Publications

We test a social media conversational agent for canvassing on the topic of anti-transgender prejudice, replicating and benchmarking treatment effects. In-person deep canvassing is the gold standard for durably changing attitudes on polarizing topics. However, door-to-door canvassing is costly, and many populations may not be feasibly reached in this manner. Campaigns are already conducting outreach using digital tools, including text messages and social media. If appropriately trained agents messaging over social media can achieve a fraction of the effect of in-person canvassing, canvassing may be scaled up to achieve large overall impacts at lower costs. Scripts used in this application are based on those used by transgender allies in the original study. To personalize messaging, the conversational agent uses natural language processing to detect conversational topics, and shares relevant pre-scripted messages of information and third-person experiences, encouraging respondents to engage in perspective-taking with respect to an outgroup. This study demonstrates the potential of automated social media messaging for deep canvassing, with possible applications by governments, public health agencies, and political organizations. Estimated effects are positive and significant under covariate adjustment and reweighting; due to important differential attrition, partial-identification bounds are also reported and include zero.

@article{offer2026deep,
  title   = {Deep canvassing with automated conversational agents: Personalized messaging to change attitudes},
  author  = {Offer-Westort, Molly and Liu, Jiehan and Feamster, Nick and Garg, Kartik and Hoang, Nguyen Phong and Hosamane, Sudhamshu},
  journal = {Research & Politics},
  volume  = {13},
  number  = {1},
  pages   = {20531680251414926},
  year    = {2026},
  publisher = {SAGE Publications Sage UK: London, England}
}

This Element offers a practical guide to estimating conditional marginal effects—how treatment effects vary with a moderating variable—using modern statistical methods. Commonly used approaches, such as linear interaction models, often suffer from unclarified estimands, limited overlap, and restrictive functional forms. This guide begins by clearly defining the estimand and presenting the main identification results. It then reviews and improves upon existing solutions, such as the semiparametric kernel estimator, and introduces robust estimation strategies, including augmented inverse propensity score weighting with Lasso selection (AIPW-Lasso) and double machine learning (DML) with modern algorithms. Each method is evaluated through simulations and empirical examples, with practical recommendations tailored to sample size and research context. All tools are implemented in the accompanying interflex package for R.

@article{liu2025practical,
  title   = {A Practical Guide to Estimating Conditional Marginal Effects: Modern Approaches},
  author  = {Liu, Jiehan and Liu, Ziyi and Xu, Yiqing},
  journal = {arXiv preprint arXiv:2504.01355},
  year    = {2025}
}

Simonsohn (2024a) and Simonsohn (2024b) critique Hainmueller, Mummolo and Xu (2019, HMX), arguing that failing to model nonlinear relationships between the treatment and moderator leads to biased marginal effect estimates and uncontrolled Type-I error rates. While these critiques highlight the issue of under-modeling nonlinearity in applied research, they are fundamentally flawed in several key ways. First, the causal estimand for interaction effects and the necessary identifying assumptions are not clearly defined in these critiques. Once properly stated, the critiques no longer hold. Second, the kernel estimator HMX proposes recovers the true causal effects in the scenarios presented in these recent critiques, which compared effects to the wrong benchmark, producing misleading conclusions. Third, while Generalized Additive Models (GAM) can be a useful exploratory tool (as acknowledged in HMX), they are not designed to estimate marginal effects, and better alternatives exist, particularly in the presence of additional covariates. Our response aims to clarify these misconceptions and provide updated recommendations for researchers studying interaction effects through the estimation of conditional marginal effects.

@article{hainmueller2025response,
  title   = {A Response to Recent Critiques of Hainmueller, Mummolo and Xu (2019) on Estimating Conditional Relationships},
  author  = {Hainmueller, Jens and Liu, Jiehan and Liu, Ziyi and Mummolo, Jonathan and Xu, Yiqing},
  journal = {arXiv preprint arXiv:2502.05717},
  year    = {2025}
}

Uncertainty in quantitative social science is usually discussed as a problem of sampling, estimation, or model selection. Text-as-data research introduces an earlier layer of uncertainty: the measurement itself. When social scientists code text into quantitative variables, they must translate a conceptual target into a rubric, which is a natural-language decision rule that maps text to scores. This translation is contestable at two levels. One source is conceptualization uncertainty: because many political concepts are inherently contested, researchers may disagree about which attributes are constitutive of the construct. The other is implementation uncertainty: even holding the concept fixed, different coding rubrics may implement that definition differently and thus yield different observed measurements. I refer to the dependence of empirical conclusions on these theoretically admissible choices as measurement uncertainty. This paper develops a framework for characterizing, decomposing, and propagating this uncertainty into downstream inference. The framework formalizes text-based measurement as a two-layer design problem — conceptualization and implementation — and models rubric choice as a partial-identification problem, where the estimand is bounded by the set of conclusions generated by a local class of admissible measurement choices.