Simulating Effects of Noise Dose Error and Sample Size on Dose-Response Models for Sonic Booms

Error in noise dose estimates can degrade results of community noise surveys. Accurate dose-response models are essential for informing regulators in their noise policy decisions. Increasing the number of survey participants, i.e., sample size, can improve precision but does not alleviate the effect of dose error. During the Quesst mission, NASA will survey community annoyance response as the X-59 aircraft flies supersonically over communities in the USA. To determine the practical impact of dose error on community noise survey results, this paper simulates dose-response data with varying levels of error and sample size. Improvements to precision are found to be minimal for a sample size greater than 2000 participants. In the absence of dose error, a sampled population of 2000 results in a precision of ±2 dB in the dose that elicits 3% high annoyance. Additionally, this study determines that when the dose error exceeds 2 dB, the estimated dose-response curve deviates outside the confidence interval of the true dose-response model.