PNB 3RM3 Lecture Notes - Lecture 19: Pocket Cube, Analysis Of Variance, Complement Factor B
Document Summary
Measuring multiple ivs can address more research qs in one study. Most common type is a factorial design. Where you manipulate more than one iv at once. Good thing about experimental studies: everything else is constant or randomized, the only thing we change is the iv, any differences can be attributed to the iv, allows us to make causal conclusions. In this case, the first factor has three levels (e. g. cognitive, behavioural, cognitive- behavioural) Main effect of a: mean (a1b1, a1b2) vs mean (a2b1, a2b2) Main effect of b: mean (a1b1, a2b1) vs mean (a1b2, a2b2) Comparison of marginal means: factor a has two levels and factor b has two levels. If you give a notation to each of the unique groups you have, you have 4 different groups: a1b1, a2b2, etc. If there"s a difference in means, we have a main effect of a.