Effects of listener bias on auditory acuity for aircraft in real-world ambient environments

Abstract

Hoglund et al. (2010) investigated the ability of listeners to detect the presence of aircraft masked by ongoing ambient sounds using a two interval forced choice (2IFC) procedure. They found that the signal-to-noise ratio required for target detection varied across the different types of ambient environments. Recordings of helicopters in flight were used as target signals and maskers were recorded in rural, suburban and urban locations. Their goal was to better approximate real-world conditions. The goal of the current study is to extend those results to include factors that may bias the listener under more realistic conditions. The 2IFC procedure is designed to minimize listener bias; however, real-world listening conditions are more typically one interval situations. The frequency of occurrence of aircraft over-flights and the costs of errors and rewards for correct responses may substantially affect some estimates of listener sensitivity. Work reported here investigated the influence of apriori probability of target occurrence and manipulation of the pay-off matrix on the acuity measures reported by Hoglund, et al., using the same target sounds and environmental maskers. Psychometric functions shifted by ~18 dB as frequency of targets varied from 20% to 80%. ROC curves display the influence of pay-off manipulations.