Matthew's research interests centre on two main areas. He has a longstanding commitment to understanding how to improve health and prevent injury through modifying the behaviour of individuals. This research links public health with psychological and behavioural research. He also maintain a strong interest in statistical methods for applied research, particularly observational studies in which statistical adjustment is required to account for the underlying complexity of the data. Matthew's program of research focuses on three broad areas. These are: - The epidemiology of suicide: describing trends over time in the lethality of the major suicide methods; understanding the substitution of one suicide method for another over successive suicide attempts; geographic and temporal variation in suicide rates; interventions to reduce suicide by jumping at hotspots; and media reporting of suicide. - The relationship between quality of care and medico-legal risk: quantifying the risk of complaints, disciplinary action and malpractice claims against medical practitioners; understanding the link between quality of care and negligence litigation in nursing homes. - Statistical methods for predicting events where the underlying risk is changing over time: using parametric and flexible parametric survival methods to model recurrent events; methods of predicting survival adjusted for covariates; and using Poisson regression models for meta-analysis.