A team of psychologists and mathematicians from Technische Universität Dresden have shown in a modeling study that consecutive value-based decisions influence each other. This result is particularly relevant for the creation and evaluation of questionnaires in value-based decision research.

Do I eat a chocolate croissant for breakfast or rather the healthy oatmeal? Every day, we are facing decisions, many of which we make unconsciously, others consciously and under consideration of the circumstances. The latter are scientifically referred to as value-based decisions, since each option is assigned certain attributes with an individual value, e.g., availability, health, taste, etc. Scientists would like to better understand the basic mechanisms behind such decision processes and are therefore collecting neural and behavioral data in value-based decision-making studies. However, the results of these data collections are often analyzed in a way as if these trials would be independent from one another. Potential patterns that extend across multiple decisions are usually ignored.

However, it has been shown that such patterns are ubiquitous in perceptual decision making. "So why shouldn't there also be inherent dynamics behind value-based decision making?" wondered Stefan Scherbaum, professor of Methods of Psychology and Cognitive Modelling at TU Dresden. With an interdisciplinary team of psychologists and mathematicians, he set up a dynamic neural-inspired attractor model that predicts the dependence of consecutive value-based decisions. "In three experiments, we show that this dependence is also reflected in human decision behavior. Furthermore, we demonstrate that in questionnaires on psychological studies, there can be errors and misjudgements of participants' decision tendencies, if this dependence is ignored," explains Scherbaum.

Original publication:

Scherbaum S, Lade SJ, Siegmund S, Goschke T, Dshemuchadse M (2022) From single decisions to sequential choice patterns: Extending the dynamics of value-based decision-making. PLOS ONE 17(4): e0267249. https://doi.org/10.1371/journal.pone.0267249

Contact:

Prof. Stefan Scherbaum
Chair of Methods of Psychology and Cognitive Modelling
SFB 940
TU Dresden
Email: