Failure to Detect Mismatches Between Intention and Outcome in a Simple Decision Task

Abstract

A fundamental assumption of theories of decision-making is that we detect mismatches between intention and outcome, adjust our behavior in the face of error, and adapt to changing circumstances. Is this always the case? We investigated the relation between intention, choice, and introspection. Participants made choices between presented face pairs on the basis of attractiveness, while we covertly manipulated the relationship between choice and outcome that they experienced. Participants failed to notice conspicuous mismatches between their intended choice and the outcome they were presented with, while nevertheless offering introspectively derived reasons for why they chose the way they did. We call this effect choice blindness.

Human–Computer Interaction in Face Matching

“…Paradoxically, however, the reported experiments indicate that the commission of errors by algorithms facilitates errors in humans, given that observers were more likely to accept a mismatch, and reject an identity match, if these were labeled as depicting the same person or different individuals, respectively. This finding aligns with evidence that facial identification processes are guided by information from trustworthy sources, such as experimenters, even when inaccurate (see, e.g., Johansson, Hall, Sikström, & Olsson, 2005; Menon et al., 2015; Sagana, Sauerland, & Merckelbach, 2016; Sauerland et al., 2016). In addition, human operators are typically expected to monitor up to seven e‐Gates concurrently (FRONTEX, 2015a).…”

A Problem for Self-Knowledge: The Implications of Taking Confabulation Seriously

“…Furthermore, the dual-method theorist needs to explain why we regard confabulations with the same degree of certainty that we do our genuine introspective insights. The findings of Johansson et al (2005Johansson et al ( , 2006 which I discussed at the end of Section 2.4 indicate that if there are two methods, they deliver absolutely similar subsequent cognitive processes in all respects. This appears to be a serious anomaly for the dual-method theorist and another area where their account currently lacks a plausible explanation.…”

“…In both Johansson et al's (2005) original study and a reanalysis conducted a year later (2006), they were unable to find any significant differences in the justifications provided in the manipulated and the non-manipulated trials. They checked a large number of factors which might differentiate confabulated justifications in terms of manner of delivery or markers of attitude.…”

“…Fiala and Nichols (2009) themselves concede that further systematic study is needed to investigate how well feelings of certainty detect confabulations. I would presume that they were not aware of the work by Johansson et al (2005Johansson et al ( , 2006 in which they conducted a comprehensive analysis and found no evidence that participants were less confident about their confabulations. Amongst other things, they checked for response time, length of statement, word frequency checks for markers of certainty, unfilled pauses, laughter, the use of the past vs. present tense, the use of first vs. third person, emotional content and word length.…”

“…Readers may think that participants did notice the manipulation but did not report it to the experimenter due to social conformity pressures. In order to rule this out, Johansson et al (2005) included a question in the debrief about a hypothetical variation to the study which would use the manipulation which was, in fact, used in the study. Of the participants who failed to notice the manipulation, 84 % claimed that they would be able to spot this kind of manipulation.…”