Dear team,

I want to test the the difference between two conditions in a within-subject design, while accounting for inter-individual differences in a subject-level score. So we compute the condition difference of that score per subject and mean-centre across subjects, and enter it as an interaction with condition. 

Would this be a reasonable set up of the design, in order to adjust the condition effect?

(let's say for two subjects as example):

1  0   1   1.20    
0  1   1  -1.20   
1  0  -1  -1.20    
0  1  -1   1.20   

contrast: [0 0 1 0] for a paired t-test of the condition effect, adjusted for the covariate

Exchange blocks: [1;2;1;2]

Many thanks in advance!