Dear all,
This week at the Applied Statistics Workshop we will be welcoming Stephen Pettigrew, a Ph.D. candidate at Harvard University. He will be presenting work entitled "The Downstream Consequences of Long Waits: How Lines at the Precinct Depress Future Turnout." Please find the abstract below and on the website. The paper is attached.
We will meet in CGIS Knafel Room 354 at noon and lunch will be provided. See you all there!
Best,
Pam
Title: "The Downstream Consequences of Long Waits: How Lines at the Precinct Depress Future Turnout"
Abstract: Political scientists have increasingly emphasized the role played by an individual's identity and life experiences in their patterns of political participation. In this paper, I explore how one particular type of experience-standing in line at a precinct to vote-shapes the turnout behavior of voters in future election. I demonstrate that for every additional hour a voter waits in line to vote, their probability of voting in the subsequent election drops by 1 percentage point. As a result, nearly 200,000 people did not vote in November 2014 because waiting in a long line in 2012 turned them off from the process. To arrive at these estimates, I analyze vote history files using a combination of exact matching and placebo tests to test the identification assumptions. I then leverage an unusual institutional arrangement in the City of Boston and longitudinal data from Florida to show that the result also holds at the precinct level. The findings in this paper have implications for our understanding of what motivates or demotivates a person from voting. They also suggest that racial asymmetries in precinct wait times are contributing to under-representation of racial minorities in the voter pool.
Hi everyone,
This week at the Applied Statistics Workshop we will be welcoming Francesca Dominici, a Professor of Biostatistics and Senior Associate Dean for Research at the Harvard T.H. Chan School of Public Health. She will be presenting work entitled "Model Averaged Double Robust Estimation." Please find the abstract below and on the website<http://projects.iq.harvard.edu/applied.stats.workshop-gov3009/home>. In addition, the paper is attached.
As usual, we will meet in CGIS Knafel Room 354 and lunch will be provided. See you all there!
-Pam
Model averaged double robust estimation
Francesca Dominici, Harvard T H Chan School of Public Health
Joint with: Matt Cefalu, Giovanni Parmigiani, Nils Arvold.
ABSTRACT. Researchers are increasingly being challenged with decisions on how to best control for a high-dimensional set of potential confounders when estimating causal effects. Typically, a single propensity score model is used to adjust for confounding, while the uncertainty surrounding the procedure to arrive at this propensity score model is often ignored and failure to include even one important confounder will results in bias. We propose a general causal framework that overcomes the limitations described above through the use of model averaging. We illustrate the proposed framework in the context of double robust estimation.The MA-DR estimator is defined as a weighted average of double robust estimators, where each double robust estimator corresponds to a specific choice for the outcome model and the propensity score. The MA-DR estimator extend the desirable double robustness property by achieving consistency under the much weaker assumption that either the true propensity score model or the true outcome model be within a specified, possibly large, class of models. Importantly, using simulation studies, we found that our MA-DR estimator dramatically reduces mean squared error by the largest percentage in the realistic situation where the set of potential confounders is large relative to the sample size. We apply the methodology to estimate the comparative effectiveness of the oral chemotherapy temozolomide on 1-year survival in a cohort of 1887 Medicare enrollees who were diagnosed with glioblastoma between June 2005 and December 2009.
Dear all,
This week at the Applied Statistics Workshop we will be welcoming Tobias Gerstenberg, a postdoctoral fellow at MIT. He will be presenting work entitled "A Counterfactual Simulation Model of Causal Judgment." Please find the abstract below and on the website<http://projects.iq.harvard.edu/applied.stats.workshop-gov3009/home>.
We will meet in CGIS Knafel Room 354 at noon and lunch will be provided. See you all there!
-Pam
Title: A Counterfactual Simulation Model of Causal Judgment
Abstract: How do we make causal judgments? In this talk, I will present a counterfactual simulation model (CSM) of causal judgments that unifies different philosophical views on causation. The CSM predicts that people's causal judgments are influenced by the extent to which a candidate cause made a difference to i) whether the outcome occurred, and ii) how it occurred. I will show how whether-causation and how-causation can be expressed in terms of different counterfactual contrasts defined over the same generative model of a domain. I will focus on applying the CSM to the domain of intuitive physics, asking people to make judgments about colliding billiard balls. The CSM accounts for participants' causal judgments to a high degree of quantitative accuracy. Causal judgments increased the more certain participants were that a ball was a whether-cause, a how-cause, as well as sufficient for bringing about the outcome. The CSM postulates that people make causal judgments by comparing what actually happened with what would have happened if the candidate cause had been removed from the scene. In direct support of this claim, I will show eye-tracking data of how people mentally simulate how the counterfactual world would have unfolded. I will conclude by discussing how the CSM may help us better understand the mapping between causal events in the world and the words we use to describe them.
Hi everyone,
We have a last minute room change for the workshop today. Instead of room 354, we are now meeting in CGIS Knafel room 450 today at noon.
Pam
Hi everyone,
This week at the Applied Statistics Workshop we will be welcoming Victoria Liublinska, Liam Schwartz, and their colleagues from the Office of Institutional Research at Harvard University. They will be presenting work entitled "Data, Data Science, and The Research University." Please find their abstract below and on the website<http://projects.iq.harvard.edu/applied.stats.workshop-gov3009/home>.
As usual, we will meet in CGIS Knafel room 354 at noon and lunch will be provided. See you all there!
- Pam
Title: "Data, Data Science, and The Research University"
Abstract: In our talk we will discuss some of the challenges that may arise while developing a "data science" approach to institutional research in the university setting. Despite these hurdles, with improved data governance and availability, a team with the right skills and outlook, and the support of senior leadership, the transition from a more traditional institutional research function to one representing a data science perspective is not only possible, but natural. We will give several examples of analytical projects carried out by Harvard University's Office of Institutional Research on issues that address strategic questions for the University.
Dear all,
Please find below the schedule for the fall semester of the Applied Statistics Workshop. We meet Wednesdays at noon in room K354. Our first meeting will be next Wednesday, September 7.
Best,
Pam
09/07/2016 Victoria Liublinski & Office of Institutional Research
09/14/2016 Tobias Gerstenberg
09/21/2016 Francesca Dominici
09/28/2016 Stephen Pettigrew
10/05/2016 Tina Eliassi-Rad
10/12/2016 Solé Prillaman
10/19/2016 Tyler VanderWeele
10/26/2016 In Song Kim
11/02/2016 Daniele Paserman
11/09/2016 Sharon-Lise Normand
11/16/2016 David Parkes
11/30/2016 Christopher Rycroft
