6. Factorial ANOVA

Question 1

Factorial ANOVA

Answer 1

A factorial design (such as a two-way between-subject’s ANOVA), is an ANOVA with multiple independent variables (e.g. Gender and alcohol consumed).

Question 2

Benefits of a factorial ANOVA

Answer 2

Factorial designs allow us to look at interactions between IV’s. This means we can see how the effects of one IV is dependent on the effects of another, this is often more interesting than a simple main effect. For example, hangover might have more effect on sleepiness in a stats lecture than in a clinical one.

Question 3

Running Factorial ANOVA as Regression Equation

Answer 3

Question 4

Theory of two-way ANOVA factorial design

Answer 4

SST (total variability between scores), SSM (how much variability can be accounted for by manipulations overall), SSN (how much variability can be explained by each model), SSAxB (How much variability explained by interaction), (how much variability is unexplained, residuals).

Question 5

A Priori Follow Up

Answer 5

You can design planned comparisons to follow up an interaction similarly to an ANOVA (see those notes). However, to get the values for coding interactions, we multiple the two variables that are interacting. See example below.

Question 6

Post Hoc Follow Up

Answer 6

Similarly again to ANOVA, we can do post hoc tests if we made no predictions about relationships/interactions prior to analysis. To do this, we use SIMPLE EFFECT ANALYSIS. This compares IV’s at each level, while planned comparisons only compares the things that the experimenter is interested in

Question 7

Simple Effect Analysis

Answer 7

looks at the effect of one independent variable at individual levels of the other independent variable. E.g. look at gender at each level of alcohol - compare avg. attractiveness of date for men compared to avg. attractiveness for women - do this for each level