Organizational Behavior and Human Decision Processes | 2019
Beyond preference reversal: Distinguishing justifiability from evaluability in joint versus single evaluations
Abstract
Abstract Extensive existing research has studied how decisions differ between joint evaluation (JE) and single evaluation (SE), but most of the research aims to demonstrate preference reversals between two alternatives that vary on two attributes simultaneously. Thus, extant research cannot tell whether the reversal occurs because one of the attributes has a greater effect in JE than in SE, or the other attribute has a greater effect in SE than in JE, or both. Going beyond preference reversals, this research examines options that vary on only one attribute and studies whether the single attribute has a greater effect in JE or SE. We posit that any single attribute has two underlying characteristics—evaluability (i.e., whether people can evaluate a given value of the attribute without having to compare it with other values) and justifiability (i.e., whether people believe they should base their decisions on the attribute). Whether the single attribute has a greater effect in JE or SE depends on both the attribute’s evaluability and justifiability. Specifically, (a) a high-justifiability/low-evaluability attribute (e.g., whether a candidate for a programming job has written 100 or 200 programs) has a greater effect in JE than in SE, and (b) a low-justifiability/high-evaluability attribute (e.g., whether the candidate belongs to a discriminated-against minority group) has a greater effect in SE than in JE. While the first proposition has been tested in prior research on evaluability, the second has not. Four experiments, including one in a naturally-occurring setting and another with orthogonal manipulation of evaluability and justifiability, tested and supported these propositions, especially the second.