PSYC 201 – NOVEMBER 20 , 2013 TH
Look at the marginal means:
Unequal means Main effect present
Equal means No main effect
Non parallel lines Interaction present
Parallel lines No interaction
Simple Effect Tests
When a significant interaction is obtained, one performs “simple effect” tests (aka
post hoc comparisons) to assess effect of one IV separately at each level of the other
This involves conducting two t-tests or one-way ANOVAs:
One test to see if IV A has an effect at level 1 of IV B, the other to see if IV A has an
effect at level 2 of IV B (or one to see if IV B has an effect at level 1 of IV A and
another to see if IV B has en affect at level 2 of IV A).
A two-way factorial design is an experiment in which two IVs are “orthogonally”
The simplest two-way design is a 2x2: each of the two IVs has two levels;
orthogonally crossing them creates 4 conditions.
In a 2x2 between-subjects --- see slide
A 2x2 factorial design tests for three effects:
1. Main effect of IV A
2. Main effect of IV B
3. Interactive effect of IV A and B
Interactions can take a variety of forms.
Significant interactions are interpreted with simple effect tests (post-hoc
comparisons) to compare the levels of the one IV while holding the other constant.
Factorial Designs with Subject Variables