|
Article Excerpt
This study proposed and tested a developmental model of impression formation based on observed behavior, prior expectancies, and additional incongruent information. Participants were 51 kindergartners, 53 second graders, and 104 college students who provided trait and liking judgments after watching a child actor engage in behaviors from three behavioral dimensions. Reaction times of judgments were also measured. Results showed that all age groups made dispositional attributions for all three presented traits. Nevertheless, young children had some difficulty conceiving of others as possessing both positive and negative attributes and provided less negative judgments. Prior expectancy strengthened judgments of the expected trait but did not suppress encoding of other traits. Judgments of likability were most strongly influenced by prosocial-antisocial behavior. Unlike prosocial and smart behavior, shyness influenced only adults' liking of the actor. Additional incongruent behaviors produced a substantial modification of impressions in all groups except for kindergartners who held positive impressions.
**********
The processes involved in causal attributions and mental representations of others are at the heart of interpersonal perception. The impressions we form of others guide our expectations, perceptions, and interpretations of others' behavior as well as our cognitive, affective, and behavioral responses to others. Not surprisingly, trait attributions and corresponding memory structures have been explicitly incorporated into leading developmental theories of social cognitions (e.g., Crick & Dodge, 1994; Yeates & Selman, 1989). Despite their utilization of these processes, current developmental theories do not specify how trait attributions are made, how information about others is represented in memory, and how these processes change with development. Although a separate line of developmental research has addressed the question of developmental changes in trait attributions (e.g., Dozier, 1991; Heyman & Gelman, 2000; Peevers & Secord, 1973), these investigations typically have been restricted to inferences about a single behavior. While certainly valuable, this approach does not provide insights into attributional processes under more complex conditions that involve prior expectancies and potentially conflicting information about multiple traits. In addition, virtually no studies specified the structure of mental representations of others in children.
This study aims to complement current developmental theories of social cognitions by providing a theoretical framework for the study of complex impressions and corresponding mental representations of others in children. This model will specify how children process multiple pieces of behavioral information that they encounter under the presence and absence of prior expectations, how they combine this information to derive a global judgment of liking, and how additional information changes existing impressions. We base our framework on a comprehensive and widely accepted model of impression formation from the adult social psychology literature (Srull & Wyer, 1989) while adding developmentally sensitive modifications derived from two mechanisms of cognitive development: processing speed and working memory (Kail, 2003). We will derive testable hypotheses from the proposed model, briefly review relevant empirical literature, and carry out empirical evaluations of the hypotheses.
Overview of the Aduh-Based Model of Impression Formation
In 1989, Srull and Wyer published a highly influential review of processes involved in person memory and judgment. In this paper, they integrated a number of theoretical perspectives and a wealth of empirical research into a coherent framework that specified how information about others is organized in memory and how these representations translate into social judgments. This review integrated and explained a variety of empirical findings in the field of person perception and has been cited by nearly 200 articles since. Because of its comprehensive nature and widespread acceptance in the field of adult person perception (see Gilbert, 1998), we will utilize this model in the current investigation.
Srull and Wyer (1989) postulate that individual representations of a target person are organized in memory in the form of associative networks that include both descriptive and evaluative aspects. In addition to encoding specific behaviors of other people, these behaviors are interpreted and encoded in terms of traits. However, trait inferences are influenced by prior expectations so that when certain traits are expected in a person, only behaviors that exemplify these concepts will be encoded in terms of traits. Behaviors with implications for other attributes will not be encoded in trait terms because all behaviors are interpreted according to the preexisting trait knowledge that does not involve these other attributes. For instance, if we expect a specific person to be smart and then observe her or his enacting behaviors that indicate both smartness and honesty, we would only infer and encode the trait of smartness but not the trait of honesty because of the influence of our prior expectations. On the other hand, if we did not have any expectations, we would infer and encode both smartness and honesty.
Besides encoding specific behaviors and traits, people attempt to form a coherent evaluative impression (i.e., liking) of the target person during the impression-formation process. This evaluative concept is based on early information that has clear and consistent evaluative implications and is resistant to modification based on later information, mainly because it influences the interpretation of additional behaviors. For instance, if a disliked person performs a positive behavior, we may discount this behavior in order to maintain our negative evaluation. In contrast to the evaluative concept, descriptive trait judgments should be influenced by early as well as later information. Srull and Wyer (1989) predict that when making judgments, people will rely on their representations of traits and the evaluative concept rather than the specific behaviors that originally served as a basis for the inference of those traits and evaluations.
