In recognition of John T. Williams’ contribution to graduate training, the John T. Williams Award has been established for the best dissertation proposal in the area of political methodology.
| 2024 Winner | |
| Recipient | Sooahn Shin |
| Work | Moving beyond a Single-Dimensional Ideological Score |
| Citation | The John T. Williams Dissertation Prize for the 2023-2024 academic year is awarded to Sooahn Shin, a Ph.D. student in the Government Department at Harvard University. Her dissertation includes three essays that introduce innovative methodologies for political science research. The first essay, "Moving beyond a Single-Dimensional Ideological Score," features IssueIRT---a hierarchical IRT model that precisely estimates issue-specific latent political stances from roll-call votes, demonstrating its efficacy through simulations and a U.S. House validation study. The second essay proposes a novel framework for testing decision-making systems, assessing the effectiveness of human-alone, human-with-AI, and AI-alone configurations. This study examines the impact of pretrial risk assessment scores on judicial decisions and their accuracy compared to human judgment. The third essay explores causal inference with panel data, tackling the challenges of list-wise deletion and offering strategies to manage data missingness using auxiliary variables. Shin's dissertation makes important contributions to enhance the fields of computational social science and causal inference. |
| Selection committee | Yamil Velez (Columbia), Erin Hartman (Berkeley), Walter Mebane (Michigan), and In Song Kim (MIT, chair) |
| 2023 Winner | |
| Recipient | Guilherme Duarte |
| Work | Essays on Automated Causal Inference for Partial Identification, Data Fusion, and Sensitivity Analysis |
| Citation | The William Prize Committee is pleased to announce Guilherme Duarte, a Ph.D. candidate in the Operations, Information, and Decisions Department at the Wharton School of the University of Pennsylvania, as the recipient of the John T. Williams Dissertation Prize for the 2022-2023 academic year. His dissertation proposal, "Essays on Automated Causal Inference for Partial Identification, Data Fusion, and Sensitivity Analysis," introduces significant methodological advances by employing robust computational techniques to automatically calculate precise bounds, termed "autobounds," for point estimates and uncertainty quantification. Duarte's proposal establishes a comprehensive framework for numerical solutions in partial identification, which he further extends to address key issues in transportability and external validity. His dissertation project advances the research agenda in algorithmic-assisted causality, demonstrating its utility in research designs that include instrumental variables, differences-in-differences, and mediation analysis. |
| Selection committee | Yamil Velez (Columbia), Erin Hartman (Berkeley), Walter Mebane (Michigan), and In Song Kim (MIT, chair) |
Past Recipients
| 2022 Winner | |
| Recipient | Melody Huang (Berkeley) |
| Work | “Three Essays on Causal Inference under Interference and Hypothesis Testing in Random Experiments” |
| Citation | This year's Williams Dissertation Prize is awarded to Melody Huang, Ph.d. candidate at the University of California, Berkeley, for her work on the credibility and generalizability of causal inference, more specifically, for her development of a new sensitivity analysis framework for weighted estimators in the generalizability setting. Unlike other frameworks, in Huang's set up the weights account for the confounding effect of sampling, not treatment. In addition, in her framework the parameters are guaranteed to be bounded across finite ranges. Hence there is no need to impose parametric assumptions on the outcome and selection models. Huang produces three sensitivity tools for generalizing experimental results: bias contour plots, a robustness value, and a benchmark that allows researchers to use observed covariates to estimate parameter values for the confounder. Her work potentially will strengthen our inferences about the effects of voter mobilization efforts and of other important public policies. |
| Selection committee | John Freeman (chair, Minnesota), Yamil Velez (Columbia), Erin Hartman (Berkeley), Walter Mebane (Michigan), and In Song Kim (MIT) |
| 2020 Winner | |
| Recipient | Ye Wang (NYU) |
| Work | “Three Essays on Causal Inference under Interference and Hypothesis Testing in Random Experiments” |
