Key points are not available for this paper at this time.
The present study addresses two limitations of the literature regarding the ethnicity-drug use relationship, namely, the absence of ethnically diverse samples and also the inadequate control of background variables. The data employed consisted of responses from 3,697 students derived from a single-stage probability survey from which eight ethnic groupings were constructed: (1) Eastern European, (2) Western European, (3) British Isles, (4) Black, (5) Oriental, (6) East/West Indian, (7) Jewish, and (8) Mediterranean. The findings reveal that, first, the significant zero-order ethnicity effect for alcohol use did not attenuate after controlling for background variables. Also, this relationship was not conditional upon other independent variables. Second, for other substance use measures, the significant zero-order ethnicity effect was found to be spurious or conditional.
Adlaf et al. (Sun,) studied this question.