Power analysis for glmer?
On Fri, 2 Dec 2011, David Duffy wrote:
On Thu, 1 Dec 2011, Adam D. I. Kramer wrote:
I have been looking online for a while and come up dry. Could anybody recommend a paper or library that would give me some idea of how many subjects I would need to detect an effect of a certain size using a mixed-effects binomial model? --I'm really just trying to get an idea of the scale involved. 10k versus 1m, for example.)
You need to simulate data from your proposed setup. See http://glmm.wikidot.com/power-analysis
Oops, and add a step to dichotomize the continuous outcome at a percentile matching your observed data.
| David Duffy (MBBS PhD) ,-_|\ | email: davidD at qimr.edu.au ph: INT+61+7+3362-0217 fax: -0101 / * | Epidemiology Unit, Queensland Institute of Medical Research \_,-._/ | 300 Herston Rd, Brisbane, Queensland 4029, Australia GPG 4D0B994A v