PSYB70H3 Lecture 10: PSYB70_Unit_10-12
Document Summary
3 conditions of causality: covarinces , temp precedence and internal validity. Experiments are best at meeting the conditions with comparison gps, longitudinal design, and isolate controls. However, experiment might not be ideal by of logistics (not possible to do random assignment), ethnics (can"t just give ppl harmful drugs), generalizability(want to be able to generalize to real world context, ecological validity). Establish the 3 conditions even though we are not doing a true experiment. Nonequivalent control gp design: can"t randomly assign ppl to conditions. Selection effects: occur when it becomes difficult to know if effects due to iv or systematic differences exist between the participants. Waitlist design: longitudinal, can have the ppl on waitlist to treatment as a natural control gp to ppl who reached treatment. Equating: ensure equal characteristics across all gps. Matching: match participants based on identified criterion, u. Equating occurred after experiment and before analysis, sue computer algorithms. For true experiments both happened before random assignment.
