David Matthews

Professor

David MatthewsContact Information:
David Matthews

Research interests

Professor Matthews' research interests encompass the fields of biostatistics, quality improvement-especially in relation to health care-and statistical consulting. He is particularly concerned with finding effective ways to communicate statistical ideas and results to clinical researchers.

In virtually every journal reporting the results of medical research, investigators rely on statistical tools to make sense of the data they have collected during the study. Biostatistics represents a collection of diverse statistical methods that are used to analyze such data, turning the variable measurements collected from study participants into key insights that reveal the essential features hidden in the study data.

Sometimes graphical tools are all that's required to summarize appropriately the results of a research study. More often, however, the investigators find it essential to collaborate with a biostatistician. This collaboration is usually much more productive if the medical researchers have at least a qualitative understanding of the tools that their statistically knowledgeable partner is using to summarize the study results.

For example, Professor Matthews and a colleague demonstrated that the tools of survival analysis were also applicable to studies of the intervals between attempts to donate blood. Thus, for possibly the first time, blood collection agencies had at their disposal quantitative methods for studying what features were characteristic of regular donors. These insights could, in turn, help to ensure a safer, more reliable supply of blood and blood products.

As clinical research becomes more complex and more costly, it has become increasingly important for biostatisticians and clinical researchers to work together right from the initial stages of any proposed project. Close collaboration is the key to ensuring that statistical principles - which may prove crucial in analyzing the study data - receive careful consideration during the design phase as well. Much of Professor Matthews' research in recent years has focused on developing appropriate statistical methods for analyzing data from carefully executed, well-designed studies.

Education/biography

Professor Matthews' interest in medical statistics was sparked initially by a leave of absence that he spent at the Fred Hutchison Cancer Research Center in Seattle, Washington in 1980-81. This exposure to the medical research environment led, ultimately, to co-authorship of a book entitled Using and Understanding Medical Statistics that has set the tone for much of the research and teaching he has done since then. For example, since 1997 he has prepared and presented various short courses on topics in medical statistics to doctoral students in public health in Finland, Thailand, and Sweden. Also, two of his sabbatical leaves were spent visiting the Health Services Research Group at the University of Newcastle in Australia.

Professor Matthews is a Fellow of the Royal Statistical Society, as well as a member of the Statistical Society of Canada, the Statistical Society of Australia, and the International Biometric Society.

Selected publications

  • Dong, B. and Matthews, D.E.  (2012).  "Empirical Likelihood for Cumulative Hazard Ratio Estimation with Covariate Adjustment."  Biometrics, in press.
  • Jurgensen, H. and Matthews, D.E. (2010). "Entropy and Higher Moments of Information." Journal of Universal Computer Science, 16, 749-794.
  • Matthews, D.E. and Farewell, V.T.  Using and Understanding Medical Statistics, 4th completely revised and enlarged edition. Basel:  S. Karger AG 2007, 342 pp.
  • Matthews, D.E.  (2005).  "Multiple Linear Regression." in Encyclopedia of Biostatistics, 5, 2nd edition,  P. Armitage and T. Colton (eds.), pp. 3428-3441.  London:  Wiley.
  • Matthews, D.E. and Farewell, V.T.  Using and Understanding Medical Statistics, 3rd Revised Edition. Tokyo:  Asakura 2005, 237 pp.
  • Lim, H., Sun, J. and Matthews, D.E.  (2002).  "Maximum Likelihood Estimation of a Survival Function with a Change-Point for Truncated and Interval-censored Data."  Statistics in Medicine, 21, 743-752.
  • Stevens, J.K., Dunn, J.M., Capatos, D., and Matthews, D.E.  "Method for Testing for Mutations in DNA from a Patient Sample." U.S. Pat. No. 6,270,963 (August 7, 2001).
  • Rai, S.N. and Matthews, D.E.  (1997).  "Discrete Scale Models for Survival/Sacrifice Experiments."  Applied Statistics, 46, 93-109.
  • James, R.C. and Matthews, D.E.  (1996).  "Analysis of Blood Donor Return Behaviour Using Survival Regression Methods."  Transfusion Medicine, 6, 21-30.
  • Gentleman, R.C., Hamada, M.S., Matthews, D.E. and Wilson, A.R.  (1994). "Statistical Quality Control of HIV-1 ELISA Test Performance." Journal of the American Statistical Association, 89, 1200-1208.
  • Matthews, D.E., Farewell, V.T. and Pyke, R. (1985).  "Asymptotic Score-Statistic Processes and Tests for Constant Hazard Against a Change-Point Alternative."  Annals of Statistics, 13,  583-591.
Affiliation: 
University of Waterloo
Contact information: