Luke Hagar
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
Luke Hagar has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Nathaniel Tyler Stevens
Department of Statistics and Actuarial Science, University of Waterloo
9 shared publications Nathaniel T. Stevens
Department of Statistics & Actuarial Science University of Waterloo N2L 3G1 Waterloo ON Canada
2 shared publications- Alex Deng
Airbnb, Seattle, WA, USA
1 shared publication - Alex Deng
Seattle Airbnb San Francisco Airbnb San Francisco Airbnb San Francisco WA CA CA CA USA
1 shared publication - Amit Gandhi
University of Pennsylvania, Philadelphia, PA, USA
1 shared publication - Amit Gandhi
University of Waterloo
1 shared publication
Latest Publications
- Open Access
- 20 November 2026
- Published by SAGE Publications in Statistical Methods in Medical Research
An efficient framework for robust sample size determination
In many settings, robust data analysis involves computational methods for uncertainty quantification and statistical inference. To design frequentist studies that leverage robust analysis methods, suitable sample sizes to achieve desired power are often found by estimating sampling distributions of p -values via intensive simulation. Moreover, most sample size recommendations rely heavily on assumptions about a single data-generating process. Consequently, robustness in data analysis does not by itself imply robustness in study design, as examining sample size sensitivity to data-generating assumptions typically requires further simulations. We propose an economical alternative for determining sample sizes that are robust to multiple data-generating mechanisms. Applying our theoretical results that model p -values as a function of the sample size, we assess power across the sample size space using simulations conducted at only two sample sizes for each data-generating mechanism. We demonstrate the broad applicability of our methodology to study design based on M-estimators in both experimental and observational settings through a varied set of clinical examples.
- 20 November 2026
- Published by SAGE Publications in Statistical Methods in Medical Research
Open AccessAn efficient framework for robust sample size determination
In many settings, robust data analysis involves computational methods for uncertainty quantification and statistical inference. To design frequentist studies that leverage robust analysis methods, suitable sample sizes to achieve desired power are often found by estimating sampling distributions of p -values via intensive simulation. Moreover, most sample size recommendations rely heavily on assumptions about a single data-generating process. Consequently, robustness in data analysis does not by itself imply robustness in study design, as examining sample size sensitivity to data-generating assumptions typically requires further simulations. We propose an economical alternative for determining sample sizes that are robust to multiple data-generating mechanisms. Applying our theoretical results that model p -values as a function of the sample size, we assess power across the sample size space using simulations conducted at only two sample sizes for each data-generating mechanism. We demonstrate the broad applicability of our methodology to study design based on M-estimators in both experimental and observational settings through a varied set of clinical examples.
- 27 August 2026
- Published by Taylor & Francis in Journal of the American Statistical Association
- p. 1-82
Group Sequential Design with Posterior and Posterior Predictive Probabilities
1 IntroductionScientific experiments drive progress across a wide range of disciplines, including medicine, manufacturing, public transportation, and consumer marketing. While exploring new interve...
- 27 August 2026
- Published by Taylor & Francis in Journal of the American Statistical Association
- p. 1-82
Group Sequential Design with Posterior and Posterior Predictive Probabilities
1 IntroductionScientific experiments drive progress across a wide range of disciplines, including medicine, manufacturing, public transportation, and consumer marketing. While exploring new interve...
- 1 August 2026
- Published by Institute of Mathematical Statistics in Statistical Science
- Vol. 41 (3) , 507-523
Posterior Ramifications of Prior Dependence Structures
Prior elicitation methods for Bayesian analyses transfigure prior information into quantifiable prior distributions. Recently, methods that leverage copulas have been proposed to accommodate more flexible dependence structures when eliciting multivariate priors. We show that the posterior cannot retain many of these flexible prior dependence structures in large-sample settings, and we emphasize that it is our responsibility as statisticians to communicate this to practitioners. We therefore overview objectives for prior specification that guide conversations between statisticians and practitioners to promote alignment between the flexibility in the prior dependence structure and the objectives for posterior analysis. Because correctly specifying the dependence structure a priori can be difficult, we consider how the choice of prior copula impacts the posterior distribution in terms of asymptotic convergence of the posterior mode. Our resulting recommendations clarify when it is useful to elicit intricate prior dependence structures and when it is not.
- 1 August 2026
- Published by Institute of Mathematical Statistics in Statistical Science
- Vol. 41 (3) , 507-523
Posterior Ramifications of Prior Dependence Structures
Prior elicitation methods for Bayesian analyses transfigure prior information into quantifiable prior distributions. Recently, methods that leverage copulas have been proposed to accommodate more flexible dependence structures when eliciting multivariate priors. We show that the posterior cannot retain many of these flexible prior dependence structures in large-sample settings, and we emphasize that it is our responsibility as statisticians to communicate this to practitioners. We therefore overview objectives for prior specification that guide conversations between statisticians and practitioners to promote alignment between the flexibility in the prior dependence structure and the objectives for posterior analysis. Because correctly specifying the dependence structure a priori can be difficult, we consider how the choice of prior copula impacts the posterior distribution in terms of asymptotic convergence of the posterior mode. Our resulting recommendations clarify when it is useful to elicit intricate prior dependence structures and when it is not.