| Citation | The William Prize Committee is delighted to announce that the John T. Williams Dissertation Prize, 2020 was awarded to Ye Wang's dissertation proposal “Three Essays on Causal Inference under Interference and Hypothesis Testing in Random Experiments.” Wang makes two major methodological innovations. First, he develops a methodological framework to identify causal relationships in time-series cross-sectional data under arbitrary within-unit (temporal) and between-unit (spatial) interference. Wang shows, under the sequential ignorability assumption, how one can obtain unbiased/consistent estimates of cumulative causal effects via inverse probability of treatment weighting (IPTW) estimators. More specifically, the proposed estimator identifies the expected average treatment effect generated by any particular treatment history of a representative unit on itself or on its neighbors. He demonstrates the usefulness of his innovation in a simulation and in a re-analysis of a published study of the impacts of a political reform in New York, accounting for temporal and spatial interference. He then applies this methodology to gauge the effect of protests in a handful of constituencies during the Umbrella Movement on electoral support for the opposition in Hong Kong. The second methodological contribution of Wang’s dissertation is to develop tools for testing nonlinear moderating effects in experiments. He adapts the evolutionary tree algorithm and sample splitting design to experimental analysis; the algorithm enables researchers to find the optimal partition of the moderator’s support on the training set for any loss function. He conducts a simulation study as well as a pilot study of the effects of a get-out-the-vote (GOTV) experiment, providing a promising direction for applying machine learning algorithms to experimental settings. |
| Selection committee | John Freeman (chair, Minnesota), Walter Mebane (Michigan), and In Song Kim (MIT) |
| Year | Recipient | Work |
| 2019 | Naijia Liu (Princeton) | "Essays on Model Selection and Honest Inference" |
| 2018 | Kevin McAlister (University of Michigan) | "Roll Call Scaling in the U.S. Congress: Addressing the Deficiencies" |
| 2017 | Naoki Egami (Princeton) | |
| 2016 | Dean Knox (MIT) | "Essays on Modeling and Causal Inference in Network Data" |
| 2015 | Drew Dimmery (NYU) | "Essays on Machine Learning and Causal Inference with Application to Nonprofits" |
| 2014 | Yiqing Xu (MIT) | "Causal Inference with Time-Series Cross-Section Data with Applications to Chinese Political Economy" |
| 2013 | Scott Cook (University of Pittsburgh) | The Contagion of Crises: Estimating Models of Endogenous and Interdependent Rare Events |
| 2012 | Adriana Crespo-Tenorio (Washington University in St. Louis) | Three Papers on the Political Consequences of Oil Price Volatility |
| 2011 | Matthew Blackwell (Harvard) | Essays in Political Methodology and American Politics |
| 2010 | Teppei Yamamoto (Princeton) | Essays on Quantitative Methodology for Political Science |
| 2009 | Xun Pang (Washington University in St. Louis) | A Bayesian Probit Hierarchical Model with AR(p) Errors and Non-nested Clustering: Studying Sovereign Creditworthiness and Political Institutions |
| 2008 | Justin Grimmer (Harvard) | A Bayesian Hierarchical Topic Model for Political Texts: Measuring and Explaining Legislator's Express Agenda |
| 2007 | Arthur Spirling (University of Rochester) | Bringing Intuition to Fruition: 'Turning Points' and 'Power' in Political Methodology |
| 2006 | Roman Ivanchenko (Ohio State) | Interactions Between the Supreme-Court and Congress: A Different Look at the Decision-Making Process |
Past Selection Committees
| Year | Committee |
| 2019 | Xun Pang (Tsinghua University), Dean Knox (Princeton, recused) and Yiqing Xu (University of California, San Diego) |
| 2018 | Xun Pang (Tsinghua, chair), Arthur Spirling (NYU), and Yiqing Xu (UCSD) |
| 2017 | Xun Pang (Tsinghua, chair), Arthur Spirling (NYU), and Yiqing Xu (UCSD) |
| 2016 | Justin Grimmer (Chicago, chair), Matt Blackwell (Harvard) and Teppi Yamamoto (MIT) |
| 2015 | Curt Signorino (Chair), John Ahlquist, Jennifer Jerit |
| 2014 | Curt Signorino (Chair), John Ahlquist, Jennifer Jerit |
| 2013 | Michael Colaresi (Chair), Guy Whitten, Irfan Nooruddin |
| 2012 | Michael Colaresi (Chair), Guy Whitten, Irfan Nooruddin |
| 2011 | Guy Whitten (Chair), Michael Colaresi, Jonathan Nagler |
| 2010 | Guy Whitten (Chair), Michael Colaresi, Jonathan Nagler |
| 2009 | Guy Whitten (Chair), Michael Colaresi, Betsy Sinclair |
| 2007 | John Aldrich (Chair), Michael Colaresi, Tse-Min Lin |
| 2006 | John Aldrich (Co-Chair), Virginia Gray (Co-Chair), Patrick Brandt, Burt Monroe |