This chart, showing electrodermal activity believed to be associated with “cognitive stress” is making the educational blog rounds recently:

The thing generating interest in the blogs, of course, is the flatlining you’re seeing in the class segment, which lends it to the catchy soundbite that “Lectures produce no more cognitive activity than TV.”

Of course, this is one of the problems of increasing abstraction. Engagement correlates with learning outcomes which correlate with learning. Certain types of engagement correlate with cognitive stress which correlates with certain skin conductance patterns. At this level of abstraction, it’s not particularly useful for education (and indeed, the paper it come from makes zero claims about education — it is instead a demonstration that the device can detect various patterns of conductance over long periods of time, which may help people with seizures). Eventually a well-designed experiment might cut out a number of links in this game of research telephone, and who knows, maybe something will come of it.

If I had to say one thing about it, I’d say the piece that would be most interesting would be the comparison of the reading portion of studying with the lecture directly. There’s a seminal study I can’t find at the moment, a very early learning study from maybe the 1940s? Soldiers were randomly assigned to a group that either saw a film, watched a lecture, or read a manual on how to do some sort of operation with a gun — when the time came to execute the task, the soldiers from the three groups performed identically (Does anybody else remember this study? Help me!).

Anyway, my question is what that sort of experiment would look like with these EDA sensors. My guess is whatever the result, we’d find out there is a lot more subtlety to this data than is initially evident.