Search This Blog

Tuesday, June 19, 2018

A Week in Olin: Not a College but a Lab in Engineering Education

Ever since I met Richard Miller, President of Olin College, in 2010 at IIT Gandhinagar, I had been dreaming of visiting Olin, a small private engineering college near Boston. Each batch of students is less than 100, and unlike its well known counterparts, it offers only under-graduate program. The total student body on campus is about 350. Fully residential. The first batch of students was admitted only in 2002. So it is a rather new college.The degrees it offers are in "Electrical and Computer Engineering," "Mechanical Engineering," and simple, "Engineering." Under "Engineering" program, students can design their own specialization and get it approved by the college.

Olin is not an ordinary college. It is really a research lab to test ideas on how to teach engineering, and the students love to be guinea pigs there. It is one of the most difficult college to get into in US.

Every year, they conduct a one-week Summer Institute, in which any university in the world can send a team of 3 faculty members or more, and they explain through a variety of workshops and personalized sessions their pedagogy and indeed, their story. This year, it was from 4th to 8th June, and I was lucky that J K Lakshmipat University, Jaipur, where I am working as an academic advisor, invited me to join two other faculty members of the university to attend this event. A dream come true for me. This year, there were 18 universities from 11 countries in 5 continents. (And a very positive side effect was the friendships we made across the globe and across the border.) They conduct this summer institute because it is one of their missions to promote quality engineering education pedagogy.

The keywords that you would hear at Olin all the time include "Project Based Learning," Design, "Student Centric," Inter-disciplinary, and so on. And these happy to be keywords that you would hear these days on any campus, in any presentation, but at Olin, they are not just keywords to be spoken, but values to be lived.

Typically, "Project Based Learning" means that we will assign a project to a student group in the beginning of the semester. This is in addition to the 40-lectures, exams, quizzes, and so on. We will tell the students to come back to us if they ever had any difficulty. And at the end of the semester, we announce a date when we will look at their presentation and demo. No student comes to see you throughout the semester, since obviously they are so smart that they can't possibly have any difficulty. But one week before the demo, you start receiving requests for extension of deadline. You agree to a one week extension. On the day of the demo, you can see that everyone is having red eyes. They have done all the work in the last 48 hours.

Olin is different. A project means that the students meet faculty at least twice a week, for 100 minutes each time. In a typical session, there may be 5-6 groups with two faculty members. Larger number of groups would mean additional manpower support in terms of under-graduate TAs called NINJAs in the local lingo ("Need Information Now; Just Ask"). In our institutes, we will keep worrying about under-graduate TAs helping unfairly their friends with better grades, hence avoid using an excellent and cheap resource that most of us have. With this kind of interaction, the faculty knows exactly what each group member is doing, where they need support, etc. There are several intermediate goals in terms of defining scope of the project, mid-term reports and reviews, etc. So the students make progress throughout the semester. Also, the projects are not assigned from the top. Students decide their projects, and hence they are really excited about doing them. And invariably, they have to be useful in real life. And in many courses, there are no scheduled lectures. You learn even complex material while working on the project, largely on your own. The faculty can guide you, tell the right material to read, may even explain an occasional concept, but largely you learn yourself. Of course, not all courses are like this. Every course has something unique, and there are some courses which are the regular 40-lecture courses.

The very first workshop was to tell us what students do on day 1 of their first semester. They have a course called "Design Nature." On day 1, they go to a nearby forest. They are told to observe nature and do a project that replicates nature. You could replicate flying, or swimming, or hopping or whatever. We were put in groups of 4, given a few cardboard pieces, rubber bands, springs, tape, scissor, etc., and asked to make something that hops. It wasn't simple. Initially, we were to do things individual, then in groups of 2, and finally in groups of 4, every time discussing with others, sharing our design and improving the design. This was a lesson not just in design but teamwork as well. They later showed us the "hopping devices" that students make in about 5 weeks. Very impressive for someone who has just joined college.

Another very interesting workshop was about how they teach a course on History and Material Science. Yes, a combination of history and Material science. Two professors jointly teach this course. Once again, we were put into the same situation as the students on first day of this course, put into groups of 4. There were many consumer goods on a table on the side. Each group was to pick up one thing. Our group picked up the bottled water. Now, we were told to think of questions that we could ask about this product from a material science point of view. Well, we started thinking why is water packaged in plastic, are there different types of plastics, could there be other materials, how do we test if a material would be good enough for this purpose. Is plastic polluting water by any chance. Is the water quality in the bottle substantially better than the quality in the tap water that we are willing to pay so much for it, and so on. They told us that each group would come up with their own questions, and would actually learn about different materials, their properties, and indeed use a wonderful material science lab to test those properties. In parallel, we were to think of questions that would be interested in from the perspective of history. We thought hard, and felt that we first need to think of the purpose that the bottled water serve today. Then we should think of whether that is a purpose unique to modern society or was it needed in historical times. If yes, how did they solve their problem. We felt that bottle was really a water transportation mechanism. Then we wanted to study how did historical societies transported water in large (river to a nearby city) and small (a person, or a family traveling) quantities. We were then told that the old Roman empire had built aqua ducts, and this would lead to studying of the Roman civilization, etc. And the small quantity was transported initially by earthen pots and vessels made of animal skin, and later, after bronze age, vessels made of copper and bronze. So when did bronze age start, what was its impact and all that. So if we had continue in this course for the entire semester, we would study all that, mostly on our own.

