Research ReportThe role of evaluation mode on the unit effect

Recent research on the unit effect has suggested that consumers tend to ignore relevant unit information and over-rely on numeric magnitudes in judgments (e.g., perceiving the difference between 700 and 900 on a 1000-point quality scale to be larger than the difference between 7 and 9 on a 10-point scale). The current work investigates the nature of the unit effect by studying the role of different modes of evaluation, and types of information processing, on the unit effect. Specifically, three studies demonstrate that the unit effect occurs when options are evaluated simultaneously and attenuated when options are evaluated sequentially. The current article builds on research concerning comparative versus selective information processing. It demonstrates that, when information is processed in a comparative rather than selective manner, common elements in the decision (i.e., units) are more likely to be edited out, resulting in the unit effect.