Research Synthesis Lab
This lab utilizes innovative computing technologies for high dimensional, complex clinical trial data to better provide answers that have been just beyond our reach based on single trials alone. This lab is devoted to developing and refining state-of-the-art quantitative methods and using them as a scaffold to pursue “hard-to-reach” research objectives. The ultimate purpose of this lab is twofold: to seek innovations and refinements in clinical diagnoses and trials aimed at reducing excessive drinking and strengthen their research standards; and to seek innovations in quantitative research methodologies to meet the challenges of analyzing clinical trial data across the data life cycle. Current projects include Project INTEGRATE, which combines individual participant data (N = 12,630) from 24 independent, brief motivational interventions (BMIs) for college students to examine the overall strength of the effectiveness of BMIs and their mechanisms of change; and the Research Diagnostic Project (RDP), which uses both proprietary and public-access datasets to explore refinements of diagnostic decision rules for alcohol, cannabis, cocaine and opioid problems.
- To develop a more rigorous set of measures based on data from different studies to ensure that each measure shares a common underlying metric using a modified item response theory (IRT) model.
- To evaluate whether and when distinctive transitions in post-intervention trajectories of alcohol and drug use occur over time, and whether individual and situational differences contribute to different trajectories post-intervention.
- To test distinctive mechanisms of change (e.g., changes in alcohol expectancies, protective behavioral strategies, readiness to change, peer norms) and to examine whether potential moderators (e.g., family history of alcoholism, gender) either facilitate or hinder post-intervention changes.
- To test the efficacy of alcohol interventions for drug use and other secondary outcomes.
- To explore refinements of diagnostic decision rules for alcohol, cannabis, cocaine and opioid problems.
- "Innovative Analyses of Alcohol Intervention Trials for College Students" NIAAA R01 AA019511, April 2010 to March 2016 (Project INTEGRATE)
- Development of item response theory models and MCMC algorithms for complex data
- Development of meta-analysis and IDA as research methods
- Understanding racial and gender differences in alcohol use, alcohol problems, and other intervening variables such as protective behavioral strategies
- In depth content analysis of brief motivational interventions (BMIs) in relation to treatment outcomes
- Meta-analysis of the average treatment effect size of BMIs for college students
- Analysis of subgroups and conditions that modify the effect of BMIs
- Research Diagnostic Project (RDP). Using proprietary and public access data sets to explore refinements of diagnostic decision rules for alcohol, cannabis, cocaine and opioid problems in DSM-5 and ICD 11
- Application of advanced quantitative techniques designed for psychological test construction to vexing problems in psychiatric diagnosis, including first applications by the RDP of survival/hazard analysis, configural frequency analysis, IRT analysis, and receiver/operator characteristic analysis to diagnostic criteria
- Sharpening the diagnostic subalgorithm for substance-specific withdrawal through the application of the above and other techniques
|Core Faculty:||Eun-Young Mun, PhD, CAS
Helene R. White, PhD, CAS
James Langenbucher, PhD, CAS
Judit H. Ward, PhD, CAS
Jimmy de la Torre, PhD, Graduate School of Education, Rutgers University
|Affiliated Faculty:||David Atkins, PhD, University of Washington
Su-Young Kim, PhD, Ewha Womans University
Mary Larimer, PhD, University of Washington
Christopher Martin, University of Pittsburgh
|Graduate Students:||Yang Jiao, MS, CAS (Department of Statistics)
Eric Garfinkle, JD, CAS (Department of Clinical Psychology, GSAPP)