I am an Associate Professor of Political Science at the University of California, Berkeley. My research sits at the intersection of the social sciences and statistics. My mission is to create a body of research that bridges these two worlds — with an emphasis on answering causal questions — within which experts from both worlds can have dialogue with one another and foster beneficial collaborations.
My research sits primarily in the field of causal inference and survey design and analysis. My main research agendas focus on external validity of experiments, falsification testing, and survey weighting.
PhD in Political Science, 2013
University of California, Berkeley
MA in Statistics, 2013
University of California, Berkeley
BS in Economics and Political Science, 2007
California Institute of Technology
Even in the best-designed experiment, noncompliance can complicate analysis. While the intent-to-treat effect remains identified, randomization alone no longer identifies the complier average causal effect (CACE). Instrumental variables (IV) approaches, which rely on the exclusion restriction, can suffer from high variance, particularly when the experiment has a low compliance rate. We provide a framework which broadens the set of design and analysis techniques political science researchers can use when addressing noncompliance…
This chapter provides an introduction to the problems of generalization and transportation and methods for addressing these concerns. The field of causal inference is one that, at its core, focuses on improving internal validity–the extent to which a study can establish a trustworthy cause-and-effect relationship between a treatment and outcome. To understand potential external validity bias,…
A pressing challenge in modern survey research is to find calibration weights when covariates are high dimensional and especially when interactions between variables are important…
Survey weighting allows researchers to account for bias in survey samples, due to unit nonresponse or convenience sampling, using measured demographic covariates. Unfortunately, in practice, it is impossible…
Generalizing causal estimates in randomized experiments to a broader target population is essential for guiding decisions by policymakers and practitioners in the social and biomedical sciences. While recent papers developed various weighting estimators for the population average treatment effect (PATE), many of these methods result in large variance because the experimental sample…
External validity of randomized experiments is a focus of long-standing debates in the social sciences. While the issue has been extensively studied…
In 2011, the Los Angeles Police Department (LAPD), in conjunction with other governmental and nonprofit groups, launched the Community Safety Partnership (CSP) in several public housing developments in Los Angeles. Following a relationship-based policing model…
Scientists are often interested in generalizing causal effects estimated in an experiment to a target population. However, analysts are often constrained by available covariate information…
Experiments have come to be a widely accepted and highly regarded method for political science research. Randomization allows for well identified causal effects that are “internally valid” to the experimental setting. However, political scientists are driven by asking big questions with broad impacts…
Regression discontinuity (RD) designs are increasingly common in political science. They have many advantages, including a known and observable treatment assignment mechanism. The literature has emphasized the need for “falsification tests” and ways to assess the validity of the design. …
We elaborate a general workflow of weighting-based survey inference, decomposing it into two main tasks….
This chapter focuses on methods for analyzing data from Internet surveys with complex survey designs in order to draw inferences that can be generalized to a target population of interest…
Recent emphasis on credible causal designs has led to the expectation that scholars justify their research designs by testing the plausibility of their causal identification assumptions, often through balance and placebo tests. Yet current practice…
Do voters like the party they already agree with or do they agree with the party they already like? Previous studies have suggested a link…
Randomized controlled trials (RCTs) can provide unbiased estimates of sample average treatment effects. However, a common concern is that RCTs may fail to provide unbiased estimates of population average treatment effects. We derive assumptions…
Response rates to surveys have declined precipitously. Some researchers have responded by relying more heavily on convenience-based internet samples. This leaves researchers asking…
Post-election audits are an integral part of the broader election audit process. Depending on the auditing method, they may detect miscounts in the official tabulation of votes, or limit the risk of certifying an incorrect outcome…
Survey nonresponse is a ubiquitous problem in modern survey research. As individuals have become less likely to respond to surveys there has been a simultaneous rise in highly granular data sources…