Xinwei Ma
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Xinwei Ma has a total of 29 co-authors. Here are the top co-authors by number of shared publications:
- Matias Damian Cattaneo
Princeton University
3 shared publications Jingshen Wang
2 shared publications- Jingshen Wang
Division of Biostatistics, University of California Berkeley
2 shared publications Matias D. Cattaneo
Department of Operations Research and Financial Engineering, Princeton University, Princeton NJ, United States
2 shared publicationsMichael Jansson
Department of Economics, UC Berkeley, Berkeley CA, United States
2 shared publications- Michael Jansson
Princeton University UC Berkeley UC San Diego
2 shared publications
Latest Publications
- Open Access
- 30 September 2026
- Published by Elsevier in SSRN Electronic Journal
Can Language Models Boost the Power of Randomized Experiments Without Statistical Bias?
Randomized controlled trials (RCTs) are widely adopted for causal inference, yet cost and sample-size constraints limit power. We introduce CALM(Causal Analysis leveraging Language Models), a statistical framework that integrates insights generated by large language models (LLMs) into the analysis of RCTs using established causal estimators to increase precision while preserving statistical validity. In particular, CALM treats LLM-generated outputs as auxiliary prognostic information and corrects their potential bias via a heterogeneous calibration step that residualizes and optimally reweights predictions. We prove that CALM remains consistent even when LLM predictions are biased and achieves efficiency gains over augmented inverse probability weighting estimators for various causal estimands. In particular, CALM develops a few-shot variant that aggregates predictions across randomly sampled demonstration sets. The resulting U-statistic-like predictor restores i.i.d. structure and also mitigates prompt-selection variability. Empirically, in simulations calibrated to a mobile-app depression RCT, CALM delivers lower variance relative to other benchmarking methods, is effective in zero- and few-shot settings, and remains stable across prompt designs. By principled use of LLMs to harness unstructured data and external knowledge learned during pretraining, CALM provides a practical path to more precise causal analyses.
- 30 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessCan Language Models Boost the Power of Randomized Experiments Without Statistical Bias?
Randomized controlled trials (RCTs) are widely adopted for causal inference, yet cost and sample-size constraints limit power. We introduce CALM(Causal Analysis leveraging Language Models), a statistical framework that integrates insights generated by large language models (LLMs) into the analysis of RCTs using established causal estimators to increase precision while preserving statistical validity. In particular, CALM treats LLM-generated outputs as auxiliary prognostic information and corrects their potential bias via a heterogeneous calibration step that residualizes and optimally reweights predictions. We prove that CALM remains consistent even when LLM predictions are biased and achieves efficiency gains over augmented inverse probability weighting estimators for various causal estimands. In particular, CALM develops a few-shot variant that aggregates predictions across randomly sampled demonstration sets. The resulting U-statistic-like predictor restores i.i.d. structure and also mitigates prompt-selection variability. Empirically, in simulations calibrated to a mobile-app depression RCT, CALM delivers lower variance relative to other benchmarking methods, is effective in zero- and few-shot settings, and remains stable across prompt designs. By principled use of LLMs to harness unstructured data and external knowledge learned during pretraining, CALM provides a practical path to more precise causal analyses.
- Open Access
- 4 September 2026
- Published by Elsevier in Journal of Econometrics
- Vol. 258
Principal component analysis for a mix of stationary and nonstationary variables
- 4 September 2026
- Published by Elsevier in Journal of Econometrics
- Vol. 258
Open AccessPrincipal component analysis for a mix of stationary and nonstationary variables
- 21 May 2026
- Published by Taylor & Francis in Journal of the American Statistical Association
- p. 1-13
Covariate-Adjusted Response-Adaptive Design with Delayed Outcomes
1 Introduction1.1 Motivation and ContributionIn clinical trials, social science field experiments, and A/B tests, participants often enroll sequentially, and responses to treatments can vary due to...
- 21 May 2026
- Published by Taylor & Francis in Journal of the American Statistical Association
- p. 1-13
Covariate-Adjusted Response-Adaptive Design with Delayed Outcomes
1 Introduction1.1 Motivation and ContributionIn clinical trials, social science field experiments, and A/B tests, participants often enroll sequentially, and responses to treatments can vary due to...
