Hi all

i would like some advice on the following issue.

I have an event-related design experiment in which i want to compare two within-subject conditions.
I already performed a mixed-effect anova direclty comparing my conditions without taking into account any individual differences nor stimulus differences and this analysis is quite noisy, becase as it turns out there are reasonably large differences among individuals in terms of how sensitive they are to my experimental manipulations, but also an item analysis computed on my experimental stimuli shows that there are differences in terms of the quality of the experimental stimuli themselves.

I have also already run a linear regression using 3dregana to use a behavioral measure across my subjects, but i also wanted to run an analysis that allows me to use the variability found among stimuli, to see if i can get ride of that source of noise on my data.

In other words, i would like to compare my two conditions taking into account a score obtained from the item analysis for each of my experimental stimuli.
However i am not exactly sure what the best way to do this would be.

Any suggestions?

Thanks
giulia