I try to find a good model to my data but I get stuck all the time as I'm not so good at statistics.
I have data on approximately 100 fishes and each fish has been incubated in one of four possible tanks. Two of the tanks contained cold (control) water temperature and two contained warm water. The fishes come from four different crosses (AA, AR, RR, RA). I want to investigate if incubation temperature influence different parameters, e.g. total swim time.
I have thought about a two way anova with crosses and incubation temperature as independent fixed factors, tank as random factor or both tank and crosses as random factors. But it seems to me that tank is nested within incubation temperature and even fish individual is nested within tank. So should I use a mixed effect nested anova instead?