Sorry just read the paper again to answer another question--I see that there was also an ANOVA on Prejudice manipulation, but the factors and levels are the same
They did the same ANOVA design with the same IVs but did it on 2 DVs:
IVs = Subject race(2)* evaluators race(2) * prejudice condition (3) on (DV) Prejudice maipulation
Anova is a tough one to get your head round-we had to do them manually when I did my first degree so you get to grips with them better. Interpreting them past a 2* 2 design gets iffy and interactions can only really be interpereted using output plots--unless your a math-head.
An independant variable is the one the experimenter manipulates (even if there is no real manipulation--eg quasi experiments, one class group compared to another). The dependant variable is what is measured--it depends upon the independent variables.
The way this study is described does lack a bit of clarity but lets have a go.
In the manipylation check, the IV for the t test was the manipulated factor--qualifications, DV evaluation of candidate.
Next to see if race had an impact (IV=race_ 2 levels) on evaluation of employabity (DV) or evaluation of qualifications (DV)
Finally the tough one
the DV was Stroop scores so the IVs that impacted on them were:
2 levels (subject’s race) * 2 levels (evaluator’s race)*3 levels (prejudice condition)
As I said a clearer description in the paper would have helped
Hope this helps