3

I want to run a multivariate regression data analysis. I have 5 predictors--two of them are binary--and one DV.

When planning to use Multivariate Regression with some binary categorical predictors, how do I find the sample size I need to justify this test using G* Power?

Is it any different than having all continuous predictors?

I know how to do it when I have predictors that are all continuous. The procedure I use: F tests -> Linear multiple regression: Fixed model -> Rsquare deviation from zero with .15 effect size and .05 err probability. When I execute/calculate this, I get a total sample size of 138.

I'm going to have to explain this to my committee for my thesis proposal soon.

Axelbrain
  • 61
  • 7
  • Isn't it true that the software has a menu you can use to obtain the required sample size? – Michael R. Chernick Mar 05 '17 at 18:04
  • Of course. However, there are many ways that you can obtain these sample sizes using the said menu, and most of them are the incorrect way to do it for my situation. I'm asking for the correct way to do it so that it yields a correct sample size for my particular analysis. – Axelbrain Mar 05 '17 at 18:11

0 Answers0