Key points are not available for this paper at this time.
Examined the structure of self-report scales designed to assess the frequency of adolescent problem behaviors. Urban (n = 988) and rural (n = 1,895) middle school students completed the Problem Behavior Frequency Scale (Farrell, Danish, & Howard, 1992a) and measures of other relevant constructs. Confirmatory factor analyses supported a model that included specific factors related to aggression, drug use, and delinquent behaviors, and a higher order problem behavior factor. Findings did not support a distinction between physical and nonphysical aggression. Results were generally consistent across settings (i.e., urban vs. rural) and gender. Other relevant constructs, including peer pressure for drug use and attitudes favoring aggression, had both specific associations with relevant first-order factors and more general associations with the second-order factor. These findings support the construction of separate scales assessing specific domains of problem behaviors in studies of adolescents' problem behaviors.
Farrell et al. (Mon,) studied this question.