Broadly, the long-term objective is to empirically answer a well documented paradox in the scientific literature that indicates racial/ethnic groups who report a lower prevalence of substance use have a higher likelihood of substance-related health and social consequences.
Three specific aims i nclude...i) identify social mechanisms in the individual, family, neighborhood, and school domains that play a central role in determining health and social substance use consequences over and above substance use itself and socioeconomic status...2) identify promotive social mechanisms that may create or sustain social and health disparities by conferring an un-equivalent advantage in fewer substance use consequences between White and Black youth...3)to test and refine scale properties and capacity to provide systematic and unbiased trait level measurement across a racially/ethnically diverse group of adolescents. The National Institute on Drug Abuse (NIDA) introduced the Strategic Plan on Reducing Health Disparities for years 2004-2008 in which goal #1 calls for scientists to assess "both individual and community/environmental...protective factors for drug use and related consequences in various racial/ethnic populations." NIDA's specific plan for action includes the, "support of secondary analysis of data obtained under NIH-supported research pertinent to understanding drug abuse in racial/ethnic populations." In addition, "better measures and designs are needed to appropriately assess drug abuse and addiction and related behaviors in racial/ethnic populations." In response to NIDA's announcement, the specific aims will be tested using secondary data analysis from an NIH supported study of 999 multiethnic adolescents (DA 07031). A structural equation modeling framework will be employed to determine if individual and contextual promotive social mechanisms are underlying determinants of substance use consequences beyond substance use and socioeconomic status. Item Response Theory techniques, specifically Mixed Rash Modeling will permit statistical analysis capable to detect heterogeneity within a single construct as evidenced by latent subgroups that differ in their use of questionnaire response scales. Differential Item Functioning will allow for reconstructive scale analysis based on item difficulty or discrimination. Items that display bias (i.e., probability of endorsement is influenced by group membership) will be identified and adjusted for prior to administering total scores on a given ability or trait.
Findings from the proposed study have direct implications for measurement employed with diverse samples, the way substance-related interventions are crafted and delivered, as well allow policymakers to make the intervention choices required to reduce drug-related health and social problems.
|Vilsaint, Corrie L; Aiyer, Sophie M; Wilson, Melvin N et al. (2013) The ecology of early childhood risk: a canonical correlation analysis of children's adjustment, family, and community context in a high-risk sample. J Prim Prev 34:261-77|