“Please sort these sounds into 2 identities”: Effects of task instructions on performance in voice sorting studies

Abstract

In the current study we investigated the effects of two types of task instructions on performance of a voice sorting task by familiar and unfamiliar listeners. In Experiment 1, listeners were asked to sort 15 naturally-varying exemplars from two voice identities into as many identities as they perceived. Results replicate the findings of previous voice and face sorting studies: unfamiliar listeners form more clusters than familiar listeners, selectively failing to “tell people together”. That is, unfamiliar listeners erroneously perceive exemplars from the same identity as coming from different identities, while being highly accurate at telling apart the exemplars from different voices. In Experiment 2, we show that a change in task instructions - forcing participants to sort exemplars into two identities only - allows unfamiliar listeners to overcome this selective failure at “telling people together”. This improvement, however, comes at the cost of an increase in errors in telling people apart. For familiar listeners, similar non-significant trends are apparent. The current study thus shows that, even when being given the correct number of identities, listeners may fail to accurately perceive identity. These findings highlight voice identity perception in the context of natural within-person variability as a challenging task for both unfamiliar and familiar listeners.

Sorting through the impact of familiarity when processing vocal identity: Results from a voice sorting task

“…Following the familiarity rating task, participants were introduced to the experimental task. This was described as a voice sorting task, with the method mimicking the free-sort face task used by Jenkins et al (2011) and Andrews et al (2015) and used more recently with voices by Lavan, Burston, and Garrido (2018) and Lavan, Burston, et al (2019) (see also Lavan, Merriman, et al, 2019). Participants were instructed that they would be presented with a set of voice clips, which appeared as loudspeaker icons on a PowerPoint slide.…”

A model of person perception from voices

“…For unfamiliar voice processing, this means that if listeners are given a number of variable samples of unfamiliar voices, their percepts of speaker characteristics may differ for each sample, regardless of the availability of information in the acoustic signal about the ground truth of the speakers' vocal anatomy. This has been demonstrated for voices for perceived identity Lavan, Burston, Ladwa, Merriman, Knight & McGettigan, 2019;Lavan, Merriman, Ladwa, Burston, Knight & McGettigan, 2019). The face perception literature furthermore attests to the influences of variability.…”