- Open Access
- 7 March 2025
- Published by The Open Journal in Journal of Open Source Software
- Vol. 10 (107) , 7241
lpcde: Estimation and Inference for Local Polynomial Conditional Density Estimators
Conditional cumulative distribution functions (CDFs), conditional probability density functions (PDFs), and derivatives thereof, are important parameters of interest in statistics, econometrics, and other data science disciplines. The package lpcde implements new estimation and inference methods for conditional CDFs, conditional PDFs, and derivatives thereof, employing the kernelbased local polynomial smoothing approach introduced in Cattaneo et al. (2024a).
- 7 March 2025
- Published by The Open Journal in Journal of Open Source Software
- Vol. 10 (107) , 7241
Open Accesslpcde: Estimation and Inference for Local Polynomial Conditional Density Estimators
Conditional cumulative distribution functions (CDFs), conditional probability density functions (PDFs), and derivatives thereof, are important parameters of interest in statistics, econometrics, and other data science disciplines. The package lpcde implements new estimation and inference methods for conditional CDFs, conditional PDFs, and derivatives thereof, employing the kernelbased local polynomial smoothing approach introduced in Cattaneo et al. (2024a).
- 24 February 2025
- Published by Oxford University Press (OUP) in Journal of the Royal Statistical Society Series B: Statistical Methodology
- Vol. 87 (4) , 1055-1084
Adaptive experiments toward learning treatment effect heterogeneity
Understanding treatment effect heterogeneity has become an increasingly popular task in various fields, as it helps design personalized advertisements in e-commerce or targeted treatment in biomedical studies. However, most of the existing work in this research area focused on either analysing observational data based on strong causal assumptions or conducting post hoc analyses of randomized controlled trial data, and there has been limited effort dedicated to the design of randomized experiments specifically for uncovering treatment effect heterogeneity. In the manuscript, we develop a framework for designing and analysing response adaptive experiments toward better learning treatment effect heterogeneity. Concretely, we provide response adaptive experimental design frameworks that sequentially revise the data collection mechanism according to the accrued evidence during the experiment. Such design strategies allow for the identification of subgroups with the largest treatment effects with enhanced statistical efficiency. The proposed frameworks not only unify adaptive enrichment designs and response-adaptive randomization designs but also complement A/B test designs in e-commerce and randomized trial designs in clinical settings. We demonstrate the merit of our design with theoretical justifications and in simulation studies with synthetic e-commerce and clinical trial data.
- 24 February 2025
- Published by Oxford University Press (OUP) in Journal of the Royal Statistical Society Series B: Statistical Methodology
- Vol. 87 (4) , 1055-1084
Adaptive experiments toward learning treatment effect heterogeneity
Understanding treatment effect heterogeneity has become an increasingly popular task in various fields, as it helps design personalized advertisements in e-commerce or targeted treatment in biomedical studies. However, most of the existing work in this research area focused on either analysing observational data based on strong causal assumptions or conducting post hoc analyses of randomized controlled trial data, and there has been limited effort dedicated to the design of randomized experiments specifically for uncovering treatment effect heterogeneity. In the manuscript, we develop a framework for designing and analysing response adaptive experiments toward better learning treatment effect heterogeneity. Concretely, we provide response adaptive experimental design frameworks that sequentially revise the data collection mechanism according to the accrued evidence during the experiment. Such design strategies allow for the identification of subgroups with the largest treatment effects with enhanced statistical efficiency. The proposed frameworks not only unify adaptive enrichment designs and response-adaptive randomization designs but also complement A/B test designs in e-commerce and randomized trial designs in clinical settings. We demonstrate the merit of our design with theoretical justifications and in simulation studies with synthetic e-commerce and clinical trial data.
- 1 November 2024
- Published by Bernoulli Society for Mathematical Statistics and Probability in Bernoulli
- Vol. 30 (4) , 3193-3223
Boundary adaptive local polynomial conditional density estimators
We begin by introducing a class of conditional density estimators based on local polynomial techniques. The estimators are boundary adaptive and easy to implement. We then study the (pointwise and) uniform statistical properties of the estimators, offering characterizations of both probability concentration and distributional approximation. In particular, we establish uniform convergence rates in probability and valid Gaussian distributional approximations for the Studentized t-statistic process. We also discuss implementation issues such as consistent estimation of the covariance function for the Gaussian approximation, optimal integrated mean squared error bandwidth selection, and valid robust bias-corrected inference. We illustrate the applicability of our results by constructing valid confidence bands and hypothesis tests for both parametric specification and shape constraints, explicitly characterizing their approximation errors. A companion R software package implementing our main results is provided.
