How choice affects and reflects preferences: revisiting the free-choice paradigm

Abstract

After making a choice between 2 objects, people reevaluate their chosen item more positively and their rejected item more negatively (i.e., they spread the alternatives). Since Brehm's (1956) initial free-choice experiment, psychologists have interpreted the spreading of alternatives as evidence for choice-induced attitude change. It is widely assumed to occur because choosing creates cognitive dissonance, which is then reduced through rationalization. In this article, we express concern with this interpretation, noting that the free-choice paradigm (FCP) will produce spreading, even if people's attitudes remain unchanged. Specifically, if people's ratings/rankings are an imperfect measure of their preferences and their choices are at least partially guided by their preferences, then the FCP will measure spreading, even if people's preferences remain perfectly stable. We show this, first by proving a mathematical theorem that identifies a set of conditions under which the FCP will measure spreading, even absent attitude change. We then experimentally demonstrate that these conditions appear to hold and that the FCP measures a spread of alternatives, even when this spreading cannot have been caused by choice. We discuss how the problem we identify applies to the basic FCP paradigm as well as to all variants that examine moderators and mediators of spreading. The results suggest a reassessment of the free-choice paradigm and, perhaps, the conclusions that have been drawn from it.