My experimental design is very simple: test cells at 0, 50, and 100 uM concentrations of drug and count for toxicity.

Using G*power, I'm pretty sure I need to select ANOVA fixed effects, omnibus, one way for the stats test I'd like to do in the end. I did a pilot study where I took triplicate technical measurements of each concentration and counted the cell population. I calculated the mean and standard deviations of those means.

How do I go about inputing this data into G*Power correctly to design an experiment with 0.8 power? I clicked on input parameters determine, put in the mean values for my 3 groups, and the size. For the standard deviation, do I just average the standard deviations between the groups and use that as the standard deviation within each group that's in G*Power? When I hit calculate, I also get an effect size that's way above 1 (6.6 to be exact). Is that normal? I though effect sizes had to be less than 1?