Convince Me: Modeling Naturalistic Decision Making

Abstract

This paper is positioned in response to a call for an exchange of dialogue between researchers in the fields of macrocognition and computational modeling. Our work encourages examination of the complementarities that exist between these fields proposing that some of the challenges associated with micromodeling perspectives may be addressed by drawing upon “midgranularity” cognitive architectures. The study documented here demonstrates the value of modeling macrocognitive phenomena using the midgranularity architecture Convince Me. Our results suggested a moderate degree of fit between fund managers' decision making and the theory of explanatory coherence. Insights into the macrocognitive processes of sense making, uncertainty management, and mental simulation are examined. We anticipate that this will demonstrate the utility of computational modeling for revealing the shortcomings of macrocognitive models and that this will not only motivate increased theoretical specification but will also assist in the legitimization of cognitive modeling methods within macrocognitive inquiry.