Paul Gustafson, SSC Gold Medalist 2020

Paul Gustafson
SSC Gold Medalist
2020

This year’s recipient of the Gold Medal of the Statistical Society of Canada is Paul Gustafson. This prestigious award is bestowed upon a person who has made outstanding contributions to statistics, or to probability, either to mathematical developments or in applied work. It is intended to honor current leaders in their field.

 

Paul was born in Birmingham, U.K. in 1968, albeit to Saskatchewanian parents. In 1978 his family relocated to Prince Rupert, BC where Paul completed high school. He enjoyed his BSc (Mathematics) and MSc (Statistics) at UBC so much that, after completing his PhD at Carnegie Mellon University in 1994, he returned to the campus, first as a postdoctoral fellow and then as an Assistant Professor.

 

During his MSc studies Paul became rather enamoured with Bayes’ theorem. He pursued this interest in his PhD thesis, written under the supervision of Larry Wasserman, in which he proposed formal schemes to assess the sensitivity of posterior inferences to the choice of prior distribution. Bayesian methods have remained front and center in Paul’s work ever since. His record demonstrates he is a wonderful scientist as well as an exceptional statistical scientist, with landmark contributions to understanding what inference is or is not possible in complex settings due to design, missing or mismeasured data, and the other ills plaguing real studies. For instance, his 2005 Statistical Science paper on model expansion advocates use of a scientifically “honest” model rather than conveniently changing the model to suit inferential limitations. Currently he is interested in Bayesian approaches within the domains of causal inference, evidence synthesis, measurement error and partial identification.  

 

Paul’s publications remarkably range from deep theoretical insights to important applied work, often in the same paper. Much of his research program is inspired by biostatistical and epidemiological applications, and he frequently collaborates with health scientists on a range of medical conditions. His 2014 American Journal of Epidemiology collaborative article on multiple sclerosis was selected by the editors as one of the best in field that year.

 

Paul’s two books, Measurement Error and Misclassification in Statistics and Epidemiology: Impacts and Bayesian Adjustments (2004) and Bayesian Inference in Partially Identified Models: Exploring the Limits of Limited Data (2015) survey a large portion of his research contributions and disseminates it to a broader audience. Both books reflect Paul’s interest in understanding what combinations of modelling assumptions and prior information will yield inferences that are usefully narrow, rather than uselessly wide.

 

At UBC, Paul has been an important mentor. He has supervised or co-supervised 13 PhD students, 24 MSc students and four postdoctoral fellows; many have gone on to academic positions. He has also taught at all levels. His considerable service to the department includes two stints as Acting Head and the current Headship since 2019. Paul also played a pivotal role as founding Co-Director of UBC’s Master of Data Science program, which welcomed its first cohort of students in 2016 and is now heavily subscribed.

 

Beyond UBC, Paul has taken on a variety of editorial and service commitments, nationally and internationally. He was Editor of The Canadian Journal of Statistics (2007-2009) and in 2014 he became the inaugural Special Editor for Statistical Methods at the journal, Epidemiology, a role he still holds.  Paul is a former chair of NSERC’s GSC 14 committee for grants in the statistical sciences and he was the Program Chair for SSC 2015 held at Dalhousie University.  

 

Paul’s work has previously been recognized with the CRM-SSC Prize in 2008 and an ASA Fellowship in 2011. Additionally, he is a two-time recipient of an NSERC Discovery Grant Accelerator Supplement.

 

Paul is enduringly grateful for support and foolishness-filtering from his wife, public health physician extraordinaire and occasional collaborator, Reka Gustafson. Whenever possible Paul and Reka can be spotted on their bicycles, either locally or in far-off lands.   

 

For better or worse, Paul and Reka seem to have passed on their fondness for west coast life and UBC education to the next generation. One son has just completed an engineering degree, the other is midway through a Computer Science and Physics program and their daughter is about to embark on her UBC adventure this fall.

 

The citation for the award reads: 

To Paul Gustafson, in recognition of his seminal contributions to the foundations of inference, particularly Bayesian robustness, sensitivity analysis, and model identifiability; for exceptional advances to recognize the limitations of observational studies due to mismeasurement and other inherent obstacles, along with effective statistical tools to counter those limitations; for fundamental contributions to methodology for biomedical applications and epidemiology; and for his outstanding record of mentorship and service to the statistical science community in Canada and internationally.

Will Welch was primarily responsible for producing this material.