This Year’s Award
We are extremely pleased to announce that Dr. Joseph W. Hogan, Carole and Lawrence Sirovich Professor of Public Health, Professor of Biostatistics, and Deputy Director of the Data Science Initiative at Brown University, will be the recipient of the 2020 Lagakos Distinguished Alumni Award.
Please join us for his virtual lecture on:
Thursday, October 15, 2020 | 1-2pm
See seminar & events page for zoom info
What’s in a Model?
We’ve all been hearing a lot about models lately. Both inside and outside of academia, there has been intense focus on what models can and cannot be relied upon to deliver. The constant flow of complex, publicly available data has placed a premium on rigorous modeling and effective communication of their outputs.
Partially identified models can pose specific challenges to both the modeler and the consumer. Loosely speaking, a model is said to be partially identified if one or more of its parameters cannot be learned from the sampled data, even at arbitrarily large sample sizes. Inference from a partially identified model therefore depends both on the observed data and on completely subjective assumptions. The assumptions are subjective in the sense that they cannot be verified, refuted, or otherwise critiqued with the observed data.
The goals of this largely expository talk are to take a closer look at statistical issues that arise when using partially identified models to learn from data, and to relate some of these issues to the statistical and mathematical models that have been developed in response to the coronavirus pandemic. Two fundamental issues that will be addressed are model specification and uncertainty quantification. Using both examples and counter-examples from missing data and causal inference, I will argue that specifications that clearly separate the observed-data model from the subjective assumptions lead to greater transparency and more effective communication of uncertainty. I also hope to make the case that, owing to their reliance on subjectivity, partially identified models practically cry out for a Bayesian approach to drawing inference. Along the way I will illustrate connections between frequentist bounds and Bayesian posteriors with flat regions.
To conclude I will discuss the role of identifiability in the specification and interpretation of compartmental and other types of models that have been ubiquitous during the pandemic.
Joseph Hogan received his ScD in biostatistics in 1995 from Harvard School of Public Health under the direction of Professor Nan Laird. His research concerns the development and application of statistical methods for missing data, causal inference, and sensitivity analysis. During his doctoral work he served as a research assistant with SDAC (now CBAR) and has collaborated with researchers in HIV/AIDS ever since. For the past 12 years the focus of his work has been on HIV/AIDS in Kenya and sub-Saharan Africa. Professor Hogan serves as Co-Director of the Biostatistics Program for AMPATH, an international consortium of universities in the US, Canada and Kenya focused on treatment and prevention of HIV in Kenya; and as co-Director of the Biostatistics Core for the Providence-Boston Center for AIDS Research. Since 2017 he has served as Deputy Director for the Brown University Data Science Initiative. He teaches both introductory and advanced courses in biostatistics and supervises both PhD and masters students in biostatistics. For the past several years he has led an NIH-funded international training program designed to build research capacity in biostatistics at Moi University in Kenya.
Reflecting the spirit of the award, Dr. Hogan’s trajectory has many similarities with Steve Lagakos’ time at Harvard: an academic career of leadership in both methodologic and collaborative work; excellence in teaching and mentorship; and a focus on in HIV/AIDS.
About the Award
The annual Lagakos Distinguished Alumni Award has been established in memory of Dr. Stephen Lagakos, a faculty member and former chair of the Department of Biostatistics who passed away in a tragic automobile accident in 2009.
Professor Lagakos was a leader in the Department, the School of Public Health, and more broadly, in the international community of quantitative biomedical researchers. Steve’s qualities of commitment, passion, intellectual brilliance, and personal generosity had a direct personal impact on our lives; and his contributions to biostatistics and to AIDS research were fundamental.
This award serves to honor Steve’s distinguished career, and to recognize Department alumni whose research in statistical theory and application, leadership in biomedical research, and commitment to teaching have had a major impact on the theory and practice of statistical science. The award will be open to all who have an earned degree through the department, regardless of length of time since graduation or type of degree.
The award recipient will be invited to the school to deliver a lecture on their career and life beyond the Department.
or by mail to:
Lagakos Alumni Award Committee
Harvard T. H. Chan School of Public Health
Department of Biostatistics
Building 2, 4th Floor
655 Huntington Avenue
Boston, MA 02115
Nominations should include contact information for yourself and your candidate, and the candidate’s curriculum vita, if available. Please include a letter describing the contributions of the candidate, specifically highlighting the criteria for the award. Supporting letters and materials would be extremely helpful to the committee, but are not required.
Nominations must be received by Friday, April 3, 2020.
Previous Award Winners
2019 Fong Wang Clow
2018 Amy Herring
2017 Nicholas Horton
2016 Judith Goldberg
2015 Victor DeGruttola
2014 Michael Daniels
2013 Jesse Berlin
2012 Melissa Begg
– renamed The Lagakos Distinguished Alumni Award –
2011 Manning Feinleib
2010 Daniel Scharfstein
2009 John Simes
2008 Robert Strawderman
2007 Takeuchi Masahiro
2006 Daniel Siegel
2005 Christl Donnelly
2004 Stuart Baker