I have an experiment where there are more factors than budget / sample to test so I am using a D-Optimal design from AlgDesign in R. (or maybe a fractional factorial, I dont think it matters). I run this procedure and end up with the design points I will use that allows me to read the main effects and interaction terms I chose. A design point I am defining as factor A - level 1 + factor B - level 3 + factor C - level 1 etc.

Now, the rub. My experiment is regarding customers and direct mail. I need to measure a 'response rate' for each of the design points and ultimately create a model (e.g. logistic regression) to find the "optimal" factor levels and response surface.

How does one then go about determining the sample size for each design point? Use a two proportion power analysis and set that required sample size to each of the design points?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.