He did what? The role of diagnosticity in revising implicit evaluations.

Abstract

Research suggests that implicit evaluations are relatively insensitive to single instances of new, countervailing information that contradicts prior learning. In six experiments, however, we identify the critical role of the perceived diagnosticity of that new information: counter-attitudinal information that is deemed highly diagnostic of the target's true nature leads to a complete reversal of the previous implicit evaluation. Experiments 1a and 1b establish this effect by showing that newly-formed implicit evaluations are reversed minutes later with exposure to a single piece of highly diagnostic information. Experiment 2 demonstrates a valence asymmetry in participants’ likelihood of exhibiting rapid reversals of newly formed positive versus negative implicit evaluations. Experiment 3 provides evidence that a target must be personally responsible for the counter-attitudinal behavior and not merely incidentally associated with a negative act. Experiment 4 shows that participants exhibit revision only when they judge the target's counter-attitudinal behavior as offensive and thus diagnostic of his character. Experiment 5 demonstrates the behavioral implications of newly-revised implicit evaluations. These studies show that newly-formed implicit evaluations can be completely overturned through deliberative considerations about a single piece of counter-attitudinal information.

Mechanisms underlying approach-avoidance instruction effects on implicit evaluation: Results of a preregistered adversarial collaboration

“…8 Overall, the current findings support the idea that propositional processes play an important role in AA instruction effects on implicit evaluation. This conclusion is consistent with the growing body of evidence showing that (a) verbal instructions can have strong, immediate effects on implicit evaluations (Castelli et al, 2004;Gregg et al, 2006;Whitfield & Jordan, 2009) and (b) instruction-based changes in implicit evaluation depend on the operation of propositional processes (Cone & Ferguson, 2015;Peters & Gawronski, 2011b;Zanon et al, 2014). The current results extend these findings by showing that propositional processes also play a major role in AA instruction effects.…”

“…According to the APE model, any information that allows participants to entertain the proposition that a stimulus is positive or negative may instigate the proactive construction of new evaluative associations, which in turn may influence implicit evaluations. In line with this idea, changes in implicit evaluations have been observed when participants are provided with verbal information about the evaluative properties of a stimulus (Castelli, Zogmaister, Smith & Arcuri, 2004;Cone & Ferguson, 2015;Gawronski, Walther, & Blank, 2005;Gregg, Seibt, & Banaji, 2006). Importantly, these models predict a specific pattern of mediation such that instruction effects on explicit evaluation should mediate effects on implicit evaluation (e.g., Gawronski & Walther, 2008;Peters & Gawronski, 2011a;Whitfield & Jordan, 2009;see Gawronski & Bodenhausen, 2006;Case 4).…”

The base rate principle and the fairness principle in social judgment

“…Much of cognition occurs unconsciously (32) and the topic of changing implicit responses has gained traction (18)(19)(20)(21)(22)(23)(24)27). As such, it will be crucial to understand the reaches of how these implicit responses might change, and Bayesian models may be highly useful.…”

“…However, other research has identified conditions under which implicit associations appear highly amenable to new information (20)(21)(22)(23)(24).…”

Reversing implicit first impressions through reinterpretation after a two-day delay

“…In the context of impression change, these ideas suggest that revision of implicit impressions may be strongest when new information is subjectively assessed as more diagnostic or important (though even propositional information that is seemingly low in diagnosticity may sometimes drive change; Van Dessel et al, 2016). In support of this possibility, Cone and Ferguson (2015) asked participants to read 100 behavioral statements about a novel person that indicated his overall goodness, and formed implicit positive impressions of him. After then learning an additional behavior that was extremely negative (e.g., that he had molested children), participants showed robust revision, switching to a strongly negative implicit impression of him.…”