Confirmative or Normative? Information Redundancy in Decision Support Systems

Abstract

Decision Support Systems (DSS) provide users with information that is combined with their existing knowledge or information from other available sources. We investigate how redundancy of the information from the DSS affects the decision to purchase decision support, as well as the use of and trust in the support. Participants performed a classification task. Depending on the condition, they could purchase advice from either an independent, partly redundant, or fully redundant DSS. Purchase decisions, classification performance and trust were recorded. Participants preferred to purchase the independent system, performed better with it and trusted it more, compared to the partly or fully redundant systems. There was no significant learning effect across blocks. Overall, the results support a model of more normative, rather than confirmative preferences for decision support.