Department of Statistical Science

Statistical Science

• Home
• Department Information
• Staff
• Current Students
• Prospective Undergraduates
• Prospective Postgraduates
• Research
• Contact Details
• Faculty of Mathematical and Physical Sciences
• Centre for Computational Statistics and Machine Learning
• Biostatistics Network
• Site Map

Tim Morris

PhD Topic	Practical use of multiple imputation
Supervisors	Ian White and Patrick Royston
Email(*)	timothy
Themes	Biostatistics

* The full email address is name@stats.ucl.ac.uk

Biographical Details

I am a PhD student with the Department of Statistical Science, based at the MRC Clinical Trials Unit (London) and the MRC Biostatistics Unit (Cambridge). After my MSc at LSHTM, I worked for three years at the MRC Clinical Trials Unit in statistics applied to research in orthopaedics, trauma, transfusion medicine, paediatrics, and stroke.

My main interests are missing data for prognostic models and treatment allocation methods for randomised trials.

PhD Research

Missing data are a pervasive problem in medical research. In clinical trials, we fail to follow up some patients, and we worry that this can cause bias. In observational studies, we fail to record all the data we need to predict disease outcomes, and we end up with data sets that are hard to analyse well.

Multiple imputation is a popular and flexible technique for handling missing data. The missing values are imputed stochastically in a way that reflects the uncertainty about the missing data. This is repeated more than once and the estimates from each individual `completed' dataset are combined using a set of rules known as Rubin's rules.

This project aims to further develop the methods used to impute missing data. Although these are well understood in some simple settings, they are much less well understood in the sort of practical settings in which multiple imputation is being applied, with data sets containing tens or hundreds of variables, mis-specified imputation models, clustered data, and complex analysis models.

Selected publications

• Clint S, Morris TP, Shaw OM, Oddy MJ, Rudge B, Barry M (2010) The reliability and variation of measurements of the os calcis angles in children. Journal of Bone and Joint Surgery, 92-B(4): 571-575.
• McGregor AH, Doré CJ, Morris TP, Morris S, Jamrozik K (2010) Function after spinal treatment, exercise and rehabilitation (FASTER): improving the functional outcome of spinal surgery. BMC Musculoskeletal Disorders, 11(17): 1-8.
• Malik AK, Morris T, Chou D, Sorene E, Taylor E (2009) Clinical testing of ulnar collateral ligament injuries of the thumb. Journal of Hand Surgery, 34E(3): 364-366.
• Gikas PD, Morris T, Carrington R, Skinner J, Bentley G, Briggs T (2009) A correlation between the timing of biopsy after autologous chondrocyte implantation and the histological appearance, 91-B: 1172-1177.