A Tribute to Kathryn Chaloner

The following was written by SAMSI director Richard Smith and former Deputy Director, Snehalata Huzurbazar.

Kathryn Chaloner

Kathryn Chaloner

We were very saddened to learn of the death of Kathryn Chaloner on October 19.

One of us (Richard) knew Kathryn for over forty years, since we were both undergraduate students of Mathematics in Oxford University. In Oxford, each student is a member of one of more than thirty colleges, which are all mixed up by subject of study but which still were, at the time we entered, segregated by sex. But, of course, there were lots of social interactions between the men’s and women’s colleges. Also, at that time Statistics was very much a minority interest within the university Mathematics curriculum, so we were two of rather a small number of students attending the advanced Statistics lectures. Thus we got to know each other both professionally and socially.

At that time, it was still rather unusual for British students to come to the U.S. to pursue graduate study, but Kathryn and I both did, in my case at Cornell (where my classmates included Luke Tierney), while Kathryn went to Carnegie Mellon where she completed a Ph.D. under the late Morrie De Groot. By this time, Luke had become an assistant professor at Carnegie Mellon, and he and Kathryn were married shortly afterwards.

Kathryn and Luke then continued their careers at the University of Minnesota, where Kathryn’s first student was Merlise Clyde, now Chair of Statistical Science at Duke (there ought to be a theorem about social networks somewhere here) and Kathryn built up an outstanding reputation in several research fields including Bayesian design of experiments, Bayesian approaches to outliers and residuals, clinical trials, and models for the HIV epidemic.

In 2002, Kathryn and Luke both moved to the University of Iowa where Kathryn became Head of the Department of Biostatistics, a position she continued to hold until her death. As her career developed, she became more and more concerned with issues related to mentoring of students and junior faculty, especially women and minorities, activities for which she was awarded the prestigious Elizabeth L. Scott award at the 2014 Joint Statistical Meetings. Amongst numerous other activities that this award recognized, she was cited for her engagement with the National Alliance for Doctoral Studies in the Mathematical Sciences, which is a network of faculty mentors at mostly undergraduate colleges to identify underrepresented minorities with the potential to earn doctoral degrees. For the past several years, Kathryn was Chair of the Statistics Initiative within the Alliance. In this connection, she became an effective advocate for women and minorities (and more generally, for all US-born students) in graduate programs in Statistics.

Our SAMSI connection with her arose from this work; a connection which started when the other of us (Snehalata) read an article by Kathryn in the IMS Bulletin and immediately contacted her. A year and half later, last June, we hosted the workshop “Recruiting and Retaining Graduate Students in the Statistical Sciences and Applied Mathematics”, which was designed for graduate program directors and faculty seeking to increase numbers of U.S. students and underrepresented minorities in their graduate programs. Presentations included one from Kathryn herself on comparing diversity in graduate Statistics and Mathematics programs, and sessions on such themes as “transition from a minority undergraduate institution to a majority research institution” and “best practices for mentoring Ph.D. students.” There had been previous workshops aimed at improving diversity in Mathematics graduate programs, but this was the first such workshop targeted primarily at Statistics and Biostatistics programs, and, of course, the first at SAMSI. It would never have come about without Kathryn’s energy and persistence. SAMSI hopes to continue the effort in the future, for example by hosting further workshops of a similar nature.

Kathryn Chaloner was a brilliant researcher but one whose contributions to the profession went well beyond her research. We extend our sincere condolences to Luke, their children Graham and Patrick, and other members of her family.

Richard Smith and Snehalata Huzurbazar


SAMSI Postdoctoral Profile: Daniel Taylor-Rodriguez

Daniel and Natalia near a mountain range

Daniel and his wife, Natalia.

Daniel Taylor-Rodriguez is a new postdoctoral fellow at SAMSI and is participating in the Ecology program this year. He came from the University of Florida. His wife, Natalia, is still in Gainesville currently working on her Ph.D. in animal science.

Daniel grew up in the bustling metropolis of Bogota, Colombia. When he was 10, his father, who was working for IBM, was temporally transferred to White Plains, New York, so Daniel and his family moved there. After year and a half, Daniel’s family moved back to Colombia to Barranquilla on the coast. Daniel entered a bilingual school to continue his skills in English. Later, his family moved back to Bogota and he continued at a school that offered an International Baccalaureate (IB) program. Tony Cleaver, an inspiring IB economics teacher, strongly influenced Daniel’s decision to pursue a degree in economics.

Daniel studied economics at the Universidad Los Andes from 1998-2003, “I especially enjoyed the econometric courses; to me it made more sense to let data speak by itself,” commented Daniel. He then went on to get a specialization (similar to a professional Master’s degree in the United States) in statistics. At the same time, he started working as a consultant under his game theory professor Luis Jorge Ferro. In particular, they developed optimal incentive and penalty mechanisms in contracts for large-scale highway concession projects required by the National Concessions Institute of Colombia.

About a year later, a group of his friends from National University of Colombia who were wildlife veterinarians were starting up a wildlife conservation organization called Fundación Vida Silvestre Neotropical (FVSN). They invited Daniel to come work with them and he decided to help them out. Their first project was to help cattle ranchers identify what management practices reduced the risk of predation from panthers (pumas) using various quantification protocols. “I really enjoyed being in the field, talking to people, collecting data and making a meaningful contribution”, said Daniel. These findings informed local governments in the implementation of programs to mitigate predation risk in livestock ranches.

