« spammed | Main | Si Se Puede! »

More about schools, and testing, and so on

Your Humble Blogger has no particular insight, but this article in this morning’s NYTimes points out the difficulties of testing and reporting the success of schools. The federal list of underperforming schools evidently includes lots of schools that the locals think are great; one problem is that ‘underperforming’ doesn’t actually mean ‘bad’. It’s all confusing, and in my own opinion, it will all stay confusing, as it’s inherently confusing.

I will, in the spirit of the times, pull out of context a magnificent quote from a Republican. “I never expected to see all these suburban schools on the watch list,” said Rep. Judy Biggert, R-IL.



Apologies for the blank comments: a couple of errant keystrokes, and disaster follows . . .

This seems an opportune moment for me to lay out a few basic principles I, as an educator, believe to be true about education that contribute to the sorry state of "scientific" educational policy. Perhaps they will provoke some useful discussion.

1) Most of the time, education does not work. If people were readily educable, we wouldn't have to spend so much time attempting to teach and learn. My inference is that meaningful education happens occasionally, in short, rich, and unpredictable bursts, which depend upon a variety of uncontrollable personal factors in the relations between teachers and students. We show up for class every day because we don't know when those bursts will occur, but we know that they must, or nobody would ever learn anything. This is not the sort of phenomenon that is easily measured. Thus, any studies that show that this or that technique of education doesn't work only show us what we could have inferred without the study and provide little useful guidance about pedagogy.

2) The chief corrolary of this first principle of education is that the teacher's first principle of practice should be the same as a doctor's: first, do no harm. If we teachers can help to keep the kids showing up with an open mind every day by not turning them off to learning, they'll probably be all right. We can't _know_ what and when they'll learn, but we can have some confidence that it will happen sometime. Building an educational system around batteries of standardized tests makes implementation of this principle more difficult than it should be.

3) Practice makes perfect. This is the other thing that justifies the amount of time we spend at teaching and learning. Aside from the short bursts of genuine, change-your-life education that takes place, most of the rest is the slow absorption of skills into our minds and bodies, which takes place according to the industry and talent of the individual student and the uncontrollable natural course of growing up. While this process of practice can be influenced by teachers setting their students tasks, the pace of learning cannot be forced into conformity to the calendrical arrival of batteries of standardized tests. Tests can provide spurs to study and can measure performance, but they can't regulate the learning process.


I deleted the empty comments, Chris. They served as a sort of announcement to me that you were about to write something, so that was nice...

I'm not sure if I'd accept the phrasing "Most of the time, education does not work." I am happier with the idea that "Practice makes perfect." Or am I mistaking your idea? Let me analogize: if success in teaching a thing (say, the six-times tables, or logical thinking) is like rolling doubles, then you may well roll nine or ten times before success happens. That doesn't mean that the first rolls didn't work, just that they weren't enough to finish. Is that more or less what you're saying? Thus it's often unclear, in the short run, whether the teacher's techniques are good or bad.

I do think that some teachers are better than others, and that my experience in elementary schools was pretty mixed. I can remember perhaps five really bad teachers between sixth and twelfth grades (not counting the fellow who passed a .45 around class, as he was a good teacher, or the guy who later shot a guy in the back in self-defense, who was creepy but not a bad teacher). I suspect I had about that number of really good teachers. That's my own recollection, and highly biased toward not-being-dull and not-being-ignorant, as opposed to my having any sense of whether the students learned what they were supposed to.

Anyway, I agree with the 'accountability' camp that we would prefer to minimize the number of bad teachers, even perhaps as a higher priority than maximizing the number of good ones. But I can't figure out how to really know if a teacher is good or bad. I completely reject the standardized test as a measure for a variety of reasons, but the only things I can think of to replace it are hijjusly expensive. And if, as I interpret your suggesting, good teaching is more or less probabalistic, testing it becomes even trickier.



The "most of the time, education doesn't work" is rhetoric selected for speaking to educators -- it gets their attention, and usually gets a laugh, so it's an effective way of opening a discussion.

Speaking to a general audience, I'd want to phrase it somewhat differently, because a general audience is likely to have a lot of people whose first response to the claim, "most of the time, education doesn't work" will be, "then why should we bother with it at all?" Whereas, it is obvious to me and to people who care enough about education that, as long as we are human, we cannot do without education.

Thinking through the rhetoric, I think I would want to keep "Most of the time, education doesn't work,' but I would move it to principle #2, and start with, "As long as we are human, we need to be educated," as principle #1.

To respond substantively to the rolling dice analogy: Yes, I think learning is probably more or less probabilistic, but as teaching and learning constitutes a much more complicated and more open system than rolling a pair of six-siders, I suspect that the application of probability theory to the study of education would be more useful in theory than in practice.

Looking at the description of education from a rhetorical point of view, I'm reluctant to talk in terms of probability because I think it important to discourage any line of study of education that assumes that a rate of learning can be straightforwardly derived from teaching, and even probability would falsely raise the hopes of the seekers after statistics.

Things like this are why I'm a fan of local control. For national standards-bearers to say "you may think your kids are getting a good education, but in fact they're not" seems supremely arrogant, and obviously wrong, to me.

There's a lot to be said for local control and consequent diversity in educational practice (I think the increasing homogenization of college education in the U.S. is weakening the quality of higher education, for instance), but there's also a lot to be said for outside intervention.

For instance, when schools in one region are impoverished, it's good for a state government or a national government to have some leverage to address funding inequities. For another instance, it's good to have some state or national body exercise some influence over standards of content, so when Flat-Earthers or some similar group pack a local school board, they can't simply impose a curriculum tailored to their particular perversion of knowledge.

There's no perfect system for administering public education; the standardized testing rigamarole of "No Child Left Behind" is surely an example of excessive and misguided Federal intrusiveness, but I wouldn't want to shut the feds out of shaping public education policy altogether in reaction to it.

Belatedly, in case Chris ever sees this: it sounds like you're proposing a sort of punctuated-equilibrium model of education. Do you think that's a more accurate model than slow steady learning?

I guess it seems to me (not a professional educator, so I don't have a lot of evidence for this) that there are multiple different kinds of learning:

There's the Aha! moment, when after months of bashing your head against a concept it suddenly becomes clear. It sounds like that's what you're talking about.

There's learning a bunch of small things that add up to bigger things. In learning how to juggle, first you learn how to toss one ball in the right way, then you add another one, and then you add a third and make it continuous and you're juggling.

There's guided practice: the teacher tells you how to do something, you try it, you don't entirely succeed, the teacher helps you make your next attempt better, gradually your skill improves until you can do it right most of the time.

There's exploring ideas, where you think about something you've been exposed to (a novel, say) and try to come to conclusions about it and/or ask interesting questions, and then you discuss it with others and you learn from their insights and they from yours.

And so on. I'd even say there's a kind of synthesis of some of those modes that's sort of like an internal combustion engine or a film: a series of small Aha! moments in close succession that combine to produce the illusion of smooth gradual learning.

So. You weren't talking about learning per se, but about education; still, it seems to me that saying education mostly only happens in short unpredictable bursts may be ignoring the modes of learning other than punctuated equilibrium.

Does that make any sense?

Comments are closed for this entry. Usually if I close comments for an entry it's because that entry gets a disproportionate amount of spam. If you want to contact me about this entry, feel free to send me email.