Allen McConnell - US grants

Although early theorists (Cooley, 1902; Mead, 1934) speculated about how people draw inferences from the information supplied by others, the question of how self-relevant knowledge is combined to form one's self- concept has received little attention. Recent work on self-concept has mainly focused on the implications of already-developed self-relevant knowledge but not on how one forms conceptions about the self. The current work attempts to fill this gap in our knowledge, while exploring the implications of self-concept formation. The current work draws on the on-line vs. memory-based (Hastie & Park, 1986) processing distinction, which has been used to understand how people form impressions about individual and group social targets (McConnell, Sherman, & Hamilton, 1994b, in press). It is proposed that when people have resources and interest in processing self-relevant information in important behavioral domains, on-line self-concepts result. Conversely, in situations where people do not possess processing resources or interest in self-relevant abilities, or perceive the behavioral domain as relatively unimportant, memory-based self-concepts result. These studies (two preliminary, five proposed) test conditions under which memory-based vs. on-line self-concepts are produced. In addition to suggesting a process account for how self-concepts are formed, it is predicted that the way in which one forms a self-concept (memory-based vs. on-line) has meaningful implications. Those who form on-line (instead of memory-based) self- concepts will be especially influenced by early feedback, hold more stable and resilient self-concepts, and not be prone to certain biases (e.g., illusory correlations) in their self-concepts. Each study (preliminary and proposed) tests the consequences of how one forms self-concepts. Thus, the current work provides a process account for how self-concepts are formed. Moreover, it predicts that self-concepts will be formed differently based on certain antecedent conditions. These differences, in turn, will have important implications for the nature of the self-concept that is formed. Although this work posits that self-concept formation shares much in common with perception for other social targets (e.g., individuals, groups), it also predicts asymmetries for the self (relative to other social target types) based on unique motives that are associated with the self (e.g., self-enhancement). Because self-concepts have important implications (e.g., behavior, personal outlook, social comparison, and experience of emotion) for social functioning, an understanding of how self-concepts are formed, and the consequences of forming them, is important.

Although some theorists (e.g., Bem, 1967, Cooley, 1902; Mead, 1934) have speculated about how people draw inferences from their own behavior and the feedback of others, the question of how self-relevant knowledge is combined to form one's self-concept has received little attention. A great deal of work on self-concept has mainly focused on the implications of already-developed self-information processing mechanisms involved in forming self-concepts. The proposed research attempts to fill this gap in our knowledge, while exploring factors that affect how self-concepts are formed. The current work draws on the on-line vs. memory-based processing distinction (Hastie & Park, 1986), which has been used to understand how people form impression about individual and group social targets (McConnell, 1999; McConnell, Sherman, & Hamilton, 1994b, 1997). It is proposed that people's expectations about the self (e.g., implicit person theories, cultural beliefs about self-construal), self-focus (e.g., self- awareness, self-monitoring), and social interaction contexts (e.g., accountability, social comparison) influence the extent to which they process self-relevant information in an integrative fashion. As expectations about the self grow more consistent and diagnostic of one's "true essence", focus on the self increases, and social contexts encourage greater self-understanding , on-line self-concepts should result. These experiments (four preliminary, ten proposed) test conditions that determine whether memory-based or on-line self concepts are produced. In addition to suggesting a process account for self-concept formation, it is predicted that the way in which one forms a self-concept (memory- based versus on-line) will have meaningful implications. Those who form on-line, instead of memory-based, self-concepts will be especially influenced by early feedback, and they may hold more stable and resilient self-concepts, and store self-relevant information in a more abstracted form. Thus, the current work provides a process account for how self-concepts are formed. Moreover, it predicts that self-concepts will formed differently based on particular situational and individual difference factors. Although this work posits that self concept formation shares much in common with social perception for other types of targets (e.g., individuals, groups), it also examines issues unique to the self (e.g., cultural theories about independent versus interdependent self-construal, social comparison as a means for self-assessment). Because self-concepts have important implications (e.g., behavior, personal outlook, social comparison, and experience of emotion) for social functioning and mental health, an understanding of how self-concepts are formed, and the consequences of forming them, is important.

