What IITs Need Is A Scaling Up Of Undergraduate Programmes À La UC Berkeley

Much has been discussed about the hyper-competitive entrance to the Indian Institutes of Technology (IITs). Much has also been discussed about the faculty crunch at some of the IITs.

What has surprisingly not received enough attention is how the IIT undergraduate programmes have perhaps functioned in a suboptimal manner, not accommodating enough students per major, all the more so in popular majors.

A poor country like India can ill-afford to under-utilise institutes which bag a disproportionate chunk of resources for higher education. Even in popular programmes such as computer science, electronics, and mechanical engineering, a significant number of elective courses at the third- and fourth-year level are run with empty classrooms and barely 20 students. This is a travesty given the kind of stampede there is at the entry gates for the undergraduate programmes.

IIT Mumbai, for example, has admitted a grand total of 26 students to its BS economics programme and eight to the BS mathematics one. IIT Kanpur admitted 38 to its BS economics programme and 50 to the BS mathematics and scientific computing ones. Now, these are majors which do not require much access to the kind of laboratories required by engineering majors in areas like mechanical and chemical.

By the third or fourth year, when students opt for electives, chances are that most of the higher-level economics or mathematics courses will be run with barely 10 students each. With the appropriate use of technology, lecturers, and teaching assistants, teaching a course with 100 or even 500 students instead of 20 need not generate significantly more work for the professor involved.

Here’s a table with the total intake to some of the popular departments this year. Note that these numbers include the intake in both four- and five-year programmes, and in certain cases, these numbers increase by 5-10 per cent after the first year, when a limited number of students is allowed to switch streams.

With some efficient course management and an organised fleet of teaching assistants to assist the professors with non-academic aspects of the course, there is no reason why such theoretical majors shouldn’t enrol 100 or 200 students per batch. Not all assignments need to be hand-graded all the time – an appropriate use of computer-graded quizzes and tests will help conduct such courses efficiently. The constraint will probably be in terms of hostel capacity. Given that there is no real shortage of space at the IITs, hostel capacity should be increased to match the academic capacity, and not the other way round.

The computer science department primarily requires computing labs, and a good proportion of electrical and electronics engineering lab work often involves simulators such as PSpice, Matlab, Simulink, and kits involving breadboards, microcontrollers, and FPGA sets. All of these have become quite affordable over the years, not just for institutes but also for interested individuals. These are different from engineering programmes such as aerospace and mechanical, which involve a large number of courses and require physical infrastructure such as big laboratories, making them harder to scale up.

Architecture

Space should be used efficiently for both hostels and academic complexes. Having more “vertical buildings” with a larger number of floors and elevators as well as vast sprawls with just two to three floors is not an efficient use of public land.

If possible, advanced electives with few takers should be offered only every alternate year to ensure that they run with 30 to 40 students. So, a student gets a chance to take it once: either in the third year or the fourth. This can be done for courses which a third-year student is equipped to handle just as well as a fourth-year student. This will also free up slots from the faculty schedule and make them available for expanding capacity in those courses which have a high demand: typically, courses in computer science, electrical engineering, mathematics, statistics, and business.

Going overboard by obsessing about faculty-student ratios doesn’t help either, when it comes to scaling the IIT system. UC Berkeley has a student-faculty ratio of roughly 18:1. With 130 faculty members in their popular EECS department, nearly a 1,000 undergraduates graduated from that department in 2017. An instant contrast can be drawn up with IIT Bombay, where the popular computer science department, with 38 faculty members, sees only about a 100 undergraduate students in each batch.

Similarly, IIT Bombay’s electrical and electronics engineering department has around 72 faculty members, but there are only about 150 undergraduate students in each batch.

India would not want to see precious resources allocated for higher education used in such an inefficient manner.

Another major issue is that IITs set a number of seats per department while universities such as Stanford, Massachusetts Institute of Technology, and UC Berkeley expand capacity in different departments, based on demand, by funding popular majors and divisions accordingly. Students typically go through a few introductory “weeder” courses in the field of their choice, and their performance in those courses determines whether they get to major in that area or not. This alignment of majors with professional or academic goals prevents the wastage of education frequently seen in a large number of IITs, where aerospace or naval engineers immediately enrol in MBA programmes or take up software jobs as they were simply corralled into departments without a real say in the matter.

The famous Structure and Interpretation of Computer Programs course at Berkeley recently ran with over 1,700 students and it effectively used videos and technology. As per that reddit thread (hyperlinked), the second computer science course ran with 1,400 students. A large number of the introductory classes at Berkeley run with 500 to 700 students. This permits students to take at least a few courses in their areas of interest, even if they’re unable to complete the full major in these areas.

The issue of letting students pick majors at IITs will obviously bring up the question: what prevents the entire batch from opting for computer science or mathematics or electrical engineering? These majors will be far more popular than others, but, after a few semesters, many students will steer clear from these majors if they sense that they lack aptitude or interest in these, as it will directly affect their score or grade.

Top-quality universities in the United States generally fall into two groups: expensive Ivy League ones and the likes of Stanford and MIT versus top-tier public universities like UC Berkeley, UCLA, and Georgia Tech. The second group is generally a lot more affordable, but is also known for large classes. Georgia Tech routinely has classes of 200 even for sophomore-level courses. But this model is far more suitable for the IITs given the disproportionate nature of higher-education funds they bag in a poor country which needs to provide tertiary undergraduate education at scale. Large classes at the sophomore level will also allow students to explore areas beyond their own major. In major IITs like Delhi, Kanpur, and Mumbai, it is almost impossible to do even a fundamental computer science course like algorithms or a basic electrical and electronics course in digital design if one happens to be stuck in another major.

Redundancy in course offerings should also be minimised. For example, the website of IIT Kharagpur shows a fluid mechanics course being offered by the chemical engineering, mechanical engineering, and mathematics departments.

Instead of running so many variants of the same course, surely it’ll be more efficient for them to come together as units of the same institute and run one course catering to all of those who have it as part of their major.

The allocation of majors should be aligned with what students want, which is often governed in part by what the job market requires. MIT sees over 40 per cent of its class graduate from electrical and electronics or computer science-related courses. Stanford sees computer science as one of the most popular majors; a few decades ago, it was history. IITs corral disinterested students into departments with a fixed number of seats and students promptly leave those fields for software or management or finance after they graduate.

Expanding class sizes for courses in popular streams will help stop this coercive way of thrusting a major on to disinterested students.

The gist of this list is that with careful planning, judicious use of teaching assistants, teachers, and technology, better-planned vertical hostel architecture, and thoughtful scheduling and timetabling of courses, the IIT system can welcome a much larger student body, and it can also ensure that everyone gets to do a few courses of their interest beyond majors that they’ve been corralled into. All of this can be accomplished without significantly increasing the workload for professors by leaving some of the purely operational complexities to professionals from outside the faculty body. Given the public resources and the land poured into this system, it is imperative to use them optimally.