This paper was written while Dr. Perkins was at Michigan State University. An earlier version of this paper was presented at the 1995 Annual Conference of the National Council on Family Relations, Portland, Oregon.

Compared to the situation in prior historical periods, current social circumstances (e.g., regarding a burgeoning of youth poverty¹), place many more of today's youth at risk for adverse developmental outcomes, such as engagement in problematic behaviors (e.g., high school dropout, unemployability, prolonged welfare dependency, delinquency and crime²). Previous research has examined several factors that are related to youth involvement in risk behaviors and those that contribute to their resiliency³. However, very little is known about how these variables interrelate within diverse populations of youth. Policies and programs aimed at preventing or reducing risks among these groups are essentially uninformed by developmental studies of the individual and contextual bases of risk behaviors among diverse adolescents.

This study explored the interrelationships of six risk behaviors: Antisocial behavior/delinquency, alcohol use, hard drug use, soft drug use, sexual activity, and school misconduct. In turn, their relationships with adolescents' characteristics--age, gender, ethnicity, involvement in extracurricular activities, religiosity, and view of the future--and contextual characteristics--family support, parent-adolescent communication, peer group characteristics, and school climate--were examined.

This study addresses the following questions:

1. How are risk behaviors interrelated, and does this interrelation vary by age, ethnicity, and gender?

2. What are the individual and contextual variables that covary with risk behaviors among adolescents, and does this differ by age, ethnicity, and gender?

3. Are there particular variables that best account for the variance in selected outcome variables and/or in sets of these variables?

Cross-sectional data were gathered by utilization of the Profiles of Student Life: Attitude and Behavior Questionnaire, a 152-item inventory developed by Search Institute⁴. During Spring and Fall of 1993 and the winter of 1994, this survey was administered to 16,375 Michigan students in grades 6 through 12 in approximately 36 communities throughout the state⁵. Sample characteristics are presented in Table 1.

To address question 1, zero-order correlations among risk behaviors were calculated for the entire sample and among gender, ethnic, and age subgroups. In addition, a series of tests for differences between independent correlations were calculated to determine whether significant gender, ethnic, or age differences occurred in any of these intercorrelations. To guard against an inflated alpha, because the number of tests a Bonferroni correction was employed.

To address Questions 2 and 3, a total of 84 multiple regression equations were computed. Using the entire sample, a multiple regression was computed for each of the six risk behaviors. Next, six multiple regressions were computed for the male subgroup (one for each of the risk behaviors), and six were computed for the female subgroup. The same six risk-related multiple regressions were computed for each of the five racial/ethnic groups in the sample (i.e., African American, Asian American, European American, Latinos/Hispanic, and Native American subgroups for a total of 30 equations). Finally, for each of the six age groups in the data set (i.e., for the 12 to 17 year olds) the same six risk-related multiple regressions were computed for a total of 36 equations.

Question 1: How are risk behaviors interrelated, and does this interrelation vary by age, ethnicity, and gender?

Finding: Interrelationships were found among the six risk behaviors examined in this study (i.e., antisocial behavior/delinquency, alcohol use, hard drug use, soft drug use, sexual activity, and school misconduct) for all the subgroups.

Generally speaking, this intercorrelation did not vary by age, gender, and ethnicity. The only exception was the intercorrelations between the African American and the European American subgroups. Of the 15 comparisons made, 9 of the 13 that proved to be significantly different (p < .0001) were higher for the European Americans than for the African Americans.

Question 2:What are the individual and contextual variables that covary with risk behaviors among adolescents, and does this differ by age, ethnicity, and gender?

Finding: Generally, the same predictors were significant in the majority of the multiple regressions (see Table 2.). They were: peer group characteristics, age, ethnicity, religiosity, school climate, family support, and involvement in extracurricular activities. Three characteristics were consistently found not to be significant in the equations. They were: parent-adolescent communication, self-esteem, and view of the future.

Overall, the R²s from the multiple regressions for the risk behaviors that included the individual and contextual predictors were significant for the entire sample, for males and females, for the five ethnic groups, and for the six age groups.

Question 3: Are there particular variables that best account for the variance in selected outcome variables and/or in sets of these variables?

