• Seminari: «The Cumulative Redundancy Bias»

  • Start: Friday, 24 September 12:00
    End: Friday, 24 September 13:00
  • Universitat de les Illes Balears, Carretera de Valldemossa, Palma, Espanya
  • A les 12 hores.

    En línia.

    A càrrec de Hans Alves.

    The formation of impressions, attitudes, judgments and decisions usually requires integration of multiple pieces of information into a summary value. Such information integration processes have for example been argued to resemble summation or averaging rules. What has been neglected by past research is that in many real-world contexts, distributed information is provided to perceivers in an already-aggregated format like running averages or running totals. Such cumulative information sequences require a drastically different way of integration. In cumulative sequences, the only relevant piece of information is the most recent one because all previous pieces of information are redundant. In other words, the best estimate of a lottery’s expected value, of a politician’s chance to win the election, or of a soccer team’s chance to win the season is given by the most recent cumulative observation of the lottery’s total wins, the politician’s average popularity or the soccer team’s standing in the league table. Hence, normative perceivers and decision makers should not be influenced by the history of a cumulative sequence of observations. However, disregarding previously-processed information is a surprisingly difficult task for humans. And consequently, human perceivers and decision makers are systematically biased by the redundancy inherent in cumulative information sequences. We present data from laboratory experiments, real-world sports team evaluations and simulations, that document the robust nature of the cumulative redundancy bias (CRB), its boundary conditions and real-world implications.