- 1 November 2024
- Published by Bernoulli Society for Mathematical Statistics and Probability in Bernoulli
- Vol. 30 (4) , 3193-3223
Boundary adaptive local polynomial conditional density estimators
We begin by introducing a class of conditional density estimators based on local polynomial techniques. The estimators are boundary adaptive and easy to implement. We then study the (pointwise and) uniform statistical properties of the estimators, offering characterizations of both probability concentration and distributional approximation. In particular, we establish uniform convergence rates in probability and valid Gaussian distributional approximations for the Studentized t-statistic process. We also discuss implementation issues such as consistent estimation of the covariance function for the Gaussian approximation, optimal integrated mean squared error bandwidth selection, and valid robust bias-corrected inference. We illustrate the applicability of our results by constructing valid confidence bands and hypothesis tests for both parametric specification and shape constraints, explicitly characterizing their approximation errors. A companion R software package implementing our main results is provided.
- 30 July 2024
- Published by Bernoulli Society for Mathematical Statistics and Probability
- Vol. 30 (4) , 3193-3223
Boundary adaptive local polynomial conditional density estimators
We begin by introducing a class of conditional density estimators based on local polynomial techniques. The estimators are boundary adaptive and easy to implement. We then study the (pointwise and) uniform statistical properties of the estimators, offering characterizations of both probability concentration and distributional approximation. In particular, we establish uniform convergence rates in probability and valid Gaussian distributional approximations for the Studentized t-statistic process. We also discuss implementation issues such as consistent estimation of the covariance function for the Gaussian approximation, optimal integrated mean squared error bandwidth selection, and valid robust bias-corrected inference. We illustrate the applicability of our results by constructing valid confidence bands and hypothesis tests for both parametric specification and shape constraints, explicitly characterizing their approximation errors. A companion R software package implementing our main results is provided.
- 30 July 2024
- Published by Bernoulli Society for Mathematical Statistics and Probability
- Vol. 30 (4) , 3193-3223
Boundary adaptive local polynomial conditional density estimators
We begin by introducing a class of conditional density estimators based on local polynomial techniques. The estimators are boundary adaptive and easy to implement. We then study the (pointwise and) uniform statistical properties of the estimators, offering characterizations of both probability concentration and distributional approximation. In particular, we establish uniform convergence rates in probability and valid Gaussian distributional approximations for the Studentized t-statistic process. We also discuss implementation issues such as consistent estimation of the covariance function for the Gaussian approximation, optimal integrated mean squared error bandwidth selection, and valid robust bias-corrected inference. We illustrate the applicability of our results by constructing valid confidence bands and hypothesis tests for both parametric specification and shape constraints, explicitly characterizing their approximation errors. A companion R software package implementing our main results is provided.
- Open Access
- 13 May 2024
- Published by Cambridge University Press (CUP) in Econometric Theory
- Vol. 41 (5) , 1129-1162
TESTING LIMITED OVERLAP
Extreme propensity scores arise in observational studies when treated and control units have very different characteristics. This is commonly referred to as limited overlap . In this paper, we propose a formal statistical test that helps assess the degree of limited overlap. Rejecting the null hypothesis in our test indicates either no or very mild degree of limited overlap and hence reassures that standard treatment effect estimators will be well behaved. One distinguishing feature of our test is that it only requires the use of a few extreme propensity scores, which is in stark contrast to other methods that require consistent estimates of some tail index. Without the need to extrapolate using observations far away from the tail, our procedure is expected to exhibit excellent size properties, a result that is also borne out in our simulation study.
- 13 May 2024
- Published by Cambridge University Press (CUP) in Econometric Theory
- Vol. 41 (5) , 1129-1162
Open AccessTESTING LIMITED OVERLAP
Extreme propensity scores arise in observational studies when treated and control units have very different characteristics. This is commonly referred to as limited overlap . In this paper, we propose a formal statistical test that helps assess the degree of limited overlap. Rejecting the null hypothesis in our test indicates either no or very mild degree of limited overlap and hence reassures that standard treatment effect estimators will be well behaved. One distinguishing feature of our test is that it only requires the use of a few extreme propensity scores, which is in stark contrast to other methods that require consistent estimates of some tail index. Without the need to extrapolate using observations far away from the tail, our procedure is expected to exhibit excellent size properties, a result that is also borne out in our simulation study.
