Hi, I'm sort of new to regression, but I was wondering how one compares two regression slopes in R. Is it just a matter of calculating the slopes and then using the SD reported by R to test for a difference. I can't imagine one can assume normal distributions in this case. :)
On Tuesday, August 19, 2003, at 09:20 PM, Simon Blomberg wrote:> You need to look at Analysis of Covariance in any basic stats book. > Basically, if you have y versus x for 2 groups ("group" is a factor), > use lm to fit the model: > > fit <- lm(y ~ group + x + x:group, data=your.dat)but, my situation is two separate regressions with different collected Y's and a predicted X.