I am testing two DV and two IV using a two-way MANOVA. I have four groups. I had used g*power to recommend a sample size, however I have not yet achieved this recommended sample size. Instead I have 11 samples from 11 random participants. I understand that this is not optimal because then I have uneven samples in groups (i.e. A=3, B=3, C=3, D=2) and DV's (DV1=6, DV2=5). However, that aside, my concern is whether there is a minimum sample size that is acceptable.

I have checked the significance of my results and found they are strongly significant. In RStudio the results are marked with three asterisk ***. Given that my alpha remains at 0.05 (in Gpower the graphs show that) I have no chance to make a Type I error. By the low power of my sample size I have a very large chance to make a Type II error (to accept the null hypothesis even though it is false). My results show a significant difference despite this low power. So in this case I won't make a Type II error, as I will reject the null hypothesis. I was wondering if this is an acceptable statistical result (so that results would be considered for publication), or whether there is a minimum sample size that should be reached. I heard there is a rule 'you need to have more cases in each cell than you have DVs'. I have two DVs. Does that mean I have covered the minimum requirements? I don't have a textbook or journal article to cite this rule either. I would appreciate if someone can help me out here.

Thank you in advance!


1 Answer 1


I have found the following reference. I would prefer Journal or Textbook references.

"Assumption #4: You should have an adequate sample size. Although the larger your sample size, the better; for MANOVA, you need to have more cases in each group than the number of dependent variables you are analysing."


Your feedback is much appreciated!

As we have seen, the first test that’s performed when we do a MANOVA is a test for homogeneity of the VCV matrices. For this test to run, we must have more subjects than variables in every cell (Tabachnick and Fidell, 1996)."



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.