Predicting prevalence of noise annoyance using the CTL method

Dose-response functions that attempt to relate the prevalence of a consequential degree of transportation noise-induced annoyance to long term cumulative noise exposure levels have typically been derived by descriptive statistical methods. These methods are not based on understanding of the effects of noise on people, but only on the correlation between predictor and predicted variables. They have no explanatory value to aid understanding of the relationship between exposure and annoyance and offer little guidance for research intended to improve the accuracy and precision of predictions. In this paper a first principles alternative to descriptive analysis is described which requires fewer assumptions than
regression analysis, suggests further research that can lead to genuine understanding of community-level effects of transportation noise, and yields average dosage-response relationships that agree closely with
modern, source-specific curve fits