Stone, C., Mattingley, J. B., & Rangelov, D. (2022).
Trends in Cognitive Sciences, 26(5), 419–431.
https://doi.org/10.1016/j.tics.2022.02.004
Abstract
The ability to change initial decisions in the face of new or potentially conflicting information is fundamental to adaptive behavior. From perceptual tasks to multiple-choice tests, research has shown that changes of mind often improve task performance by correcting initial errors. Decision makers must, however, strike a balance between improvements that might arise from changes of mind and potential energetic, temporal, and psychological costs. In this review, we provide an overview of the change-of-mind literature, focusing on key behavioral findings, computational mechanisms, and neural correlates. We propose a conceptual framework that comprises two core decision dimensions – time and evidence source – which link changes of mind across decision contexts, as a first step toward an integrated psychological account of changes of mind.
Highlights
- Changes of mind are observed during decision-making across a range of decision contexts.
- While changes of mind are relatively infrequent, they can serve to improve overall behavioral performance by correcting initial errors.
- Despite often improving performance, changes of mind incur energetic and temporal costs which can bias decision makers into keeping their original responses.
- Computational models of decision-making have demonstrated that changes of mind can result from continued evidence accumulation in the post-decisional period.
- Brain regions involved in metacognitive monitoring and affective processing are instrumental for change-of-mind behavior.
Concluding remarks
Changes of mind have received less attention in the scientific literature than the decisions which precede them. Nevertheless, existing research reveals a wealth of compelling findings, supporting changes of mind as a topic worthy of further exploration. In this review, we have covered changes of mind from a behavioral, computational, and neural perspective, and have attempted to draw parallels between disparate lines of research. To this end, we have proposed a framework comprising core decision dimensions relevant to change-of-mind behavior which we hope will foster development of an integrated account. These dimensions conceptualize changes of mind as iterative, predominantly corrective behavioral updates in the face of newly arriving evidence.
The source of this evidence, and how it is integrated into behavior, depends upon both the decision context and stage. However, the mechanisms underlying changes of mind are not equally well understood across the entire decision space. While changes of mind for perceptual decisions involving accumulation of sensory evidence over short durations have been well characterized, much work is needed to extend these insights to the complex decisions we make in everyday life.
One conclusion, ignoring contradictory evidence can account for "confirmation bias".
