Improving Decisions Using Nontraditional, Nonfinancial Accounting Information : Evaluation Mode Considerations When Benchmark Data Give Unclear Signals

How are decisions influenced by the presentation of nontraditional and nonfinancial accounting information, such as environmental measurements, when benchmark data provide vague signals for analysis? Generally, alternatives are analyzed either jointly (called joint evaluation mode, or JE) or separately (SE). Prior research suggests more focus on difficult-to-evaluate data, such as unfamiliar accounting information, in JE than SE (Hsee, 1996). Also, analyses involving benchmark data showing all alternatives performing better than a benchmark (a vague signal) receive higher evaluations in SE, yet higher evaluations in JE occur when all alternatives perform worse than benchmarks (Hsee & Leclerc, 1998). Two reported experiments use a 2X2 mixed factorial design; evaluation mode was manipulated as a between-subjects variable, i.e., either JE or SE, while benchmark data indicating alternatives’ performances were manipulated as a within-subjects variable, i.e., better or worse performance than the benchmarks. In Study 1 (n = 80), participants evaluated two equally weighted environmental accounting metrics with benchmark data for two factories and indicated how much they would invest in each based on success in achieving environmental objectives; one factory outperformed the other. Results indicate higher investment levels as well as more desirable investments in JE than SE when the benchmark data indicate worse-than-average alternatives (but not better-than-average alternatives). Study 2 (n = 53) was similar to Study 1, except 1) benchmark data were available for only one of the two metrics, and 2) the two factories were equally attractive. Investment levels were higher in JE than SE mode when alternatives performed worse than average (but not better-than-average alternatives). These results showcase cognitive consequences to evaluating nontraditional, nonfinancial accounting information in different evaluation modes. This awareness should inform designers of information systems, particularly those designing systems with environmental accounting attributes