Comparing multiple matrices using multcompare?

I'm trying to compare several large matrices for differences in power. I want to retain the variability, so that I can assess significant differences.

each matrix is 32x1216, and I'm trying to compare them. Neither multcompare, mattest, anova etc don't tolerate these datasets.

What am I missing?

[Scott MacKenzie]

This question is too vague. If I recall, P = VI (power equals voltage times current), but perhaps you are talking about some other form of "power". Explain. Give an example.

[Adam Danz]

My guess is that David is describing statistical power, or, the risk of committing type II errors. I agree with Scott that this is quite vague. Power is described by an effect size, sample size, and significance level. Effect size is measured in many ways depending on the statistical test you're using. And the choice of a statistical test depends on the question you're asking. Way too often the problem is approached in reverse order: students choose a test, perform the stats, and then figure out what the question is. That's really bad science (I'm not saying you're doing this).

• Start with the question you're asking.
• Then decide which test addresses the question. Usually only 1 test will fit your question.
• Then confirm all of the assumptions related to the test (I strongly urge everyone to use non-parametric statisticts that have less rigid assumptions).
• Then perform the test and interpret the results.

If you ever find yourself trying another test that usually either means the question is not clear or you're p-hacking and trying anything to get the result you want. Both are really really really bad idea.

That being said, I'm not accusing you of any of this. It's unclear to us what the question is which is why it's not really possible to help out.