against strategies, part 2

from Dan Willingham:

I have written (on this on this blog and elsewhere) about the importance of background knowledge and about the limited value of instructing students in reading comprehension strategies.

To be clear, I don’t think that such instruction is worthless. It has a significant impact, but it seems to be a one-time effect and the strategies are quickly learned. More practice of these strategies pays little or no return. You can read more about that here.

Knowledge of the topic you’re reading about, in contrast, has an enormous impact and more important, there is no ceiling—the more knowledge you gain, the more your reading improves.

In a recent email conversation an experienced educator asked me why, if that’s true, there has been such emphasis on reading strategies and skills in teacher’s professional development.

There’s a temptation to say “Oh, Americans don’t value knowledge; they don’t think that people need to know stuff.” But I rarely run into a teacher or administrator who believes that, and I actually have (as yet unpublished) survey data to support that impression. So I don’t think that’s behind it.

Doubtless there is more than one reason, but as a researcher, I have a hypothesis: People think strategies are important because most of the reading research is on strategies. But that’s an accident of the way research is done.

Behavioral research (and educational research in particular) is a more conservative enterprise than you might think. When a researcher decides to conduct a study in classrooms, he or she typically commits at minimum two years of his or her life to the project. What if nothing much happens?

A researcher is therefore motivated not to conduct studies that break utterly new ground, but rather to conduct studies in which one is fairly confident that something will happen.

One takes an intervention that has worked in the past, and adds an interesting tweak.

[snip]

[T]o be competitive for funding the proposal has to be something for which there is already evidence that it will work.
Now consider what it takes to do research on strategy instruction versus knowledge instruction. Teaching children reading strategies is quick. A research project might call for 10 or 20 lessons in total, each lasting 30 minutes or less. One can imagine getting a school administrator’s permission to do such a study in his or her district.
But the hypothesis for knowledge instruction is that it takes years to make a broad impact on students’ knowledge.
Willingham: The Zeitgeist of Reading Instruction
The Answer Sheet