Data Science Research Group

Statistical Methodology theme

Statistics plays a key role in bridging research between mathematics and applied data science within the department.

Research problems in statistics are relevant to a range of areas in mathematics, from measure theory to analysis and algebra, and are linked with the statistical and probabilistic foundations of Data Science and Operational Research.

Statistical methodology theme members work on a broad range research areas, including Bayesian statistics, longitudinal and survival analysis, causal inference and applied probability.

Areas of expertise

Members of this research theme have several key areas of expertise within the field:

  • Methodologies in computationally intensive Bayesian Modelling and Monte Carlo methods - Theme member Dr Dai is particularly interested in the theoretical aspects of Bayesian computation, particularly in the exact Monte Carlo algorithms, such as coupling from the past and its application of graphical models and complex queuing models, and path space rejection sampling for diffusions and its application on modern Bayesian big-data inference (Bayesian Fusion methods for unification of distributed statistical analysis).
  • Nonparametric and semiparametric methods for survival data - All theme members are interested in survival analysis, particularly in nonparametric/semiparametric methods for multivariate length-biased and censored data, nonparametric bivariate survival analysis, couplas, cure rate models, joint modelling of survival events and longitudinal data, covariance modelling.
  • Causal inference - Theme member Dr Bao has interests on instrumental variable methods, including Mendelian Randomization. She also works on structure mean model for longitudinal data analysis.

Theme members are also interested in the practical problems proposed by industrial partners. An important goal for the theme is to build on our existing collaborations with business partners via KTP projects (for example the KTP with Ocado), as well as links with local councils via the Catalyst project funded by HEFCE.

Our department has always understood the importance of statistics and methodology. Our late colleague, Professor George Alfred Barnard, was President of the Royal Statistical Society during his time in our department, and was awarded the Society’s Guy Medal in Gold in 1975.

We will continue to build on this dedication to statistical methodology and its impacts, by engaging in further collaborations with researchers in other areas, such as biology, sociology and computer science (for example the BIAS project currently funded by ESRC).

Recent publications




A photo of a woman, with glasses and dark hair, working at a computer, with piles of paper and folders on her desk on her left.
Project: BIAS - Responsible AI for Labour Market Equality

The Department of Mathematical Sciences is collaborating with Lancaster University in this ESRC funded project that will create new methodologies for the creation of responsible and trustworthy algorithms used by companies for job applicant selection, reducing the bias shown by existing algorithms.

Read more about the BIAS project