I can probably help...some general notes - like any regression, you're looking for an R-squared for overall fit/significance, and B's for effect of various variables. For any variables that are factors, the B will tell you which level of this variable has a significant effect compared to (likely) the first level. Eg. if a factor is race, you (or the program) might set whites as the 'control' (I'm not using the right terminology here, but it's not coming to me...), and thus your B's for other races will tell you if there's a larger or smaller effect for African Americans or Hispanics than for whites. If that doesn't help enough, write again and I'll send you my email.




Reply With Quote