Developmental Modifications of the Model
In children, two general mechanisms of cognitive development--processing speed and working memory--limit the amount of information that can be effectively processed. We hypothesize that while young children encode small amounts of information about others in a manner similar to adults, they will have difficulties processing more complex information due to their limited cognitive capacity. When confronted with too much information or with stimuli of greater complexity, young children may simplify the input by focusing on parts that are most salient for them. These more salient perceptions will then bias their impressions of less-salient information. We argue that the salience of different information will be closely tied to personal meaning. In other words, behaviors and traits that have the greatest potential for personal impact in a given situation will be considered most salient. By this definition, the salience of various behaviors and traits will depend on a particular situation. For instance, when choosing players for one's basketball team, athletic ability will be highly salient. By contrast, intellectual ability will gain in saliency when selecting a peer for an academic task. In general social situations, however, the prosocial-antisocial dimension should be the most salient, thanks to the highly emotional consequences these behaviors produce in both the enactor and the recipient (Arsenio & Ford, 1985; Arsenio & Kramer, 1992). In fact, research confirms that prosocial and antisocial behaviors are more important for young children, while other traits (e.g., shyness or intellectual ability) become more important and cohesive later in childhood (Benenson & Dweck, 1986; Bukowski, 1990; Younger, Schwartzman, & Ledingham, 1985).
Testable Hypotheses Derived from the Model
1. If no expectations are present, even young children will form dispositional attributions of others based on relatively simple observed behaviors. Srull and Wyer (1989) postulate that in the absence of prior information, all behaviors are encoded in terms of corresponding traits. A number of developmental studies have demonstrated that children as young as 5-6 years old make dispositional attributions when presented with vignettes portraying a single trait (e.g., Alvarez, Ruble, & Bolger, 2001; Cain, Heyman, & Walker, 1997; Dozier, 1991; Fergusson, van Rozendaal, & Rule, 1986; Heller & Berndt, 1981; Smetana, 1985, Yuill & Pearson, 1998). In these studies, young children demonstrated their abilities to generate correct trait labels, to correctly rate degrees to which the traits are present, and to predict behavior consistent with the inferred traits over time and across different situations.
2. If expectations are present, only the expected traits will be inferred and encoded in both children and adults. Any other behaviors will not be encoded in terms of traits. This effect of expectancies follows from the influence of prior information on the interpretation and encoding of new stimuli (Srull & Wyer, 1989). Although no study examined the effect of expectancies on encoding of unrelated traits in children, a few studies suggest that prior expectations influence children's impressions and behavior. For instance, when children expected to interact with peers labeled as not smart, they rated their partners more negatively than if they believed their partners to be smart (McAninch et al., 1996). In another study, manipulation of shy versus outgoing expectations caused better liking of the outgoing actor (McAninch, Manolis, Milich, & Harris, 1993).
3. When no expectancies are present and information about several traits is observed, young children are less likely to encode the traits that are less salient to them than are older children and adults. For adults, Srull and Wyer's (1989) model predicts that in the absence of expectancies, all observed behaviors will be encoded together with their corresponding traits. However, given developmental changes in processing speed and working memory capacity, we expect that young children will be more likely to encode traits related to the prosocial-antisocial dimension than less-salient types of traits (Alvarez et al., 2001).
4. A global evaluation will be derived from the combined valence of all encoded traits. According to Srull and Wyer (1989), a general evaluative concept is formed early in the impression-formation process by combining the valence information of all encoded traits that have evaluative implications. Two studies with children indicate that when presented with two behaviors or personality traits, children as young as 5 years of age use the valences of both of these behaviors to arrive at a final evaluative impression (Dozier, 1991; Hendrick, Franz, & Hoving, 1975). In the absence of empirical research, we speculate that children will use information about all presented traits to produce a global evaluation.
5. Later information will affect trait judgments more than evaluative judgments. When...
|
