I am conducting a basic preference test experiment to determine whether subjects spend a longer time investigating sample A vs. B, and for logistical reasons I have to present the samples separately. (The data are paired, each subject is presented with both A and B, which I'd like to keep since high between-subject variability is typical.)
Presentation order has a strong effect (subjects investigate first sample longer than second) that may be masking a milder overall preference for sample type (subjects investigate A more than B.) Both possible effects are interesting to me, but my primary question is whether subjects investigate A equally to B or not.
My very basic stats training and the websites I've looked at provided suggestions for paired data or for 2-factor analysis, but I haven't come across a single test that can deal with both. I am also not sure how to test for normality.
I would appreciate any feedback you have-thanks!
Advertise on Talk Stats