Not only it was very interesting, but also, every group in the class was studying different things. They said that they would make sure that every one learnt a very small core of material science. But beyond that, they would learn different things. I can't imagine in our universities (or any other ones), the curriculum to be so flexible. But come to think of it, if one student learns Indus Valley civilization and the other one studies Roman or Greek civilization, what is wrong. Learning of history teaches us critical thinking, and that would happen to both groups. And what both groups have also learnt is to learn oneself. Those are important skills. And, of course, communication, and team work.

Another interesting class was the "User Centric Design." Again, groups of 3-4 students. They will brainstorm in the beginning of the semester as to what is the target audience for whom they are designing some dummy product. Then they will actually go and meet many such persons. Find a problem to be solved. Design a product that solves it. Discuss that design with the target group. They told us that students are keen to actually build the product right from the beginning of the semester, and they have to be told that there would be many other courses and projects to build things and that design is an equally important challenging component of finally coming out of the product.

We hear many times statements such as "we want every student to succeed," and we feel that these are some of those motherhood statements that we should keep speaking without meaning them. But at Olin, they really care for their students. Continuous feedback throughout the semester, evaluation of motivation levels, and if one sees the motivation levels going down, then researching into reasons for it, and fixing the way the course is taught (or not fixing, if it is felt that handling stress is a learning outcome in the course).

I found the final year project to be very exciting. They have decided that it has to be something that makes a real impact and not the toy projects that we often do in IITs. There are two kinds of projects. One, working on a problem in collaboration with a company. Typically a large team of 5-6 students and 1-2 faculty members, and they take up really challenging problems. They had a lot of posters about the kind of problems they have taken up in the past few years in their corridor and it made a very impressive list. The other possibility is to join an ongoing multi-year project under the category, "Affordable Design and Entrepreneurship" (ADE). Projects under ADE are typically to design solution to a problem faced by disenfranchised class whether in US or abroad. The team formation for industry supported projects is very interesting. Depending on the skills required to do a project, the faculty will decide the team. However, each student can give a list of two students with whom they don't want to be in the team. So teams are formed keeping the small negative lists in mind. They seem to have done a lot of research into team formation - when to form teams randomly, when to ask students to form teams, and so on. (Like they have done research on every aspect of teaching/learning process.)

Is this project based learning causing them to not understand the basics very well. Their own answer is that several students are going for PhDs in top research universities and are able to do very well in the research programs which require an understanding of basics. Also, I met a student who had spent her 6th semester in MIT. I asked her if she had any difficulty in doing courses at MIT which has huge focus on basics. She said that she had no difficulty and performed well in MIT. The industry seems to love their graduates.

I had a detailed discussion with a faculty member of Maths regarding a course on linear algebra, which is considered absolutely crucial for engineering in most places. And he argued that it is not necessary to do a 40-lecture course when all you need to learn is which library routine to call for what kind of problem. If we give up on understanding of basics, we can use that time to learn higher order skills or different skills which are more relevant for an engineer, we would be making a better engineer out of our students. And depend more on just-in-time learning for whatever has not been taught. So your program needs to ensure that students can learn on their own whenever they need that knowledge.

And Olin is not just about experimenting in the classroom. They have experimented with other things too. For example, none of their faculty members is permanent (tenured in US parlance). When they started, it was said that they won't be able to attract and retain quality faculty, since they would prefer to go to other universities who provide permanent jobs. Olin has been able to attract and retain absolutely outstanding faculty.

And finally, I think Olin is where it is because of the leadership of Prof. Miller. The founding President (Vice Chancellor or Director in our case) is so important to set up the right processes and culture.


PK said...

Thanks for the detailed description. It is good to see that Olin is taking project based learning seriously. In some of the cases I have seen, it is used mainly to save money as the academics/administrators think there are no need for lectures. The end result is that the students do not learn any principles and cannot really solve any new problem.

ratan said...

The question is whether the Olin experiment is scalable?

Waqar Saleem said...

Hello from across the border :) You have done a good job summarizing the event and your thoughts resonate to a large extent with mine. Our team from Habib University plans to deliver a workshop to the rest of the faculty about our Olin experience and your blog just made our job easier!

