Abstract
The information available to our senses is limited, making multiple interpretations possible. The brain decides for the most probable interpretation to provide stable and reliable percepts. Depending on the quality of the sensory information, this perceptual decision can become difficult. Recent studies indicate the existence of a meta-perceptual evaluation system, rating the reliability of perceptual decisions. In the current study, we investigated whether the necessity to explicitly report reliability of the perceptual decision changes processing of the observed stimulus. Methods: We presented ambiguous and unambiguous Necker cubes in random order. In Condition 1, participants indicated the perceived 3D cube orientations. In Condition 2, participants additionally reported the confidence of their perceptual decision. 17 participants were included (24 to 32 years; 12 females). All participants had a corrected vision of at least 0.8 visual acuity and no neurological problems or eye diseases. EEG was recorded with 32 active silver/silver-chloride electrodes. EEG data was sorted for condition and stimulus and averaged to ERPs. Results: We found overall larger ERP amplitudes in Condition 1 compared to Condition 2, starting with an occipital component at 104 ms (E1), followed by a frontopolar component at 172 ms (E2), a parietal P300-like signal at 324 ms (E3), and a temporally sustained positivity lasting for more than 400 ms (E4). Discussion: The a priori knowledge about the evaluation task may amplify visual processing units (E1). Perceptual decision and top-down evaluation steps may start immediately after the lattice-gestalt construction (E2). In Condition 2 the evaluation result needs to be kept in working memory until the evaluation response is executed 1600 ms after stimulus onset (E4). In summary, the necessity to rate our perceptual decisions starts to alter perceptual processing remarkably early. We seem to perceive the world differently if we need to evaluate our percepts.