Publications

1. Shiro Kuriwaki, Stephen Ansolabehere, Angelo Dagonel, and S. Yamauchi. 2023+. The Geography of Racially Polarized Voting: Calibrating Surveys at the District Level. American Political Science Review, Forthcoming.

   We estimate vote choice by race at the Congressional District level using survey data.

   [supplemental material] [software]

2. Jong Hee Park and S. Yamauchi. 2023. Change-Point Detection and Regularization in Time Series Cross-Sectional Data Analysis. Political Analysis 31(2): 257-277.

   We develop a Bayesian regression model with shrinkage priors for analyzing longitudinal data with multiple change-points.

3. Naoki Egami and S. Yamauchi. 2023. Using Multiple Pre-treatment Periods to Improve Difference-in-Differences and Staggered Adoption Designs. Political Analysis, 31(2): 195-212.

   We propose an estimator for estimating the causal effect that improves upon the standard difference-in-differences design in terms of identification and estimation accuracy by exploiting multiple pre-treatment periods.

   [arXiv] [replication material] [software] [appendix]

4. Diana Stanescu, Erik H. Wang, and S. Yamauchi. 2019. Using LASSO to Assist Imputation and Predict Child Wellbeing. Socius, 5:1--21.

   Final prize (the best score for material hardship) for Fragile Family Challenge.

   [supplemental material]

Manuscripts

1. Matthew Blackwell and S. Yamauchi. Adjusting for Unmeasured Confounding in Marginal Structural Models with Propensity-Score Fixed Effects.

   We propose a method to adjust for unmeasured time-invariant confounders in marginal structural models.

   [abstract] [slides]

2. Shiro Kuriwaki and S. Yamauchi. Synthetic Area Weighting for Measuring Public Opinion in Small Areas.

   We propose a weighting method to estimate small area quantities.

   [abstract] [software]

3. S. Yamauchi. Difference-in-Differences for Ordinal Outcomes: Application to the Effect of Mass Shootings on Attitudes towards Gun Control.

   I propose a method for drawing causal inferences using ordinal outcomes in the difference-in-differences setting.

   [abstract] [software] [slides]

4. S. Yamauchi. Nonparametric Sensitivity Analysis for Randomized Experiments with Missing Outcomes.

   I propose a method to conduct a nonparametric sensitivity analysis and conduct inference for missing outcome in randomized experiments. The method provides bounds and confidence intervals for treatment effects that account for non-ignorable missing and attrition.

   [software]

Dissertation

• S. Yamauchi. 2022. Essays on Political Methodology. PhD Dissertation, Harvard University.




Statistical Software

1. Kuriwaki and Yamauchi. bmlogit: R package for implementing multinomial logistic regression with prediction constraints.
2. Yamauchi. attritionCI: R package for constructing sensitivity-aware confidence interval for causal effect estimates under nonignorable missingness in the outcome. The method is proposed in Yamauchi (2021).
3. Egami and Yamauchi. DIDdesign: R package for implementing the double difference-in-differences method proposed in Egami and Yamauchi (2019).
4. Yamauchi. dyRank: R package for implementing the hierarchical dynamic rating model for estimating the dynamic rating with rank-ordered data.
5. Yamauchi. emlogit: R package for implementing the multinomial logistic regression based on the Expectation and Conditional Maximization (ECM) algorithm.
6. Yamauchi. orddid: R package for implementing the difference-in-differences for ordinal outcomes proposed in Yamauchi (2020+).
7. Park and Yamauchi. BridgeChange: R package for implementing the Bayesian regularization regression with multiple change points for time-series and panel data.



Miscellanea

• Yamauchi. (2022+). Hanada Beamer Template.
  [example slides]
• Yamauchi. (2020+). EM Algorithm for Multinomial Logit.
  [software]
• Yamauchi. (2020+). Dynamic Plackett-Luce Model for Ranked Data: Application to the Rating of Formula One Drivers.
  [software]