I have a study that I have conducted on resilience that I conducted a MANCOVA on as it has 6 categories of DVs (factors of resilience). All of my DVs are significant. I have chunked in my demographics as my co-variates and some have significant findings whereas some do not. For instance, the co-v's of age, gender, and sexual orientation are significant. How do I go about reporting these significant co-variates? Also, do I report that the others such as ethnicity and education level were not significant predictors of resilience in this study?
https://www.statisticssolutions.com/multivariate-analysis-of-covariance-mancova/
https://www.statisticssolutions.com/one-way-mancova/
Hello Nuchelle,
The brief answer is, give the results for all tests of interest. The presence of a "significant" covariate informs you that scores on that covariate were significantly associated with a linear combination of scores on your DVs.
When you have multiple covariates, it is possible that one or more which show as non-significant may be such because their relationship with the DVs is redundant or overlapping with that of other or another covariate in the set (e.g., some collinearity exists). As well, it may be even the simpler situation that the specific covariate is unrelated to the DV ensemble.
Good luck with your work.
