Social Goal Development during Middle School: Normative Changes and Prediction by Self-Esteem and Narcissism

• In Event: 1-116 - Circle Up: Using Interpersonal Theory and The Interpersonal Circumplex to Study Interpersonal Relationships Across Development

Integrative Statement

Social goal orientations encompass motives for social interaction and like other trait-like cognitions, may be activated from long-term memory to affect social behavior and adjustment (Crick & Dodge, 1994). Accordingly, social goals are related to social-emotional development. For example, agentic (status, influence) goals are related to peer-reported aggression, peer rejection, perceived popularity and predict increasing aggression over time, whereas communal (relationship, closeness) goals are related to prosocial behaviors and peer acceptance, and buffer against aggression across time (e.g., Caravita & Cillessen, 2012; Ojanen & Findley-Van Nostrand, 2014; Sijtsema et al., 2009). However, while correlates and outcomes of social goals are well understood, development of goals themselves and factors predicting individual variation in goals are not.
Using the Interpersonal Goals Inventory for Children (IGI-C; Ojanen, et al., 2005), a well-validated measure across countries, we examined trait-like agentic and communal goals among peers, and four goals representing blends of these predominant interpersonal orientations (see Figure 1) during middle school in Finland (N = 384; 12-14 years at T1; 53% girls). During adolescence, time spent with peers and clarity and autonomous pursuit of personal goals increase (e.g., McAdams & Olson, 2010) and adjustment among peers becomes critical for well-being (Bukowski et al., 2011).
We focused on mean-level changes in goals and self-esteem (Rosenberg, 1962) and narcissism (NPI; Ames et al., 2006) as predictors of goals. Self-esteem has been linked to agentic goals and aggression (Salmivalli et al., 2005), but we argue that inflated, rather than secure self-views account for these links and links to agentic-separate (dominance) goals. Following initial research in adults (Findley & Ojanen, 2013), these variables can help to explain individual variation in goal development.
All constructs were assessed in three occasions, across two 6-month intervals (T1 = 7th and 8th grade; T2 = 8th and 9th grade; longitudinal attrition = 28%). Missing data was imputed using FIML. Latent growth modeling was used to examine mean-level changes in goals and path modeling was used to examine self-esteem and narcissism as predictors of goals across two intervals (Mplus 5.2).
As expected based on existing research on adolescent development, we observed average increases in agentic-communal goals (change factor M = .10, p < .01; M = 16, p < .001 for boys and girls, respectively) and decreases in submissive-separate goals (change factor M = -.09, p <.01 for both genders) across time. Also, when controlling for each, only narcissism predicted gains in overarching agentic goals and in agentic-separate (dominance) goals (β = .24, p < .001; β = .15, p < .01), whereas self-esteem predicted gains in agentic-communal goals (β = .13, p < .01), consistently across time. These findings are among the first to document normative developments in interpersonal motivation and have implications for the study of social goals, self-perceptions and adolescent social and personality development.

Authors

• Tiina Ojanen, University of South Florida
  Presenting Author

• Danielle Findley-Van Nostrand, Roanoke College
  Non-Presenting Author