Dear all, First of all apologies, I'm pretty newbie, and secondly I know this is not a question for the forum but I'd be very grateful if you'd help me. I had problems with normality assumptions, and I tried with weights in the linear regression this way: modelw <-lm(log(v1)~v2+log(v2)+log(v3)+v4, data=dat, weight=1/group.var^2) I'd like to know how interpret/compute adj R-sq/mse from modelw, in order to compare it to the adj R-sq/mse from model (model <- lm(log(v1)~v2+log(v2)+log(v3)+v4, data=dat) Thanks in advance, user at host.com pd I set the weights following this notes: (I was reading here: statistics.unl.edu/faculty/bilder/stat870/schedule/Chapter%2011.doc http://statistics.unl.edu/faculty/bilder/) -- View this message in context: http://r.789695.n4.nabble.com/lm-weight-adj-r-sq-tp3498020p3498020.html Sent from the R help mailing list archive at Nabble.com.