Alex is concerned that education can be implemented via lecture-remotely. Alex is very simply wrong about what constitutes value in education. Value in college consists of:
- Contacts.
- Signalling.
- Practice.
How much work did you do is FAR better of a predictor than professor quality on learning. And so...star lecturers, dear Dr. Tabarrok, do not matter a bit.
The ever-insightful Dr. Caplan is wrong in a different way.
He is stuck in the book/abstract model of learning, as opposed to the action model of learning. Incentives matter, and marginal value matters a lot. The marginal cost and in particular the marginal weirdness factor of online colleges is dropping rapidly. 10% in 10 years is easy. The question for me is 50% in 20 years? Bryan's complacency reminds me of the newspaper industry 15 years ago. How many folks in 1994 realized that the business model of the newspaper industry would be dead in 15 years? Education is in the same spot now as the newspaper industry was back then. There are glimmers of a path that makes the model obsolete. But the fundamentals are thoroughly rotten, and there is room to exploit them. Someone, soon(and if I'm lucky, me) will build a different model of education that massively improves upon the current one. And college will in 15 years, be in the same dire straits that newspapers are in now.
The decline will start (has started -- death count: 2) with small, low marginal value, undergraduate liberal arts colleges, and then leak into community colleges and low-tier state schools. As soon as it leaks into schools with masters/doctoral level programs, the university financing models fall apart entirely, as undergraduates finance huge portions of the education system. High schools as holding pens will be impacted later, as the new model is accepted more. Harvard, other high-endowment doctoral programs, and other high-prestige programs will remain (anyone with several Nobels/equivalent on staff). But the prestige difference between Podunk State and UoPhoenix is not large, and indeed may not even be positive for Podunk State.
Second part of the decline is that evaluation systems will improve. Both IQ tests (Wunderlich) and Personality tests (Big 5-conscientiousness) are legal in many hiring decisions. Someone who tests well on real, deception indexed conscientiousness and IQ will be worth hiring out of high school instead of hiring a college grad at some level of test quality. Indeed, if Dr. Caplan takes his own education signalling posts seriously, he should be expecting some level of verifiable test quality to break the system, with some lag.
Posts
8 comments:
I tend to agree, but I don't think it will happen as fast as it did with the newspaper industry. I don't think high schools will ever move to online in the next 50 years. With that assumed, the culture of "going to school" will continue to be physically going to a building and sitting in a classroom. So students will tend to lean that way.
{humor}Plus most of us kids are dumb and we want to "go away" to college to drink, do drugs and have sex.{/humor}
There's a glitch in your data. Namely, me.
I agree with you almost entirely about education. Yet, to first order, I have never taught in my life.
What do you make of that?
(The only thing is my relentless autodidactism. I don't think that should really count.)
Alrenous,
I am not following your question.
I do not think it's necessary that someone has taught to get to my positions, but I am a teacher, and so have spent a LOT of time on education decisions. I'm glad that you, an autodidact, agrees.
To understand? Sure, they don't need to have taught. But to independently generate?
I, who have spent no time on (other-focused) education decisions, ended up coming to the same conclusions.
To be honest I'm also baffled. I didn't expect my ideas to survive contact with the student, even contact-by-proxy.
Ahh...ok now I see.
I think that almost everyone with +1stdv IQ goes through at least 17 years of schooling in the USA. I think it's hard to underestimate the amount of time an average, pretty smart person will spend thinking about how badly it's done.
I think my position as a teacher is actually a detriment to this policy. If I were not a teacher, it would be much clearer that action dominates talk...and so I'm not shocked that a clearly thoughtful student would have figured it out. I'm honestly more surprised when teachers do...it violates their sense of worth.
Personally, when teaching, I spend HUGE piles of energy working to not teach as much, show more, and have the student do more. In teaching corporate programming classes for a dozen years, the intro classeswent from 60% talk, 40% lab down to 15% talk, 85% lab...with support on the lab portions. And it has been a LOT of work.
Incidentally...if you're talking to highly knowledgeable folks...about an advanced topic on which they're already intermediate/advanced students, AND they're conditioned to ask questions...lecture is actually a decent to good channel for learning. It's just that college is no longer that.
I thought it might be that kind of practice-by-proxy that got me all knowledged up. Nice to know I wasn't building a mountain out of a molehill
If I ever try my hand I'm going to have to watch myself for over-talking as well. It'll be tempting to use the captive audience for ego gains.
Hmm I wonder if this has any effect on schoolteachers. /sarcasm.
I have a small tip you're helping to support, which is being explicitly goal-directed. If I were a lecturer, I'd start by writing on the board the hardest, most comprehensive question they'll be able to answer if the class is successful.
This also links into the reports I keep reading from you about failure leading to more learning. "Hey try this...it won't work, but it will give you an idea of how far you need to go and what you need to get there."
Partly I would want to do something like this so that I can get the students policing me. If I go too far off course, they could yank me back by reminding me to stay relevant to the goal.
I still need to down-regulate how much I expect to transmit by talking, though. Limit it to just enough that they can ask useful questions, I would guess.
It's kind of ironic that when the knowledge gap is the greatest, the students learn the least by explicit instruction.
Alrenous,
Here's problem #2...
Some amount of over-talking is useful if you're a big-personality type. Motivation dominates (almost??) everything else in learning, and big-personality (high-status) folks being interested is a MASSIVE social motivator. Most of the teachers who are really effective are motivators first, and teachers 2nd. But motivation has a fair amount of talk involved...
Goal-based education is huge. I find that in programming instruction, I get lots of mileage out of showing them what their current approach can't do.
Problem #3: Well designed exercises with fast feedback kick serious butt over talk (order of magnitude on effectiveness, IMO). HOWEVER, they also take about 2 orders of magnitude more work. So we live in a crappy talk-based learning world, because no one has put in the time/$. I'm working to fix that. January of next year, I'll unveil my (already started) math software (elementary ed first) that actually solves the problem.
My physics teacher in high school was hilarious. He was this huge, six foot plus guy, built like a football star, with a crew cut. Used to work developing submarine laser weapons for DoD.
He would get childishly excited about the Doppler effect.
Voice walking its entire range, speeding up, slowing down, gesticulating garrulously...
He would even sometimes say, "And there's Alrenous, sitting in the back, laughing..." because he really did get that excited. It was impossible not to like his class.
Problem #2.1.
Am I a big personality type?
*shrug*
Problem #3.1
Fast feedback can only be done on well-defined problem sets. Most real world problems are problems precisely because they're poorly defined and open ended. You're not sure how far the nearest solution is, and even if you find one, you're not sure you've solved the problem.
So fast feedback can teach skills but not scholarship. I think. No, perhaps a set of messy data problems...even include problems where the right answer can't be worked out with the given data...
Also just to say, bad as current 'best' practice is, it does provide plenty of practise with messy data and slow feedback. It's not purely horrible!
Post a Comment