Factorial ANOVA

What is ANOVA like in comparison to t-tests?

more flexible bc it compares more than 2 groups simultaneously

What are conditions like in 1 way anova?

levels of the same IV

What is a factorial design?

bt subjects design with more than 1 IV

What will factorial ANOVA have?

different pps in each condition

What is a benefit of a factorial design?

we can look at how variables interact (how variable work in combination)

What is an interaction?

effect of 1 IV may depend on the level of the other IV - may be more interesting than main effects which are effects of a variable in isolation

What are marginal means?

the means for 1 IV averaged across all levels of other IV

What is the effect of 2 variables like?

not additive

What may the presence of an interaction affect?

the generality of the main effects even if they are significant

What is total variability bt all scores and grand mean split into?

variability bt grp averages (variability explained by our IVs) and error variability not explained by our IVs

What increases sensitivity of designs?

adding extra factors and interaction term which shrinks error term

What happens if p>.05 when checking for homogeneity of variance?

we assume the groups have similar variances and you can continue with ANOVA

What could happen with 2 IVs?

we could have main effect of IV and nothing else (below parallel horizontal lines)

main effects of IV and IV² but not interaction (lines parallel but not horizontal)

interaction but no main effects - lines form a cross

What are ordinal interactions?

non parallel lines slope in same direction but do not cross

treatment has a more intense effect in 1 condition than another

When may an ordinal interaction be spurious (not accurate)?

data is ordinal than interval

floor/ceiling effects

What do ordinal interactions assume?

interval data

What is a ceiling effect/

reading has got to a point where the readings/variables can’t go any higher

What is a cross over interaction?

when a factor has 1 effect in 1 condition and opposite effect in another condition

not vulnerable to ordinal data issues

When do we use posthocs in factorial ANOVA?

to interpret main effects when IV has more than 2 levels

What do interaction posthocs aid?

interpretation of a significant interaction

compare cell means across factors

What design is used when the same pps are tested across multiple levels of 2 IVs?

2 way RM design

What is the design called if the same pps are tested across multiple levels of 1 IV but contribute to only 1 level of another IV?

mixed design