Evidence or Confidence: What Is Really Monitored during a Decision?

Area of Science:

• Cognitive Science
• Neuroscience
• Decision Science

Background:

• Decision-making models typically rely on accumulated evidence, treating confidence assessment as secondary.
• Standard models like the drift-diffusion model (DDM) do not integrate confidence into the core choice mechanism.

Purpose of the Study:

• To propose and validate a novel confidence-driven decision-making model (cDDM).
• To demonstrate that monitoring evolving confidence, rather than raw evidence, can drive adaptive choices.
• To show the generalizability of the cDDM to multi-option decisions.

Main Methods:

• Development of a confidence-driven drift-diffusion model (cDDM).
• Reinterpretation of existing findings on perceptual and value-based decisions through the lens of the cDDM.
• Theoretical extension of the model to accommodate decisions with multiple alternatives.

Main Results:

• The cDDM offers a more parsimonious explanation for previously observed decision-making phenomena.
• The model naturally generalizes to decisions with more than two options, a limitation of traditional DDMs.
• Confidence monitoring provides a framework for understanding the trade-off between decision accuracy and resource allocation.

Conclusions:

• Confidence, not just evidence, is a crucial factor monitored during decision-making.
• The confidence-driven drift-diffusion model (cDDM) provides a unified and more generalizable framework for understanding choice behavior.
• Future empirical research is needed to further differentiate between evidence-accumulation and confidence-driven models.