Repeated measures at many different time points


I have logged movement data from a number of subjects (n=21) who were moving during 3 different conditions. Each subject was presented with a counterbalanced order of stimuli (conditions). Each condition was presented 6 times for every subject, giving a total of (6 times *3 conditions =18) observations per subject.

I would now like to investigate if there is any statistical difference between these three conditions. This is a repeated measurements experiment, but I am confused since I have so many different observations (6 per condition and subject). Also, the 6 measurements are actually continuous position data (not single points). Would it be OK to merge all my positiondata for respective condition and subject and then put the resulting mean into an ANOVA?

If not, I have been reading about mixed models, GEEs and OLS. Don't understand which is the best for my case though. If anyone has an idea on the best way to analyze this, that would be highly appreciated!