Asymmetric decoy effects on lower-quality versus higher-quality brands: Meta-analytic and experimental evidence

Prior research demonstrates that adding decoys to choice sets can increase choice shares of brands similar to decoys while reducing shares of brands dissimilar to decoys. Such effects have been dubbed attraction effects and violate the principles of independence of irrelevant alternatives (IIA) and regularity. We report a metaanalysis that, in addition to revealing heretofore unsupported range effects, demonstrates an effect of brand quality. Decoys reduce shares of lower-quality competitors more than they reduce shares of higher-quality competitors. Moreover, whereas IIA is violated throughout, regularity is violated only when higher-quality brands are targeted. Decoys increase shares of higher-quality brands but typically do not increase shares of lower-quality brands. To assess the generalizability of the mete-analytic pattern, we tested decoy effects on two distinct populations in a large experiment. The more traditional population replicated the mete-analytic pattern (standard asymmetry) while the more nontraditional population reversed it. These findings suggest that while the standard asymmetry is replicable, it may not generalize to all market segments.