PHSP workshop: Statistical approaches for modelling highly correlated exposures
The Doctoral Programme in Public Health Science (PHSP) welcomes all doctoral students and researchers at KI to our workshop on Monday May 22, with Andrea Bellavia.
The aim of this workshop is to provide an overview of statistical approaches for correlated exposures. Methods based on ordinary regression (e.g. multiple testing correction), classification and prediction (e.g. structural equation model, principal component analysis), and exposure-response surface estimation (e.g. Bayesian kernel machine regression) will be presented and compared.
Illustrative examples from simulated data will be used, and real-data examples in the field of chemicals mixtures will be presented. R code and simulated datasets will be made available to the participants. Basic knowledge of linear regression is recommended.
About the workshop leader
Andrea Bellavia is postdoctoral fellow in Epidemiology and Biostatistics, Harvard T.H. Chan School of Public Health and affiliated to the Biostatistics Team at the Department of Public Health Sciences.
His methodological research concerns the development of statistical methods for mediation and interaction analysis, for chemical mixtures, and for time-to-event outcomes.
Bellavias current empirical interests range from health disparity research to the areas of perinatal, environmental, and psychiatric epidemiology.
To attend the workshop, register by May 18 to firstname.lastname@example.orgContact person: Sara Fritzell