Psychology 2800E Lecture Notes - Lecture 14: Internal Validity
Document Summary
More than 2 groups: drawing a conclusion is easy when we only have 2 groups. If the ratio [bt groups / within groups] is > 1 then treatment is different from control. But what does the ratio tell us when we have more than 2 groups. With > 2 treatment conditions, if ratio > 1. 0, then at least 1 group mean is diff from another. More than 2 of our groups may differ but at least 2 differ. We determine which group means are different through analytical comparisons (comparing treatment means in pairs) T1 vs t2 t1 vs t3 t2 vs t3 (3 treatment groups; analogous pairings would be work w > How to form groups: having 2 qualities random assignment & equal sample sizes. Block randomization: rst make blocks groups of participants. Then randomly assign participants within blocks to treatments. # participants in each block = # of treatments. Eg. 4 treatments > 4 participants per block.