Hierchal multiple linear regression Flashcards
what is a hierarchal multiple linear regression
You have multiple ‘models’ containing different variables
so predictors are entered in ‘blocks’ (models)
they are entered hieratically
the assumption of multicollinearity in hierarchal regression
VIF needs to be less than 10
tolerance needs to be above 0.5
this is on the coefficients table
checking standardised and cooks distance
std. residuals should be between +- 3.3
cooks distance needs to be below 1
this is on the residuals table
How to report
similar to standard but you talk about each model as if its a separate regression
so you need to report significance for each model, and state which variables in each model are most significant. Also need to report R2 for each model, to say how much the dv was explained by Iv in each model.
Also state which model is most significant, and explains the most %