TY - JOUR
T1 - Cognitive costs of decision-making strategies
T2 - A resource demand decomposition analysis with a cognitive architecture
AU - Fechner, Hanna B.
AU - Schooler, Lael J.
AU - Pachur, Thorsten
N1 - Funding Information:
This work was supported by the Center for Adaptive Behavior and Cognition (ABC) and the Center for Adaptive Rationality (ARC) at the Max Planck Institute for Human Development, as well as the International Max Planck Research Network on Aging (MaxNetAging). We thank Gregor Caregnato, Julia Cleeman, Flora Klie, Marianne Ritthausen, Annika Walinski, and Swantje Wenzel for the data collection, Rebecca Müller and Jürgen Rossbach for help with the design of the stimulus materials of the empirical study, and Dan Bothell from the research group of John R. Anderson at Carnegie Mellon University and Klaus Oberauer and Peter Shepherdson at the University of Zürich for helpful comments on this project and manuscript. We thank Anita Todd for editing the manuscript.
Publisher Copyright:
© 2017 Elsevier B.V.
PY - 2018/1
Y1 - 2018/1
N2 - Several theories of cognition distinguish between strategies that differ in the mental effort that their use requires. But how can the effort—or cognitive costs—associated with a strategy be conceptualized and measured? We propose an approach that decomposes the effort a strategy requires into the time costs associated with the demands for using specific cognitive resources. We refer to this approach as resource demand decomposition analysis (RDDA) and instantiate it in the cognitive architecture Adaptive Control of Thought–Rational (ACT-R). ACT-R provides the means to develop computer simulations of the strategies. These simulations take into account how strategies interact with quantitative implementations of cognitive resources and incorporate the possibility of parallel processing. Using this approach, we quantified, decomposed, and compared the time costs of two prominent strategies for decision making, take-the-best and tallying. Because take-the-best often ignores information and foregoes information integration, it has been considered simpler than strategies like tallying. However, in both ACT-R simulations and an empirical study we found that under increasing cognitive demands the response times (i.e., time costs) of take-the-best sometimes exceeded those of tallying. The RDDA suggested that this pattern is driven by greater requirements for working memory updates, memory retrievals, and the coordination of mental actions when using take-the-best compared to tallying. The results illustrate that assessing the relative simplicity of strategies requires consideration of the overall cognitive system in which the strategies are embedded.
AB - Several theories of cognition distinguish between strategies that differ in the mental effort that their use requires. But how can the effort—or cognitive costs—associated with a strategy be conceptualized and measured? We propose an approach that decomposes the effort a strategy requires into the time costs associated with the demands for using specific cognitive resources. We refer to this approach as resource demand decomposition analysis (RDDA) and instantiate it in the cognitive architecture Adaptive Control of Thought–Rational (ACT-R). ACT-R provides the means to develop computer simulations of the strategies. These simulations take into account how strategies interact with quantitative implementations of cognitive resources and incorporate the possibility of parallel processing. Using this approach, we quantified, decomposed, and compared the time costs of two prominent strategies for decision making, take-the-best and tallying. Because take-the-best often ignores information and foregoes information integration, it has been considered simpler than strategies like tallying. However, in both ACT-R simulations and an empirical study we found that under increasing cognitive demands the response times (i.e., time costs) of take-the-best sometimes exceeded those of tallying. The RDDA suggested that this pattern is driven by greater requirements for working memory updates, memory retrievals, and the coordination of mental actions when using take-the-best compared to tallying. The results illustrate that assessing the relative simplicity of strategies requires consideration of the overall cognitive system in which the strategies are embedded.
KW - ACT-R
KW - Cognitive costs
KW - Computational modeling
KW - Decision making
KW - Strategy
KW - Working memory
UR - http://www.scopus.com/inward/record.url?scp=85030460927&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85030460927&partnerID=8YFLogxK
U2 - 10.1016/j.cognition.2017.09.003
DO - 10.1016/j.cognition.2017.09.003
M3 - Article
C2 - 28987923
AN - SCOPUS:85030460927
SN - 0010-0277
VL - 170
SP - 102
EP - 122
JO - Cognition
JF - Cognition
ER -