[taking the liberty of forwarding this back to r-sig-mixed-models] I'm looking for the multilevel logistic regression equivalent of this: http://www.quantpsy.org/interact/hlm2.htm with my scenario being as per case 2 in the spreadsheet cheers Chris
On 19/05/2014 23:00, Ben Bolker wrote:
On 14-05-19 04:37 PM, Chris Stride wrote:
Hi all To avoid re-inventing the wheel I was wondering if anyone could point me towards some R code to plot simple slopes and the region(s) of significance for an interaction between two continuous higher-level predictors in a 2-level multilevel logistic regression model? cheers Chris
Could you be a little more specific and possibly provide a reproducible example? Can you point to/have you found solutions of your problem for either (1) a plain linear model (2) a LMM (i.e. non-logistic but multilevel) (3) a plain logistic regression (i.e. binary response but single-level)? (These would all help because they would clarify precisely what you want and because the general strategies might be extendable to your case ...) Ben Bolker