Hi everyone I've looked around for clear answer about sample size needed for binary logistic regression but cannot find one. I'm trying to get my head around g-power as new to it.
I have two dependent variables outcomes.
Predictors:
4 IVs each with two levels each plus 1 additional which is a continuous measure.
Sample size 104. Is this enough?
Many many thanks
I would simulate data and see if the achieveable precision in the relevant model coefficient(s) is sufficient for the purpose.
I would simulate data and see if the achieveable precision in the relevant model coefficient(s) is sufficient for the purpose.
Thanks for the paper Sanjoy Datta
Thanks very much for the suggestion Jochen Wilhelm Im afraid its more technical than my mediocre stats brain can handle though!
Actually, simulations need less brain than understanding power calculations. You you do need is some programming environment that allows you practically doing these simulations. Although it's possible even in Excel, I wouldn't recommend it. How about cooperating with a local statistician?
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size