Computational Modeling and Simulation of Attitude Change. Part 1, Connectionist Models and Simulations of Cognitive Dissonance: an Overview

Abstract

Cognitive Dissonance Theory is considered part of the cognitive consistency theories in Social Psychology. They uncover a class of conceptual models which describe the attitude change as a cognitive consistency-seeking issue. As these conceptual models requested more complex operational expression, algebraic, mathematical and, lately, computational modeling approaches of cognitive consistency have been developed. Part 1 of this work provides an overview of the connectionist modeling of cognitive dissonance. At their time, these modeling approaches have revealed that a Computational Social Psychology project would acquire the community recognition as a new scientific discipline. This work provides an overview of the first computational models developed for the Cognitive Dissonance Theory. They are connectionist models based eitheron on the constraint satisfaction paradigm or on the attributional theory.Three models are described: Consonance Model (Shultz and Lepper, 1996), Adaptive Connectionist Model for Cognitive Dissonance (Van Owervalle and Joders, 2002), and the Recurrent Neural Network Model for long-term attitude change resulting from cognitive dissonance reduction (Read and Monroe, 2007). These models, and some others, proved from the very beginning the considerable potential for the development of cognitive modeling of the theories of cognitive dissonance. Revisiting the Cognitive Dissonance Theory once again only proves that this potential is even larger than expected.