Dr. Steele's primary statistical research interests are in the analysis of missing data and model selection using computationally intensive Bayesian approaches. Dr. Steele also has worked on problems in applying statisics to other fields, particularly applications in medical and population health research. He is a member of the Canadian Scleroderma Research Group (CSRG) and has collaborated closely with several physicians on problems related to analyzing data from the CSRG patient registry. He has also worked on epidemiology methodological problems relating to meta-analysis and the analysis of sports injury data.
Steele, R.J., Wang, N., and Raftery, A.E. Inference from Multiple Imputation for Missing Data Using Mixtures of Normals, Statistical Methodology (7) 351–365 (2010).
Hudson, M., Impens, A., Baron, M., Siebold, J.R., Thombs, B.D., Walker, J.G., CSRG, and Steele, R. Discordance between Patient and Physician Assessments of Disease Severity in Systemic Sclerosis, Journal of Rheumatology. To appear. (2010).
Shrier, I., Steele, R.J., Hanley, J., Rich, B. Analyses of injury count data: some dos and some donts, American Journal of Epidemiology (170) pp. 1307–15 (2009).