People confabulate with high confidence when their decisions are supported by weak internal variables

People can introspect on their internal state and report the reasons driving their decisions but choice blindness (CB) experiments suggest that this ability can sometimes be a retrospective illusion. Indeed, when presented with deceptive cues, people justify choices they did not make in the first place, suggesting that external cues largely contribute to introspective processes. Yet, it remains unclear what are the respective contributions of external cues and internal decision variables in forming introspective report. Here, using a brain–computer interface, we show that internal variables continue to be monitored but are less impactful than deceptive external cues during CB episodes. Moreover, we show that deceptive cues overturn the classical relationship between confidence and accuracy: introspective failures are associated with higher confidence than genuine introspective reports. We tracked back the origin of these overconfident confabulations by revealing their prominence when internal decision evidence is weak and variable. Thus, introspection is neither a direct reading of internal variables nor a mere retrospective illusion, but rather reflects the integration of internal decision evidence and external cues, with CB being a special instance where internal evidence is inconsistent.

Choice blindness and choice-induced preference change in groups

Contrary to common belief, our preferences do not only shape our decisions but are also shaped by what decisions we make. This effect, known as choice-induced preference change, has been extensively studied in individuals. Here we document choice-induced preference change in groups. We do so by using the choice-blindness paradigm, a method by which participants are given false feedback about their past choices. Participants are given a second round of choices following the choice blindness manipulation´measuring preference change resulting from accepting the manipulation. In Experiment 1 (N=83), we introduce a roommate selection task used in this paper and use it to replicate choice-induced preference change using choice-blindness in individuals. In Experiment 2 (N=160), dyad members made mutual choices in the roommate selection task and then receive either veridical or false feedback about what choice they made. The majority of the false feedback trials were accepted by the dyads as their own choices, thereby demonstrating choice blindness in dyads for the first time. Dyads exhibited choice-induced preference change and were more likely to choose the originally rejected option on trials where they accepted the manipulation compared to control trials. In Experiment 3 (N=80), we show that the preference effect induced by the choice blindness manipulation at the group level does not generalize back to follow up choices made by individual participants when removed from the group. In all studies, response time analyses further support our conclusions. Choice-induced preference change exists for both individuals and groups, but the level at which the choice was made constrains the influence of that choice on later preferences.

The Magician’s Choice: Providing illusory choice and sense of agency with the Equivoque forcing technique.

Forcing techniques allow magicians to subtly influence spectators’ choices and the outcome of their actions, and they provide powerful tools to study decision-making and the illusory sense of agency and freedom over choices we make. We investigate the Equivoque force, a technique that exploits semantic ambiguities and people’s failure to notice inconsistencies, to ensure that a spectator ends up with a pre-determined outcome. Similarly to choice blindness paradigms, the Equivoque forces participants to end up with an item they did not choose in the first place. However, here, the subterfuge is accomplished in full view. In three experiments, we showed that the Equivoque is highly effective in providing participants an illusory sense of agency over the outcome of their actions, even after 2 repetitions of the trick (experiment 2), and using items for which pre-existing preferences can be present (experiment 3). Across all experiments, participants were oblivious to inconsistencies in the procedure used to guide their decisions, and they were genuinely surprised by the experimenter’s matching prediction. Contrary to our prediction, the Equivoque force did not significantly change participants’ preference for the chosen item. We discuss the results with regards to other illusions of agency (e.g. forcing, choice blindness), failures in noticing semantic inconsistencies (e.g Moses illusion), and issues surrounding choice-induced-preference literature.

Choice awareness and manipulation blindness: A cognitive semiotic exploration of choice-making

Within cognitive science, “blindness” to choice is commonly treated as typical of human cognition, implying unreliable agents who essentially lack any awareness of their own choices (e.g. Johansson et al., 2005, 2008; Hall et al., 2010, 2013). Within cognitive semiotics, however, choice awareness is seen as a continuous phenomenon, which is susceptible to the influence of a variety of factors. Manipulation blindness is proposed as a more adequate term for what is known in the literature as “choice blindness”, referring to participants’ tendency to accept a choice as if it were their own. This suggests that “blindness” is strictly limited to the level of detection (of the switch of the preferred choice to a non-chosen one), and not to the level of choice. Using a cognitive-semiotic framework, I examine manipulation blindness as an “indicator” of choice awareness by employing the factors of memory, consequence, and affectivity, and introduce a two-level hierarchy of choice-making. 43 participants were assigned two tasks combining choices with a) two degrees of consequence (more/less) – based on task instructions, and b) two degrees of affectivity (high/low) – based on stimuli with different degrees of abstractness. Participants were first asked to state their preference for one of two alternatives (choice) . After that they were shown chosen as well as non-chosen pictures and asked to confirm whether the picture presented was the one of their choice (memory). Lastly, they were asked to justify their choice, although some of the trials had been manipulated (i.e. the chosen card was switched with the non-chosen one) (manipulation) . Half of the manipulations were detected, and 75% of these detections occurred for the choices participants remembered correctly. While the consequential impact of the choice did not seem to influence detection, affectivity did. Unlike other experiments that investigate “choice blindness”, the results indicate that manipulation blindness is subject to memory and affectivity, suggesting that we are aware of our choices and that we have, to various degrees, access to our intentional acts.

Is choice blindness a case of self-ignorance?

When subject to the choice-blindness effect, an agent gives reasons for making choice B, moments after making the alternative choice A. Choice blindness has been studied in a variety of contexts, from consumer choice and aesthetic judgement to moral and political attitudes. The pervasiveness and robustness of the effect is regarded as powerful evidence of self-ignorance. Here we compare two interpretations of choice blindness. On the choice error interpretation, when the agent gives reasons she is in fact wrong about what her choice is. On the choice change interpretation, when the agent gives reasons she is right about what her choice is, but she does not realise that her choice has changed. In this paper, we spell out the implications of the two interpretations of the choice-blindness effect for self-ignorance claims and offer some reasons to prefer choice change to choice error.