This study examined the reliability and construct validity of a modified version of the Colorado Symptom Index (MCSI), a brief, self-report measure of psychological symptomatology, in a study of interventions to prevent homelessness. Eight projects in a national, cooperative study collected new data at baseline, 6, and 12 months using a set of common measures as well as site-specific instruments. The pooled sample consisted of 1,381 persons in treatment for mental illness or substance abuse (or both), of which 84% had a history of homelessness. The analyses employed classical and Rasch methods to examine the MCSI's content validity, internal consistency and item quality, test/retest reliability, dimensionality, appropriateness for the sample, construct validity, and responsiveness to change. This 14-item scale was found to be a reliable and valid measure of psychological symptoms in this sample. Its content was consistent with other symptom measures, its high internal consistency and test-retest coefficients supported its reliability, its relationships to other measures indicated that it had good construct validity, and it was responsive to change. We conclude that the MC

Given the dearth of information concerning outcomes for adolescents who have been in treatment for substance abuse and related problems, the purpose of the present article is to examine the posttreatment status of adolescents who have been in residential therapeutic communities (TCs). One-year posttreatment outcome data are described for 485 adolescents, and separately for treatment "completers" and "noncompleters." The majority of the sample had been mandated to treatment by the criminal justice system. Most reported marijuana as their main drug of abuse. There were significant reductions in drug use and criminal activity, and the most consistent predictors of positive outcomes were completion of treatment and not associating with deviant peers posttreatment. The findings were obtained on a treatment sample for whom there is relatively little research to date; they provide important evidence for the effectiveness of the therapeutic community for this population.

The sample presents with extensive psychopathology and a history of physical and sexual abuse. Gender differences indicate that, except for antisocial personality, females yield higher rates on measures of both psychiatric disturbance and abuse. The relationship between psychopathology and abuse also appears to be much stronger for females than for males. However, the relationship between abuse and adult homelessness appears to be similar for men and women. The gender differences in the relationship between histories of abuse and manifestations of psychiatric disturbance support a hypothesis that has been proposed elsewhere: Females internalize the trauma associated with abusive experience, while males externalize it. The findings suggest that, although there may be a need for gender-specific targeted interventions, treatment providers must also recognize that the impact of abuse seems to transcend gender within this population.

No abstract

HIV risk behaviors of adolescents (N = 938) admitted to residential therapeutic communities for drug abuse and related problems are described. Approximately 95% of the sample was sexually active. The adolescents reported that they had used drugs or alcohol about half of the time that they engaged in sexual relations and that half of their sexual activity was unprotected. Males and females differed in their self-perceptions of risk for HIV infection. Three dimensions of risky behaviors were identified by factor-analytic procedures: Risky Sex With Men, Risky Sex With Women, and Risky Drug Use Behaviors. Separate regression equations for males and females identified common and unique predictors of risky behaviors. A comparison of 1-year pretreatment with 1-year posttreatment risky behaviors for a subsample of the full cohort revealed significant positive changes (i.e., reductions) on some, but not all, measures of risky behaviors.AIDS, along with drug-related violence, is becoming the leading cause of death among adolescents (Strunin & Hingson, 1992). One fifth of all AIDS cases in the United States have occurred among those 16-29 years of age, the majority of whom contracted the virus during adolescence.Most studies reporting seroprevalence rates offer limited information about adolescents' sexual behaviors. Heterosexual intercourse, as the route of HIV transmission, is more common for 13 to 19 year olds (12%) compared with older adults (5%; Office of Technology Assessment [OTA], 1991). Data reported in recent studies indicate that sexual activity among