Department of Statistics

Thomas Lumley

Thomas Lumley

Job title: Professor
Phone: +64 9 373 7599 ext 83785
Office: 303.325 Science Centre


Thomas Lumley attended Monash University (B.Sc.(Hons) in Pure Mathematics), the University of Oxford (M.Sc. in Applied Statistics) and the University of Washington, Seattle (PhD in Biostatistics). He spent twelve years on the faculty of the Department of Biostatistics at the University of Washington, and then moved to Auckland in 2010. He is still an Affiliate Professor at the University of Washington. His research interests include semiparametric models, survey sampling, statistical computing, foundations of statistics, and whatever methodological problems his medical collaborators come up with -- currently, the design and analysis of a DNA resequencing study.

Further Information

  • The survey package for R is a fairly comprehensive system for analysis of data from complex probability samples.
  • I have written a book on survey analysis, published by Wiley.
  • My CV. (Eventually the department will provide an automated list of recent papers below, but it's not quite right yet.)
  • Potential student projects.
    • Likelihood of the empirical distribution function as an approach to Bayesian analysis of survey data.
    • Survey software: design and implementation of various things for the survey package in R. Graphics, probability distributions, regression models, multivariate methods...
    • Survey integration: combining clinical data from NHANES and self-report data from BRFSS for better estimation of trends
    • Software for teaching:design and implementation of interface to R for teaching introductory biostatistics
    • How many categories? Data on, say, hospital visits, can be categorized coarsely (eg lung problem) or much more finely (eg infection by pencillin-resistant Strep. pneumoniae). When the categories are too fine the sample size in each one is too small to see patterns; when they are too coarse, the patterns are masked by events that don't really belong. The idea is to build a tree structure that uses all levels of categorization simultaneously in a Bayesian model.


Please give us your feedback or ask us a question

This message is...

My feedback or question is...

My email address is...

(Only if you need a reply)

A to Z Directory | Site map | Accessibility | Copyright | Privacy | Disclaimer | Feedback on this page