The Long, Slow, But Inevitable Death of the Classroom

If you read enough about Robert Solow, this quote coming up is but a matter of time:

You can see the computer age everywhere but in the productivity statistics

http://www.standupeconomist.com/pdf/misc/solow-computer-productivity.pdf

Much the same could be said about internet based learning technologies if you tried to measure it in colleges and universities before March 2020. We had lip service being paid to MOOC’s and all that, but if we’re being honest, that’s all it was: lip service.

Things have changed around a bit since then, I think.

We’ll get to that later on this post, but let’s go back to the seeing computers everywhere but in the productivity statistics bit for the moment. Paul David, an American economist, wrote a wonderful essay called “The Dynamo and the Computer: An Historical Perspective on the Modern Productivity Paradox“, back in 1990.

I think of this essay as an attempt to respond to the question Robert Solow had posed – why isn’t the data reflecting the ubiquitousness of the computer in the modern workplace? Read the essay: it’s a very short, very easy read.

Paul David draws an analogy between the move away from steam as a source of power, back at the end of the 19th century.

In 1900, contemporary observers well might have remarked that the electric dynamos were to be seen “everywhere but in the productivity statistics!”

David, P. A. (1990). The dynamo and the computer: an historical perspective on the modern productivity paradox. The American Economic Review80(2), 355-361.

Adjusting to a new technology, it turns out, takes time.

Steam-powered manufacturing had linked an entire production line to a single huge steam engine. As a result, factories were stacked on many floors around the central engine, with drive belts all running at the same speed. The flow of work around the factory was governed by the need to put certain machines close to the steam engine, rather than the logic of moving the product from one machine to the next. When electric dynamos were first introduced, the steam engine would be ripped out and the dynamo would replace it. Productivity barely improved.
Eventually, businesses figured out that factories could be completely redesigned on a single floor. Production lines were arranged to enable the smooth flow of materials around the factory. Most importantly, each worker could have his or her own little electric motor, starting it or stopping it at will. The improvements weren’t just architectural but social: Once the technology allowed workers to make more decisions, they needed more training and different contracts to encourage them to take responsibility.

https://slate.com/culture/2007/06/what-the-history-of-the-electric-dynamo-teaches-about-the-future-of-the-computer.html

Again, please read the whole thing, and also read this other article by Tim Harford from the BBC, “Why didn’t electricity immediately change manufacturing?” The article, by the way, is an offshoot of a wonderful podcast called “50 Things That Made The Modern Economy“. Please listen to it!

But here’s the part that stood out for me from that piece I excerpted from above:

“Eventually, businesses figured out that factories could be completely redesigned on a single floor. Production lines were arranged to enable the smooth flow of materials around the factory. Most importantly, each worker could have his or her own little electric motor, starting it or stopping it at will.”

https://slate.com/culture/2007/06/what-the-history-of-the-electric-dynamo-teaches-about-the-future-of-the-computer.html

Colleges and universities are today designed around the basic organizational unit of a classroom, with each classroom being “powered” by a professor.

Of the many, many things that the pandemic has done to the world, what it has done to learning is this:

each worker learner could have his or her own little electric motor personal classroom, starting it or stopping it at will.

In fact, I had a student tell me recently that she prefers to listen to classroom recordings later, at 2x, because she prefers listening at a faster pace. So it’s not just starting or stopping at will, it is also slowing down or speeding up at will.

Today, because of the pandemic, we are at an extreme end of the spectrum which describes how learning is delivered. Everybody sits at home, and listens to a lecture being delivered (at least in Indian universities, mostly synchronously).

When the pandemic ends, whenever that may be, do we swing back to the other end of the spectrum? Does everybody sit in a classroom once again, and listens to a lecture being delivered in person (and therefore synchronously)?

