Contextualizing Statistical Suppression Within Pretest-Posttest Designs
Full text PDF
Cited references information:
, Weiss, Jonathan A.
, Tajik-Parvinchi, Diana
, Cribbie, Robert A.
statistical suppression; Lord's paradox; pretest-posttest designs
(no sample data)
Statistical suppression occurs when adjusting for some third variable enhances or substantially modifies the association between an initial predictor and an outcome. Although many methodologists have discussed this phenomenon, very little work has examined suppression in longitudinal regression models such as the pretest-posttest design. This research addressed this gap with two separate studies. Study One was a literature review that reviewed 80 articles from a variety of fields within psychology. Study Two was an analysis of a large longitudinal clinical dataset via 925 statistical models. Both studies revealed consistent results: in approximately 20% of instances, suppression effects were observed and were attributable to the inclusion of a pretest measure. Results underscore that controlling for pretest measures when assessing change may be of value, as this can help clarify the associations between predictors and posttest outcomes.