Towards a deeper understanding of impression formation—New insights gained from a cognitive-ecological perspective.

Abstract

Impression formation is a basic module of fundamental research in social cognition, with broad implications for applied research on interpersonal relations, social attitudes, employee selection, and person judgments in legal and political context. Drawing on a pool of 28 predominantly positive traits used in Solomon Asch's (1946) seminal impression studies, two research teams have investigated the impact of the number of person traits sampled randomly from the pool on the evaluative impression of the target person. Whereas Norton, Frost, and Ariely (2007) found a "less-is-more" effect, reflecting less positive impressions with increasing sample size , Ullrich, Krueger, Brod, and Groschupf (2013) concluded that an-independent averaging rule can account for the data patterns obtained in both labs. We address this issue by disentangling different influences of on resulting impressions, namely varying baserates of positive and negative traits, different sampling procedures, and trait diagnosticity. Depending on specific task conditions, which can be derived on theoretical grounds, the strength of resulting impressions (in the direction of the more prevalent valence) (a) increases with increasing for diagnostic traits, (b) is independent of for nondiagnostic traits, or (c) decreases with when self-truncated sampling produces a distinct primacy effect. This refined pattern, which holds for the great majority of individual participants, illustrates the importance of strong theorizing in cumulative science (Fiedler, 2017) built on established empirical laws and logically sound theorizing. (PsycINFO Database Record

Learning too much from too little: False face stereotypes emerge from a few exemplars and persist via insufficient sampling

“…The novel manipulation of granting the power of seeing more or fewer samples to participants enabled us to identify the underlying mechanism. Consistent with past Running head: FACE STEREOTYPES FROM INSUFFICIENT STATISTICAL LEARNING work (e.g., Klein and O'Brien, 2018;Prager, Krueger, & Fiedler, 2018), we found that participant-generated samples are much smaller than the samples needed to gain an accurate picture of the data. As a result, such insufficient sampling makes spurious beliefs sticky, and resistant to updating even when faced with the complete environmental data.…”

“…Similarly, when judging a person's intellect or character, people thought they would choose to view five pieces of information, but only chose to view three pieces. Likewise, recent impression formation research has shown that participantdetermined samples are substantially smaller than experimenter-determined samples (Prager, Krueger, & Fiedler, 2018). When forming impressions, if the first few traits happened to convey a clear-cut impression (e.g., Bob is friendly, nice, trustworthy, and helpful), then people cut the sample short.…”

Quo Vadis, Methodology? The Key Role of Manipulation Checks for Validity Control and Quality of Science

“…Almost ideal MC conditions are implemented in so-called sampling approaches to judgment and decision research ( Fiedler & Kutzner, 2015 ). Because the results of information search and environmental sampling processes are assessed independently of the final judgment and decision phenomena to be explained, the experimental task renders the hypothesized IV (i.e., sampling error and bias) amenable to direct empirical assessment, independently of the DVs measuring judgments and decisions ( Denrell & Le Mens, 2012 ; Prager et al, 2018 ). It may be no coincidence that sampling theories are so illuminating and satisfactory and that the resulting empirical findings are so robust.…”

Prediction and Explanation in a Postmodern World

“…Much progress has been made in theory development and theory testing, although it is not always clear what is meant by "theory." Rigorous hypothetico-deductive processes are not the rule in experimental design (Prager et al, 2018), and many researchers settle for vague verbal descriptions of phenomena (Gigerenzer, 1998(Gigerenzer, , 2010.…”