Jerald G. Bachman
Published: 2007
Total Pages: 37
Get eBook
The Education-Drug Use Connection focused primarily on a nationwide sample of adolescents, first surveyed when they were nearing the end of 8th grade in the years 1991, 1992, and 1993, and followed biennially for eight years thereafter. (Full details on samples and methods are provided in Bachman et al., 2008, and are not repeated here.) The analyses in that book were based on the total samples of males and females, analyzed separately because of important gender differences in educational experiences, delinquency, and substance use. Most of the analyses included a race/ethnicity dimension (distinguishing between Whites, African Americans, Hispanics, and a remaining category that combined all others); it was a background dimension and statistically controlled any additive effects of race/ethnicity on the relationships between educational factors and various dimensions of substance use. Important subgroup differences were evident in the analyses, and were often noted--albeit briefly. Statistically controlling additive effects does not, however, fully guard against spuriously attributing racial/ethnic differences to other effects. As a further check against that risk, we repeated key analyses with only the White respondents included, and satisfied ourselves that racial/ethnic differences did not produce erroneous results in our findings for the total samples of males and females. We did not, however, attempt to explore interactions involving race/ethnicity. That remaining task, examining whether the same patterns of relationships found for the total samples are evident in each of the subgroups, is the main focus of the present paper. (Contains 1 footnote, 8 figures, and 12 tables.) [For "Monitoring the Future Occasional Paper Paper 65," see ED508288.].