Virtual Seminar: Modeling Measurement Error in Dose-Response Models of Community Annoyance to Low-Noise Supersonic Flight

We are fortunate to have Dr. Nathan Cruze of NASA to deliver his seminar "Modeling Measurement Error in Dose-Response Models of Community Annoyance to Low-Noise Supersonic Flight" on April 30 at 12:00. 

If you are in Houston at SPH and would like to use our facilities to attend, the seminar will be broadcast to RAS 102A.

The primary research goal of the forthcoming NASA Quesst mission community test campaign is to collect representative community response data in support of the development of supersonic overflight noise certification standards. Beginning in 2026, NASA will fly the novel X-59 demonstrator aircraft over select communities in United States in order to demonstrate the possibility of low-noise supersonic flight over land and to collect objective measurements and subjective data on the perceptual experience of this new noise source. It is believed that a regression of a binary perceptual response (‘highly annoyed’ or ‘not’) on estimated noise levels (doses, measured in decibels) will provide a useful dose-response relationship for regulators. However, as these estimated doses will be subject to measurement error, naïve estimators of regression coefficients are inconsistent and slopes may be subject to attenuation bias. In this presentation, I contrast functional modeling of measurement error via simulation extrapolation (SIMEX) with structural Bayesian measurement error models.  These methods are applied to available data collected during two NASA risk reduction studies in California in 2011 and Texas in 2018.  I’ll conclude noting that in the presence of nonnegligible measurement errors, probabilities of annoyance may be overpredicted for low noise levels and underpredicted for high noise levels, therefore, methods of correcting for measurement error will be necessary to improve the utility of the dose-response relationship for policy-making purposes.

Dr. Nathan Cruze joined NASA Langley Research Center in 2021 as a senior statistician in the Engineering Directorate supporting the planning and execution of community testing with the X-59 demonstrator aircraft during the Quesst mission.  Prior to joining NASA, he served as a research mathematical statistician at USDA’s National Agricultural Statistics Service for more than eight years, where his research focused on improving crop and economic estimates programs by combining survey and auxiliary data through small area estimation approaches.  During his tenure there, Dr. Cruze led the agency’s successful implementation and conversion of official crops county estimates, county-level cash rental rate estimates, and farm labor estimates products to a system of model-based estimates published with estimates of uncertainty.  His joint publications on these topics have appeared in Statistical Science, JRSS-A, and Survey Methodology, among others.  His interdisciplinary doctoral studies were co-directed by Dr. Prem Goel (statistics) and Dr. Bhavik Bakshi (chemical engineering) at Ohio State University.  Dr. Cruze holds bachelor’s degrees in economics and mathematics and master’s degrees in economics and statistics, also from Ohio State University.  He is part of the 2024 Symposium on Data Science and Statistics conference organizing committee, and he serves as a member on the American Statistical Association’s Committee on Career Development.