163 – Learning as teaching

Posted on September 3, 2012 by

0


Blogging is at its best when it generates a conversation that elicits new ideas and garners new perspectives. Earlier this month, Diplo’s Hannah SLAVIK did just that when she posted a blog asking herself and others whether teaching 20’000 students at the same time is possible. Her post offers Coursera as an example. It recently registered 680’000 students in 43 courses. She writes: “When comparing this to Diplo, we are overshadowed and yet we have almost 100% completion rate. Compare this to the small percentage of 160’000 students enrolled in Stanford’s MOOC on artificial intelligence that completed the course.”

SLAVIK decided to enroll in one of these mega courses…alongside 19’999 others. Week 1 generated some 400 discussion threads, some with over 50 postings. And so began her search to marry limited time with the desire to distill the best of what was being discussed. And what of evaluation? Can two tutors accurately assess the progress made by 20’000 students? SLAVIK ended her post by asking readers whether they thought this methodology (or elements of it) could work for professional training. The replies Hannah got made for interesting reading – and this comment.

I recently came across a description a middle-brow American Food chain – Cheesecake Factory – which provides three-course, “fork-and-knife” sit-down restaurant meals that most people across the country couldn’t previously find or afford[1]. The firm serves 80 million people per year.

Unlike fast food chains, which serve the same fare (burp!) year in year out Cheesecake renews its menus significantly twice yearly. It faces the challenge of ensuring uniformity and quality of experience throughout their chain of restaurants in regular cycles. How does it go about teaching their staff, which is scattered nationwide, new dishes (over ten dishes for each cycle)?

“Once the chefs in the test kitchens figure out how to make each recipe reproducible, appealing, and affordable, they teach the new recipe to the company’s regional managers and kitchen managers. (…) The Cheesecake Factory’s instructors also train the attendees how to teach what they were learning.” Not only do they rate the quality of the prepared dish; the rate severely the attendee’s ability to transmit his knowledge to the next level. One has “learned” only when one has proven that he can teach others – and this all the way down to the last server. The success of the chain is based on this vital element of speedy and fool-proof transmission from the experimental kitchen to each and every restaurant countrywide: if the chain of learning-teaching is broken anywhere, the reputation of the whole chain suffers. In a twitter world it can be terminal. Learning as teaching – the firm’s survival depends on it. Is there a lesson on which to reflect?

Learning as teaching – suddenly I remembered my own university years. After the lectures we would congregate in self-help groups and teach each other the subject matter. We would simulate the exam environment – questions would fly. As soon as one of us got stuck, others chipped in, and no one was left behind, nor was a subject left untreated. Confronting difficulties of others made me aware of difficulties I might have overlooked; it also enabled me to give better presentations. More fundamentally: teaching led me to a sense of empowerment: I felt I mastered the essence of the material and its structure.

I knew “more”, but also I knew how to use it in a concrete and immediate context: “learning as teaching”. Ah, yes, this system was fairly unfair. Fair ladies always got to be part of the group – without having to ask for admission. This was compensated by the fact that the exam tested absolute knowledge levels, not relative knowledge among students. As we helped each other, we all managed to pass.

I still practice “learning as teaching”: my blog entries are my way of ordering and organizing reading and reflections into memorable (in the literal sense of remembering) mental states. I might say, by writing I transform floating knowledge into a personal APP (see my 159) which I know how to use. There is a bonus too. It is also a creative process (see my impending 164). Entries are mostly born inchoate, with only a general sense of how they should end. As I develop the text, new ideas crop up, words connect to new thoughts, and suddenly the “general sense” congeals into what appears to me to be a powerful conclusion. When correcting policy papers I’d drafted my boss used to say: “It is not about dribbling – you have to shoot the goal” and back I went to redraft the conclusions as something more than mere summary. That’s what education is truly about: until I’ve successfully shot the goal, I have not learned. Teaching as learning is as close to “shooting the goal” as it comes without being in the game proper.

What does this all mean for learning and e-learning  in particular?

We all live in an age of “economism” – economic concepts pervade our thinking, and we hardly notice it, nor see the weakness of analogies that derive from it. Much recent economics has centered on “supply side” measures. As they enable economic actors – the line of argument goes – the subjects will inevitably and predictably act on the new opportunity. Some kind souls have described this as “voodoo economics” or “pushing on a wet noodle”.