- Open Access
- 23 March 2026
- Published by Taylor & Francis in Journal of Business & Economic Statistics
- Vol. 44 (3) , 1-10
Design of Bayesian A/B Tests Controlling False Discovery Rates and Power
- 23 March 2026
- Published by Taylor & Francis in Journal of Business & Economic Statistics
- Vol. 44 (3) , 1-10
Open AccessDesign of Bayesian A/B Tests Controlling False Discovery Rates and Power
- Open Access
- 12 March 2026
- Published by Wiley in Statistics in Medicine
- Vol. 45 (6-7) , e70488
Design of Bayesian Clinical Trials With Clustered Data
In the design of clinical trials, it is essential to assess the design operating characteristics (e.g., power and the type I error rate). Common practice for the evaluation of operating characteristics in clinical trials that employ Bayesian analysis and decision procedures relies on estimating the sampling distribution of posterior summaries via Monte Carlo simulation. It is computationally intensive to repeat this estimation process for each design configuration considered, particularly for clustered data that are analyzed using complex, high‐dimensional models. In this paper, we propose an efficient method to assess operating characteristics and determine sample sizes for Bayesian trials with clustered data. We prove theoretical results that enable posterior probabilities to be modeled as a function of the number of clusters. Using these functions, we assess operating characteristics at a range of sample sizes given simulations conducted at only two values for the number of clusters. These theoretical results are also leveraged to quantify the impact of simulation variability on our sample size recommendations. The applicability of our methodology is illustrated using an example Bayesian cluster‐randomized clinical trial.
- 12 March 2026
- Published by Wiley in Statistics in Medicine
- Vol. 45 (6-7) , e70488
Open AccessDesign of Bayesian Clinical Trials With Clustered Data
In the design of clinical trials, it is essential to assess the design operating characteristics (e.g., power and the type I error rate). Common practice for the evaluation of operating characteristics in clinical trials that employ Bayesian analysis and decision procedures relies on estimating the sampling distribution of posterior summaries via Monte Carlo simulation. It is computationally intensive to repeat this estimation process for each design configuration considered, particularly for clustered data that are analyzed using complex, high‐dimensional models. In this paper, we propose an efficient method to assess operating characteristics and determine sample sizes for Bayesian trials with clustered data. We prove theoretical results that enable posterior probabilities to be modeled as a function of the number of clusters. Using these functions, we assess operating characteristics at a range of sample sizes given simulations conducted at only two values for the number of clusters. These theoretical results are also leveraged to quantify the impact of simulation variability on our sample size recommendations. The applicability of our methodology is illustrated using an example Bayesian cluster‐randomized clinical trial.
- Open Access
An Efficient Framework for Robust Sample Size Determination
In many settings, robust data analysis involves computational methods for uncertainty quantification and statistical inference. To design frequentist studies that leverage robust analysis methods, suitable sample sizes to achieve desired power are often found by estimating sampling distributions of p-values via intensive simulation. Moreover, most sample size recommendations rely heavily on assumptions about a single data-generating process. Consequently, robustness in data analysis does not by itself imply robustness in study design, as examining sample size sensitivity to data-generating assumptions typically requires further simulations. We propose an economical alternative for determining sample sizes that are robust to multiple data-generating mechanisms. Applying our theoretical results that model p-values as a function of the sample size, we assess power across the sample size space using simulations conducted at only two sample sizes for each data-generating mechanism. We demonstrate the broad applicability of our methodology to study design based on M-estimators in both experimental and observational settings through a varied set of clinical examples.
Open AccessAn Efficient Framework for Robust Sample Size Determination
In many settings, robust data analysis involves computational methods for uncertainty quantification and statistical inference. To design frequentist studies that leverage robust analysis methods, suitable sample sizes to achieve desired power are often found by estimating sampling distributions of p-values via intensive simulation. Moreover, most sample size recommendations rely heavily on assumptions about a single data-generating process. Consequently, robustness in data analysis does not by itself imply robustness in study design, as examining sample size sensitivity to data-generating assumptions typically requires further simulations. We propose an economical alternative for determining sample sizes that are robust to multiple data-generating mechanisms. Applying our theoretical results that model p-values as a function of the sample size, we assess power across the sample size space using simulations conducted at only two sample sizes for each data-generating mechanism. We demonstrate the broad applicability of our methodology to study design based on M-estimators in both experimental and observational settings through a varied set of clinical examples.
- Open Access
- 16 April 2025
- Published by Taylor & Francis in Journal of the American Statistical Association
- Vol. 120 (552) , 2559-2568
An Economical Approach to Design Posterior Analyses
- 16 April 2025
- Published by Taylor & Francis in Journal of the American Statistical Association
- Vol. 120 (552) , 2559-2568
Open AccessAn Economical Approach to Design Posterior Analyses
- Open Access
- 23 January 2025
- Published by Wiley in Statistics in Medicine
- Vol. 44 (3-4) , e10321
Bioequivalence Design With Sampling Distribution Segments
In bioequivalence design, power analyses dictate how much data must be collected to detect the absence of clinically important effects. Power is computed as a tail probability in the sampling distribution of the pertinent test statistics. When these test statistics cannot be constructed from pivotal quantities, their sampling distributions are approximated via repetitive, time‐intensive computer simulation. We propose a novel simulation‐based method to quickly approximate the power curve for many such bioequivalence tests by efficiently exploring segments (as opposed to the entirety) of the relevant sampling distributions. Despite not estimating the entire sampling distribution, this approach prompts unbiased sample size recommendations. We illustrate this method using two‐group bioequivalence tests with unequal variances and overview its broader applicability in clinical design. All methods proposed in this work can be implemented using the developed dent package in R.
