Nested design

I am conducting a study investigating the effects of different media treatments on protein expression in cells. I have four levels of media treatment which are crossed with valproic acid or no valproic acid (ie a 4x2 factorial design).

The question I have is with regards to nesting. Each observation (in the form of a picture) is one of a set of 4 taken from a single well. These are replicated in 2 different plates, which are from one of 3 subjects in either of two groups. Should I be nesting these variables in my ANOVA, or is the effect of this irrelevant?



TS Contributor
I would use a nested design because otherwise the analysis would treat the data points as independent samples which they are not.