Dr Andrew Smith
Biography | I am an associate professor of statistics: a researcher in applied statistics working in health and wellbeing, and a lecturer in statistics and research methods. My research focuses on the development and application of smart statistical methods. Methods that are used in challenging modelling situations, such as working across the human epigenome, dissecting the roles of determinants of health, or simultaneously choosing and testing scientific hypotheses. Because I work in health and wellbeing, my research needs to be interdisciplinary and collaborative, with a global reach. I focus on adding value by making sure our methods are suitable for these challenging situations and making sure we can clearly interpret and explain their numerical results to those on whom our research has impact. This frequently involves developing new methods. I lead UWE Bristol's collaboration with The Dunn Lab, part of a five-year project funded by the US National Institutes of Health (total $3.7m), investigating sensitive periods in development and how stress is embedded in epigenetics. I am a head of the Mathematics and Statistics Research Group. My work on cumulative live-birth rates over a course of IVF treatment has informed changes to Scottish Government policy regarding NHS provision of IVF and planning of resumption of services after the Covid-19 lockdown. My research informs my teaching. I see my position as ensuring students are ready and able to process the increasing amounts of data they will encounter in their futures. I have been a Fellow of the Higher Education Academy since 2016. I joined UWE Bristol in 2015, having previously been a postdoctoral researcher at the University of Bristol, at which I completed my PhD in 2011. |
---|---|
Research Interests | Statistical methods used in observational settings in health and wellbeing, with additional expertise in: The structured lifecourse modelling approach (SLCMA), which explores the time-dependent relationship between determinants and measures of health as a person grows and ages. Requires simultaneously choosing and testing scientific hypotheses (post-selective inference). Causal analysis methods, for dissecting the roles of determinants of health. High-throughput applications, such as those across the human epigenome. Prognostic modelling of in vitro fertilisation (IVF) and assessment of best practice. Cluster analysis and principal components analysis of dietary patterns. |