Two case studies of perceived age, perceived speech quality, and speech acoustics

Abstract

Physiological processes involving speech anatomical structures change with age and affect voice and speech production. Additionally, those with high voice demand (e.g., singers, teachers, coaches) have also demonstrated voice and speech production related changes. From a listener’s perspective, speech changes can be used to give insight into the vocal health and the age of a talker. This study compared perceived production quality, perceived age, and speech acoustics of two elderly individuals. Speech samples were used from two public figures. The first person (F, 78 y/old) gave an 8-h continuous public address. The second person (M, 49-98 y/old), spoke publicly for nearly 50 years. Using speech samples from both individuals, the following analysis were completed: (1) direct age estimation of samples; (2) production quality ratings trained listeners rated the samples voice quality using the GRBAS; and (3) acoustic analysis of the samples was performed. As would be expected, actual age and estimated age were related for the longitudinal samples representing 50 years of aging. Also, speech production quality and speech acoustic characteristics also changed with age. Interestingly, analyses of estimated vocal age was shown to decrease with prolonged vocal speaking showed a negative relationship (younger estimated age after hours of speaking).