Susie Bayarri Lecture

The Susie Bayarri Lecture is delivered at the ISBA World Meetings by an outstanding young researcher under 35 years of age. The lecture is supported by an endowment fund started in November 2014 by Susie’s colleagues and friends, Jim Berger’s family, and the Duke Department of Statistical Science, who contributed with generous donations. The lecture is named to honor the memory of lifetime member, M.J. Susie Bayarri.

M.J. (“Susie”) Bayarri

Susie Bayarri (2014)Susie Bayarri passed away at age 57 on August 19, 2014 in Valencia, Spain, after an eighteen month battle with a brain tumor, during most of which she remained highly active professionally. Susie was a past president of ISBA, and one of the most prominent Bayesian statisticians in the world.  She was born on September 16, 1956 in Valencia, Spain. She studied at the University of Valencia, receiving Masters and Ph.D. degrees in mathematics in 1979 and 1984, respectively, with theses on Bayesian statistics under the direction of Jose Bernardo.  She started her career in 1978 as an Assistant Professor in the Department of Statistics and Operations Research at the University of Valencia, becoming Full Professor in 1998 and remaining at the department for the rest of her career.  As one of the few early female Bayesian statisticians, Susie served as a role model for many students and younger researchers (both male and female). She instilled her passion for research in the five Masters and seven PhD students she directed, most of whom have gone on to have prominent research careers of their own.

Susie spent the 1985-1986 academic year at Carnegie-Mellon on a Fulbright fellowship, and significant time there during the ensuing years, doing wonderful research on selection models and foundations with Morrie Degroot.  She decided to make visits to the United States a permanent part of her schedule. She would visit the US from August through December each year, focusing on research, while concentrating her teaching in Valencia from January through June. Her US visits were first at Carnegie Mellon University, then Purdue University (also spending the entire year 1994 as a visiting professor) and, from 1998 onward, at Duke University and the Statistical and Applied Mathematical Sciences Institute (SAMSI).

Susie made major contributions to both the theory and methodology of Bayesian statistics, helping it become the prominent part of the scientific landscape that it is today. There is almost no area of Bayesian research in which Susie was not involved, but her continuing long-term interests (spanning at least a decade) were selection models and weighted distributions; objective Bayesian methods; Bayesian analysis of queueing systems; Bayesian robustness; model criticism and p-values; model uncertainty and multiple compa- risons; and calibration and validation of complex computer models.

Susie was author or editor of five books and nearly 70 scientific research papers. For an introduction to the breadth and profundity of Susie’s work, one could start with her three papers that won awards: “Statistical inverse analysis for a network micro simulator” won the 2006 Frank Wilcoxon Award for the best applied paper published in Technometrics in 2005; “A framework for validation of computer models” won the 2008 Jack Youden Prize given to the best expository paper published in Technometrics in 2007; and “Criteria for Bayesian model choice with application to variable selection” was selected for presentation at the Special Annals of Statistics Invited Session at the Joint Statistical Meetings in 2014. For a more philosophical/critical tone, try her discussion papers “P-values for composite null models” in the Journal of the American Statistical Association in 2000 and “Bayesian checking of the second level of hierarchical models” in Statistical Science in 2007.

She was also highly active in major interdisciplinary collaborations including pharmaceutical research and risk assessment. As one recent example of the latter, she worked with mathematicians, engineers and volcanologists to develop a quantitative methodology for assessing the risk associated with catastrophic pyroclastic flows from volcanoes. This pushed the limits of what Bayesian methodology could do and would have been impossible without it.

Susie had major leadership roles in statistical societies. In addition to serving as President of ISBA in 1998, she was Presidenta de la Sociedad Español de Biometría from 2001-2003.  Susie also served on thirteen editorial boards, including being Coordinating Editor of the Journal of Statistical Planning and Inference from 2001 – 2007, and was an organizer of 38 international meetings and conferences.

Susie was a central figure to the development of statistics in Spain and the world. Her extensive funding (as principal investigator) by the Ministry for Science and Education in Spain provided an umbrella for dozens of researchers to collaborate and develop their own research careers in epidemiology, biostatistics, spatial methodology, survival analysis, and foundations. Susie was also central to the research project “Consolider i-MATH” from 2006-2012, which involved over 300 research teams from Spanish Universities. One of her last leadership roles in Spain was serving as the principal investigator of the BIOSTATNET project, a web-based collaboration of 180 investigators at universities inside and outside Spain and at biomedical institutions. Her international stature was reflected by her also being principal investigator of major grants from the European Community and the US National Science Foundation.  Susie received numerous honors during her career, including being elected Fellow of the American Statistical Association in 1997, elected to the International Statistical Institute in 1997, elected Fellow of the Institute of Mathematical Statistics in 2008 and, in a bittersweet ceremony during the World Conference of ISBA in July, 2014, was in the first elected class of ISBA Fellows. The outpouring of admiration and love for her that was expressed by the attendees at the conference was inspiring.

Susie was an inveterate traveler, and loved the explosion of Bayesian conferences that happened around the world in recent decades. She would often be one of those who would close down the bar or the dance floor at the conferences.  Those who had the privilege of knowing Susie will remember her not only as an outstanding statistician and tireless advocate for Bayesian statistics, but also as someone lively, funny, and extremely generous. She entered the hearts of everyone who came to know her, and will remain in our hearts forever.

Adapted from her obituary published in the ISBA Bulletin (Vol. 21, No. 3), written Jim Berger, Duke University.

James Scott (2016)

“Empirical Bayes and Penalized Likelihood”

2016 ISBA World Meeting
Sardinia, IT