Withdrawal versus reversal: A necessary distinction?

Abstract

In the behavior analysis literature, the terms withdrawal and reversal are widely used to refer to various iterations of the basic A-B-A or A-B-A-B experimental design structure. Although these terms were initially used to label two distinct but closely related experimental designs, it now appears that many researchers have come to use these terms interchangeably. A review of major behavior analytic journals published between 2009 and 2013 reveals a strong preference among researchers for use of the term reversal in labeling the design described in the publication, whether or not that design is characterized by the specific features initially associated with the reversal design. This paper describes the origin of this definitional distinction and reviews the case for abolishing that distinction.