A Longitudinal Study of School Dropout and Substance Use

Introduction.  The Flint Adolescent Study (FAS) is a longitudinal study of 850 ninth graders four main public schools in the second largest school district in Michigan.  The sample includes youth at risk for deleterious outcomes because of low school achievement status. The study provides a useful sample to study resiliency theory because a large number of youths with similar baseline risks were recruited. The goal of the study is to explore the promotive factors associated with school dropout and alcohol and substance use.

Method.  Individuals with an 8th grade GPA of 3.0 and below were selected and followed for four years beginning in Fall 1994. Data were collected once a year starting with respondents’ ninth grade using 50-60 minute face-to-face interviews. After the interview, respondents were asked to complete a pencil-and-paper questionnaire about more sensitive information such as substance use and sexual behavior. We obtained a 90% response rate from Year 1 to Year 4. The Year 1 sample includes 679 African‑American youth (80%), 145 white youth (17%) and 26 mixed African‑American and white youth (3%).  It is evenly divided by males and females.  

Key variables include: psychological well-being; relationships with family and peers; school experiences and attitudes; school and community involvement; delinquent and violent behaviors; alcohol and substance use; sex behavior and child bearing; and family structure and relationships.  Year 3 and 4 also included information about driving behavior, attachment style, stress, mentoring, and racial identity.  We have also collected information about parental education and occupation. 

Results.  We consistently find support for resilience model across different risk and protective factors. Representative findings include: 1) father involvement is associated with less drug use and other problem behaviors longitudinally; 2) mother and father support protects against risks for violent behavior; 3) depressive symptoms predicted later marijuana use for males4) youth with adult male role models reported less drug use and problem behavior and adult female role models less psychological distress than youth with no role model; and 5) racial identity was associated with less alcohol use. (A list of publications can be found at: http://prc.sph.umich.edu/research/flint-adolescent-study/)

Conclusion.  Our sample is unique because it is predominantly African-American, includes longitudinal data, and contains data on a wide variety of variables that are not often all in one dataset (e.g., racial identity, participation in community and school activities, family and school data, sexual behavior, and substance use).  The use of resiliency theory in our work also enables us to not only identify risk factors, but to examine factors in adolescents’ lives that may help them avoid the negative consequences of those risks.