Proxy-Guided Measurement Calibration

Saketh Vishnubhatla, Shu Wan, Andre Harrison, Adrienne Raglin, Huan Liu
Proceedings of the Fifth Conference on Causal Learning and Reasoning, PMLR 323:1604-1634, 2026.

Abstract

Aggregate outcome variables collected through surveys and administrative records are often subject to systematic measurement error. For instance, in disaster loss databases, county-level losses reported may differ from the true damages due to variations in on-the-ground data collection capacity, reporting practices, and event characteristics. Such miscalibration complicates downstream analysis and decision-making. We study the problem of outcome miscalibration and propose a framework guided by proxy variables for estimating and correcting the systematic errors. We model the data-generating process using a causal graph that separates latent content variables driving the true outcome from the latent bias variables that induce systematic errors. The key insight is that proxy variables that depend on the true outcome but are independent of the bias mechanism provide identifying information for quantifying the bias. Leveraging this structure, we introduce a two-stage approach that utilizes variational autoencoders to disentangle content and bias latents, enabling us to estimate the effect of bias on the outcome of interest. We analyze the assumptions underlying our approach and evaluate it on synthetic data, semi-synthetic datasets derived from randomized trials, and a real-world case study of disaster loss reporting. Our code will be publicly available.

Cite this Paper

BibTeX
@InProceedings{pmlr-v323-vishnubhatla26a, title = {Proxy-Guided Measurement Calibration}, author = {Vishnubhatla, Saketh and Wan, Shu and Harrison, Andre and Raglin, Adrienne and Liu, Huan}, booktitle = {Proceedings of the Fifth Conference on Causal Learning and Reasoning}, pages = {1604--1634}, year = {2026}, editor = {Mazaheri, Bijan and Hanson, Niels Richard}, volume = {323}, series = {Proceedings of Machine Learning Research}, month = {06--08 Apr}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v323/main/assets/vishnubhatla26a/vishnubhatla26a.pdf}, url = {https://proceedings.mlr.press/v323/vishnubhatla26a.html}, abstract = {Aggregate outcome variables collected through surveys and administrative records are often subject to systematic measurement error. For instance, in disaster loss databases, county-level losses reported may differ from the true damages due to variations in on-the-ground data collection capacity, reporting practices, and event characteristics. Such miscalibration complicates downstream analysis and decision-making. We study the problem of outcome miscalibration and propose a framework guided by proxy variables for estimating and correcting the systematic errors. We model the data-generating process using a causal graph that separates latent content variables driving the true outcome from the latent bias variables that induce systematic errors. The key insight is that proxy variables that depend on the true outcome but are independent of the bias mechanism provide identifying information for quantifying the bias. Leveraging this structure, we introduce a two-stage approach that utilizes variational autoencoders to disentangle content and bias latents, enabling us to estimate the effect of bias on the outcome of interest. We analyze the assumptions underlying our approach and evaluate it on synthetic data, semi-synthetic datasets derived from randomized trials, and a real-world case study of disaster loss reporting. Our code will be publicly available.} }
Endnote
%0 Conference Paper %T Proxy-Guided Measurement Calibration %A Saketh Vishnubhatla %A Shu Wan %A Andre Harrison %A Adrienne Raglin %A Huan Liu %B Proceedings of the Fifth Conference on Causal Learning and Reasoning %C Proceedings of Machine Learning Research %D 2026 %E Bijan Mazaheri %E Niels Richard Hanson %F pmlr-v323-vishnubhatla26a %I PMLR %P 1604--1634 %U https://proceedings.mlr.press/v323/vishnubhatla26a.html %V 323 %X Aggregate outcome variables collected through surveys and administrative records are often subject to systematic measurement error. For instance, in disaster loss databases, county-level losses reported may differ from the true damages due to variations in on-the-ground data collection capacity, reporting practices, and event characteristics. Such miscalibration complicates downstream analysis and decision-making. We study the problem of outcome miscalibration and propose a framework guided by proxy variables for estimating and correcting the systematic errors. We model the data-generating process using a causal graph that separates latent content variables driving the true outcome from the latent bias variables that induce systematic errors. The key insight is that proxy variables that depend on the true outcome but are independent of the bias mechanism provide identifying information for quantifying the bias. Leveraging this structure, we introduce a two-stage approach that utilizes variational autoencoders to disentangle content and bias latents, enabling us to estimate the effect of bias on the outcome of interest. We analyze the assumptions underlying our approach and evaluate it on synthetic data, semi-synthetic datasets derived from randomized trials, and a real-world case study of disaster loss reporting. Our code will be publicly available.
APA
Vishnubhatla, S., Wan, S., Harrison, A., Raglin, A. & Liu, H.. (2026). Proxy-Guided Measurement Calibration. Proceedings of the Fifth Conference on Causal Learning and Reasoning, in Proceedings of Machine Learning Research 323:1604-1634 Available from https://proceedings.mlr.press/v323/vishnubhatla26a.html.

Related Material