I would like to compare how different species use their habitat.
Habitat use is measured as the number of times an individual is observed inside the habitat, close to that habitat, or far from it out of 300 observations (so proportions).
If my response variable was just inside/outside (success or failure), I guess I could use beta or logistic regression to test if there are differences between species (is that correct?) and if my data were not constrained between 0 and 1 I could, for example, use multivariate ANOVA, but I don't think this would be correct for proportion data.
Any idea how to analyze this type of data?
(I usually use R for data analysis, so something easily implementable in R would be welcome^^)
You can use a Z-test, Chi-square test or Fisher-exact test to compare proportions, or you can transform the data to use a parametric test, but I prefer not to transform :)
To add to Andrew's answer, you can use Chi-square with Bonferroni adjustment to test for habitat preferences. See for example Anderson (2009):Article Ground Versus Canopy Methods for the Study of Birds in Tropi...
And we borrowed Anderson's approach to look at something similar in our 2018 paper too: Article Canopy bird assemblages are less influenced by habitat age a...
- Badges
- Science topic
- Similar topics
- Correction