Interdisciplinary Seminar in Quantitative Methods (ISQM)

The goal of the Interdisciplinary Seminar in Quantitative Methods is to provide an interdisciplinary environment where researchers can present and discuss cutting-edge research in quantitative methodology. The talks will be aimed at a broad audience, with more emphasis on conceptual than technical issues. The research presented will be varied, ranging from new methodological developments to applied empirical papers that use methodology in an innovative way. We welcome speakers and audiences from all fields in the social, natural, and behavioral sciences.

Organizers: Matias Cattaneo and Rocio Titiunik
To be added to the email list, please contact us at:

Location: Eldersveld Room, 5670 Haven Hall

Time: Wednesdays, 4:00 - 5:30pm

Note: Please see event listing for particular changes in location or time.

Quantifying Complexity

September 10, 2014: Scott Page, Complex Systems, Political Science, and Economics, University of Michigan

How do Climate Models Compare with Reality Over the Tropics from 1958-2012? HAC-Robust Trend Comparisons Among Climate Series with Possible Intercept Shifts

September 24, 2014: Timothy J. Vogelsang, Economics, Michigan State University

Why Does the American National Election Study Overestimate Voter Turnout?

October 8, 2014: Simon Jackman, Political Science, Stanford University

Estimating the Impacts of Program Benefits: Using Instrumental Variables with Underreported and Imputed Data

October 22, 2014: Mel Stephens, Economics, University of Michigan

Using Experiments to Estimate Geographic Variation in Racially Polarized Voting

November 5, 2014: Kevin M. Quinn, School of Law, University of California at Berkeley

Mitigating the Usual Limitations of the basic Regression-Discontinuity Design: Theory and Three Empirical Demonstrations from Design Experiment

November 19, 2014: Thomas D. Cook, Sociology, Psychology, and Education and Social Policy, Northwestern University

Statisticians (Social Science) and Data Scientists (Machine Learners): Let’s Talk

December 3, 2014: Neal Beck, Department of Politics, New York University

Essential Ideas of Causal Inference in Experiments and in Observational Studies

February 11, 2015: Don Rubin, Statistics, Harvard University

Measuring Political Knowledge in the Mass Public: Calibrating a Useful Instrument

February 25, 2015: William G. Jacoby, Political Science, Michigan State University

New Developments in Mediation Analysis

March 11, 2015: Tyler J. VanderWeele, Epidemiology and Biostatistics, Harvard School of Public Health

Assessing and enhancing the generalizability of randomized trials to target populations

March 25, 2015: Elizabeth A. Stuart, Mental Health and Biostatistics, Johns Hopkins Bloomberg School of Public Health

Optimal Multilevel Matching in Clustered Observational Studies: A Case Study of the School Voucher System in Chile

April 8, 2015: Luke Keele, Political Science, Penn State

Strong Control of the Family-wise Error Rate in Observational Studies

April 22, 2015: Dylan Small, Statistics, Wharton School, University of Pennsylvania

Add to Your Google Calendar


An effect modifier is a pretreatment covariate such that the magnitude of the treatment effect or its stability changes with the level of the covariate. Generally, other things being equal, larger treatment effects and less heterogeneous treatment effects are less sensitive to unmeasured biases in observational studies. It is known that when there is effect modification, an overall test that ignores an effect modifier may report greater sensitivity to unmeasured bias than a test that combines results at different levels of the effect modifier. This known combined test reports that there is evidence of an effect somewhere that is insensitive to bias of a certain magnitude, but it does not draw inferences about affected subgroups. If there is effect modification, one would like to identify specific subgroups for which there is evidence of effect that is insensitive to small or moderate biases. We propose an exploratory method for discovering effect modification combined with a confirmatory method of simultaneous inference that strongly controls the family-wise error rate in a sensitivity analysis, despite the fact that the groups being compared are defined empirically. A new form of matching, strength k matching, permits a search through more than k covariates for effect modifiers, yet no pairs are lost providing at most k covariates are selected to group the pairs. In a strength k match, each set of k covariates is exactly balanced, though a set of > k covariates may exhibit imbalance. We apply the method to study the effects of the powerful earthquake that struck Chile in 2010. This is joint work with Jesse Hsu, Jose Zubizarreta and Paul Rosenbaum.

An empirical model of network formation: detecting homophily when agents are heterogeneous

May 6, 2015: Bryan S. Graham, Economics, University of California at Berkeley