Proceedings of the Annual Meeting of the Cognitive Science Society

Leveraging people's proficiency in extracting summary statistics from ensembles, we conducted two studies in which we presented rating information from consumer feedback systems through color-coded icon arrays. The investigation aims to explore how different icon array arrangements (ascending, descending, random) influence decision making and average estimation across varying levels of ensemble means (average ratings) and ensemble sizes (review volume). Our results revealed four key insights: 1) Preferences for rating variance differed, particularly at the extremes of the average rating spectrum when ensembled were dominated by one or two rating categories. 2) Structured information yielded greater certainty in responses, with confidence increasing when the task setup aligned with task goals. 3) Ensemble size prompted individuals to adapt strategies based on contextual needs. 4) Unstructured presentations led to higher estimation accuracy, suggesting that a lack of structure may encourage heightened processing effort.