Everyone wants education. So “reaching” tends to be identified with “teaching” – Kerron Ramganesh has made this very useful distinction in response to Hannah’s blog.

On the “supply side” of teaching there is no dearth of effort – though there is an issue of quality control (as the saying goes – “there is a sucker born every second”). The efforts that have gone into making the learning experience more appealing and enjoyable have been successful (though we still blather “no pain, no gain”). This is good so, for the amount of knowledge we need to incorporate is growing every day and wrapping it in silk eases the swallowing.

An immediate problem arises when students perceive education as “laying on of hands”. The argument goes somewhat like this: the teacher has a contractual obligation to deliver. If the student fails the exam, he’ll sue the teacher for failing to do so (no kidding – this has happened already). The exam has transmogrified from validation of the student’s effort into one of the teacher’s.

Then there is another issue: “attendance” may be taken as proxy for learning. The number of comments is taken as measure of commitment. These criteria are all predicated on the view that quantity is a good proxy for quality. My experience with e-learning foundered in part on such “objective” criteria. There was significant variance between what I subjectively perceived as a student’s “knowledge” based on the exam essay and what these figures were telling me – yet the attendance numbers had the last word.

The shift from elite to mass education after WWII threw up the problem of fair testing of large numbers. Multiple choice tests were introduced. I never did experience them, so I have no hands-on familiarity, but it seems to me that the shift made collegial preparation more difficult. Rather than teaching each other “winning” narratives that would impress the professor with our mastery of the subject as a whole, one was now training to “spot the error” in the text[2]. Testing oneself in preparation for a multiple choice exam focused on going through past exams and measuring one’s score – a procedure that involved checking the correct answer against the lecture text. When there are too many detailed questions, group preparation becomes inefficient, because transaction costs increase. Multiple choice testing also allowed increase in the material for the exam: cramming and sharing don’t go very well together. Multiple choice testing probably contributed to a decline of group learning[3]. Grading in accordance with a set distribution curve (and guaranteed failure %) reinforced the trend toward solitary preparation: helping others may make one’s position along the distribution curve insecure.

Admittedly – the system “reached” more – but was it “teaching”?

e-Learning strengthens this trend toward isolation. Students are scattered around the world. They can e-mail each other – but only across time zones, with ensuing delays. This is hardly the face-to-face experience of my youth. In my experience on-line sessions are undisciplined, and one-liners are no way to teach each other anything – or learn from the tutor. My first encounter with webinars was no better.

I’d argue here: It’s time to change direction and concentrate less on “reaching” and more on “teaching” – as measured by the success of participants teaching others – as in the Cheesecake Factory.

How to do this? Actually there were experiments in the 60s along these lines – like having older students teach younger ones in school. It got involved in politics, and got dropped. The kernel, I think, is sound. Learning is not so much a personal as it is a social process: “I know because I am able to transmit my knowledge”. Short of hands-on experience with acquired knowledge teaching is probably the best indicator that we have properly and usefully acquired knowledge. So let’s do more of it.

One approach might be to replace the current “one-many” module (a virtual class of 20’000?) by a cascade of many small clusters of students. Each group would include both “teachers” and “learners” (an optimal ration would have to be determined). All would be graded. Final grade in the course would be predicated on two components. First the student has to pass the usual knowledge test and second he must have participated successfully in a second phase that involves him teaching a group the same material. True it would slow down the “throughput”, but what time is lost up-front may be recovered by the enhanced validation of knowledge through teaching. Knowledge acquired in this way endures.

In an office setting this may be achieved by injecting the learning experience in a “quality circle”. Rather than giving diplomats leave to take courses individually, MFA would manage such a learning experience involving a self-help group.

This is just an opener: let’s find new ways of “learning by teaching”.

Education is an “enabler” – it does not follow that empowerment automatically follows acquisition. We must test for empowerment – the best measure I can see is testing the teaching success. Diplo may do worse than reflect on this. Do you have suggestions?


[1]           Atul GAWANDE (2012): Big Med. The New Yorker, August 13.

[2]           I won’t enter into a discussion whether “spotting errors” is a good proxy for mastery of a subject. It seems to me, from my armchair, that there is a difference between aiming to master a subject coherently (as I did) and spotting incoherencies.

[3]           Other factors were involved. Mass throughput forced universities to place exams at the immediate end of the teaching period, so the process could be reset. At university I was operating on a yearly cycle.