chp 10: Moderation, mediation and multicategory predictors. Flashcards
What is moderation?
the combined effect of two plus predictor variables on an outcome.
The statistical term for moderation is…
an interaction effect.
How do you solve the problem of variables that cannot equal zero such as heartrate of gamers.
Grand mean centreing
How do you do grand mean centreing
transform the variable into deviations around zero.
What effect does grand mean centreing have on the highest order predictors?
No effect.
Does grand mean centreing have an effect on the lowest order predictors?
Yes. It effects the b
Two interpretations of predictors b’s with centreed variables.
effect of that predictor at mean value of the sample.
avaerage effect of the predictor across the range of scores.
Procedure for follow up for an interaction effect to work out what the nature of the effect is.
simple slopes analysis
What is simple slopes analysis?
Compare slopes: direction, significance and values
Does the relationship between predictor and outcome change at different levels of the moderator.
What is mediation?
When the strength of the relationship between a predictor variable and outcome variable is changed by another predictor variable.
Relationship is explained by a third variable.
How is mediation tested?
By assessing the size of the indirect effect and its confidence interval.
If the mediation confidence interval includes zero assume…
A genuine mediation effect does not exist.
When is dummy coding used?
With categorical variables that have more than two categories
How do you dummy code a categorical variable?
Code the variable into more variables with values of zero and one.
1982 method for testing for mediation effect.
Sobel test.
What does sobels test do?
Estimates indirect effect of a variable on outcome variable and its significance.
How to interpret a sobel test?
If significant the predictor significantly affects the outcome via the mediator.
Works well with large samples.
A method better than sobels for assessing mediation effects?
Bootstrap method and calculate confidence interval for indirect effect.
Moderation example
a lot of stress, + good social support, = fewer symptoms of depression, anxiety, fatigue…) than someone with low social support. Social support is the moderating variable.
mediation example
pizza increases positive morale which increases amount of work done
Moderator variables influence
Strength and direction of the relationship between two variables.
Mediator variables explain…
The relationship between two variables. The independent variable influences the mediator variable influences which influences the dependent variable