Finding: An examination of the unique variance accounted for by the predictors produce similar findings for the entire sample, for males and females, for the five ethnic groups, and for the six age groups (see Table 2). Peer group characteristics accounted for the most variance followed by age, gender, religiosity, and school climate. Although the predictor of ethnicity was significant in over one-third of equations (35.2%) in which it was involved, it accounted for the most variance in only one of the multiple regression equations.

Future Direction and Implications

Future research needs to extend longitudinally the scientific understanding of the co-occurrence of risk and their predictors. Moreover, future research needs to be systematic, change-sensitive research that includes a larger array of risk and predictor measures, among more diverse participants. Such research should pursue a multitrait-multimethod approach to measurement such that triangulation of the elements of the measurement model can occur.

Data from such future research would be used to further scholarship about adolescents, risk, and resiliency, as well as to inform policies and programs directed to the promotion of positive youth development. This Future research could better identify the individual and contextual characteristics that distinguish youth who, across their adolescent years, do and do not engage in risk behaviors. The findings of such research would more fully inform scientists and practitioners about individual differences in the interrelatedness of risk behaviors. Such research would provide the basis for prevention programs that are sensitive to contextual and individual diversity.

References

1. Hernandez, D. J. (1993). America's children: Resources from family, government, and the economy. New York: Russell Sage Foundation.

Huston, A. C. (1992). Children in poverty: Developmental and policy issues. In A. C. Huston (Ed.), Children in poverty: Child development and public policy (pp. 1-22). Cambridge: Cambridge University Press.

2. Feldman, S. S., & Elliott, G. R. (1990). At the threshold: The developing adolescent. Cambridge, MA: Harvard University Press.

Schorr, L. B. (1988). Within our reach: Breaking the cycle of disadvantage. New York: Doubleday.

3. Boyer, D., & Fine, D. (1992). Sexual abuse as a factor in adolescent pregnancy and child maltreatment. Family Planning Perspective, 22, 252-255.

Dryfoos, J. G. (1994). Full service schools: A revolution in health and social services for children, youth and families. San Francisco: Josses-Bass.

Hawkins, J. D., Catalano, R. F., Jr., Barnard, K. E., Gottferdson, G. D., Holmes, A. B., & Miller, J. Y. (1992). Communities that care: Action for drug abuse prevention. San Francisco, CA: Jossey-Bass.

Hogan, D. P., & Kitagawa, E. M. (1984). The impact of social status, family structure, and neighborhood on fertility of black adolescents. American Journal of Sociology, 90, 825-855.

Luster, T., & Small, S., A. (1994). Factors associated with sexual risk-taking behaviors among adolescents. Journal of Marriage and the Family, 56, 622-632.

Rutter, M. (1987). Psychosocial resilience and protective factors. American Journal of Orthopsychiatry, 57, 316-331.

Rutter, M. (1989). Pathways from childhood to adult life. Journal of Child Psychology and Psychiatry, 30, 23-51.

Werner, E. E. (1986). Vulnerability and resiliency in children at risk for delinquency: A longitudinal study from birth to young adulthood. In J. Burchard and S. Burchard (Eds.), Prevention of delinquent behavior (pp. 16-43). Newbury Park, CA: Sage.

Werner, E. E. (1989). High-risk children in young adulthood: A longitudinal study from birth to 32 years. American Journal of Orthopsychiatry, 59, 72-81.

Werner, E. E. (1992). The children of Kauai: Resiliency and recovery in adolescence and adulthood. Journal of Adolescent Health, 13, 262-268.

4. Benson, P. L. (1990). The troubled journey: A portrait of 6th-12th grade youth. Minneapolis, MN: Search Institute.

5. Keith, J. G., & Perkins, D. F. (1995). 13,000 Adolescents speak: A profile of Michigan youth. East Lansing, MI: Institute for Children, Youth, and Families.

6. Pittman, K. J., & Zeldin, S. (1994). From deterrence to development: Shifting the focus of youth programs for African-American males. In R. B. Mincy (Ed.), Nurturing young black males: Challenges to agencies, programs, and social policy (pp. 45-58). Washington DC: Urban Institute.

Children, Youth and Families Education & Research Network (CYFERNet). Permission is granted to create and distribute copies of this document for non-commercial purposes provided that the author and CYFERNet receive acknowledgement and this notice is included. Phone: 612.626.1111; E-mail: cyf@umn.edu

Return to CYFERNet