In the spring and summer of 1987, 1,150 youth aged 17 to 22 and a parent of the youth are being interviewed in the third wave of the National Survey of Children. The child and his or her parent were initially interviewed in 1976 and re-interviewed in 1981. Consequently, a wealth of information exists about the early lives of these young people, including demographic data; measures of personality; a test of intellectual ability; information about family relationships, closeness to parents, religion, and school progress; attitudes about school; behavior problems; and relationships with siblings and peers. In 1987, information was again obtained on a variety of topics, including if and when the youth initiated use of alcohol, of marijuana, and of other illicit drugs; on drug use during the past year; and on the occurrence of varied problems associated with drug use. It is proposed to use these data to examine the etiology of adolescent drug abuse. The antecedents of age at initial use of each drug would be assessed using proportional hazards life table models. Multivariate logit models would be used to investigate the antecedents of current use, frequency of current use, and problems associated with current use. Hypotheses to be tested are based on a synthesis of previous research with elements drawn from personality theory, alienation theory, social learning theory, and developmental theory. It is hypothesized that youth will be more likely to be drug users if they manifested personality deficits as children, if they were alienated from family, school, and religious institutions as children, and if they were exposed to models of substance use in the home and among friends. Based on developmental theory, we anticipate experimentation with substance use as a part of the identity formation process; however if youth experienced personality problems or alienation as children or if they were exposed to models of substance use, it is further hypothesized that youth will have initiated use at younger ages and that they are more likely to be heavy users experiencing problems related to drug use.
