Participants in randomized trials sometimes prefer to be assigned to one arm rather than another. While randomization aims to ensure that participants with different preferences are spread equally among the arms of the trials, it cannot always address the possible effects of participant preferences on outcomes, compliance, and attrition. This paper discusses the conceptual, design, and statistical complications associated with preference effects and evaluates a number of experimental designs that have been proposed to gauge and minimize the effect of participant preferences.
The Annals of the American Academy of Political and Social Science 02/2010; 628(1):176-188. DOI:10.1177/0002716209351524 · 1.01 Impact Factor