Or does society begin to ask if we could retain some parts of virtual classrooms? Should the semester than be, say, 60% asynchronous, with the remainder being doubt solving sessions in classroom? Or some other ratio that may work itself out over time? Should the basic organizational unit of the educational institute still be a classroom? Does an educational institute still require the same number of in person professors, still delivering the same number of lectures?

In other words, in the post-pandemic world…

How long before online learning starts to show up in the learning statistics?

Additional, related reading, for those interested:

  1. Timothy Taylor on why “some of the shift to telecommuting will stick
  2. An essay from the late, great Herbert Simon that I hadn’t read before called “The Steam Engine and the Computer
  3. The role of computer technology in restructuring schools” by Alan Collins, written in 1990(!)

Colleges: A Framework for Efficiency

I wrote an essay yesterday about unbundling college. You might want to read that first before tackling this essay.

In today’s essay, we’re going to decide how we’re going to judge the efficiency of college. What framework should we be using and why?

A Framework for Judging the Efficiency of College

Efficiency is actually fairly easy to define: maximal output for minimum input.

And input is also fairly easy to define: minimum resources to be used in terms of time and money.

It is the output bit that is rather more difficult to specify and define.

What is the output of a college? Here are some candidates:

  1. The number of students who graduate in a given year
  2. The marks these students score
  3. Particularly in the Indian scenario, the placement record of the college
  4. Number of seminars/workshops/outreach programmes conducted by the college
  5. Research output of the faculty/students in the college

Let’s work our way through each of these, and highlight the problems that present themselves:

  1. Number of students graduating per year:

    Do large classrooms – that is, classes with a very large number of students in them – work well?
    The obvious reason I ask this question is if you’re going to graduate more students, the number of students per class must go up.
    There is a large amount of academic research on the subject, if you feel like going through it. I haven’t read all of the papers on the subject, but the consensus seems to be that larger classes are necessary from an economic viewpoint, but don’t work as well as small classes would. Key points being, there ideally needs to be some sort of an opportunity to have a discussion with the professor, and that doesn’t necessarily work out well in a large class.
    I’ll speak of my own personal experience in this blog post, rather than cite academic studies. The largest class I have taught included 250 students, while the smallest included just two students. Neither of those extremes is ideal: 250 is more of a speech than a lecture, while two students is economically infeasible. My personal preference would be for a class size of not more than thirty.
    More students graduating per dollar spent by the university isn’t a great way to judge the output, or the educational outcome, of a college, because students don’t necessarily learn better in a large classroom. And in any case, if you are going to use this is a measure, online classes have offline classes beat!
  2. The marks these students score:

    Here’s a Hindustan Times article for your reading pleasure.
    ..
    ..
    I quote:
    “Compared to last year, the number of students scoring 95% and above has increased by 118.6% and those getting 90% and above by 67.48% this year in the Central Board of Secondary Education (CBSE) Class 12 results announced on Monday.
    A total of 38,686 students have scored 95% and above marks in aggregate of all subjects, up by 118.6% from 17,693 in 2019. As many as 1,57,934 students have scored 90% and above, 67.48% higher than last year’s tally of 94,299.”
    ..
    ..
    Are students this year twice as good as they were last year? If so, why? If not, are we just giving more marks this year than we did last year?
    Personally, I don’t think this year’s cohort is any better (or worse) than last year’s. It is quite likely that marks have been given more liberally this year than they were the last year. You can work your way through to what the equilibrium will be, but here’s my assessment: grades become mostly meaningless in the long run. Especially when they’re being handed out like this.
    Judging the output of a college by taking the average marks scored by the graduating batch doesn’t make for a great outcome either.
  3. Placement record of the college:

    I should state at the outset that I am the faculty-in-charge for placements at Gokhale Institute.