- Open Access
Boundary Adaptive Local Polynomial Conditional Density Estimators
We begin by introducing a class of conditional density estimators based on local polynomial techniques. The estimators are boundary adaptive and easy to implement. We then study the (pointwise and) uniform statistical properties of the estimators, offering characterizations of both probability concentration and distributional approximation. In particular, we establish uniform convergence rates in probability and valid Gaussian distributional approximations for the Studentized t-statistic process. We also discuss implementation issues such as consistent estimation of the covariance function for the Gaussian approximation, optimal integrated mean squared error bandwidth selection, and valid robust bias-corrected inference. We illustrate the applicability of our results by constructing valid confidence bands and hypothesis tests for both parametric specification and shape constraints, explicitly characterizing their approximation errors. A companion R software package implementing our main results is provided.
Open AccessBoundary Adaptive Local Polynomial Conditional Density Estimators
We begin by introducing a class of conditional density estimators based on local polynomial techniques. The estimators are boundary adaptive and easy to implement. We then study the (pointwise and) uniform statistical properties of the estimators, offering characterizations of both probability concentration and distributional approximation. In particular, we establish uniform convergence rates in probability and valid Gaussian distributional approximations for the Studentized t-statistic process. We also discuss implementation issues such as consistent estimation of the covariance function for the Gaussian approximation, optimal integrated mean squared error bandwidth selection, and valid robust bias-corrected inference. We illustrate the applicability of our results by constructing valid confidence bands and hypothesis tests for both parametric specification and shape constraints, explicitly characterizing their approximation errors. A companion R software package implementing our main results is provided.
- Open Access
- 4 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 5867
Causal Inference and Machine Learning in Practice: Use Cases for Product, Brand, Policy and Beyond
The increasing demand for data-driven decision-making has led to the rapid growth of machine learning applications in various industries. However, the ability to draw causal inferences from observational data remains a crucial challenge. In recent years, causal inference has emerged as a powerful tool for understanding the effects of interventions in complex systems. Combining causal inference with machine learning has the potential to provide a deeper understanding of the underlying mechanisms and to develop more effective solutions to real-world problems. This workshop aims to bring together researchers and practitioners from academia and industry to share their experiences and insights on applying causal inference and machine learning techniques to real-world problems in the areas of product, brand, policy, and beyond. The workshop welcomes original research that covers machine learning theory, deep learning, causal inference, and online learning. Additionally, the workshop encourages topics that address scalable system design, algorithm bias, and interpretability. Through keynote talks, panel discussions, and contributed talks and posters, the workshop will provide a forum for discussing the latest advances and challenges in applying causal inference and machine learning to real-world problems. The workshop will also offer opportunities for networking and collaboration among researchers and practitioners working in industry, government, and academia.
- 4 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 5867
Open AccessCausal Inference and Machine Learning in Practice: Use Cases for Product, Brand, Policy and Beyond
The increasing demand for data-driven decision-making has led to the rapid growth of machine learning applications in various industries. However, the ability to draw causal inferences from observational data remains a crucial challenge. In recent years, causal inference has emerged as a powerful tool for understanding the effects of interventions in complex systems. Combining causal inference with machine learning has the potential to provide a deeper understanding of the underlying mechanisms and to develop more effective solutions to real-world problems. This workshop aims to bring together researchers and practitioners from academia and industry to share their experiences and insights on applying causal inference and machine learning techniques to real-world problems in the areas of product, brand, policy, and beyond. The workshop welcomes original research that covers machine learning theory, deep learning, causal inference, and online learning. Additionally, the workshop encourages topics that address scalable system design, algorithm bias, and interpretability. Through keynote talks, panel discussions, and contributed talks and posters, the workshop will provide a forum for discussing the latest advances and challenges in applying causal inference and machine learning to real-world problems. The workshop will also offer opportunities for networking and collaboration among researchers and practitioners working in industry, government, and academia.