Tuesday, January 15, 2013
Turning In to the Skid
Growing up in western New York, I learned to drive on snow. Since I didn’t come from money, I learned to drive small cars on snow.
Small cars are relatively light, and the snow and ice around Rochester could be impressive. A small car braking on black ice is pretty much a hockey puck. So to survive, I had to learn the counterintuitive truth that Northern drivers learn early: when you start to skid, turn in the direction of the skid. You get control back much more quickly that way. If you refuse to acknowledge the skid, or fight it, you lose control completely and crash.
Correcting for climate, I was reminded of that when reading the latest from San Jose State.
Apparently, San Jose State University has contracted with Udacity to run credit-bearing basic algebra classes -- both developmental and college-level -- at a cost to students of $150.
Some folks are already manning the battle stations. My favorite, from the vice president of the San Jose State faculty union chapter:
“My personal opinion is that it’s not by accident that this is being announced at a time when most faculty are not on campus, but I have no evidence for that,” said Preston Rudy, a sociology professor at San Jose State who serves as vice president of the chapter.
It has to be sinister. It just has to! What other explanation could there possibly be?
And that’s where the conversation should start. What other explanation could there be? What’s the appeal?
Cathy Davidson claimed earlier this week that “if we (profs) can be replaced by a computer screen, we should be.” Her point was that it’s no longer plausible to argue that face-to-face instruction is clearly the only possible way to convey information. If the best instruction that a college can offer is a sage on a stage lecturing to 300 freshmen, whom that sage will then duck afterwards to get back to writing, then it’s hard to argue that a video presentation would be markedly worse. If anything, it may be better; at least with a video, you can play back parts you missed the first time. And the cost advantage is not to be ignored, particularly when tuition and student loan burdens are the highest they have ever been, even after inflation.
Davidson is gracious enough not to say so, but the dirty little secret we all know is that the massive lecture was only ever an economic expedient; it was never a particularly effective way to teach. Replacing one economic expedient with another, more effective one hardly constitutes an outrage.
The limits of the traditional approach are particularly clear when we look at student pass rates in developmental and lower-level classes. Nationally, there’s nothing unusual about a 50 percent fail rate for a developmental math class. Early MOOCs have had even worse attrition rates, but that’s hardly an apples-to-apples comparison; most enrollees in the first wave of MOOCs had nothing at stake. Motivation matters. San Jose already ran its “circuits and electronics” course as a blended MOOC, and found that pass rates were actually higher than in the traditional class. Whether the same will be true on the “lower” end of the curriculum isn’t obvious, but it isn’t preposterous, either. And if it turns out to be higher, I’d like to hear the argument against it.
And here’s where I remembered what it felt like doing my first donut on an icy hill in Mom’s Ford Escort in 1985.
If you read the earlier paragraph carefully, you’ll notice the word “blended.” Students in the blended class did better than students in the traditional class. They also did better than students in pure MOOCs.
To the extent that results matter -- as opposed to tradition, politics, Luddism, or technophilia -- it looks like we get the best results when we turn into the skid. When faculty use MOOCs as resources, rather than attack them as threats, students thrive. MOOCs could offer one way to ‘flip the classroom,’ to move exposition outside so the people inside could focus on understanding, applying, and questioning. They can free up faculty to work with students on the more interesting (and idiosyncratic) process of helping students internalize knowledge, come to grips with it, and sometimes even attack it.
In a sense, I’m suggesting using MOOCs in ways similar to the ways professors have long used books. They can be wonderful outside-of-class-time resources for introducing new material. (They can also be wonderful in-class resources for closely guided analysis.) But unlike books, they come with real-time data analytics, so they can be refined as they go. And they’re a hell of a lot cheaper for students, which is no small thing.
TechCrunch opined yesterday that San Jose State’s move “spells the end of higher education as we know it.” I suspect that higher education will outlast TechCrunch. But the teaching side of higher education will only thrive if it’s able to turn into the skid and use the new resources to its advantage. This is not the time to jam the brakes. If we do, we’ll crash, and be replaced -- rightly -- by Davidson’s screens. This is the time to use some unexpected momentum to get back on the track we should have been on in the first place.
"This article has been revised to reflect the following correction: The article also misstated the nature of the courses that San Jose State University will offer with Udacity. The courses will involve students watching videos and taking interactive quizzes; they will not be blended courses with students first watching videos on their own, and then coming to class to work on assignments with a professor."
When completing my graduate work years ago, I had a class which met only to work on our compositions. The readings, research, and rough drafts were completed outside of class. If there was a practice or concept I didn't understand, I could bring that up with the professor when he was editing my writing assignment. MOOC would have worked in that instance, if it had been around.
My writing and understanding of the various types of writing improved immensely. This blended MOOC might be good stuff.
MOOCs can't be the answer for every course, but they won't be stopped. We're at the very beginning of the adoption curve. Once there are a few more success stories, there will be a stampede.
Although I have expressed some reservations about MOOCs in the past, I think that they can be beneficial in certain circumstances. I think that the most effective use of MOOCS will probably be in a mixed format, where the online material largely replaces the “sage-on-the-stage” in the conventional classroom, but with there also being small face-to –face recitation sections in which students who are having difficulty in understanding the material can have their questions answered by a real live instructor. The homework could be handed in and graded in these recitation sections, and even the tests and exams could be proctored there in order to ensure that cheating is not taking place. If everything is done exclusively online, it will be difficult to make sure that some other student is not actually doing the homework or taking the tests.
But I fear that such a change would result in a rapid de-skilling of the entire teaching profession. College instructors who are used to organizing, planning, and teaching their own courses will now be reduced to the status of performers who are forced to read the works of others to their students. They won’t be able to design their own classes, they won’t be able to choose their order of presentation, and they won’t even be able to choose the homework assignments. This design work will all be done by a specalized group of educational consultants or by some superstar professor from the outside.
Perhaps in the future, MOOCs will be designed and operated by a small group of superstars, all knocking down six-figure salaries, and supported by a much larger cohort of largely temporary faculty doing piecework for low pay and no benefits.
It must be recognized that this is an experiment being done by SJSU, so I hope they are doing it in a controlled fashion to be sure there isn't any selection bias. They had great success with a high-level blended class (these classes are neither high level or blended), but no indication if the students were assigned at random to the different modes of delivery. If not, many factors could contribute to a higher success rate. I've had huge differences just based on when a class met during the day.
I question the fear that just any random person can generate success in a blended environment. Sure, there are plenty of examples where a TA in that role can produce really good results without even having to speak English, but that is when the students don't need a TA to help them learn. I'd argue that the top schools are where this would work the best. Marginal students are the ones who need a real expert helping them get over the hump.
However, that said, the program being tested as SJSU would be a success even if the failure rate did not change, simply because it is cheaper and less financial aid money would get wasted as a result. The real measure of its success is if students who pass this program do as well in the next class as those on the traditional path.
[For people who aren't up on biostatistics: Johns Hopkins is one of the three programmes that pretty much everyone agrees are the top three]