Testing the Role of Aspirations, Future Expectations, and Strain on the Development of Problem Behaviors across Young and Middle Adulthood

Abstract

Using prospective data collected across 14 years by the National Youth Survey Family Study (N = 1,436), we assessed change in aspirations, future expectations, and strain, as well as contemporaneous and longitudinal effects of these predictors on drug use, offending, and the perpetration of intimate partner violence. Growth curve models showed that, although time significantly predicted change in these variables, trajectories remained relatively stable. Contemporaneous random-intercept regression models revealed that all three measures were associated with all three problem behaviors. The effects were predictive longitudinally but some associations and gender differences were attenuated when controlling for prior involvement in outcomes.

Notes

¹ We include life-course theories here because, indirectly, they focus on aspiration- and expectation-based achievements that are traditionally operationalized as positive transitions or turning points, like starting a marriage or career.Color versions of one or more of the figures in this article can be found online at www.tandfonline.com/udbh.

² The primary variables of interest were collected in W1–W5 but only 2–4 items per measure were assessed. Additional items were asked in W6–W7 but skip instructions limit the interpretation of the findings. More specifically, the skip instructions were such that the chances of achieving a given goal (i.e., future expectations) question was only asked if the importance of the goal (i.e., aspirations) was “very important.” These skip instructions were dropped in W8–W11.

³ While the current study is neither a formal nor a complete test of traditional strain theory, we use the term strain for brevity.

⁴ In W10, a 6-month measure was assessed instead of an annual measure. This limitation is mitigated by using a prevalence scores and the difference in drug use between W10 (9%) and W11 (10%) is minimal.

⁵ We also created variety scores and tested Poisson and negative binomial models, many of which were not an appropriate fit to the data. Furthermore, transforming all dependent variables to prevalence scores allowed us to present findings across different models in a consistent and concise manner.

⁶ As shown in Table 4, the Model 5 effect of expectations on offending is marginally significant (p = .08). The effect, however, is significant (p = .02) when the model is trimmed to exclude non-significant controls, specifically, age (p = .65) and education (p = .45).

⁷ We also explored the effects of predictors on outcomes from W8 to W9 (across a 3-year period) and from W9 to W10 (across a 9-year period). We are hesitant to report these findings, given that they are derived from particularly conservative models. Methodologically, we argue that these models are conservative because of the three- to nine-year lag between waves and because when lagged dependent variables are included, the models test the effect of the predictor on change in the outcome between waves (i.e., residual gain) and not the effect on the outcome itself. Ultimately, getting the lag correct is critically important. We also argue that even though the predictors are relatively stable over time, theoretically, their effects are likely to be more proximal, to be mediated by other factors, and to have cumulating consequences that unfold over the life course. Nevertheless, supplemental tables are available showing that almost all effects were significant in bivariate models and that over half of the effects were significant in models controlling for the lagged dependent variable. When demographic factors and the lagged dependent variables were added to each model, however, only two of the full models were significant.

Additional information

Notes on contributors

Kelly E. Knight

KELLY E. KNIGHT is an Assistant Professor in the Department of Sociology and Anthropology at Montana State University. She received her Ph.D. from the Department of Sociology at the University of Colorado at Boulder and has held positions at the Center for the Study and Prevention of Violence and at the National Council on Crime and Delinquency. Her research examines developmental, life course, and intergenerational trajectories of victimization, substance use, and offending.

Colter Ellis

COLTER ELLIS is an Assistant Professor in the Department of Sociology and Anthropology at Montana State University. He received his Ph.D. from the Department of Sociology at the University of Colorado. His research brings a social–psychological approach to the study of trauma, animal studies, and natural resources.

Jennifer Roark

JENNIFER ROARK is an Assistant Professor in the Social Work Program at Utah State University. She received her M.S.W. from Portland State University and her Ph.D. from the Department of Sociology at the University of Colorado at Boulder. Her research focuses on criminal justice program and policy evaluation, with an emphasis on intimate partner abuse.

Kimberly L. Henry

KIMBERLY L. HENRY is an Associate Professor of Psychology and Public Health at Colorado State University. She received her Ph.D. in Biobehavioral Health from the Pennsylvania State University. Her work is focused on adolescent and young adult development, prevention science, and longitudinal methodology.

David Huizinga

DAVID HUIZINGA is a Senior Research Associate at the Institute of Behavioral Science, University of Colorado and holds graduate degrees in mathematics and psychology. For over four decades he has conducted basic and evaluation research on developmental life-span issues and social problem interventions. He is the co-author of four books and numerous book chapters, journal articles, and government reports on issues surrounding delinquency, gangs, drug use, victimization, and mental health. Dr. Huizinga is the Principal Investigator of the Denver Youth Survey (1986–2015), has been Co-Principal Investigator and Principal Investigator of the National Youth Survey (1976–2006), and Principal Investigator of two cross-national projects investigating the effects of the transition from school-to-work and the effects of justice system processing on delinquency, gang membership, crime, and drug use. He frequently serves as a methods/statistical consultant for basic and evaluation research projects.