Medical and Social Statistics
Design and Analysis of Clinical Trials for Complex Interventions (Rachel Appleton)
01/10/2022 → 30/09/2024
Statistical Methods for Health Technology Assessment (Joseph Price)
01/10/2022 → 30/09/2024
Statistical Methods for Evaluating Quality and Length of Life in Post-stroke Patients and Other Applications (Alice Fletcher)
01/10/2021 → 30/09/2023
Input Uncertainty Quantification for Large-Scale Simulation Models
01/10/2019 → 31/03/2023
Illness-Death Modelling on Oncology Clinical Trials: Leveraging the Semi-Markov Assumption
01/07/2019 → 23/08/2019
Improving the design and analysis of trials for efficacy and mechanisms evaluation: workshop and training days
01/07/2018 → 31/10/2019
EU: TRAFFDAT - study on the data analysis of trafficking in human beings
11/06/2018 → 10/10/2018
Secondary School Choice and Academic Attainment
01/01/2017 → 30/09/2019
Testing New Models of Care: An Evaluation of the Lancashire and Cumbria Innovation Alliance NHS Test Bed
01/03/2016 → 30/06/2018
The Medical and Social Statistics group specialises in methodology for clinical trials, statistical epidemiology, quantitative criminology and social research.
Our main areas of interest are:
- Statistical Epidemiology
- Methodology for the design and analysis of clinical trials
- Social research
- Quantitative Criminology
Our methodological research is often conducted in close collaboration with companies and public sector research institutions.
Adaptive designs for clinical trials
Our research in this area focuses on multiple testing procedures that ensure control of the proportion of false positive findings. A particular area of focus is multi-arm studies in which several active treatments are compared to a common control within one study. Additional work is looking at estimation at the end of such studies as well as simultaneous confidence intervals.
This work concerns first-in-man studies of new drugs or combinations of drugs. Past research has considered the design of Phase I oncology trials, where subjects are patients and responses are toxicities, and healthy volunteer studies, where responses are pharmacokinetic and pharmacodynamic assessments. Work is focused on developing novel Bayesian decision procedures for determining the optimal dose to be taken forward for subsequent evaluations.
Methods for time to event data
We work on the development of methods for event history data with a particular focus on the analysis of multi-state survival data with intermittent and possibly informative observation times. Further work considers techniques for joint modelling of longitudinal processes in conjunction with time to event data.
Our social research focuses on developing methods for the analysis of social data and applying those methods to social and epidemiological problems in the health and social sciences. Application areas focus both on health (such as drug use and issues of escalation); foetal development and movement; psychology (child development) and crime (studies on various aspects of interpersonal violence such as homicide, domestic violence, threats to kill; modern slavery and human trafficking). Members of the group often work in close collaboration with other departments at Lancaster, and with police and government organisations.
Statistical areas of development include missing data, new latent class analysis methods, statistical disclosure control, instrumental variable studies using Mendelian randomisation, high-frequency count data in surveys; score building through log-linear modelling; estimation of hidden populations; longitudinal and survival analysis methods, and confidence interval estimation for high dimensional data.
Current work is on facial recognition, and specialists' ability to correctly identify suspects from CCTV images. We are participating in a European effort to bring advanced statistical methods to forensic trace evidence to the wider forensic community, the application of statistics to the analysis of lead from bullets, and spectral methods for the identification and provenance of inks.