Flashcards
Hierarchical Regression
Cards in this deck (8)
- 1Primary reason for Hierarchical Regressoininterested in change in r^2 over and above some other set of variables.
- 2Variables tend to becovariates
- 3Always put the coeffients from final model,coefficents from previous models are not interesting because they are based on misspecified models based
- 4Interaction with 2 continuous variablesYhat = b0 + b1x2 + b2x2 + b3x1x2
- 5Yhat = b0 + b1x1 + b2x2 + b3x1x2= (b0 + b1x1) + (b2 + b3x1)x2 (what’s the difference of x1 at b2
- 6When effects are highly colinear, huge problem:tolerance, want to be able to still see main effects
- 7If you don’t center:regression coefficients don’t change a lot
- 8Simple slope =slope at specific values of another variable (ex SD+1 and SD-1)
Information
- Description
- Psychology Flashcards on Hierarchical Regression, created by Robert Gray on 12/12/2016.
- Created by
- Robert Gray
- Views
- 0
- Rating