Is it a probability? Is it a rate? It’s event data! A look at some different ways of modelling events on R and how to make sense of the output

Is it a probability? Is it a rate? It’s event data! A look at some different ways of modelling events on R and how to make sense of the output

Researchers are sometimes interested in exploring associations between the occurrence of some event (think categorical health outcomes, classification of species behaviour, etc.) and some other measured variables or groupings. Modelling approaches, such as generalised linear models, can answer these questions by accommodating the outcome variable with a variety of categorical distributions. While the choice of distribution typically depends on the properties of your outcome variable, we sometimes have a little more flexibility for event data. There are a variety of closely related ways they can be considered (and encoded): did the event occur? (as a binary variable); how often did the event occur? (as an integer count); when/in whom/in what did the event occur? (times/subjects/trials/experiments at/in which the events occurred). This is handy since we can choose a distribution for which the model assumptions are met and, with a little wrangling, can often reframe the model output to specifically address the research question at hand – even when the natural interpretation might not. In this seminar I will demonstrate such an approach to analysing event data, via examples on R.

Date: 20 November 2025 

Time: 2.30pm - 3.30pm 

Speaker: Elliot Dovers, Statistical Consultant, UNSW Stats Central 

Delivery Mode: Online

Location: AGSM LG06 - Our team will be there, please feel free to join us!

FAQs

How can I contact the organiser with any questions?

Please send us an email at: stats.central@unsw.edu.au