Restoring sight in choice blindness: pupillometry and behavioral evidence of covert detection

Pablo Rodrigo Grassi^1,2,3*Lena Hoeppe⁴Emre Baytimur^4,5Andreas Bartels^1,2,3

• ¹Department of Psychology, Faculty of Mathematics and Natural Sciences, University of Tübingen, Tübingen, Baden-Württemberg, Germany
• ²Centre for Integrative Neuroscience, Tübingen, Germany
• ³Department for High-Field Magnetic Resonance, Max Planck Institute for Biological Cybernetics, Tübingen, Baden-Württemberg, Germany
• ⁴Graduate Training Centre of Neuroscience, International Max Planck Research School, University of Tübingen, Tübingen, Baden-Württemberg, Germany
• ⁵Zurich Center for Neuroeconomics, Department of Economics, University of Zurich, Zurich, Switzerland

Intriguing results from “choice blindness” (CB) experiments have shown that when people make choices, but are presented with a false outcome, many seem not to notice the mismatch and even provide reasons for choices they never made. They appear to be “blind” about their intentions. Yet, this effect goes against decision-making accounts and experience, in which we regularly notice outcomes that do not match our choices (e.g., when ordering food). Here, we ask whether participants really fail to detect the manipulation, or whether CB can be accounted for by covert detection, in that participants detect changes, but do not report them. To test this, we measured pupil dilation during the experiments to quantify objective responses in addition to reports by participants. In both experiments, we consistently observed that participants failed to report detected mismatches. Moreover, we observed increased pupil dilation during all manipulated trials, irrespective of whether they were reported or not. Thus, we provide conclusive evidence of covert detection in CB. In addition, we show that CB is strongly modulated by the idiosyncrasies of the experimental design. Our results cast doubt on the general validity of CB, and with that on key conclusions of previous studies. Instead, our results suggest no failure of detection, but instead higher-level, cognitively or socially driven hesitance of reporting. Our evidence leads us to a cautious discussion of CB and provides an account that no longer violates our intuitions about human intentionality and rationality, in that participants are less introspectively blind than originally portrayed.