Journal of Consulting and Clinical Psychology 2003, Vol. 71, No. 3, 516 –527

Copyright 2003 by the American Psychological Association, Inc. 0022-006X/03/$12.00 DOI: 10.1037/0022-006X.71.3.516

When Syndromal Similarity Obscures Functional Dissimilarity: Distinctive Evoked Environments of Externalizing and Mixed Syndrome Boys Jack C. Wright

Audrey L. Zakriski

Brown University

Connecticut College

This research examined how syndromal approaches to assessment can obscure differences between children in the quality of their social interactions. Mixed boys (high for externalizing and internalizing problems) showed heterogeneity in the responses they evoked from others: For one subgroup, aggression and withdrawal evoked aversive responses from others, and even prosocial behavior evoked hostile peer responses; for the other, aggression and withdrawal evoked positive peer responses. Externalizing boys also showed heterogeneity in the patterning of their evoked responses. Within-group heterogeneity was not explained by boys’ syndrome scores but was linked to their reactions to specific antecedent events. The results illustrate how a contextualized analysis of behavior can reveal distinctive social interactional patterns that underlie similar overall rates of problem behaviors.

number of traits to describe personality and to operationalize a person’s standing on a trait as the summation over a set of correlated behavioral units. Contemporary approaches to personality remain rooted in this nomothetic tradition. The act frequency approach defines dispositions in terms of act trends—the overall frequency of dispositionally relevant acts over a period of observation (Buss & Craik, 1983). Big Five and Little Five approaches, so named because they focus on the five trait dimensions that best describe personality, generate trait scores that are sums over related behavior statements or adjectives (Costa & McCrae, 1995; John, Caspi, Robins, & Moffitt, 1994) and, thus, provide little information about context (see Block, 1995). As Barkley (1988) noted, instruments that are widely used to assess child psychopathology filter out situational variation in an effort to identify children’s “stable and enduring” properties, often by summing context-free behavior statements (e.g., Achenbach, 1993; Conners, 1999). Psychiatric classification systems (e.g., Diagnostic and Statistical Manual of Mental Disorders, 4th ed.) have also been characterized as syndromal, as they too emphasize the frequency of behaviors (see Scotti, Morris, McNeil, & Hawkins, 1996). The continued reliance on syndromal methods underscores the need to examine what may be lost when one focuses on overall behavioral output. As Gresham and Noell (1993) noted, a key concern is that syndromal approaches obscure psychological differences between people who are “topographically” similar in their syndrome scores. First, syndromes can obscure differences between individuals in how they react to events. Using a popular syndromal instrument (Teacher Report Form; TRF; Achenbach, 1993) and teachers’ reports of children’s reactions to events, Wright and Zakriski (2001) illustrated this point for two widely studied syndrome groups—pure externalizing boys and mixed (or comorbid) boys who display both externalizing and internalizing behaviors (see McConaughy & Achenbach, 1994). Each group consisted of one subgroup that showed problem behaviors primarily in response to aversive events (e.g., peer threat) and another that showed a contextually deviant pattern of problem behaviors in response to positive stimuli (e.g., friendly peer talk). Despite their

Floyd Allport (1924) provided a social interactional analysis of personality that remains remarkably fresh considering its vintage. He defined personality as a person’s characteristic reactions to stimuli—what he termed the “quality of the adaptation to the social features of the environment”—and he proposed that personality is both a result and a cause of other people’s behaviors, with “the responses of each person being re-evoked or increased by the reactions which his own responses called forth from others” (p. 149). Elements of his “characteristic reaction” concept can be found in the view that personality is better revealed in the patterning of people’s responses to stimuli than in their overall behavioral output (Wright & Mischel, 1987). Elements of his “circular social behavior” principle can be found in views that stress dynamic behavior– environment relations, including Bandura’s (1986) reciprocal determinism, Patterson’s (1997) social interactional approach to aggression, and interactional models of development (Sameroff, 1995; Thomas & Chess, 1984). Like Allport’s, these views underscore the reciprocal interactions between an individual’s personality and the social environment. Contemporary theorizing about personality has incorporated some of Allport’s (1924) themes, but the practice of assessment has pursued other objectives. As MacKinnon (1944) noted, the tendency of many American psychologists was to seek a small

Jack C. Wright, Department of Psychology, Brown University; Audrey L. Zakriski, Department of Psychology, Connecticut College. We thank the students, teachers, and school administrators whose cooperation and support made it possible to collect the data reported here. We are especially grateful to Harry Parad, Director of Wediko Children’s Services, Boston, for his continuing support; to Kristen Lindgren for her assistance in data collection; and to numerous colleagues who made comments on earlier versions of this article. Correspondence concerning this article should be addressed to Jack C. Wright, Department of Psychology, Brown University, 89 Waterman Street, Providence, Rhode Island 02912, or Audrey L. Zakriski, Department of Psychology, Connecticut College, 270 Mohegan Avenue, New London, Connecticut 06320. E-mail: [email protected] 516