مقاله سال ۲۰۱۶

 مجله تحقیقات بازاریابی – Journal of Business Research

Numerous empirical studies underpin that consumers’ buying decisions depend on the context in which they are embedded (Celedon, Milberg, & Sinn, 2013; Dhar & Simonson, 2003; Simonson & Tversky, 1992). The compromise effect, according to which consumers prefer options positioned—as a compromise—in the middle position of a perceived product space to more extreme options, ranks among the most prominent of these context effects (Ryu, Suk, Yoon, & Park, 2014; Simonson, 1989). Specifically, researchers have confirmed that adding a high-tier premium option does promote the choice share of a compromise option of medium quality and price compared to that of a low-tier option (e.g., Neumann, Böckenholt, & Sinha, 2016; Simonson, 1989).

Recently, however, Lichters, Sarstedt, and Vogt (2015) have emphasized that the experimental designs of practically all compromise effect studies are subject to serious limitations, which hinder the generalization of their findings to real buying decisions. The authors’ review of 47 compromise effect studies published over a 30-year period shows that such studies mostly rely on hypothetical choices that do not entail real economic consequences and use imaginary items, or unrealistic product descriptions. Further, the participants could not evaluate the products prior to making a choice and seldom had the option to defer buying. Testing whether the products under research have the same relevance for the decision makers in the sample as for those in the population to which the effect should apply is usually ignored, as is controlling for the participants’ perception of the choice alternatives. Instead, prior research designs have fostered artificial learning processes triggered by a high number of repeated choices. In light of the biases that potentially result from such design-related limitations, Lichters et al. (2015, p. 14) conclude that researchers should consider context effects, such as the compromise effect, “from a perspective that takes the basic conditions of real-world settings into account” and that “future research should systematically evaluate the effects of the identified background factors.”

Tying-in with Lichters et al.’s (2015) call, the present work (1) examines the compromise effect’s robustness in an experimental design that fosters the external validity (Table 1) and (2) investigates the impact of the product type as a potential moderating background factor on the effect’s occurrence and magnitude. Specifically, to date, researchers know only a little about the extent to which the compromise effect’s magnitude would differ depending on whether, in realistic research designs, consumers choose between sets of durables or fast-moving consumer goods (FMCG). Purchase decisions on durables evidently induce fundamentally different mental processes than on FMCG, leading to more elaborated information processing and decision making (e.g., Duncan, 1972; Milliken, 1987), which should also affect the compromise effect’s magnitude (Neumann et al., 2016).