Dear List members, I am using MAANOVA to fit a multi-factor ANCOVA-type model for my data. However, I encounter the below error when attempting to use the matest function to run the F tests of my continuous covariates. I don't have any problems with this function when testing fixed factors using the same design matrix, and model-fitted object. The traceback indicates a problem with the makeContrast function, but I am not doing any contrasts of the continuous covariates. In the method description for matest, it says "For testing a covariate, use a one by one contrast matrix of 1." So I also tried the below code using a 1x1 matrix, which didn't work either. I am confused as to why a contrast matrix should be needed for testing continuous covariates. Any suggestions? test = matestdata.ma, fit.fix3, term = "v", n.perm = 100) Error in apply(tmpL, 1, any) : dim(X) must have a positive length > traceback() 5: stop("dim(X) must have a positive length") 4: apply(tmpL, 1, any) 3: which(apply(tmpL, 1, any)) 2: makeContrast(model, term) 1: matestdata.ma, fit.fix3, term = "v", n.perm = 100) c = matrix(1,1,1) > c [,1] [1,] 1 > test = matestdata.ma, fit.fix3, term = "v", Contrast = c, n.perm = 0) Doing F-test on observed data ... Doing permutation. This may take a long time ... Error: dims [product 48802] do not match the length of object [0] In addition: Warning messages: 1: In any(parsed.formula$random) : coercing argument of type 'double' to logical 2: In any(parsed.formula$random) : coercing argument of type 'double' to logical Kristen Dang Postdoctoral Associate Duke Human Vaccine Institute [[alternative HTML version deleted]]