    Colleges, in an ideal world, should have skin in the game. Is learning a means to an end in itself, or is it the means to an end? If it is the means to an end, and that end is gainful employment, then should college not be paid only if the end is achieved? Income sharing agreements are, you might argue, a natural next step in this regard.
    But that being said, there is a part of me that dies a little when I think about the whole means to an end thing. Yes, employment matters, and yes, colleges should be held to higher standards than they are right now in this regard. But I would hate to live in a world where you go to college only in order to get employed. There’s surely more to life and education than that!
  4. Number of seminars/workshops/outreach programmes conducted by the college

    Goodhart’s Law. Goodhart’s Law. Goodhart’s Law.
  5. Research output of the faculty in the college

    See 4. above.

Honestly, I don’t quite know how to measure educational outcomes, and I’m not the only one. By the way, the first chapter from “The Case Against Education” is worth reading while thinking about some of the issues we’re speaking about here.

But this raises an important, and potentially problematic issue. Change is worth pursuing if what is new is objectively better than the old. But as it turns out, we find ourselves unable to settle on an objective measure by which to differentiate between the old and the new!

Consider the range of services that a college provides:

The range of services provided by college

There are four outcomes you get from college: grades, placements, peer networks and the degree. Of the four, we have spoken about and discarded grades, placements and number of degrees awarded as useful/appropriate measures of outcomes in education. That leaves peer networks, and this is perhaps even more nebulous and difficult-to-measure than the other three.

Of the lot, employability and the quality of peer networks are (to me) the most important. But the latter is literally immeasurable, and the former ought to not be the only benchmark for judging the outcomes from education.

There is, as it turns out, a problem waiting to be solved, and it is a fundamental one.

If I’m wrong, please let me know how!

Large classes, small groups

You might want to read my previous posts about online education and learning before reading this. See this essay about the state of higher education in India, this about signaling and bundling in higher education, this about unbundling college and this about measuring efficiency in education.

In addition, Aadisht had a great comment about optionality and higher education, which really deserves a post in its own right – but you can click on the link in this paragraph and scroll to the bottom to read it for now.

All that being said, today’s post ties together the thoughts and deeds of three people whose thinking I try to follow very carefully when it comes to online education.

The role of community in education

Let’s begin with this tweet from one of them, David Perell.

Both the thread of which 4. is a part, and the Twitter thread referenced in 4. are worth reading.

But today, I wanted to focus on the community bit.

A quick reminder: my thesis is that college sells you three things. The education itself, the access to peer networks and the credentialing. If there is to be an online model that will work for colleges, it must successfully provide all three (and more) at the same price (or less) as college does today.

When it comes to peer networks, can they ever be as successful online as they have been offline?

That begs the question: have they been successful offline? And that is really two separate questions.

About Peer Groups

  1. Are peer groups worth the effort in the first place?
  2. Is there something special about peer groups you form in college?

With regard to the first, I’m going to take a pass on answering it in depth for at least two reasons. First, I know nowhere near enough sociology to be able to speak about this sensibly for any length of time. And second, isn’t the answer obvious?

About the second question, you might want to read this essay – a part of which is excerpted below:

As external conditions change, it becomes tougher to meet the three conditions that sociologists since the 1950s have considered crucial to making close friends: proximity; repeated, unplanned interactions; and a setting that encourages people to let their guard down and confide in each other, said Rebecca G. Adams, a professor of sociology and gerontology at the University of North Carolina at Greensboro. This is why so many people meet their lifelong friends in college, she added.

https://www.nytimes.com/2012/07/15/fashion/the-challenge-of-making-friends-as-an-adult.html

The entire essay is worth your time, but the crux of it is those three points above: proximity; repeated, unplanned interactions; and college life.

Anecdote Time

Two of the best years of my life were spent while studying for my Masters degree at the Gokhale Institute in Pune. There was a fair bit of reading/learning involved, but most of those two years were spent in just hanging out with a group of people I am still close friends with.

And of the three things that Gokhale Institute gave me when I purchased a Masters degree from it, it is this group of friends that I value the most. Then comes the degree, and the least important – as it turns out – was the learning itself.

