04-04-2014 01:41 PM
hello all, I am trying to calculate the Effective sample size for my generalized linear mixed model. I used the traditional formula of , where observations are nested within each cluster and intraclass correlation is . I have read that it is possible to calculate the intraclass correlation correlation can be calculated through the variance explained by The random factor divided the total variance. But my question is, how does the this work when you have more than ONE random factor clustering your data, how do you calculate the intraclass correlation and Neffective then?
Thank you so much in advance