Although he continued to participate in FVSN as an honorary member, Daniel went to work as a quantitative risk analyst at Bancolombia, Colombia’s largest banking institution. There he developed models to quantify credit risk components. He also built in-house models to forecast macroeconomic series affecting risk behavior of client portfolio. Through Bancolombia he applied to a Fulbright Scholarship, which he was awarded to pursue a Master’s degree in statistics at the University of Florida, at the end of which he would have to return.

group of seven people from University of Florida

Daniel (lower right) along with former SAMSI postdoc Kenny Lopiano (lower row, center) at the University of Florida

While at the University of Florida, he studied under George Casella, who made him aware of the wonderful opportunity he would miss if he left without doctoral training. Daniel made arrangements with the bank’s directives to stay for the PhD; however, a semester before concluding his Master’s, a new CEO was appointed to the bank bringing along his own team. Due to the new conditions Daniel was requested to return to Colombia. Intent on pursuing his doctoral studies, he turned down this request. Fortunately enough the new bank’s directives reconsidered and decided that it was not their place to truncate Daniel’s academic aspirations. They terminated his contractual obligations with the bank, but left the doors open for him to come back if he ever wished to do so.

Daniel in the Gators stadium

Daniel at a Gators game.

He stayed at University of Florida to receive his Ph.D., but eager to learn more about ecology, he enrolled in the interdisciplinary ecology program with concentration in statistics. Daniel truly appreciates the strong bonds and collaborative approach to research that George Casella fostered within his working group. After Dr. Casella’s passing he received the continued and insightful supervision of Professors Linda Young and Nikolay Bliznyuk.

In the course of his studies, he contributed to projects with researchers in the Animal Sciences, Ophthalmology, and Horticulture Departments at UF, and took part in the NSF funded program IGERT Quantitative Spatial Ecology, Evolution, and Environment (QSE3). In his collaborations he worked jointly with scientists from diverse backgrounds to develop interdisciplinary solutions to important applied problems.

Daniel’s doctoral research focused on developing Bayesian procedures for variable selection with good frequentist properties, and adapting these methodologies to models widely used in population ecology. The Jaguar Corridor Initiative, a large-scale conservation effort lead by Panthera foundation, inspired Daniel’s work. Panthera is a nonprofit organization that helps large felines through scientific research and global conservation. The corridors established by Panthera across South and Central America, aim to ensure an active pathway that preserves the link between the last two jaguar populations in the world. A preliminary but crucial step in building the corridor in Colombian territory is to produce accurate density maps for jaguars and their prey, which can be obtained using models that account for imperfect detection. Surveys conducted to fit these models collect information about a large number of predictors. However, there is need to improve the methods used to identify the relevant predictors and to assess uncertainty in the model parameter estimates due to selection.

At SAMSI Daniel will study questions associated with the dynamics of infectious diseases, and will expand his work on population ecology, modeling the joint behavior of species sharing the same ecosystem. His two years at SAMSI will give him the opportunity to meet and collaborate with researchers from different backgrounds and to engage in new and exciting projects.

Daniel at a marathon

Daniel enjoys running.

Daniel also enjoys running, mountain biking, watching movies and, of course, spending time with Natalia.

Become a SAMSI Postdoc Fellow!

a collage of the postdocs from 2013-14

Our 2013-14 SAMSI postdocs.

SAMSI has a great opportunity for six people to become a postdoctoral fellow next year! Get the chance to collaborate, and mingle with top researchers in your field of interest.

You can apply for either of the two SAMSI Research Programs: Challenges in Computational Neuroscience (CCNS) and Statistics and Applied Mathematics in Forensic Science (Forensics). Appointments (up to 2 years) will begin in August 2015, and will offer competitive salaries, travel stipend and health insurance.

The Challenges in Computational Neuroscience (CCNS) program will develop mathematical and statistical methods for neuroscience applications to understand the underlying mechanisms that bridge multiple spatial and temporal scales, linking the activity of individual components (e.g., molecular biology, genetics, and neuron networks), and their interactions to the complex dynamic behavior of the brain and nervous system. Brain theory, modeling, and statistics will be essential to turn data into better understanding of the brain. The CCNS program will address the underlying methodological, theoretical, and computational challenges. Probability and statistics, dynamical systems, geometry, and computer science will be combined with respect to theory and in applications.

SAMSI’s program on Statistics and Applied Mathematics in Forensics is focused on strengthening the statistical and applied mathematical foundations of forensic science. Forensic science is fundamentally based upon statistical comparisons of the characteristics of materials left at a crime scene to characteristics of possible sources or suspects. These comparisons are often acknowledged by forensic scientists to be highly subjective. A series of reports by the National Research Council (NRC) has raised deep questions about major forms of forensic evidence and has made a clear case for heeding statistical underpinnings for forensic procedures. Evidence from a crime include fingerprints, patterns and impressions (footprints and tire tracks), toolmarks and firearms, hair, fibers, documents, paints and coatings, bloodstains, and fire debris.

To apply, go to mathjobs.org, SAMSIPD2015, Job #6133. In your cover letter, please indicate which of the two research programs you are interested in. The deadline for full consideration is December 15, 2014, although later applications will be considered as resources permit.

Criteria for selection of SAMSI Postdoctoral Fellows include demonstrated research ability in statistical and/or applied mathematical sciences, computational skills along with good verbal and written communication abilities, and finally, a strong interest in the SAMSI program areas.

SAMSI is an AA/equal opportunity employer All qualified applicants are encouraged to apply, especially women and members of minority groups.