Awareness of implicit attitudes: Large-scale investigations of mechanism and scope.

People can predict their scores on the Implicit Association Test with remarkable accuracy, challenging the traditional notion that implicit attitudes are inaccessible to introspection and suggesting that people might be aware of these attitudes. Yet, major open questions about the mechanism and scope of these predictions remain, making their implications unclear. Notably, people may be inferring their attitudes from externally observable cues (e.g., in the simplest case, their demographic information) rather than accessing them directly. This problem is exacerbated by the fact that, in past work, predictions have been obtained only for a small set of targets, attitudes toward which are demonstrably possible to infer. Here, in an adversarial collaboration with eight preregistered studies (N = 8,011), we interrogate implicit attitude awareness using more stringent tests. We demonstrate that people can predict their implicit attitudes (a) across a broad range of targets (many of which are plausibly difficult to infer without introspection), (b) far more accurately than third-party observers can based on demographic information, and (c) with similar accuracy across two different widely used implicit measures. On the other hand, predictive accuracy (a) varied widely across individuals and attitude targets and (b) was partially explained by inference over nonintrospective cues such as demographic variables and explicit attitudes; moreover, (c) explicit attitudes explained considerably larger portions of variance in predictions than implicit attitudes did. Taken together, these findings suggest that successful predictions of one's implicit attitudes may emerge from multiple mechanisms, including inference over nonintrospective cues and genuine introspective access. (PsycInfo Database Record (c) 2023 APA, all rights reserved).