This course will provide an overview of the theory underlying the use of multi-level models, using a parametric framework, as well as teach the basic application of methods necessary to conduct and interpret multi-level analyses of epidemiologic data. After reviewing statistical and theoretical principles of linear and logistic regression models using only one level of organization, lecture and lab sessions will be focused on the use of random-effects models and generalized estimating equation (GEE) models for the analysis of data with two levels. Hands-on exercises will use data from an investigation of the influence of neighborhoods on obesity, focusing on the application and interpretation of regression models that account for clustered observations and group-level covariates. Stata code will be provided for all in-class exercises and participants will learn how to implement multi-level models; R and SAS code will also be provided.
This course is suitable for academics/professionals and advanced postgraduate students who are undertaking or have completed Masters-level training in epidemiology and/or biostatistics related disciplines. Intermediate-level understandings of study design, measurement concepts, and regression models are required to complete the course.
The course takes place over 4 half-days, from 13.30-17.00 each day, with theory and practical sessions each day.
• April 10th (day 1): principles of clustered data in epidemiology
• April 11th (day 2): multilevel research designs: rationale and examples
• April 12th (day 3): multilevel research designs: data structures and analytic approaches
• April 13th (day 4): complex survey designs in epidemiology
About the lecturer
Dr. Keyes is an expert in methodological issues in age-period-cohort effect estimation, and her empirical work in age-period-cohort effect has examined a range of outcomes including obesity, perinatal outcomes, substance use disorders, and psychological distress. She is the author of more than 170 peer-reviewed publications as well as two textbooks published by Oxford University Press with co-author Sandro Galea: “Epidemiology Matters: A New Introduction to Methodological Foundation” published in 2014 and “Population Health Science” published in 2016.
Venue at the UCT Faculty of Health Sciences: Seminar Room 2, SPH&FM, Falmouth Building, Entrance 5
Registration is free of charge but spaces are limited.
To request more information, please contact Tando Mpikwa ([email protected])