Apr 22, 2016 Three Different t tests using SPSS with Dr Ami Gates. Mod-01 Lec-14 Hotelling's T2 distribution and various confidence intervals. TUTORIAL MENGOLAH DATA QUISIONER DENGAN EXCEL DAN.
Posted by8 years ago
Archived
I was hoping somebody could give me some help with some stats. I'm working on my Honours thesis (in psychology) and I have been running some multiple regression using two predictors. What I would like is to compare the Beta values of these two predictors to determine if they are significantly different from each other (as opposed to just different from zero).
I've been told about Hotelling's t and (alternately) Steiger's Z as a way of comparing correlations composed of overlapping samples. But I'm not sure if I can use Beta values in place of the zero-order correlations. Does the fact that the Beta value controls for the other variables in the regression model make them 'non-overlapping', or should I be using the zero-order correlations, or what?
Any help would be appreciated. I'm in way over my head here. I've been using this page for info, along with the FZT Computator it mentions.
4 comments