Calculate power for logistic regression

[John Havens]

Hi everyone,
I want to calculate power for a logistic regression with one dependent and one independent variable in Stata. It looks like there is no built-in function in Stata to do that. Can I use the calculation for two proportion instead? What do I have to adjust if I calculate power for a comparison of two proportion and use it to estimate for power for a logistic regression? I think they should be related.
Thank very much!

[Carlo Lazzaro]

John:
you may be interested in: http://www.ats.ucla.edu/stat/stata/dae/logit_power.htm

[carole fantini]

is this is also relevant for logistic mixed random model?

[Carlo Lazzaro]

Carole:
I do not think so; see http://stats.stackexchange.com/quest...r-mixed-models and http://samplesizeshop.org/files/2013...2pmTo330pm.pdf.

[John Havens]

Thank you very much, Carlo!

[Dick Campbell]

For mixed designs you might be interested in a user-written routine named ipdpower (findit ipdpower).

[John Havens]

Thank you very much Dr. Campbell!