Decisions are based on less information than metacognitive judgments in multialternative contexts

Area of Science:

• Cognitive Neuroscience
• Decision Science

Background:

• Conflicting evidence exists on information accessibility from unchosen alternatives during decision-making.
• Understanding information processing in multi-alternative choices is crucial for cognitive models.

Purpose of the Study:

• To quantify the accessibility of information from unchosen alternatives after an initial decision.
• To investigate the dissociation between decision-making and metacognitive systems.

Main Methods:

• Employed a second-guess paradigm in two preregistered experiments with varying numbers of alternatives (4 and 12).
• Utilized computational modeling to fit behavioral data and assess evidence accessibility.
• Extended analysis to two previously published datasets with different stimuli and alternative counts (6 and 3).

Main Results:

• Evidence from initially unchosen options is corrupted by noise after the first decision, limiting accessibility for subsequent choices.
• Computational models indicate that the decision system has access to less evidence than the metacognitive system.
• Metacognitive computations benefit from a separate, noise-unaffected evidence channel.

Conclusions:

• Reconciles conflicting findings regarding information availability in multi-alternative decisions.
• Demonstrates a dissociation between the decision-making and metacognitive systems.
• Suggests inherent constraints in decision processes and highlights the robustness of metacognition.