Registration is required (limited places available; please see the registration tab). It is no longer possible to register via the Webshop for a pre-conference workshop. To register for a workshop, please email ISEEyoung2015Utrecht@uu.nl. Limited places (around 20 per workshop) are offered on a first-come, first-served basis. The fee for a workshop is 10€. Please bring 10€ in cash to the Workshop on the date of the workshop, 1 November.
Workshops will be held at the Nieuw Gildestein building (close to the main conference venue), Yalelaan 2, Utrecht Science Park, on Sunday, 1 November 2015 from 13:00-17:00. Registration fee includes refreshments but not lunch.
(1) Introduction to land use regression modelling: Exposure assessment for epidemiological studies of long-term exposure to ambient air pollution is challenging because of substantial small-scale spatial variation. In large population studies, it is impossible to assess individual exposure to ambient air pollution by performing personal measurements. Instead, air pollution exposure modeling is used. Land-use regression modeling, combining air pollution monitoring and modeling using Geographic Information System (GIS) data is becoming increasingly popular. More information on the workshop can be found here.
(2) Statistical strategies for multi-pollutant modelling: Analyzing mixtures of exposures is a recognized challenge in environmental epidemiology. In this workshop, we will introduce several statistical approaches for multi-pollutant modeling and modeling of exposomics data; models which can simultaneously model multiple, potentially correlated, exposures in one model, including multivariate and variable selection approaches such as sparse partial least squares regression and penalized lasso regression. These modelling approaches generally yield less biased exposure-outcome effect estimates and improved selection accuracy (fewer false positive discoveries) than conventional single-pollutant linear and logistic regression modelling. More information on the workshop can be found here.
The workshop "Statistical strategies for multi-pollutant modelling" is full. Note that a course covering similar material will be held at Imperial College London in December.
(3) Understanding bias through directed acyclic graphs: The observational nature of many (environmental) epidemiological studies, make these studies susceptible to different forms of bias. Some of these biases can be controlled for in the analysis of a study, while others may actually be induced by (incorrectly) controlling for particular variables. Directed acyclic graphs (DAGs) provide an intuitive way of representing the ideas one has about the causal structure (s)he tries to unravel. DAGs can help to understand the nature of bias, but also how to control for it. In this workshop, the ideas behind DAGs will be discussed, and what the added value of such graphs can be in the design and analysis of an epidemiological study will be shown. More information on the workshop can be found here.