In an effort to explain the persistence or discontinuation of childhood psychopathology, factors associated with the increase, decrease, or stability of conduct problems (CP) in low-risk and high-risk boys across a 5-year period will be examined using linear and nonlinear models. Boys were first seen at ages 6 - 12 as participants in naturalistic summer research programs in 1991 through 1995. Two-thirds of participants met DSM-III-R criteria for ADHD, one-third did not. Risk status will be determined by level of CP at baseline assessment. It is expected that by early 2001 approximately 141 of the original participants will have been seen for the 5-year follow-up assessment. Multi-method and multi- informant data in the following domains are relevant to this proposal were collected at baseline and follow-up: CP, comorbid psychopathology, intelligence, peer status and friendships, self- perceptions of competence. Some of these variables will be coded from audiorecorded semi-structured interviews conducted with children. Change in level of parent- and teacher-reported CP across time will constitute the dependent variable in analyses. Variables from baseline and follow-up assessments that for theoretical or empirical reasons are hypothesized to be associated with developmental outcome will be tested as predictors of change measured continuously and categorically. The risk vs. protective effect of variables associated with change in level of CP will be considered by examining the form of relations between predictors and outcome, and two models of the nature of relations between risk, protective factors, and outcome will be tested and compared.
