Bowlby's attachment theory is a theory of psychopathology as well as a theoryof normal development. It contains clear and specific propositions regarding the role of earlyexperience in developmental psychopathology, the importance of ongoing context, and the natureof the developmental process underlying pathology. In particular, Bowlby argued that adaptationis always the joint product of developmental history and current circumstances (never eitheralone). Early experience does not cause later pathology in a linear way; yet, it has specialsignificance due to the complex, systemic, transactional nature of development. Prior history ispart of current context, playing a role in selection, engagement, and interpretation of subsequentexperience and in the use of available environmental supports. Finally, except in very extremecases, early anxious attachment is not viewed as psychopathology itself or as a direct cause ofpsychopathology but as an initiator of pathways probabilistically associated with laterpathology.

A model of the effects ofchildren's temperament (negative and positive emotionality, impulsivity and attentionfocusing) on post-divorce threat appraisals, coping (active and avoidant), and psychologicalsymptoms (depression and conduct problems) was investigated. The study utilized a sample of223 mothers and children (ages 9 to 12 years) who had experienced divorce within the last twoyears. Evidence was found of direct effects of child-report negative emotionality onchildren's threat perceptions and of child-report positive emotionality and impulsivity onchildren's coping. Indirect effects of negative emotionality on active and avoidant copingthrough threat appraisal were found. Direct effects of the temperament variables on symptomswere also found. Cross group analyses indicated that the models were robust to age differences,but gender differences were found in the relation between negative emotionality and depression.The results of this study indicate that temperament and threat appraisals are important predictorsof children's post-divorce symptoms, and that temperament is a predictor ofchildren's appraisal and coping process.

The hypothesis was tested that weak theory of mind (ToM) and/or emotionrecognition (ER) abilities are specific to subjects with autism. Differences in ToM and ERperformance were examined between autistic (n = 20), pervasive developmentaldisorder—not otherwise specified (PDD-NOS) (n = 20), psychiatric control(n = 20), and normal children (n = 20). The clinical groups were matchedperson-to-person on age and verbal IQ. We used tasks for the matching and the contextrecognition of emotional expressions, and a set of first- and second-order ToM tasks. Autisticand PDD-NOS children could not be significantly differentiated from each other, nor could theybe differentiated from the psychiatric controls with a diagnosis of ADHD (n = 9). Thepsychiatric controls with conduct disorder or dysthymia performed about as well as normalchildren. The variance in second-order ToM performance contributed most to differencesbetween diagnostic groups.

The prediction ofyoung-adult adjustment from early adolescent conduct problems and depressive symptoms wasexamined for an at-risk sample of approximately 200 males. Conduct problems and depressivesymptoms were expected to show stability to young adulthood. It was predicted that earlyadolescent conduct problems would be associated with a broad range of adjustment problems inyoung adulthood due to cumulative adjustment failures. Early adolescent depressive symptomswere expected particularly to predict poor relationships with parents and peers. Additive andinteractive effects of the two predictors were examined. Conduct problems and depressivesymptoms showed significant stability to young adulthood. Conduct problems were associatedwith a broad range of adjustment problems including continuing problems in peer associations,substance use, self-esteem, relationships with parents, and new problems in noncompletion ofeducation, unemployment, driver's license suspensions, and causing pregnancies.Depressive symptoms predicted particularly to problems in social relationships. Higher levels ofboth conduct problems and depressive symptoms in early adolescence did not predict toincreased difficulties for any one outcome over either problem alone, either due to main orinteraction effects. Such co-occurrence, however, did result in problem outcomes in multipleareas, thus, the poorest adjustment overall.

The role of peer relations in childhoodand behavioral and family characteristics in early adolescence as risk factors for adolescentchildbearing was investigated. Sociometric surveys across third, fourth, and fifth grade andparent and child measures of behavioral and family functioning at sixth and eighth grade werecollected in a lower income, urban sample of 308 African American females. Results replicatedearlier findings on the role of childhood aggression as a predictor of teen motherhood. Inaddition, girls who displayed stable patterns of childhood aggression were at significantly higherrisk not only to have children as teenagers but to have more children and to have children atyounger ages. Results also indicated that females who were depressed in midadolescence were atgreater risk to become parents between age 15 and 19 years. These findings demonstrate the needto take a differentiated approach to understanding teen childbearing and varying developmentalpathways in the prediction of teen motherhood.

