Hi! I constructed a model where most of the variables are linearly related to each other. However, the dependent variable is dichotomous, which means I should not try to fit a linear relationship between the other variables, which are continuos, and this variable. I'm wondering how I would have to go about addressing this issue in sem?
Additionally, this post is relevant to my question: http://r.789695.n4.nabble.com/Link-functions-in-SEM-td859182.html . The way I understood this post is that I should use hetcor to compute the correlations between the variables and then feed these correlations into sem. While the partial regression coefficients of the continuos variables associated with the dichotomous variable did change, so did the partial regression coefficients of all of my other relationships, which should not be the case. In addition, I am not sure how I should interpret the partial regression coefficients of the association between the continuos and dichotomous variable, i.e., do they represent odds? Thank you in advance. Best, ~Kino [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list 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.
