This advanced webinar on multilevel modeling (MLM)includes a brief conceptual overview and demonstrates the application of MLM for public health questions. The session is intended for public health systems researchers, health services researchers, and others who are interested in using multilevel models to investigate the health of populations.

Free
301
* Please complete an evaluation of this seminar once you have finished viewing the presentation.

Dr. Adam C. CarleFaculty: Adam Carle, Ph.D., M.A., Assistant Professor of Pediatrics, University of Cincinnati School of Medicine

Presentation Slides (PDF)

Suggested Readings (PDF)

Join Discussion

Examples and Data: Throughout the presentation Dr. Carle will refer to the following* example datasets (Mplus and Stata), code to run models in Stata (xtmixed and GLLAMM) and Mplus, and spreadsheet containing results for each model, including calculations for additional statistics.

*Clicking on these file links will bring you to a zip folder containing the exercise materials. For best results, after downloading the folder right click on the files, choose "open with" and select the appropriate program on your system.

Course Level: 301 (Advanced)

Duration: 90 min. 

Overview: This advanced webinar on multilevel modeling (MLM) includes a brief conceptual overview and demonstrates the application of MLM for public health questions. The session is intended for public health systems researchers, health services researchers, and others who are interested in using multilevel models to investigate the health of populations.

In 2009, the Public Health Systems Research (PHSR) program at AcademyHealth conducted a needs assessment of the methods and training needs in this emerging field. Based on key informant interviews and a survey of researchers, policymakers, and practitioners, the PHSR community expressed a need for additional training in qualitative and quantitative methodologies.  Among these, the application of MLM was considered a high priority. For more information on the field of PHSR, please see, "A Data and Methods Needs Assessment for Public Health Systems Research."

*Due to generous support from the Robert Wood Johnson Foundation, the cost of participation in this webinar is subsidized to enable a broader participation from the PHSR community.

Abstract: Health services research often seeks to investigate the relationships between individuals and their contexts. While some traditional regression approaches allow investigators to include variables at both levels, they do not allow one to investigate what predicts variance between and within contexts. As researchers become increasingly interested in the simultaneous influences of individual and contextual variables, analysts need to adopt methods that allow fuller investigations of predictors at both levels. Multilevel models, also called hierarchical linear models, allow this. This webinar provides an introduction to multilevel modeling through an applied approach, emphasizing interpretation. A software-based example using MLwiN and public data are included, with access to example SAS, Stata, and Mplus code. By the end of this session, participants will:

  1. Understand fundamental concepts in multilevel models.
  2. Know the types of research questions and hypotheses appropriately answered with multilevel models.
  3. Have a basic knowledge for conducting multilevel model analyses in MLwiN.

Faculty Bio: Adam C. Carle, Ph.D., M.A., is a clinically and quantitatively trained investigator who uses advanced statistical methods to study health disparities among adults and children, especially children with special health care needs. Dr. Carle's publications use latent variable models (e.g., item response theory, confirmatory factor analysis, structural equation modeling) and multilevel models to examine health disparities across traditionally underserved and underrepresented groups. He seeks to better understand individual and contextual variables' influences on health and health disparities at individual, local, state, and national levels. Dr. Carle received a Ph.D. in 2003 from Arizona State University, followed by a a two-year post-doctoral appointment at the U.S. Census Bureau. Dr. Carle spent three years as an assistant professor of psychology at the University of North Florida before joining Cincinnati Children's Hospital and Medical Center and the University of Cincinnati's Medical School.

System Requirements: The latest version of Adobe Flash Player is required to view on-demand seminars. Download Adobe Flash Player for free.

Please check back for updates on new online training opportunities. For more information or to sign up for notification about new opportunities please contact Adam Barnhart at 202.292.6700 or send a message to hsrmethods@academyhealth.org.