PSY 250 Lecture Notes - Lecture 3: Helping Behavior, Effect Size, Conscientiousness
15 Oct 2016
School
Department
Course
Professor
Document Summary
Lecture outline: other types of designs- quasi experimental designs. Quasi-experimental designs: desig(cid:374)s (cid:449)he(cid:396)e the(cid:396)e"s o(cid:374)l(cid:455) partial control over ivs. Ex) participants are put into one or more conditions by some means other than random assignment. Why use a quasi-experimental designs: cannot always do true experiment, sometimes random assignment is impossible. Ex) cannot assign natural disasters: random assignments are often unethical. Hiv, smoking, pregnancy, abortion, diet, exposure to toxic, (cid:272)he(cid:373)i(cid:272)als, p(cid:396)ejudi(cid:272)ial attitudes . Quasi-experimental designs vary in quality: some allow for stronger conclusions than others. But possible to eliminate many alternative explanations. Psy250 test 3: we need to know about the bad designs so we can avoid using them and design better quasi-experiments. Bad quasi-experimental designs: one group posttest. Why: because this design is of virtually no value, one group pretest-posttest. If response gets better is it due to treatment: no, rival hypotheses (threats to internal validity) in one group pretest-posttest design.
