I saw a line graph the other day that showed learning decay over time. The research was credited to something called the American Research Institute–not a research body I was familiar with but, hey, interesting research is interesting.
The graph, which looked like this:
raised all sorts of questions in my head. What was the learning event? How was it designed? What was the expertise level of the participants? What was the nature of the skills taught? What age?
So I did some digging. And, as far as I can tell, the original research doesn’t exist. Often cited in infographics and books, but never in research journals, I imagine that at some point someone either overgeneralized from some real study (perhaps casually credited to “an American research institution”) or made up to make a directionally correct point in a particular context.
My guess is that everything can be traced back to this:
Looks eerily similar, doesn’t it? The numbers have drifted a little, but that’s what you’d expect from a few rounds of internet telephone.
In any case, the original study was about people trying to memorize lists of arbitrary words–hardly a good analog (I hope) of what goes on in most classrooms. So, interesting, and important, but a big leap to conclude that classroom knowledge will reliably decay at a certain rate. (In any case, I believe the most compelling work done with learning decay has been in the context of spaced learning.)
This sort of thing–propagating research of indeterminate or suspect origin–really bugs me. That said, if I’m wrong and the Research Institute of America really has done and published this research, please comment below to help set the record straight.