- 23 January 2025
- Published by Wiley in Statistics in Medicine
- Vol. 44 (3-4) , e10321
Open AccessBioequivalence Design With Sampling Distribution Segments
In bioequivalence design, power analyses dictate how much data must be collected to detect the absence of clinically important effects. Power is computed as a tail probability in the sampling distribution of the pertinent test statistics. When these test statistics cannot be constructed from pivotal quantities, their sampling distributions are approximated via repetitive, time‐intensive computer simulation. We propose a novel simulation‐based method to quickly approximate the power curve for many such bioequivalence tests by efficiently exploring segments (as opposed to the entirety) of the relevant sampling distributions. Despite not estimating the entire sampling distribution, this approach prompts unbiased sample size recommendations. We illustrate this method using two‐group bioequivalence tests with unequal variances and overview its broader applicability in clinical design. All methods proposed in this work can be implemented using the developed dent package in R.
- Open Access
- 24 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 4885-4895
Metric Decomposition in A/B Tests
More than a decade ago, CUPED (Controlled Experiments Utilizing Pre-Experiment Data) mainstreamed the idea of variance reduction leveraging pre-experiment covariates. Since its introduction, it has been implemented, extended, and modernized by major online experimentation platforms. Despite the wide adoption, it is known by practitioners that the variance reduction rate from CUPED utilizing pre-experimental data varies case by case and has a theoretical limit. In theory, CUPED can be extended to augment a treatment effect estimator utilizing in-experiment data, but practical guidance on how to construct such an augmentation is lacking. In this article, we fill this gap by proposing a new direction for sensitivity improvement via treatment effect augmentation whereby a target metric of interest is decomposed into components with high signal-to-noise disparity. Inference in the context of this decomposition is developed using both frequentist and Bayesian theory. We provide three real world applications demonstrating different flavors of metric decomposition; these applications illustrate the gain in agility metric decomposition yields relative to an un-decomposed analysis.
- 24 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 4885-4895
Open AccessMetric Decomposition in A/B Tests
More than a decade ago, CUPED (Controlled Experiments Utilizing Pre-Experiment Data) mainstreamed the idea of variance reduction leveraging pre-experiment covariates. Since its introduction, it has been implemented, extended, and modernized by major online experimentation platforms. Despite the wide adoption, it is known by practitioners that the variance reduction rate from CUPED utilizing pre-experimental data varies case by case and has a theoretical limit. In theory, CUPED can be extended to augment a treatment effect estimator utilizing in-experiment data, but practical guidance on how to construct such an augmentation is lacking. In this article, we fill this gap by proposing a new direction for sensitivity improvement via treatment effect augmentation whereby a target metric of interest is decomposed into components with high signal-to-noise disparity. Inference in the context of this decomposition is developed using both frequentist and Bayesian theory. We provide three real world applications demonstrating different flavors of metric decomposition; these applications illustrate the gain in agility metric decomposition yields relative to an un-decomposed analysis.
- Open Access
Bioequivalence Design with Sampling Distribution Segments
In bioequivalence design, power analyses dictate how much data must be collected to detect the absence of clinically important effects. Power is computed as a tail probability in the sampling distribution of the pertinent test statistics. When these test statistics cannot be constructed from pivotal quantities, their sampling distributions are approximated via repetitive, time-intensive computer simulation. We propose a novel simulation-based method to quickly approximate the power curve for many such bioequivalence tests by efficiently exploring segments (as opposed to the entirety) of the relevant sampling distributions. Despite not estimating the entire sampling distribution, this approach prompts unbiased sample size recommendations. We illustrate this method using two-group bioequivalence tests with unequal variances and overview its broader applicability in clinical design. All methods proposed in this work can be implemented using the developed dent package in R.
Open AccessBioequivalence Design with Sampling Distribution Segments
In bioequivalence design, power analyses dictate how much data must be collected to detect the absence of clinically important effects. Power is computed as a tail probability in the sampling distribution of the pertinent test statistics. When these test statistics cannot be constructed from pivotal quantities, their sampling distributions are approximated via repetitive, time-intensive computer simulation. We propose a novel simulation-based method to quickly approximate the power curve for many such bioequivalence tests by efficiently exploring segments (as opposed to the entirety) of the relevant sampling distributions. Despite not estimating the entire sampling distribution, this approach prompts unbiased sample size recommendations. We illustrate this method using two-group bioequivalence tests with unequal variances and overview its broader applicability in clinical design. All methods proposed in this work can be implemented using the developed dent package in R.