Lecture 9: Moderation in ANOVA (Alt 3) Flashcards
(36 cards)
What does ANOVA examine in moderation analysis?
Mean differences across categorical IVs on a continuous DV.
When should regression be used instead of ANOVA?
When IVs are continuous, to avoid arbitrary categorisation and loss of information.
How is total variance partitioned in ANOVA?
Into between-groups variance (systematic variance due to group membership) and within-groups variance (random/unexplained variance or error).
What does the F-ratio express in ANOVA?
The ratio of systematic variance to error variance.
What is the basis for inferential testing in ANOVA?
The F-ratio.
What does factorial ANOVA allow?
Inclusion of multiple IVs (factors) in the same model to be considered simultaneously.
What are the two types of effects tested in factorial ANOVA?
Main effects and interaction effects.
What do main effects in factorial ANOVA represent?
The impact of each factor averaged across levels of the other factor.
What do interaction effects in factorial ANOVA test?
Whether the effect of one IV varies across levels of another IV.
How is moderation operationalised in ANOVA?
As an interaction effect.
When does a variable moderate the relationship between an IV and DV?
When the strength or direction of that relationship depends on another IV.
In the facial masculinity and attractiveness study, what did disease primes predict?
Preference for masculine faces (good genes).
In the facial masculinity and attractiveness study, what did scarcity primes predict?
Preference for feminine faces (good providers).
How are main effects evaluated in ANOVA?
Using marginal means (average DV scores across all levels of the other IV).
What does a non-parallel line in an interaction plot indicate?
Moderation (interaction present).
Can interactions occur without main effects?
Yes, interactions can occur even without main effects.
Are main effects and interactions dependent on each other?
No, they are conceptually and statistically independent.
What is the role of omnibus tests in ANOVA?
Detect whether any group differences exist but do not specify where those differences lie.
What is the purpose of follow-up tests in ANOVA?
To localise significant effects identified by omnibus tests.
What follow-up tests are used for main effects?
Simple comparisons, including pairwise t-tests and linear contrasts.
What follow-up tests are used for interaction effects?
Simple effects analyses.
What do simple effects analyses test?
The effect of one factor at each level of the other factor.
What follow-up procedures are needed if factors have more than two levels?
Further simple comparisons may be required.
What is a characteristic of between-subjects designs in ANOVA?
All participants appear in only one condition.
