Editors' Introduction: Best Papers From the 2018 International Conference on Cognitive Modeling

Document Type

Article

Publication Date

2-2019

Abstract

Modeling and simulation are powerful techniques for explaining phenomena, uncovering new ones, and testing formal theories in all sciences, including cognitive science. Each year, modelers from across the globe gather at the annual International Conference on Cognitive Modeling (ICCM) to discuss progress on developing models of the human mind for different uses. We present the top three papers from ICCM 2018, which include applications of computational cognitive modeling to supplementing results from fMRI (Rice & Stocco), providing a formal demonstration that the theoretical construct of selective attention is not necessary to explain patterns of response times in visual search tasks (Kieras), and formally specifying a theory of working memory (Glavan & Houpt).

The selected papers demonstrate the diversity of areas with which modeling and simulation are involved in studying human cognition. Models of visual search often rely on a covert selective attention mechanism to explain task performance. David Kieras calls into question the necessity of assuming such a mechanism to explain response time effects by implementing a formal model that includes basic visual mechanisms, saccades, and simple task strategies. Across the feature search, conjunction search, and shape search tasks, Kieras’s model was capable of accounting for response times for both target‐absent and target‐present trials as well as the proportion of errors observed in human participants, including effects of item crowding in the visual stimulus.

Patrick Rice & Andrea Stocco used computational cognitive models to provide alternative hypotheses about the cognitive functions affected by the application of repetitive transcranial magnetic stimulation. Their model simulated the effect of stimulation of the left dorsal premotor cortex right as participants provide a response to an uninstructed stimulus. The authors used the ACT‐R cognitive architecture to demonstrate that the increased variability in response times observed in humans can result from interference in replanning during the process of responding to the uninstructed stimulus.

Finally, Joseph Glavan & Joseph Houpt formalized a theory of working memory, time‐based resource sharing (TBRS), within the ACT‐R cognitive architecture. Instantiating the theory within ACT‐R allowed the authors to predict task accuracy and response times when an articulatory rehearsal mechanism was integrated with the TBRS mechanism. Simulation results indicated the integrated mechanisms of TBRS and articulatory rehearsal provided a decent explanation of human response times and accuracy within a dual‐task environment. Further, Glavan & Houpt investigated model predictions under higher secondary task load conditions than is typically tested on human subjects, providing novel predictions and supporting the hypothesis that working memory capacity is a linear function of cognitive load. The authors received the Allen Newell Award for the best student‐led paper submitted to ICCM2018 for their research efforts.

These best papers apply computational cognitive modeling in different ways, demonstrating the diversity of application for cognitive modeling. Where they share common ground is their attention to the details of processing and knowledge and in using modeling and simulation to better understand, account for, and predict human cognitive processes. We hope you enjoy the papers as much as we did.

DOI

10.1111/tops.12409

Link to Full Text


Share

COinS