How Can I Vary The Number Of Replicates In A Power Analysis

Question

[pummy]

Hi,

I am new to using statistics and I am trying to do a power analysis to determine the sample size I might need in my research. All the packages I see use the effect size, significance and power to determine the sample size. Some packages let me enter the Standard deviation I have from my test data. I would also like to see the effect of varying the number of replicates on the effect size. Does anyone know what package I should be using to do this?

Thanks very much for your help in advance!

Answer 1

[Alex Paciorkowski]

I've found this straightforward to do using basic methods in R. You can read more for example here.

Also, I like the discussion of power analysis for sample size in Michael Crawley's Statistics: An introduction using R.

[pummy]

Thanks so much Alex!

Answer 2

[Devon Ryan]

G*Power is one of the nicer programs I've seen for such things. You can vary your sample sizes (i.e. replicates) or whatever else you want. There are mildly useful tutorials for many of the use cases on their website.

[pummy]

Thanks dpryan. I have started using GPower for my analysis.