The overall aim of this course is to present approaches to statistical design and analysis that enable researchers to design more efficient epidemiological studies and to better utilize available data from well-defined cohorts (such as those deriving from national health registers, electronic medical records or clinical studies).
The course will compare and contrast different sampling designs and the various parameter estimates they can yield by careful “reconstruction” of the underlying cohort or extensions to the regression models used. In particular, the course will show how variations of the case-control design can produce efficient and unbiased estimates of the hazard ratio and other quantities, and how the concepts of matching from classical epidemiological studies can be extended to studies of continuous outcomes
The course will demonstrate the application of these methods in designing studies to make efficient use of costly data and to conduct more flexible and informative analysis.
Lectures will be interspersed with tutorials consisting of exercises, “journal club” sessions and workshops. In the workshops, participants may (i) develop and refine a study design to address a clinical/epidemiological research question or (ii) implement some method(s) on their own data in a supervised laboratory session.
The University of Milano-Bicocca might provide one scholarship for participation of a PhD student. To apply, please, upload your CV and a short letter (max 200 words) describing your PhD project and how it relates to the topic of the course by 15 April 2018.
Αναλυτικές πληροφορίες - Δηλώσεις συμμετοχής: summerschoolbicocca.com