I am analyzing quantitative susceptibility mapping (QSM) data using randomise; t-test with covariates. My QSM data contains both positive and negative numbers. My very basic understanding of randomise t-test with covariates is that it permutes the regression coefficients. My question is whether the presence of negative numbers (QSM values) will "mess up" the stats or am I okay to use positive and negative numbers for dependent variables even with t-test with covariates.
Thanks
raj
|