Comparing size distribution between 3+ populations? Can I incorporate into ANCOVA?

#1
So, I have 20 distinct populations and the size distribution of each population. There are two things I would like to do:

  • Is the size distribution between the populations significantly different?
  • Add this size distribution data to an ANCOVA as a covariate.

Would either of these be possible? The size distribution data is not the main factor that I'm interested in but I would like to make sure that it's not influencing my response variable. Is there a way for me to test for this? As part of the ANCOVA model or separately on it's own?

Edit: I was talking with a friend briefly and he mentioned that I could run an ANOVA of the residuals of the sizes to see if they're even. Would this be a satisfactory way to address the first bullet point?
 
Last edited:
#2
Re: Comparing size distribution between 3+ populations? Can I incorporate into ANCOVA

Do you have the size distributions as raw data, counts in bins, or as percentages in bins?