By doing an experiment, I would like to analyze the effect of 3 categorical independent variables (with 2 levels each) on three constructs measured with ordinal variables. I also have 2 covariates (with ordinal variables). Among the findings, I expect correlations among the three constructs. Can I use MANCOVA to run such analysis with this type of data? In case it is possible, what could be the main issue(s)? I have not run any test so far. Thanks for your help!