DESCRIPTION (provided by applicant): Although early theorizing about the self conceptualized it as a unitary structure (e.g., AIIport, 1955; Rogers, 1951), more recent treatments have proposed that it is multifaceted and context-dependent (Baumeister, 1998; Linville & Carlston, 1994). But how does the self operate in this manner? Are there qualitative differences among multiple selves and their attributes that affect mental and physical health? And most important, what processes explain the functions of multiple selves? This project uses a self-complexity framework to examine these issues. People are greater in self-complexity when they report having more self-aspects (e.g., roles, relationships) that are comprised of a greater degree of unique attributes. It has been proposed that those lower in self-complexity are more likely to reveal spillover effects. That is, self-relevant feedback has greater impact for those lower in self-complexity because the feedback implicates a relatively larger proportion of the overall self that is comprised of relatively more interrelated attributes. For instance, research has shown that those lower in self-complexity are more vulnerable to depression and stress-related illnesses when facing life stressors (Linville, 1987) and have difficulty in suppressing unwanted self-relevant thoughts (Renaud & McConnell, 2002). However, some findings have been inconsistent and even contradictory (e.g., Rafaeli-Mor & Steinberg, 2002; Woolfolk et al., 1995), suggesting that our understanding of self-complexity is far from complete. Critical assumptions about self-complexity (e.g., Linville, 1985) remain untested, and no comprehensive model speaks to the processes by which multiple selves form, function, and influence mental health and social behavior. These experiments (3 preliminary, 10 proposed) explore three lines of work to improve our understanding of how self-complexity relates to well-being. First, three studies assess the mechanisms underlying spillover effects. A second series of studies examine important qualities of self-aspects and their attributes to develop a more complete account of how self-complexity relates to depression, stress-related illnesses, self-esteem, and mood swings. Preliminary data indicate that perceptions of personal control moderate the relations between self-complexity and well-being, and additional work explores how the importance of multiple selves, self-certainty, and personality factors shape the relation between self-concept representation and well-being. Finally, the third line of research proposes a brand new model of self-representation derived from Associated Systems Theory (AST; Carlston, 1994), providing a very different and novel conceptualization of multiple selves, measure of self-complexity, and process account for spillover effects. Preliminary data suggest this AST-based approach has utility, and additional work explores whether AST provides a more comprehensive approach to self-complexity than traditional treatments. This work will increase our understanding of the interplay between multiple selves and one's overall sense of self. Moreover, we will develop better measures of multiple selves and self-complexity, test and extend existing theory on self-concept representation, develop clearer insights into the relations between multiple selves and well-being that can address inconsistent findings in the literature, and suggest new avenues for research that illustrate the importance of understanding how the representation of multiple selves mediates affective responses to life events.

Stereotype threat occurs when the awareness of a negative stereotype about a social group in a particular domain produces suboptimal performance by members of that group (see Steele, 1997). For example, stereotype threat leads African-Americans to perform worse on tests described as assessing intelligence, Whites to perform worse on tasks described as assessing natural athletic ability, and women to perform worse in math. Indeed, stereotype threat is wide-ranging and affects many social groups in a variety of critical situations. Because the implications of stereotype threat are so considerable and are often revealed by members of underrepresented populations in society, it is important to understand how stereotype threat leads to less-than-optimal performances. In other words, why does stereotype threat occur? One hypothesis for how stereotype threat exerts its impact is that one's worries about confirming a negative performance stereotype reduce the working memory capacity necessary for successful skill execution. But reductions in working memory may not be the only explanation for how stereotype threat operates, especially for proceduralized skills (i.e., skills that run as multi-step procedures largely outside of working memory) such as high-level athletics, surgical procedures, or writing computer programming code. Previous research has demonstrated that proceduralized skills fail when performers are prompted to pay attention to skill execution in a manner that disrupts the fluent execution of automated behaviors, rather than by reducing the working memory capacity available for performance (Beilock & Carr, 2001). In the current work, we propose that stereotype threat can harm performance by multiple mechanisms, and that how particular skills are implemented and cognitively represented in memory (i.e., working-memory intensive skills vs. proceduralized skills) dictate how they will be harmed by one's awareness of a negative performance stereotype. Specifically, we examine how stereotype threat effects are realized as a function of the working memory demands of the skill being performed, the working memory capacity of the individual performing the skill, and the skill level of the performer. By focusing on understanding the cognitive mechanisms underlying stereotype threat, a more comprehensive account of this detrimental phenomenon can be developed, which in turn generates novel predictions about how, when, and for whom stereotype threat effects will be revealed, and, ultimately, how they may be prevented.