Impact of Family-Based Secondary Prevention Programming on Risk, Resilience, and Delinquency: A 6-Month Follow Up within a Randomized Control Trial in Honduras

Abstract

The present study examines the impact of Proponte Más, a family-based, multiple gated prevention program in Honduras. The primary goal of Proponte Más is to reduce the risk factors for problem behavior associated with youth in Honduras through family-based intervention services. At risk youth were randomly assigned to a treatment and control group. Data for the study were sourced primarily from the Behavior Measurement Instrument and FACES IV instrument given to study participants prior to and following programming. Our study finds that Proponte Más succeeded in improving family functioning and decreasing risk factors and increasing protective factors. Our findings pertaining to the impact of the program on delinquency, however, were mixed. Our findings suggest that RCTs and family-based interventions can be effectively carried out in the most violent and at-risk communities in the Western Hemisphere and can result in reduced risk and improved resiliency among at-risk youth.

Keywords:

• Family-based secondary prevention
• randomized control trial
• Central America
• Honduras
• risk and protective factors
• delinquency

Acknowledgements

The authors would like to thank Robyn Braverman and Guillermo Céspedes for their leadership and support throughout the project, Axel Rivera for all of his assistance on matters related to data collection and interpretation, Erik Alda for his comments and suggestions on earlier versions of this manuscript, Lidia Nuno for her work on earlier phases of the project, and Cher Stuewe-Portnoff for copy editing the manuscript.

Disclosure statement

The manuscript is the result of research conducted under Award No. AID 522‐TO‐16‐00001, which was a contract from USAID to Creative Associates, for which the authors served as the evaluators.

Notes

1 A variable may operate as both a direct and buffering protective factor, and its effect may be contingent on developmental factors, environment, and other characteristics ( Lösel & Farrington, 2012; also see Herrenkohl et al., 2012 ). These issues are beyond the scope of the current study. Additional examination of direct and buffering protective factors, using the same sample and data, is needed.

2 Programs have rarely adopted the multiple gating approach, however, perhaps due to the inherent complexities of treating impoverished families in high-risk neighborhoods, with adults already under-involved in their children’s lives and mistrustful of service providers (Hogue et al., 2002).

3 We examined demographic characteristics, family adaptation and functionality, risk and protective factors, and delinquency. We found no effect size differences greater than 0.2; none was statistically significant.

4 The present study is restricted to the analysis of direct protective factors. Researchers generally agree, however, on the need for more examination of direct and buffering protective factors using the same sample and data, that a variable may operate as both a direct and buffering protective factor, and that its effect may be contingent on developmental factors (e.g., age), environment (e.g., school, neighborhood), and other characteristics (e.g., gender; Lösel & Farrington, 2012).

5 Initially, 38 risk and protective factors were available for analysis, including eight scales determining eligibility. We excluded four risk and protective factors associated with the school domain; some youth were not enrolled . We dropped six items in the family history of antisocial behavior scale; they were not time varying and would not show meaningful change between pre/post-treatment.

6 The Percent of the Maximum Possible (POMP; Cohen et al., 1999) scoring method assigns a percent to an item as its score. It reflects the response position on the scale as a percent of the maximum possible score achievable on that scale. The POMP formula: [(observed – min) / (max - min)] * 100, where observed = the observed score for a single case, min = the minimum possible score on the scale, and max = the maximum possible score on the scale.

7 We did not convert measures to dichotomous variables; we lacked the capacity to identify extreme or risky values as we lacked population norms of these scales for this population. More problematic, dichotomizing risk and direct protective factors would invite problems associated with regression to the mean because our change scores for those extreme (of extreme) cases would have been even more susceptible to these statistical adjustments.

8 Bootstrapping is a rigorous and powerful method requiring fewer statistical assumptions about the shape of the sampling distribution of the indirect effect, while maintaining reasonable control over the Type I error rate (Hayes, 2013).

9 We found similar results using weighted factor scores, other than for family management. Treatment significantly decreased general delinquency by improving family management.

Additional information

Notes on contributors

Charles M. Katz

Charles M. Katz is a professor in the School of Criminology and Criminal Justice and Watts Family Director of the Center for Violence Prevention and Community Safety at Arizona State University. His research involves collaborating with agencies to increase their organizational capacity to identify and respond to crime effecting local communities. He currently serves as a research partner to the Phoenix Police Department to evaluate their BJA sponsored projects related to SMART policing and a Crime Gun Intelligence Center. He served as one of the two primary authors of the U.S. Department of Justice Body-Worn Camera Toolkit and currently serves as a senior advisor to the Bureau of Justice Assistance on its Body-Worn Camera Training and Technical Assistance Team. He recently completed sponsored research evaluating the impact of family-based prevention programming on delinquency and gang involvement in the Caribbean and Honduras, and continues to collaborate with the United Nations Development Program on police capacity building in the Caribbean.

Hyunjung Cheon

Hyunjung Cheon is an Assistant Professor in the Department of Criminal Justice at the University of Texas at El Paso. Her research interests focus on study of race/ethnicity and crime and criminal justice, comparative criminology, police-community relations, and quantitative methods.

E. C. Hedberg

Eric Christopher (E. C.) Hedberg earned his PhD in Sociology from the University of Chicago in 2009. Most known for his work on evaluation design, Hedberg is an interdisciplinary quantitative methodologist with solid sociological training. Hedberg’s research interests include areas of methodology related to evaluation and analysis of ego-centric networks, and relatedly, the substantive areas of education, criminology, and social capital. Currently, Hedberg is a Senior Associate at Abt Associates, an Accredited Professional Statistician® by the American Statistical Association, and a Sociologist. Hedberg's current areas of research include investigating the design of evaluations in education and criminology, in addition to measuring social capital through social network contextual effects. He is best known for estimating and publishing important experimental design parameters. Hedberg has authored or co-authored over 30 methodology-focused papers and books that have appeared in education, medical, and criminological journals, while also contributing to numerous reports and presentations at major research conferences such as APPAM, AERA, SREE, ASA, and ISNA. He has also authored a SAGE “little green” book on statistical power analysis. Much of his research is funded by grants from federal and foundational organizations.

Scott H. Decker

Scott H. Decker is Foundation Professor Emeritus at Arizona State University. He currently works as Senior Scientist at CNA. A 6-month follow up within a Randomized Control Trial in Honduras.