Biology, images, analysis, design...
Use/Abuse Principles How To Related
"It has long been an axiom of mine that the little things are infinitely the most important" (Sherlock Holmes)

Search this site



Just a note

For example, you would need to construct a quite complicated simulation model to predict the null distribution of F when group sizes vary under restricted randomization.

In practice therefore analyses of such studies generally ignore that source of variation, and their inferences are therefore conditional upon the observed group sizes.