The Missing Treatment Design Element

Abstract

We rely a great deal on the schematic descriptions that represent experimental and quasi-experimental design arrangements, as well as the discussions of threats to validity associated with these, provided by Campbell and his associates: Stanley, Cook, and Shadish. Some of these designs include descriptions of treatments removed, removed and then reintroduced, or reversed with use of multiple postobservations. The focus of this article is to consider what happens to the intervention between and among postobservations and how that relates to threats to validity and data modeling and analysis. Evaluation designs must consider such questions as “Has the intervention been removed, or is it still in place? If so, has it changed in intensity, or are there persistence or carryover effects?” This article adds to the evaluation methodology literature by filling a gap regarding the issue of treatment or intervention continuity between and among postobservations.