Measuring Talker Age Estimates Through Crowdsourced Listeners' Ratings: A Pilot Study for Voice Research

Area of Science:

• Speech and Hearing Sciences
• Psychology
• Human-Computer Interaction

Background:

• Auditory-perceptual voice research traditionally relies on trained listeners.
• Online crowdsourcing is emerging for behavioral data collection but is underutilized in voice perception studies.
• This study explores the viability of crowdsourcing for voice-specific age estimation.

Purpose of the Study:

• To assess the feasibility of online crowdsourcing for auditory-perceptual voice ratings.
• To replicate a lab-based voice age estimation study using an online platform.
• To evaluate the accuracy of untrained listeners in estimating voice age.

Main Methods:

• Fifty untrained, crowdsourced listeners estimated the age of a single talker.
• Audio samples from 20 speeches spanning 48 years were used.
• Age estimation accuracy was the primary outcome measure.

Main Results:

• Crowdsourced age estimations closely aligned with previous laboratory findings.
• Listeners tended to overestimate younger ages and underestimate older ages.
• The estimated age matched the talker's chronological age at 54 years.

Conclusions:

• Online crowdsourcing is a feasible method for auditory-perceptual voice ratings.
• This approach provides a low-cost, high-participant alternative to traditional methods.
• Further research is needed to validate crowdsourced ratings for complex voice parameters.