Data Set 6-II: Mixed-Model ANOVA of the Effects of Host Species and Size on Parasitoid Size Statistical Setting

This Handout illustrates traditional mixed-model ANOVA analyses of the data from my study of host effects on parasitoid larvae (Data Set 6) when the host species is treated as a random factor. Note: Treating species as random would not be reasonable in reality, given that these two moth species are the only common hosts for this parasitoid in North Carolina, where I conducted this study. It also rarely will be appropriate to have only two levels of a random factor. Only the inferential results are shown for these mixed-model analyses; preliminary data exploration and model diagnostics would be the same as presented in the previous handout. I first present results for the balanced data set (part A of Data Set 6), then for the unbalanced data set (part B). A later handout will present linear mixed model (ML and REML) analyses; for the unbalanced data these are preferable to the ANOVA method shown here.