Can Researchers and Practitioners Dance Together?

Think back to the 5th grade dance. Boys are standing on one side of the room. Girls are on the other side. Nobody is dancing together.

This is my perception of education research. Real teachers are standing to one side of the gym. Quant-oriented researchers are on the other. No one is dancing.

After a while, 5th grade boys leave the gym and go into the hallway and talk about video games. Similarly, quant researchers give up on doing research about teacher actions because the high transaction costs of setting up experiments. Since they find it really hard to study teachers in a mathematically sound way, they take their problem-solving marbles and find other data sets to crunch. These data sets are invariably about education policy, not about teacher moves.

Girls see that the boys aren't around anymore. So they dance with each other. It's not that fun. It's just something to do.

Similarly, since schoolteachers' day-to-day work is generally ignored by the quant research community, they end up dancing with one another. That is, teachers are on the receiving end of advice from other educators who have self-proclaimed “best practices.” It's not that fun, sitting around in professional development sessions, hearing the latest fads. It's just something to do.

There's no New England Journal of Medicine for teachers: empirical evidence about teacher choices. The closest thing is the federally funded What Works Clearinghouse.

The WW Clearinghouse has a list of topic areas including beginning reading, high school math, and English language learners.

Let’s say I’m a high school math teacher. I click around on WWC. First, I find a lot of products for which “no studies meeting evidence” are available. Second, if I find a study that “works”, I – like most high school teachers – am still constrained because I work in a school district that has a set curriculum. It’s not like I can personally get rid of all our existing books and buy something new for just my class.

What if I want ideas on actually what I (an individual teacher) should do with the kids?

For example, let’s say I’m an elementary school teacher. Is it good to read aloud to the kids? If so, is it better to do for two sessions of 20 minutes or one of 40 minutes? Does it matter if I have the kids at their desks, or crowded on the rug? Is it important that first-graders all have their own copy of a book to follow along? Or can I just read it out loud and hold up the pictures? Is it more effective to have pairs of kids read alternating paragraphs to each other? If a couple kids aren’t paying attention, do I need to “herd them” aggressively back to the flock, or does that sap energy from the whole group and diminish the total utility of what we’re doing in class?


WW Clearinghouse has lots of empirical studies about policy choices (typically made by district superintendents, state commissioners, or school boards). For example, measuring the effect of class size reduction in an entire state. Or: measuring the effect of charter lottery winners and losers across an entire city. Or: The effect of a merit pay bonus offer to the current cohort of teachers in a school district.

I'm not blaming WWC. All they're doing is sorting through what's published. Not much rigorous research about teacher moves and teacher actions gets published. The cupboard is nearly bare. A study here on strategies for teaching fractions, a study there on how to encourage girls in science.

But they've got so little on practical questions. I eagerly clicked on the 87-page research brief for "Reducing Behavior Problems In The Elementary Classroom." Know what I found? Stuff like:

Encourage parents and other family members to participate as active partners in teaching and reinforcing appropriate behavior.

Um, duh! A teacher's question is more likely to be HOW. Not WHETHER. As in: "How do I activate parents who specifically and categorically reject all my current entreaties?"

For goodness sake, a 63-page brief of the best teaching techniques has precisely two with "strong evidence" -- giving lots of quizzes, and asking deep questions. Well, amen. I concur. But that's all ya got?

An individual teacher often gets to decide:

-what his typical hour-long lesson looks like -where kids sit -how he’ll grade work that the kids produce -how he’ll try to flip a kid who won’t try hard -what he’ll assign as homework -what will be on tests and quizzes -how to explain a particular concept -how he’ll respond to small potatoes misbehavior -how many out of school hours he’d like to work, if any -what he does with that out-of-school time

and so forth.

So from a teacher’s point of view, most of the published empirical research does not relate to choices he gets to make.

Hence no dancing.

How can we have dancing?

I'll tackle that manana.