Sensitivity of implicit evaluations to accurate and erroneous propositional inferences

Abstract

is a member of the Scientific Advisory Committee of Project Implicit, a 501(c)(3) non-profit organization and international collaborative of researchers who are interested in implicit social cognition.

How do explicit and implicit evaluations shift? A preregistered meta-analysis of the effects of co-occurrence and relational information.

“…Effect sizes were not included in the meta-analytic database if (a) two types of co-occurrence information but no relational information were presented (certain conditions of Hu et al, 2017b); (b) individual difference measures were used to assign participants to groups (Kurdi & Dunham, 2021); (c) a manipulation involved different combinations of co-occurrence and relational information such that the unique contribution of each could not be determined, for example, the same target was paired both with positive behaviors revealed to be characteristic (A) and negative behaviors revealed to be uncharacteristic (D; Bading et al, 2020; Brannon & Gawronski, 2017; Calanchini et al, 2013; Kurdi & Dunham, 2021; Rydell et al, 2007; Rydell & McConnell, 2006); (d) the same task was used for learning and testing (Kawakami et al, 2000); or (e) the study investigated generalization to a trait (Förderer & Unkelbach, 2016) or target (Hughes, Barnes-Holmes, et al, 2018) that was not included in the learning phase.…”

“…The same reviewer also noted that statements (such as "Bob continually yells at his wife in public") also have relational content and should therefore be removed from the meta-analysis as sources of co-occurrence information. We decided to retain these cases for the following reasons: (a) designs involving behavioral statements are procedurally highly similar to evaluative conditioning designs, the only difference being spatial separation between the CS and US; (b) behavioral statements can shift implicit evaluations by virtue of their co-occurrence structure only (Kurdi & Dunham, 2021); (c) primary authors (e.g., Peters & Gawronski, 2011) often explicitly treat behavioral statements solely as a source of co-occurrence information; and (d) we did not find any difference between behavioral statements and single words as a source of co-occurrence information in moderator analyses ( p = .225), suggesting that the minimal relational content of the former is unlikely to be responsible for the effects obtained. In addition to the name of the variable and the corresponding variable name in the meta-analytic database, the table also provides definitions and information on interrater reliabilities (calculated via Gwet's AC 1 ) and the levels of categorical variables.…”

The fragility of implicit evaluation updating: The role of cognitive and ecological constraints

“…Contrary to these influential early conceptualizations, empirical research since the 1990s has provided ample support for the idea that implicit social evaluations can exhibit flexibility in response to relevant information (Blair, 2002;Cone et al, 2017;De Houwer et al, 2020;Forscher et al, 2019;Kurdi & Dunham, 2020;Lai et al, 2014). Specifically, implicit evaluations of both individual social targets and social categories have been shown to shift as a result of a wide variety of manipulations, including approach/avoidance training (Van Dessel et al, 2015, 2020, behavioral statements (Boucher & Rydell, 2012;McConnell et al, 2008), evaluative conditioning Rydell & Jones, 2009), propositional reasoning (Kurdi & Dunham, 2021;Moran et al, 2015), reinforcement learning (Hackel et al, 2021;, and many others. Given the amount of evidence available at this time, the momentary malleability of implicit social evaluations seems beyond reasonable doubt.…”

Advanced testing of the LoT hypothesis by social reasoning

“…First, social situations of logical reasoning are highly complicated by experiential social learning and self-referential projection. For illustration let us turn to an example: Quilty-Dunn et al iterate an experiment by Kurdi and Dunham (2021) in which participants were presented with, among other statements, the following simple logical statement: “If you see a green circle, you can conclude that Ibbonif is malicious” (target article, sect. 6.2, para.…”

“…First, social situations of logical reasoning are highly complicated by experiential social learning and self-referential projection. For illustration let us turn to an example: Quilty-Dunn et al (2022, p. 36) iterate an experiment by Kurdi & Dunham (2021) in which participants were presented with, among other statements, the following simple logical statement: "If you see a green circle, you can conclude that Ibbonif is malicious." Adapting this straightforward statement to a context of social inference -for instance: "If you see a smirk on the face, you can conclude that Peter is malicious," -can quickly ascend individuals into a rabbit hole of applying their (1) own social learning and (2) induction from introspection about their self.…”