Recent research has suggested that there are twodistinct trajectories for the development of antisocial behavior in boys: a childhood-onsetpathway and an adolescent-onset pathway. After reviewing the limited available research onantisocial girls, we propose that this influential method of conceptualizing the development ofsevere antisocial behavior may not apply to girls without some important modifications.Antisocial girls appear to show many of the correlates that have been associated with thechildhood-onset pathway in boys, and they tend to show impaired adult adjustment, which is alsosimilar to boys in the childhood-onset pathway. However, antisocial girls typically show anadolescent-onset to their antisocial behavior. We have proposed that these girls show a thirddevelopmental pathway which we have labeled the “delayed-onset” pathway. Thismodel rests on the assumption that many of the putative pathogenic mechanisms that contributeto the development of antisocial behavior in girls, such as cognitive and neuropsychologicaldeficits, a dysfunctional family environment, and/or the presence of a callous andunemotional interpersonal style, may be present in childhood, but they do not lead to severe andovert antisocial behavior until adolescence. Therefore, we propose that the delayed-onsetpathway for girls is analogous to the childhood-onset pathway in boys and that there is noanalogous pathway in girls to the adolescent-onset pathway in boys. Although this model clearlyneeds to be tested in future research, it highlights the need to test the applicability of currenttheoretical models for explaining the development of antisocial behavior in girls.

The relationship between conduct problems at age 8 years and teenage pregnancy by the ageof 18 years was analyzed in a birth cohort of 491 girls. A statistically significant association wasfound between early conduct problems and later risk of teenage pregnancy, with girls in the mostdisturbed 10% of the cohort having a pregnancy rate that was 5.3 times higher (p <.001) than the rate found in the least disturbed 50% of the cohort. The elevated risk of teenagepregnancy amongst girls with early conduct problems was in part, explained by social and familyfactors that were correlated with early conduct problems, and in part, by a causal chain process inwhich early conduct problems were associated with increased rates of risk taking behaviors inadolescence, which in turn led to an increased risk of teenage pregnancy. These results suggestthat the higher rate of teenage pregnancy among girls with early conduct problems reflected boththeir relatively disadvantaged family backgrounds and their tendencies to risk taking behavior inadolescence.

Competent outcomes in late adolescence were examined in relation to adversity over time,antecedent competence and psychosocial resources, in order to investigate the phenomenon ofresilience. An urban community sample of 205 (114 females, 90 males; 27% minority) childrenwere recruited in elementary school and followed over 10 years. Multiple methods andinformants were utilized to assess three major domains of competence from childhood throughadolescence (academic achievement, conduct, and peer social competence), multiple aspects ofadversity, and major psychosocial resources. Both variable-centered and person-centeredanalyses were conducted to test the hypothesized significance of resources for resilience. Betterintellectual functioning and parenting resources were associated with good outcomes acrosscompetence domains, even in the context of severe, chronic adversity. IQ and parenting appearedto have a specific protective role with respect to antisocial behavior. Resilient adolescents (highadversity, adequate competence across three domains) had much in common with theirlow-adversity competent peers, including average or better IQ, parenting, and psychologicalwell-being. Resilient individuals differed markedly from their high adversity, maladaptive peerswho had few resources and high negative emotionality. Results suggest that IQ and parentingscores are markers of fundamental adaptational systems that protect child development in thecontext of severe adversity.

The reliability and validity of acomprehensive, integrative strategy for assessing adversity exposure in childhood andadolescence was examined in a community-based longitudinal study of 205 children. Extensivedata from multiple measures completed by adolescents and their parents regarding chronic andacute life experiences were aggregated into a Life Chart record of lifetime adversity exposure.Severity of adversity in three nonoverlapping time spans was rated by trained judges withexcellent interrater reliability. As hypothesized, Family adversity predominated over all othertypes of adversity in childhood, remained salient into midadolescence, and showed considerablestability over time. Adversity related to an individual's own behavior and psychologicalfunctioning rose to equal prominence by late adolescence. Adversity arising from physicalailments that were independent of the individual's psychological functioning declined overtime, while adversity arising from the community rose; physical and community events wereuncommon for most individuals but substantial for some. Results suggest that coherentdevelopmental trends in adversity exposure may be identified effectively through judges'ratings of severity of adversity over multiple-year time spans.