Final-lecture 14 Flashcards
Define mediation
- extension of general linear model
- explaining why relationships occur between variables
- goes beyond simply describing relationships
- more rigorous test of theory
What is the conceptual model of mediation using X, M and Y?
- X–>M–>Y
- series of relationships
What is M?
-mediator variable
What is a mediator variable?
- Any variable that comes in between X and Y
- without M, X and Y would not be related
Would X and Y be related without M?
No
What is the difference between c and c’?
- c is the relationship between X and Y without M controlled
- c’ is this relationship with M controlled
If mediation occurs, what should c and c’ equal?
- c should be greater than c’
- c’ should not be significant
Why should we see non-significance in a mediator relationship?
- think of the wheel analogy where X turns M which turns Y
- we should not see a relationship between X and Y
C’ which controls for M results in what using the wheel analogy?
- stops M from moving and just allows X to show its effects on Y
- thus we shouldn’t see significance as X should not effect Y
What happens if the significance is decreased for c’ but not not-significant?
-partial mediation
What is necessary for it to be a complete mediation?
-c’ must be non-significant
Is a change in the slope for c and c’ enough for partial mediation?
- no, you need a difference in the level of significance
- this could be due to random error
In a mediator diagram, do you use standardized or unstandardized betas as slopes?
- standardized betas
- but you get your stars from unstandardized beta
What is the direct effect?
-X on Y
What is the indirect effect
-X on M on Y