Probing doors to visual awareness: Choice set, visibility, and confidence

ABSTRACT Visibility and confidence are two subtly (sensation-based and intuition-based) different ways to assess our visual experiences. We investigated how different choice sets affect the ability of confidence and visibility judgments to retrieve conscious information from perceptual decision-making processes. Six participants made introspective judgments on sinusoidal gratings at close to chance signal-to-noise values. We used signal detection theory and mixed regression models to analyze choices and pupilometric recordings. The results suggest that allowing null responses (e.g., “I saw nothing”) asymmetrically affects the signal distribution variances associated with null and non-null responses; confidence is more effective than visibility in accessing unconscious information; and bias-free methods hinder accurate introspective judgment by forbidding free decision criteria. Our findings underscore the importance of the context afforded by choice and stimulus sets, which determines the extent to which elicited introspective judgments are based on the actual information from the past events they are intended to measure.