There are certain points on which I would like to share my thoughts. I am breaking this into 2 parts to meet Blogger's word limit for comments.
- I like how you define Olin as a lab to research ideas on teaching. That is very apt. My university, Habib University, now 4 years old, has also been set up with similar goals, though I think we are floundering a bit on the implementation.
- I am sure every teacher wants to be the most effective that she can be for her students. That is why she got into teaching in the first place. But I think where things begin to fail is the amount of work required to reach that level of effectiveness. The university also wants effective teaching but has not devoted time to exploring what exactly that means, i.e. quantifying effective teaching, and then how exactly to achieve it. Olin, on the other hand, spent a lot of time planning and experimenting before students were inducted and continues to experiment with the students. In this, it makes sure to recruit the right kind of student, i.e. one who agrees to be experimented on.
- Useful work in the initial stages requires the right leadership. Olin's first employee was an engineer and faculty member himself, so he was intimately aware of the changes required in engineering education. He could then recruit accordingly.
- Faculty coming out of a rigorous PhD views research as the pinnacle of academic achievement. And that is what they themselves aspire to. Doing anything else will be heretical and a betrayal to their professors and their discipline. I feel it takes a great deal of courage for such a PhD, especially a recent one, to let go of that and make effective teaching their goal. Olin and a few other institutes provide a safe environment for such outliers and USA has a sufficiently large pool of PhD's for the existence of a reasonable number of such outliers. But they are constantly questioned by the larger academic community. One of the Olin faculty members told me that he is constantly asked if he is into teaching because he could not cut it in research.
- Teaching the Olin way, i.e. with a great deal of student interaction and with a lot of thought devoted to course design, takes a lot of work. The required annual teaching load at Olin is 3 courses. Faculty at other places do not have the role modes nor the sufficient background, neither personally nor institutionally, to teach that way. Their institutes will choke at the thought of an annual load of 3. Pressed for time, faculty defaults to teaching the same way they were taught. I remember when I first started teaching, I had lots of cool ideas, some of which I implemented in my first few courses. With time, I find less and less time to do so - the ideas may well be lost by now. I would design one bland course due to lack of time thinking that I will improve it next time. That next time never came and the bland courses are now my default.

Waqar Saleem said...

- Experimentation necessitates failure. Olin is very much open to this. My institute, Habib University, also is but we are not as explicit about it as Olin. I am not sure if that is a stance that established institutes do or even can take. They would thus feel secure in traditional, tried and tested measures.
- Coming up with good projects that place the course content in context is hard. My Data Structures course, for example, is mostly a delivery of the material in the book. I would love to motivate the content through real life use cases but I do not have them; my initial, feeble efforts to get some bore no fruit; and I now have a set, boring formula to teach the course. Faculty is not willing, trained, or motivated to put in hard work of this nature.
- Faculty fall in the trap of thinking of their students in the same way as their recollection of their own student life. "I regularly coded till 3a.m. when I was a student.", "I loved proving the theorems of Linear Algebra from first principles when I was a student." And they only see that kind of student to be worthy of success and of faculty attention. They may not see the need for any newfangled approaches to teaching because the process they underwent as students worked great for them. However, faculty are statistically a rare bunch - how many from their batch went on to do a PhD and become faculty members?
- In the third world, we are not innovators but followers of the first world. CS is big in Pakistan not because computing is solving major problems in Pakistan but because it allows access to a good number of jobs mainly servicing the first world. For the CS student, the best case would be the opportunity to ultimately use their skills to move abroad. As such, we are training "code monkeys" and are products of the same system ourselves. In such an environment, neither the need is felt nor is there any genuine motivation to innovate in education.
- I like your comment on the use of math. I could not attend the "adventure" on quantitative analysis at the Olin workshop but your comments remind me of this TED talk. I think you will like it.

Again, thanks for penning down your thoughts so eloquently!

Dheeraj Sanghi said...

Sorry for late publishing of comments and response. Somehow blogger was not informing me of new comments.

Dheeraj Sanghi said...

@PK, the way Olin does Project Based Learning, it requires more faculty time and not less. So it is more expensive for a university to replicate.

Dheeraj Sanghi said...

@Ratan, for me the question is not whether the Olin experiment is replicable. For the sake of argument, let us agree that it is not replicable. I would think that the question is what can we learn from such a model and I think there are several things that we can learn. (Also, our specific project at Olin was about how to do as many things as Olin does without having as many faculty or a lot of money. And we learnt that there are indeed many things we can do.)

Dheeraj Sanghi said...

@Waqar, Thanks for your comments. It was really nice meeting you at Olin. And the TED talk is awesome. I am going to share that on my timeline on FB/Twitter. And I fully agree with your comments on how a typical university is unable to do what Olin can do, and why the enthusiasm of a young PhD soon gives way to more research and less teaching. Thanks a lot.

Rahul Agrawal said...

Thank you Prof. for such a detailed blog. Olin really shows the way ahead. I especially liked their 'History and Material Science'. Maybe some of our new institutes can give this type of learning a try. Thanks a lot.