Vices and Virtues of Misguided Replications: The Case of Asymmetric Dominance

Abstract

The likelihood of replicating an effect such as asymmetric dominance (AD) largely depends on other, usually more important choice drivers (e.g., attributes, values). Accordingly, it is not surprising that the AD effect is often not observed when other choice drivers have greater impact and/or when the AD configuration is unlikely to be perceived. However, when price is an attribute (and in many other cases), the AD effect is often observed in both properly designed studies and the real world. The author notes that Frederick, Lee, and Baskin (2014) raise important questions, but they would have made a greater contribution had they (1) systematically studied the drivers of AD perceptions, (2) more accurately tried to replicate previous AD effect demonstrations, and (3) systematically studied the repulsion effect. The author also briefly comments on the Yang and Lynn (2014) studies, which he asserts did not test the AD effect properly in most cases.