Publications

Published Articles

Representativeness in Statistics, Politics, and Machine Learning

Published in ACM Conference on Fairness, Accountability, and Transparency, 2021

An exploration of the concept of representativeness.

Recommended citation: Kyla Chasalow and Karen Levy. 2021. Representativeness in Statistics, Politics, and Machine Learning. In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency (FAccT '21). Association for Computing Machinery, New York, NY, USA, 77–89. https://doi.org/10.1145/3442188.3445872
Download Paper

Algorithms and Decision-Making in the Public Sector

Published in Annual Review of Law and Social Science, 2021

This paper surveys how algorithms are used in public sector decision-making and explores the social implications.

Recommended citation: Karen Levy, Kyla E. Chasalow, and Sarah Riley. (2021). "Algorithms and Decision-Making in the Public Sector." Annual Review of Law and Social Science, 17(1), 309–334. https://doi.org/10.1146/annurev-lawsocsci-041221-023808
Download Paper

Preprints

Probabilistic Race and Ethnicity Prediction Using Group-Specific Name Lists

Published in arXiv, 2026

This paper proposes lBISG, a method for obtaining group membership probabilities from names and geography when only lists of distinctive or common names for each group are available as well as possibly some prior geographic information. These probabilities are important for estimating social disparities when group membership is missing. Previous work on BISG probabilities has generally assumed access to name-frequency data, which can be limiting.

Recommended citation: Chasalow, Kyla, Noah Dasanaike, and Kosuke Imai. 2026. Probabilistic Race and Ethnicity Prediction Using Group-Specific Name Lists. arXiv preprint arXiv:2610.06273.
Download Paper

Improving Minority Population Sampling with BISG Probabilities: Evidence from a Survey of Jewish Americans

Published in arXiv, 2026

This paper proposes a method to improve minority population sampling by integrating surname- and geography-based BISG probabilities into a Poisson sampling design and demonstrates its effectiveness in a survey of Jewish Americans.

Recommended citation: Chasalow, Kyla, Eitan Hersh, Kosuke Imai, and Laura Royden. 2026. Improving Minority Population Sampling with BISG Probabilities: Evidence from a Survey of Jewish Americans. arXiv preprint arXiv:2605.05384.
Download Paper

Missing Data Multiple Imputation for Tabular Q-Learning in Online RL

Published in arXiv, 2025

This is an unpublished working paper that originated in Skyler Wu and my project for Professor Susan Murphy’s Reinforcement Learning class at Harvard. We would like to further develop it.

Recommended citation: Chasalow, Kyla, Skyler Wu, and Susan Murphy. 2025. Missing Data Multiple Imputation for Tabular Q-Learning in Online RL. arXiv preprint arXiv:2510.10709.
Download Paper