The Architecture of Prototype Preferences: Typicality, Fluency, and Valence

Abstract

A classic phenomenon known as prototype preference effect (PPE) or beauty-in-averageness effect is that prototypical exemplars of a neutral category are preferred over atypical exemplars. This PPE has been explained in terms of deviance avoidance, hedonic fluency, or preference for certainty and familiarity. However, typicality also facilitates greater activation of category-related information. Thus, prototypes rather than atypical exemplars should be more associated with the valence of the category, either positive or negative. Hence, we hypothesize that the evaluation of a prototype depends on the valence of its category. Results from three experiments crossing a standard PPE paradigm with an evaluative conditioning procedure support our hypothesis. We show that for positive categories, greater typicality increases liking. Critically, for negative categories, greater typicality decreases liking. This pattern of results challenges dominant explanations of prototype evaluation. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Full text

BRIEF REPORT
The Architecture of Prototype Preferences: Typicality, Fluency,
and Valence
Tobias Vogel and Moritz Ingendahl
University of Mannheim
Piotr Winkielman
University of California, San Diego, and SWPS University of
Social Sciences and Humanities
A classic phenomenon known as prototype preference effect (PPE) or beauty-in-averageness effect is that
prototypical exemplars of a neutral category are preferred over atypical exemplars. This PPE has been
explained in terms of deviance avoidance, hedonic fluency, or preference for certainty and familiarity.
However, typicality also facilitates greater activation of category-related information. Thus, prototypes
rather than atypical exemplars should be more associated with the valence of the category, either positive
or negative. Hence, we hypothesize that the evaluation of a prototype depends on the valence of its
category. Results from three experiments crossing a standard PPE paradigm with an evaluative condi-
tioning procedure support our hypothesis. We show that for positive categories, greater typicality
increases liking. Critically, for negative categories, greater typicality decreases liking. This pattern of
results challenges dominant explanations of prototype evaluation.
Keywords: beauty-in-averageness, categorization, evaluative conditioning, fluency, valence
Supplemental materials: http://dx.doi.org/10.1037/xge0000798.supp
A classic phenomenon in psychology is the prototype prefer-
ence effect (PPE): preference for typical category exemplars over
atypical exemplars. Closely related is the beauty-in-averageness
effect: preference for the category average over its constituent
exemplars (Galton, 1879;Langlois & Roggman, 1990). The PPE
occurs with natural and artificial stimuli, across different popula-
tions, ages, and even species. As reviewed next, the phenomenon
has been explained by multiple influential frameworks across
psychology. Building on categorization theory, we present an
account in which the PPE crucially depends on the valence of
attributes associated with prototypes. This account generates a
novel empirical prediction—a reversal of the classic effect, with
lower preference for typical than atypical exemplars within nega-
tively valenced categories.
Prototype Evaluation: Major Explanations
PPE From Deviance Detection
Evolutionary accounts suggest that the PPE results from “devi-
ance detection”—negative sexual selection for atypical organisms
(Langlois & Roggman, 1990;Rhodes & Tremewan, 1996;Sy-
mons, 1979;Thornhill & Gangestad, 1993). However, the PPE
also occurs with objects (e.g., Halberstadt & Rhodes, 2003), in-
cluding consumer goods (e.g., Landwehr, Labroo, & Herrmann,
2011), and even abstract graphical patterns (e.g., Winkielman,
Halberstadt, Fazendeiro, & Catty, 2006). Accordingly, many
frameworks focus on basic cognitive and affective processes.
PPE From Preference for Certainty, Familiarity, or
Efficient Coding
Note that the aggregate similarity to all category exemplars is
highest for prototypes and they are most unambiguously identified
as category members (Markman & Ross, 2003). As such, proto-
types are the most certain, most “familiar” category members. The
PPE could then occur because uncertainty triggers negative affect
(Friston, Adams, & Montague, 2012;Hsu & Preuschoff, 2015;
Klein, Cosmides, Tooby, & Chance, 2002). Or because familiarity
triggers positive affect (Titchener, 1915). Related arguments em-
phasize that prototypes are statistically typical and thus more
efficiently coded (Dotsch, Hassin, & Todorov, 2016;Vogel, Carr,
Davis, & Winkielman, 2018). Thus, prototypes could be preferred
because they are simpler and less energy-demanding (Ryali & Yu,
2018).
This article was published Online First June 25, 2020.
XTobias Vogel and XMoritz Ingendahl, Department of Consumer
and Economic Psychology, University of Mannheim; Piotr Winkielman,
Department of Psychology, University of California, San Diego, and De-
partment of Psychology, SWPS University of Social Sciences and Human-
ities.
The authors want to thank Andy Arnold for helpful comments on the
article. Additional information is available at https://osf.io/qyf58/. Findings
from Experiment 1 have been presented at the Tagung experimentell
arbeitender Psychologen (TeaP) 2017 in Dresden, Germany.
Correspondence concerning this article should be addressed to Tobias
Vogel, Department of Consumer and Economic Psychology, University of
Mannheim, A5, 6 B117, 68159 Mannheim, Germany. E-mail: tobias
.vogel@uni-mannheim.de
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Journal of Experimental Psychology: General
© 2020 American Psychological Association 2021, Vol. 150, No. 1, 187–194
ISSN: 0096-3445 http://dx.doi.org/10.1037/xge0000798
187