Applied Research Topics
Risk and protective factors for substance use and HIV often cannot be randomized to individuals. For example, researchers are unable to randomly assign levels of parental knowledge of youth risky behavior (Lippold, Coffman, & Greenberg, 2013). Center researchers use inverse propensity weighting, an approach to adjust for confounders, to address key questions about the effects of non-randomized predictors on a variety of substance use and HIV behavioral outcomes.
We have used inverse propensity weighting to address questions about the effects of a variety of risk factors on substance use outcomes, as well as the effects of prevention program implementation on these outcomes. For example, we have used inverse propensity weighting to determine whether there is a causal relation between parental knowledge and youth risky behavior among rural adolescents (Lippold, Coffman, Greenberg, 2013). After adjusting for many confounders (e.g., parent-child relationship, demographic variables, earlier levels of problem behavior) this work suggests that parental knowledge is causally related to substance use during middle school.
We have also examined the critical role that prevention program implementation plays in successful prevention efforts (Crowley, Coffman, Feinberg, & Greenberg, 2014).
We integrated inverse propensity weighting with latent class analysis to determine the causal effects of predictors on latent class membership (Lanza, Coffman, & Xu, 2013). In this work, we discuss a step-by-step approach for determining the causal effects of predictors on a latent class outcome, and we examine the causal effect of college enrollment on adult substance use measured as a latent class variable.
In Coffman, Melde, & Esbensen (2014), we used generalized boosted models to estimate propensity scores in causal mediation, in order to assess the impact of gang membership on later drug use.
Coffman, D. L., Melde, C., & Esbensen, F.-A. (2014). Gang membership and substance use: Guilt as a gendered causal pathway. Journal of Experimental Criminology. doi: 10.1007/s11292-014-9220-9
Crowley, D. M., Coffman, D. L., Feinberg, M. & Greenberg, M. (2014). Evaluating the impact of implementation factors on family-based prevention programming: Methods for strengthening causal inference. Prevention Science, 15(2), 246-255. doi: 10.1007/s11121-012-0352-8 PMCID: PMC3859719
Crowley, D. M., Jones, D. E., Coffman, D. L., & Greenberg, M. T. (2014). Can we build an efficient response to the prescription drug abuse epidemic? Assessing the cost effectiveness of universal prevention. Preventive Medicine, 62, 71-77. doi: 10.1016/j.ypmed.2014.01.029. PMCID: PMC4131945
Lanza, S. T., Coffman, D. L., & Xu, S. (2013). Causal inference in latent class analysis. Structural Equation Modeling, 20(3), 361-383. PMC Journal-In Process
Lippold, M., Coffman, D., & Greenberg, M. (2014). Investigating the potential causal relationship between parental knowledge and youth risky behavior: A propensity score analysis. Prevention Science, 15, 869-878. doi: 10.1007/s11121-013-0443-1
We have applied inverse propensity weighting to address several important questions about HIV and sexual risk behavior. For example, we applied cutting-edge work on causal mediation with inverse propensity weighting to examine how the Reducing Risky Relationships HIV intervention works to reduce risky sexual behavior among women being released from prison by decreasing incorrect and dangerous thoughts about relationships (Coffman & Kugler, 2012). Read a detailed description of this study.
We have also used our approach to integrating inverse propensity weighting with latent class analysis to examine the causal effect of early sex on subsequent delinquency latent classes in high school (Butera, Lanza, & Coffman, 2012).
Butera, N. M., Lanza, S. T., & Coffman, D. L. (2013). A framework for estimating causal effects in latent class analysis: Is there a causal link between early sex and subsequent profiles of delinquency? Prevention Science. doi: 10.1007/s11121-013-0417-3 PMCID: PMC3888479
Coffman, D. L., & Kugler, K. C. (2012). Causal mediation of a human immunodeficiency virus preventive intervention. Nursing Research, 61(3), 224-230. PMCID: PMC2646486
We have examined how the causal effects of leisure boredom on cigarette smoking initiation are moderated by level of risk for initiation (Coffman, Caldwell, & Smith, 2012).
Coffman, D. L., Caldwell, L. L., & Smith, E. A. (2012). Introducing the at-risk average causal effect with application to HealthWise South Africa. Prevention Science, 13, 437-447. PMCID: PMC3405190