Multiple Regression

Question 1

What is an overview of this lecture

Answer 1

Overview
* Multiple Regression
* Collinearity
* Mediation
* Moderation

Question 2

What is the difference between moderation and mediation?

Answer 2

• Mediation
  – Can a third variable explain why two variables are associated?
• Moderation
  – Does a third variable change the size of an association between two variables?
  Same as ‘interaction’ in ANOVA

Question 3

What is the procedure for analysing interactions?

Answer 3

Procedure for analysing interactions
* SPSS
1. Put DV in first column
2. Put IVs in in next columns
3. Centre the IVs (take away their mean)
4. Use the ‘Compute’ option on the dialogue
menu to calculate interactions (i.e. X1X2).
5. Conduct a multiple regression to determine
whether DV is predicted by any of the IVs or
interactions.
6. If the interaction terms are not significant
(and there is no reason to keep them) then
they can be removed from the analysis.

Question 4

When to use correlation vs multiple regression

Answer 4

2 IVs – correlation

More than 2 – can look at multiple regression, moderation and mediation

People do not tend to look for multiple interactions, more for ones you think are there theoretically

Question 5

What are some guidelines for using regression?

Answer 5

• Include all IVs that you think will predict
  your dependent variable.
• Avoid IVs that are strongly correlated with
  each other.
• Ideally, your IVs should be experimentally
  manipulated.
• Take care in inferring causality.

Question 6

Difference between multiple regression and regression

Answer 6

Multiple regression
* This is just like regression but with more than one independent variable.

Question 7

NEED TO DO MORE ON THIS