I'm often startled anew by how pre-Newtonian learning remains. You just made a new lesson / activity / explanation. Is it effective? Is it engaging? What parts work best/worst? Getting high-signal answers is very slow—and evaluation is often many times as expensive as creation.
Conversation
As I recall, would say that usability testing, even with just one or two people, usually uncovers plenty of flaws, and one can then focus on the glaring, facepalm-inspiring ones.
Maybe teaching is too different from web design for the same principles to apply?
1
1
Unfortunately in my experience, between-student variance usually drown out between-condition effect sizes when evaluating individual lessons / activities / explanations.
I imagine the same is true in web usability testing, though, and it still seems effective.
Do you actually quantify things, or just qualitatively identify specific problems? (I’d guess the latter is more useful, and also cheaper.)
1