Don’t misunderstand me – learning was and is important! It’s just that for me, sitting in a class and listening to professors talk wasn’t the best way to learn. I have learnt much more by speaking one-on-one with some professors, arguing heatedly and passionately about random topics with friends, and by reading/listening/viewing to stuff on my own time.

But therein lies a dilemma.

How to reconcile online education with forming your college gang?

Random bike rides, conversations at three in the morning sitting on a ledge on the hostel terrace, giggling at a joke while sitting towards the back of a classroom is not just an important part of college. In my personal experience, this pretty much was college.

And not just during the pandemic, but even beyond, the key challenge is to figure out ways and means to achieve something approaching the same experience in this brave new online world of ours.

What might be an answer to this conundrum? That brings me to the second person whose thoughts about online education matter to me, Tyler Cowen

Small Group Theory, via Tyler Cowen

If you are seeking to foment change, take care to bring together people who have a relatively good chance of forming a small group together. Perhaps small groups of this kind are the fundamental units of social change, noting that often the small groups will be found within larger organizations. The returns to “person A meeting person B” arguably are underrated, and perhaps more philanthropy should be aimed toward this end.

Small groups (potentially) have the speed and power to learn from members and to iterate quickly and improve their ideas and base all of those processes upon trust. These groups also have low overhead and low communications overhead. Small groups also insulate their members sufficiently from a possibly stifling mainstream consensus, while the multiplicity of group members simultaneously boosts the chances of drawing in potential ideas and corrections from the broader social milieu.

https://marginalrevolution.com/marginalrevolution/2018/06/best-analyses-small-innovative-productive-groups.html

If you are going to run an online course, or are going to be a student enrolled in an online course, the most important thing you can do is think long and hard about forming groups.

If you are the person running the course, you need to make the process of forming a group as friction-less as you possibly can. Without these groups, not only are drop-outs more likely, but the groups themselves are perhaps the bigger point!

Here’s Tyler Cowen again, in a separate post:

Remember Lancastrian methods of education from 19th century England? Part of the idea was to keep small group size, and economize on labor, by having the students teach each other, typically with the older students instructing the younger.

https://marginalrevolution.com/marginalrevolution/2020/05/my-weird-lancastrian-method-for-reopening-higher-education.html

The post I quoted from is about how college might reinvent itself in the era of the pandemic, but the larger point he is making – or at any rate, the point I choose to take away – is about how learning in small groups is better than classrooms.

And on a related note, the third person whose thoughts on online education I choose to take very seriously, Seth Godin:

Great guy. Chip and I went to business school together. He was the third youngest person in the class and I was the second youngest person in the class. He got five of us together and every Tuesday night, we met in the Anthropology Department for four hours. We brainstormed more than 5,000 business ideas over the course of the first year of business school. It was magnificent. It wasn’t official, it wasn’t sanctioned. It was just Chip said let’s do this, and we did. And he picked the Anthropology Department because he knew someone there and could get the conference room.

https://tim.blog/wp-content/uploads/2018/09/138-seth-godin.pdf

That is from an episode from Tim Ferriss’ podcast, in which he interviewed Seth (the whole episode is well worth your time), but the point that I remembered was about small groups.

Anybody who is going to try and do education online is going to have to get small groups going. Without it – in my opinion – it simply will not work.

But how do you form these groups?

I’m still thinking about the how, and the more I think about it, the more it seems as if there is never going to be a perfect answer. Forming groups is hard, and I think we need to make peace with the fact that groups may not always work out.

People won’t get along, people will drop out, quarrels will take place even among groups that develop close bonds – there are many, many things that can go wrong. But it doesn’t matter how long it takes and how many times groups have to be formed and re-formed – it is unlikely that you’ll get an education worth the name without the formation of a group, or community.

And what do these groups do?

… will be the topic of tomorrow’s essay, for I was part of an experiment that tried to answer this question – and I really liked the answer!