214 (2) Lecture 4 Flashcards
(16 cards)
What is moderation?
A variable that affects the strength of a relationship between an IV and a DV.
Same as an interaction in ANOVA
What are the 3 prototypical types of moderation?
- IV, Moderator, & DV = dichotomous/ categorical
- IV & DV = continuous; moderator = dichotomous
- IV, Moderator, & DV = continuous
How do you test for a type 1 moderation?
IV & Moderator = dichotomous
- 2x2 ANOVA
- 2 IVs and a DV
- Third effect- the interaction
- interaction tells us if we have moderation
How do you test for type 2 moderation?
IV and DV = continuous, Moderator= dichotomous
- data split by moderator
- create interaction term between IV and moderator
- Multiply the centered variable with the moderator
- conduct hierarchical regression
- significance of the difference in R2 represents sig of interaction effect (also look to see if interaction variable is sig)
How do you create interaction term between IV and moderator?
Mean-center the continous IV. By subtracting the mean of the IV from each participant’s score on the IV
What does model 1 in the spss output show you when testing for type 2 moderation?
The results without the interaction term
What does model 2 in the spss output show you when testing for type 2 moderation?
The results with the interaction term
How do you test for type 3 moderation?
IV, Moderator, DV = continuous
- same as type 2, except we mean center both the IV and the Moderator to create an interaction term
- main effects and interaction term entered into hierarchical regression
How do you mean center a score?
IV score - mean IV score
Moderator IV score- mean moderator IV score
How do you calculate the interaction term for type 3 moderation
centered IV x centered moderator
what is simple slope analysis?
The regression of the outcome (y) on the predictor (x) at a specific value of the moderator (z)
What does simple slope analysis do?
Tests the relationship between IV and DV under 3 leveld (slopes) of the moderator:
- low group (mean - 1SD)
- moderate (mean)
High group (mean + SD)
Why do you use simple slope interaction?
To explore the nature of the interaction
How do you work out the simple slope interaction?
Using Kris Preacher’s simple slopes online calculator
When is moderation tested through ANOVA?
When both IVs are categorical
When is moderation tested through regression?
When you have a dichotomous moderator or a continuous moderator
