for other jobs? If so, you weren't good enough for any other job. If not, did you use this time to improve your learning, or did you continue to enjoy this as an extension of college life.

Unfortunately, what happened in Infosys will keep repeating in near future.Thanks to AI, number of jobs in the low level IT services will go down and companies in this segment will reduce their hiring and increase their firing. So if you don't learn fast enough, you will be in trouble sooner or later.

If you are still in college, the lesson in this episode is that you must learn your basics well in your college years. Otherwise, AI will eat your job and you will be left to criticize the software industry.

 

Are Competetive Exams (like NEET, JEE) Fair?

There is a big debate about the fairness of NEET given all the allegations of paper leak, favoritism by some center superintendents, grace marks for loss of time, two correct answers for one question, and so on. They are all implementation issues which I am sure will keep improving year to year.

I think the bigger question is whether such exams like NEET and JEE are fair even if the implementation is perfect and there are no charges of wrongdoing. Almost everyone in this country (except a few in Tamilnadu) seems to believe that if the same question paper is given to two students (may be in their respective languages) then it is a fair exam. And the believers include most academicians as well.

What is fairness. What type of student should, in general, be ranked higher? Do we have an answer to this question? No. In the past, I have asked IITs whether they have a definition of what kind of student they want to select and why, and whether they have at any point in time checked if JEE (and now JEE Advanced) is selecting those kinds of students. If you don't even know what kind of students you are trying to select, how can you declare fairness.

Let us assume a definition of fairness. An exam will be fair if a student who has learnt what has been taught to him/her very well because this gives a university hope that the student will learn well in the programs to which s/he will be admitted.

Now think about it. A student from board of state 1, and another student from board of state 2 are given the same exam which is based on the syllabus of the board of state 1. Is this a fair exam.

This will be a problem with any exam under the concept of "One Nation, One Test." The syllabus of the exam cannot be the intersection of all state boards since that would be too small a set. And if you are going to test it on any other syllabus, its overlap with different state boards will be to a different extent causing unfairness.

Such a test is not just unfair to students, but it is also unfair to universities. Every program is slightly different and this may need different inputs. A university may have a program which requires a greater focus on communication and hence they may want to admit students with decent language skills. Should they not have the right to do so. Admission is basically a system of predicting who will succeed in the program or benefit most from the program. Can we say that the same test will predict success of all programs in the country.

Unfortunately, in a country with very few good colleges, the competition for those few colleges is so intense and the incentive to game/hack the system is so high, that fairness takes the back seat. Any agency conducting the exams will only care that the exam can result in a ranking of students and the process passes the judicial test of fairness. Who cares if it does not pass the academic test of fairness.

Given the reality of few good seats being chased by large number of students, is there no hope of at least improving fairness. There is. One can go back to having multiple exams, may be one based on each state board syllabus. May be there is an exam conducted by a private party which will include language and other skills. A student can try taking those exams which are closest to examining his/her strengths. And a university can choose the exam that it wants to use for admissions. It can even have some seats through one exam and other seats through other exam and so on.

There are two perceived problems in this. One is that students will have to take a large number of exams, and two, the earlier system prior to NEET was roughly this and it had a lot of corruption including capitation fee.

My answer to this will be that students will need to take 2-3 exams only, may be 4, which is not a problem. In fact, it is good that they will have something to fall back upon if they don't do well in one exam. Second, the corruption happened because the admissions took place at the college who could manipulate the merit list based on capitation fee. We can easily have a centralized counseling for admission. Each university can tell their criteria of ranking applicants and a centralized application server will get applications from students for all colleges they are interested in, and the seats will be allocated centrally. In fact, universities may be allowed any parameter which is objective in nature (to stop manipulation). They may even use grace marks for under-privileged students as long as the criteria for deciding under-privilege status is objective.

So the real solution to NEET is to have several smaller exams. That will be more fair to both students and universities.