The human and financial costs associated with medical research motivate efforts to make the best possible use of the data collected. Current medical research requires the analysis of data that can be very complex, with much of it derived from recording information for individuals at various points in time. Such data are of interest to many of BSU’s scientists, and pose significant methodological challenges. We aim to apply available methodology and, where necessary, develop new statistical methodology to analyse such data. A particular emphasis is on modelling the relationships between different types of information collected for individuals and on methods to deal with the situation when some potentially desirable information is missing.
We have major involvement with substantive medical research in areas such as ageing, rheumatology, lung function, heart disease, cystic fibrosis and epidemics. This involvement motivates much of our work. In light of this close link with specific applications, key components of our work are: 1) specification of appropriate questions of interest, 2) consideration of general principles and scope for the translation and adaptation of methods across different application areas and 3) development of new methodology when necessary. Thus, methodology developed within one specific application will be typically applied to another to establish generic applicability and will also be used to inform design and sampling aspects of studies.