Main effect – with variable one, add both and divide by two to get the mean and
compare to other variable.
Interaction – analogy for non-parallelism
1. B (There can be no main effect of A and B but there is an AxB interaction)
Example: cross-over interaction (each mean of each variable are equal)
3. C (Simple effects test is done to analyze data from factorial design that yields a
Why believe a claim?
-Strength of evidence varies
-Cannot always do true experiments. Sometimes settle for quasi experiments.
-Ex post facto (After the fact) studies are particularly weak (e.g. I use Pantene and
my hair is great!)
-Adding a control (comparison) condition of some kind can increase internal