I am the Head of Statistics within the School of Mathematics and Statistics. My research aligns with the Biostatistics and Ecology research group with a focus on statistical methods relevant to epidemiological, psychological and population health data. My main application area is in road safety and injury research, although I routinely collaborate with cancer, psychological and cardiovascular researchers.

Research Goals

• Conducting reasonable statistical analyes using minimal information
• Improving methods for assessing population-based interventions
• Developing methods for estimating and accounting for regression to the mean
• Estimating survivability from crashes involving motor vehicles, micromobility users and pedestrians

Research in Detail

There are many statistical challenges in quasi-experimental designs due to lack of randomisation. This makes it difficult to assess the impact of population-based interventions due to unmeasured confounding. Some examples of interest are assessing the effects of bicycle helmet legislation and measures to limit access to guns. Interrupted time series, with or without a control, offer an improvement over other methods. However, little work has been done in this area and there is little standardization of analytic approaches across various fields such that varying analyses can lead to disparate conclusions using the same data source.

It is well known that statistical significance is a function of sample size. This is problematic because important effects can go unnoticed in studies with small sample sizes, and unimportant effects can give small p-values from large studies. Effect sizes can complement significance testing by removing the influence of sample size. A lot of work on this topic can be found in psychological research but little has been done on epidemiological measures like the odds ratio, relative risk or hazards ratio.