Tom Britton: Pitfalls when analysing emerging epidemic outbreaks: a case study of Ebola

New infectious disease outbreaks have great impact on different communities worldwide, as most recently manifested by the Ebola outbreak. An important statistical task in such situations is to (quickly) predict the future scenario both without and with preventive measures put into place. In the talk we describe how such analyses were done for the Ebola outbreak and show some potential biases that may arise in these situations. The main reason for biases occurring comes from events with "short waiting times" being over-represented early on in an outbreak, which can be explained using branching process theory. (Joint ongoing work with Gianpaolo Scalia Tomba).