PNB 3RM3 Lecture Notes - Lecture 31: Internal Validity, Demand Characteristics
12 Jul 2017
School
Department
Course
Professor
Document Summary
Design confounds selection effect - whether you have created equivalent groups in within groups experiment. Order effects - effect about rather than iv. One group pre test post test design - a bad design. Change on your dv, not sure if spontaneous or due to iv (did your treatment cause people to change?) i. e. depression program where clients improved. Don"t know if they got better anyway due to maturation or because of your treatment. Solution: if you had a control group (comparison group), you could know if a change actually occurred or not. Something changed everyone, so it looks like it is your iv. But here, its because of an external event that changed everyone. Did anything affect anyone systematically? i. e. going green, but actually its because temperature is going down - history thing: that you stopped using ac; not saving. Extreme scores at time 1 likely to be normalized at time 2.
