This 12-year longitudinal study followed a sample of predominantly African American youth for 4 years in high school (1994-1998), 4 year after high school as they experienced the transition to young adulthood (1999-2003), and another 4 years during their late twenties as they experienced middle adult transition (2008-2012).
We used stress-coping, and social influence and integration conceptual models to guide our study. We also applied life-span and resiliency perspectives to inform our longitudinal analyses and identify assets (e.g., self esteem, organized activity involvement) and resources (e.g., adult mentors, parent support) in youth’s lives that helped reduce the negative effects of risk exposure (e.g., poverty, family conflict).
The specific aims of the first 4 years of this study were to examine the risk/protective effects of psychosocial factors (e.g., meaningful instrumental behavior, social influence, and psychological well being) for school apathy (e.g., school dropout, failure, absenteeism and disinterest).
The specific aims of the second 4 waves of this study were to examine the effects of alcohol and other drug use (AOD) on adult transitional events such as marriage and family formation, education achievement, and employment. We also studied other psychosocial outcomes such as violent behavior, higher-stage AOD, sexual behaviors, psychological well-being, social relationships, and community involvement.
During the last 4 years of this study, we added several new measures to provide more in-depth information about parenting, employment qualities, and neighborhood context. The specific aims were to identify factors that increase risk for AOD and its negative consequences during the middle adult transition, examine the effects of young adult developmental tasks on young adult AOD and other outcomes, and investigate factors that help compensate and protect youth from the negative and positive outcomes associated with the transition to adult roles.
We sampled all youth in the four main public high schools in Flint, Michigan in 1994, who had a grade point average (GPA) of 3.0 or lower. GPA was a selction criteria because the original study was designed to examine whether school drop out and GPA was a consistent factor correlated with dropout, and self identified as African American, White, or mixed background. Race was an inclusion criteria because only 3% of the study body was not African American or White (or mixed race).
The Flint community high schools had 910 students who met this criterion. Of the 60 students who were not enrolled, most had moved out of the district but did not identify. Fewer than 10 of these 60 youth either refused to participate or their parents refused consent.
Our sample comprised 50% females. Approximately 80% of the participants were African American and 17% were European American. The demographics of our sample matched the student body in the school district.
Attrition from a longitudinal study is a major source of bias that can limit the validity of conclusions. We made various efforts throughout the study to keep the level of attrition as low as possible. Year 1 of the study included interviews with 850 ninth-grade youth. Year 4 involved interviews with 770 youth (91% response rate from Year 1).
Our overall response rate from Year 1 to Year 8 is 68%. Given the five-year gap between the wave 8 (2003) and wave 9 (2008), we confirmed the contact information for 465 participants in wave 9. Sixty-participants were ineligible for interview (N = 21; 13 deceased, seven incarcerated and one active in military duty) or refused to continue participating in the study (N = 34). Of the remaining 410 eligible participants, 341 interviews were completed in wave 9.
In the following years, we used the LexisNexis Accurint database to locate participants not found through other tracking methods and were able to complete 384 interviews in wave 12.
The table below shows the number of participants in each wave.
|Wave/Year||# of participants|
Questionnaires, Scales and Data
Our study questionnaires, scales, and data for the first four waves (1994-1997) were deposited as digital data in the University of Michigan Inter-university Consortium for Political and Social Research (ICPSR).
- List all 2083 variables in this study (Waves 1-4)
Questionnaires and data for additional waves are available upon request.