Research Project
6523386
DESCRIPTION: (provided by applicant) Human immunodeficiency virus (HIV) and other sexually transmitted diseases (STDs) are major health concerns among adolescents. The adolescent health field has repeatedly requested development of a brief, standardized, reliable and validated adolescent HIV/STD transmission risk screening instrument to be used across sites, settings, interventions, and populations. Thus, the goal of this research is to develop this reliable and valid self-report HIV/STD-risk screen that is developmentally appropriate for administration to adolescents 12 through 19 years of age. This screen will be configured like the widely used, NIDA-developed Problem-Oriented Screening Instrument for Teenagers (POSIT). Toward this goal, a 5-year, multi-site psychometric study will be conducted to: (1) translate the preliminary screen into Spanish; (2) reduce the number of items on the currently available 30-item scale; (3) accumulate diverse evidence for the reliability and validity of the English screen; (4) identify optimal cut scores to differentiate high/medium-from low/medium-risk youth; and (5) accumulate similar psychometric evidence and identify optimal cut scores for the Spanish- language version. The 30-item screen and accompanying battery of validating instruments will be administered using audio, computer-administered self-interview (audio-CASI) technology by a collaborative team of researchers at field sites across the US, initially in English and, after refining, in Spanish. Sites will include outpatient adolescent health clinics and pediatric clinics to target low/medium-risk youth and drug treatment settings, outpatient STD clinics, counseling clinics, and juvenile detention centers to target high/medium-risk youth. Data will be aggregated nationally and reliability analysis and validity analysis-construct, concurrent, and discriminant validity-performed, along with item analysis to reduce the item set so that the most efficient and psychometrically sound screen is retained. Optimal cut scores will be derived using the Minimum Loss function cut score determination method along with estimations of the relative seriousness of making Type I and Type 11 errors.
