Hello Sir I came across an old post of your discussing the singularity of the model matrix, where the data matrix suffers from multicollinearity. You mentioned that using a control variable with binary indicator is a possible option for that.
I wanted to ask you what is the correct remedy in case where we do not want to drop the collinear variables? Actually, I’m trying to estimate the coefficient of the binary variable, and it’s the same in both cases: including and not including interaction terms. Should I interpret the coefficient of the binary variable as the treatment effect or do I find another way to produce different results in both cases? ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
