Abstract
A voice parade enables collection of earwitness identification evidence, but the current procedure is costly and time-consuming. Hence, this study investigates whether the number of foils or duration of samples could safely be reduced. Listeners attempted to recognize a voice amongst same-accented foils in target-absent or target-present parades (constructed as per Home Office guidelines, 2003). Six male English speakers were chosen as targets; three SSBE, and one each from York, Bradford and Wakefield. Listeners were exposed to one target for 60s, before hearing a parade of nine voices (Experiment 1) or six voices (Experiment 2) in either 15s, 30s or 60s samples. Overall performance was low, but higher in target-present versus target-absent. Performance also varied by target-speaker. Results suggest samples can safely be reduced from 60s to between 15s and 30s, but parade size should remain at nine-person.
