Resumen de Learning to individuate versus categorize people: the role of attention in social learning
Faces are essential stimuli in social perception, as they comprise a wealth of information regarding a person’s individual identity, social category, emotional state, gaze, and so on (Hugenberg & Wilson, 2013). This information is readily encoded at early stages of face processing, and further integrated to make sense of others. The information extracted from faces substantially influences impression formation processes, especially at zero acquaintance, when nothing is known about the target. However, all faces are not equally attended and the strategies used to form impressions of others lie on a continuum stretching from social categorization to individuation (Fiske & Neuberg, 1990).
Social categorization consists of using information diagnostic of a person’s social category to make inferences about him or her. For instance, skin color, lips shape or hair texture are informative of a person’s race. Once these features have been encoded, the target person is classified in a social category and the knowledge associated with this category, based for instance on ethnic stereotypes, is used to make inferences about this particular individual. This strategy allows perceivers to integrate and organize efficiently the diversity of information that may be extracted from social stimuli. Because of its remarkable resources-saving function, social categorization is the default strategy to make sense of others (Macrae & Bodenhausen, 2000). Alternatively, perceivers may further their level of analysis beyond social categories by using the idiosyncratic attributes of a person to predict his or her behaviors. However, as compared to social categorization processes, individuation requires increased attention to integrate the variety of individual attributes contributing to the person’s uniqueness (Gilbert & Hixon, 1991). Therefore, perceivers do not engage in individuation processes unless they are both cognitively able and motivated to acquire an individuated judgment about a target person (Pendry, 1998). Such motivation may arise from different situational and personal variables. For instance, if a perceiver is outcome-dependent on understanding a target (e.g., the perceiver is economically rewarded according to the accuracy of his or her predictions), he or she will likely engage in individuation processes (Neuberg & Fiske, 1987). In that sense, perceivers are often perceived as “motivated tacticians” who possess different cognitive strategies from which they chose the one that best fits their interests (Fiske & Taylor, 1991). Cognitive efficiency is therefore achieved by means of a trade-off between cognitive cost and goals attainment.
Beyond the cognitive economy, social categorization is associated with a series of attentional and behavioral biases with important consequences on social judgments. For instance, categorizing a target as an ingroup member (i.e., a person who belongs to the same social category as oneself) may promote a more positive attitude towards this person than toward an outgroup member, even at zero acquaintance. This effect, known as ingroup favoritism, may be related to trivial misattributions of traits, but also to more serious intergroup discriminations (Brewer, 1979, 2001). Another consequence of social categorization is reflected in the outgroup homogeneity effect, that is, the perception of higher intragroup similarity among outgroup than among ingroup members. Such biased perception may be related to impaired learning about outgroup members’ traits (Park & Rothbart, 1982). Broadly speaking, both effects reflect how overgeneralizations based on social group membership may negatively impact social relationships. Understanding the specific circumstances promoting individuation or social categorization strategies in social perception is crucial to reduce the negative implications of category-based judgments, including stereotypes and prejudice.
Considering this framework, the aim of the present dissertation was to evaluate the impact of perceivers’ motivational, emotional and cognitive states on the differential use of social categorization and individuation strategies, and their consequences on social judgments regarding the principal dimensions of impression formation (i.e., gender, ethnicity, age and emotional expression). To achieve this general goal, we developed an adaptation of the trust game paradigm (Berg, Dickhaut, & McCabe, 1995) allowing us to achieve the following specific goals: 1. To explore impression formation at zero acquaintance in relation to motivational, cognitive and emotional factors.
2. To analyze how individuation and categorization strategies contribute to update first impressions across repeated interactions in a learning process.
3. To examine the impact of learned associations between specific social categories and group behaviors on the perception of new individuals from these social categories.
With this adaptation of the trust game, we conducted seven experimental series distributed in five chapters in which different cognitive, emotional and motivational factors were manipulated. In Experimental Series 1 and 2, we explored impression formation processes according to the main dimensions of social categorization, namely gender, ethnicity and age. In Experimental Series 3, we examined the potential effect of social power on social perception. In Experimental Series 4 and 5, we directly manipulated motivation and cognitive resources to understand their impact on social judgments. Finally, in Experimental Series 6 and 7, we analyzed the relationship between emotions and social perception.
In our adaptation of the trust game, participants had to predict the cooperative behaviors of unfamiliar partners to behave accordingly, cooperating with equitable partners, and not cooperating with non-equitable partners. Moreover, accurate predictions in the task were rewarded with real monetary outcomes. The task was divided in three phases with the following specific manipulations: First, in the baseline, participants played with partners from two social categories, all partners being equitable in half of the trials and non-equitable in the other half. This phase allowed us to verify whether participants were spontaneously biased to cooperate more with one of the two social groups.
Second, in the learning phase, we manipulated the group behavior by associating the two social categories with opposite cooperative behaviors, one being equitable and the other one being non-equitable. Crucially, within each social category, a small proportion of individuals were inconsistent regarding their group, displaying a pattern of cooperative behavior opposite to the group behavior. This phase allowed us to explore learning across repeated interactions and more importantly, to verify whether participants were impacted by their partners’ category or individual behavior. In this regard, cooperation decisions with inconsistent partners were informative of the strategy used. A categorization strategy would be reflected in a similar pattern of cooperation with consistent and inconsistent individual within the same social group. Conversely, an individuation strategy would be reflected in opposite patterns of cooperation with consistent and inconsistent individuals within the same social group.
Finally, in the transfer phase, participants were presented with new individuals from the two social categories manipulated, but with whom they lacked individual experience, all of them being equitable in half of the trials. This final phase allowed us to verify whether the associations established in the learning phase between group behavior and social categories would be applied to new individuals in a categorical way.
Across seven experimental series, we never observed a complete categorization pattern with human partners, as participants did not apply the category-related knowledge to make predictions about inconsistent individuals. This finding suggests that the outcome-dependency intrinsic to the game settings may have motivated participants to predict their partners’ behaviors as accurately and as individually as possible. Moreover, and importantly, the few hints of categorical thinking found in the present dissertation are better accounted by motivational than by cognitive factors. Simply stated, participants rely on categorical information only when their motivation to individuate was reduced, but never because of the cognitive cost of individuation processes, as attempts to individuate people were observed even in situations where individuation was impossible. Altogether, these results challenge the trade-off approach of impression formation theories. Moreover, and importantly, they suggest that with the adequate motivation, perceivers may fully explore their cognitive capacity to individuate, offering an encouraging perspective on motivation-oriented interventions to reduce over-categorization and its negative consequences.
