Family attention and youthful drug use: Protection against involvement

This thesis research utilizes nationally representative samples that were drawn to represent school attending adolescents in Panama, Central America, and the Dominican Republic (the PACARDO project). It focuses on the influence family attention may have on youthful drug involvement and how youthful drug involvement occur in geographical clusters that might be consistent with a "contagion effect" by region and by school.;The greatest inverse association between family attention and drug experience was found for marijuana involvement, in terms of magnitude of association. Youth with middle levels of family attention were 0.6 as likely to have the chance to try marijuana and were 0.5 times as likely to initiate marijuana use compared to youth with the lowest levels of family attention. Youth with the highest levels of family attention were 0.4 times less likely to have exposure opportunity and 0.3 times as likely to initiate use than youth with the lowest levels of family attention.;For alcohol and tobacco involvement, the magnitudes of the associations between the chance to try and initiation of drug use were similar but somewhat smaller than was noted for marijuana involvement. For inhalant involvement, family attention exerts greater influence on initiation of use than for the chance to try inhalants.;Alternating logistic regressions (ALR) also were performed to shed light on the clustering of youthful drug involvement within regions of residence and within schools. In general, the school-wise PWOR estimates were higher and less likely to entrap the null value than were region-wise PWOR estimates. The PWOR school-wise estimates were above 1.0 when both the one nested and the two-nested models were used. Region-wise PWOR estimates entrapped the null value for more countries and for all four drug groups (alcohol, tobacco, inhalants and marijuana) when the two nested model was used. In sum, a pattern of school level clustering of drug involvement, for actual initiation of drug use and for the first chance to try each of the four drugs under study was noted. There was an especially consistent pattern of school-level clustering for the first chance to try tobacco and the first chance to try marijuana. (Abstract shortened by UMI.).