Statistical regression as threat to internal validity refers to the fact that extreme scores for a scale (i.e., very high or very low scores) tend naturally to move toward the mean with repeated measurements; thus, it would not make sense to take only the clients with the extremely low pretest scores as your clients for an evaluation study and employ the one-group pretest-posttest design.

Question

Quizplus · Accepted Answer

Statistical regression is a threat to internal validity because it suggests that individuals with extreme scores on their first measurement will likely score closer to the mean on subsequent measurements. This natural tendency can falsely suggest an intervention effect in a one-group pretest-posttest design, especially if only individuals with extreme scores are selected.

Statistical Regression as Threat to Internal Validity Refers to the Fact