My dependent variable is abundance of a bird species across 22 sites. I have 8 independent variables, out of which 2 are continuous and 6 are discrete.
Use a generalized linear model (GLM).
If your response variable is "abundance" (in animals/habitat for instance), a Poisson- or Negative-binomial error model is likely appropriate.
I understand that you only have 22 observations. If your 6 discrete variables are cuantitative (they are not cualitative), you can simply use "multiple linear regression."
If your 6 discrete variables are cualitative 0-1, you can still use Multiple linear regression.
But, if any of your discrete variable is cualitative with more than 2 levels, the problem becomes a little bit more complicated. You need to search on internet: multiple linear regression with cualitative(dummy) variables.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics