“Fact or artifact? Empirical evidence on the robustness of compromise effects in binding and non-binding choice contexts”

In this article, the compromise effect (CE) is investigated according to which the choice share of a particular product is expected to increase when it is in an intermediate position in a subset of a product line. In a laboratory-based replication study, we tested CE both in hypothetical non-binding settings most commonly used in previous research on CE as well as in settings where experienced consumers make unforced and binding purchase decisions between real brands of two particular categories. While results prove robustness of the compromise effect even in the binding choice context, its magnitude is significantly reduced, hence indicating a hypothetical bias. Thus, since the CE is evidently less prevalent in the more realistic binding choice setting, compromise effects must be considered overrated to a certain degree in the previous studies on hypothetical decisions.