Abstract:
The ability of listeners to perceive the age of an unknown and unseen speaker through the voice has been well-documented in the literature. However, variation in study methods and low sample sizes have made it difficult to determine exactly how well listeners can judge the age of a person through the voice alone. The aim of the present study was to investigate the extent to which a listener can perceive speaker age, and identify any factors, perceptual or acoustic, which play a role in determining the accuracy of listener estimation. Data was collected through an online survey. The data from 213 respondents were used to analyse the perceived age of the speakers in the survey. The stimuli used in the present study were lists of words in an h-vowel-d (hVd) frame and the Rainbow Passage by five ‘younger’ speakers (20–26 years, M=23.3, SD=1.8) and five ‘older’ (56–71 years, M=62.5, SD=5.6) male New Zealand English speakers recorded in a previous study. Data were analysed using Pearson’s correlation coefficients, paired sample t-tests, and analyses of variances. The majority of respondents were able to estimate the speakers’ age within two decades of the speakers’ chronological age. The mean absolute error of estimation was 9.2 years across all respondents for all speakers. While voice training and employment were not found to play a role, speaker and listener age, the number of years lived in New Zealand, and the first language of the listener were all found to influence the accuracy of listeners. The acoustic properties of the voice were also found to be a good predictor of perceived age. Increased fundamental frequency and decreased contact quotient tended to correspond to increases in the perceived age of speakers in the present study. The findings of the present study support the theory that listeners can estimate chronological age, but to a lesser degree of accuracy than reported in previous studies. This study also broadens the scope of knowledge on vocal ageing through a unique analysis of the correlations between the acoustic features of the voice and the perceived age of the speaker.