Post-decision biases reveal a self-consistency principle in perceptual inference

Abstract

Making a categorical judgment can systematically bias our subsequent perception of the world. We show that these biases are well explained by a self-consistent Bayesian observer whose perceptual inference process is causally conditioned on the preceding choice. We quantitatively validated the model and its key assumptions with a targeted set of three psychophysical experiments, focusing on a task sequence where subjects first had to make a categorical orientation judgment before estimating the actual orientation of a visual stimulus. Subjects exhibited a high degree of consistency between categorical judgment and estimate, which is difficult to reconcile with alternative models in the face of late, memory related noise. The observed bias patterns resemble the well-known changes in subjective preferences associated with cognitive dissonance, which suggests that the brain’s inference processes may be governed by a universal self-consistency constraint that avoids entertaining ‘dissonant’ interpretations of the evidence.

Reference Repulsion Is Not a Perceptual Illusion

“…Furthermore, the response histograms of the average observer did not show any clear signs of bimodality (see Supplemental Material), in line the findings by Zamboni et al This is likely because the current reference repulsion bias, albeit clear and highly significant, was too subtle to lead to clearly bimodal response distributions and emphasizes the importance of analyzing the data in a different way to detect and quantify these subtle reference repulsion biases. Nevertheless, given the much smaller magnitude of the bias compared to previous studies which presented the reference during the reproduction phase (Jazayeri & Movshon, 2007;Luu & Stocker, 2018;Zamboni et al, 2016), the current results are in line with the observation by Zamboni et al that the reference repulsion bias is greatly reduced when removing the reference during reproduction. These findings beg the question whether the smaller bias, occurring when the reference was absent during reproduction, reflects a perceptual bias, which would exist alongside a stronger post-perceptual bias that is dependent on the presentation of the reference during the reproduction phase.…”

“…This stands in stark contrast to previous findings of reference repulsion, where the strongest bias for stimuli oriented was observed close to the discrimination boundary (Jazayeri & Movshon 2007;Luu & Stocker, 2018; see also Experiment 1 in this study).…”

“…In order to shed light on this issue, we first examined whether reference repulsion in reproduction responses would occur even when the discrimination boundary was presented only before the presentation of the actual stimulus and crucially not during the reproduction phase. Focusing on orientation perception, for which similar biases as in motion perception have been reported (Luu & Stocker, 2018), we indeed found a robust reference repulsion bias.…”

“…We simulated the observer with the generating parameters listed in Table T2. The observer was designed to have a small reference repulsion bias that deceased with increasing orientation difference between probe stimulus and discrimination boundary, in accordance with previous studies (Jazayeri & Movshon, 2007;Luu & Stocker, 2018; Experiment 1 of current study). In particular, the reference repulsion bias was computed on a trial-by-trial basis, based on the difference between the sampled boundary and probe stimulus orientation, ∆ !"…”

“…Crucially, this theory assumes that reference repulsion occurs as a consequence of being engaged in the discrimination judgment task and arises during the process of perceptual processing, therefore resulting in a perceptual illusion. An alternative theory proposed that the repulsion could arise from Bayesian inference with the initial discrimination judgment constraining the prior over motion directions when forming a perceptual decision about the overall motion direction (Stocker & Simoncelli, 2008;Luu & Stocker, 2018). Under these theories it is not clear whether reference repulsion would occur at a perceptual (Stocker & Simoncelli, 2008) or post-perceptual stage (Luu & Stocker, 2018).…”