From: Lionel G. <guy...@gm...> - 2008-10-16 05:36:26
|
Hi Thomas, I guess you sent this email to the wrong list. You are probably referring to this Amos: http://www.spss.com/AMOS/ whereas the present list is dedicated to that AMOS: http://amos.sourceforge.net/ Hope you'll find help somewhere else ! Best, Lionel On 16 Oct 2008, at 4:57 , Thomas J. Callahan wrote: > I am working on a grant project that is attempting to predict return > to > college by individuals with 'some college.' In every analysis I do, > Var A, is a very strong predictor (correlate) of the dependent > variable. In binary logistic regression, correlations, etc. it is > strongly related to the dependent variable. When I put Var A into an > Amos model, the effects of Var A on the dependent variable basically > evaporate. The difference between Var A's effects in the binary > regression and the SEM model is the one that puzzles me most. Anyone > else had a similar experience or know what I can do? Thank you. > > I have checked 'operator' error, but cannot rule that out. > |