ISCA Workshop on Plasticity in Speech Perception (PSP2005)

Senate House, London, UK
June 15-17, 2005

Effects of Adult Ageing Oo Adaptation to Time- and Frequency-Compressed Speech

Jonathan Peelle, Arthur Wingfield

Brandeis University, Waltham, MA, USA

Healthy aging is typically accompanied by declines in efficacy of both peripheral and central auditory processing. Age-related declines are also present in several domains of cognitive ability. Such changes result in decreased comprehension accuracy for older adults relative to their younger counterparts when listening to altered speech, including timecompressed speech, frequency-compressed speech, and speech in broadband noise. These difficulties raise the question of whether older adults may also be differentially impaired in their ability to adapt to these altered stimuli over time. In the current set of experiments we investigated the degree to which older adult listeners were able to adapt to altered speech compared to young adults. Time compression was performed using a variation of the sampling technique as implemented in SoundEdit software (Macromedia, Inc., San Francisco, CA). Small portions of the speech signal were removed at regular intervals, with the remaining segments abutted in time. The resulting signal retains pitch, amplitude, and relative timing cues, but is reproduced in less time. Frequency compression was accomplished by noise-vocoding the speech signal using 16 analysis bands, and shifting the output of each band downwards after vocoding. The large number of bands ensured that the vocoding process itself only minimally reduced intelligibility, and that most of the perceptual difficult was due to the spectral shifting of the speech information. In Experiment 1, we replicated previous findings with young adults and showed that older adults, when equated for starting accuracy, show similar patterns of perceptual adaptation to time-compressed speech, but fail to transfer this learning to a different speech rate. In Experiment 2, participants received a greater amount of practice than previous studies, which benefited only young adults, and then only at later trials. In Experiment 3 participants adapted to spectrally-shifted noise-vocoded speech, demonstrating that age similarities in perceptual learning are not specific to one type of stimulus. Control experiments demonstrated that there were no age differences in interference effects, and that changes in improvement in either group were not a result of simple strategy change or practice with the recall task. We conclude that fast perceptual learning is comparable in young and older adults, but maintenance and transfer of this learning decline with age.


Bibliographic reference.  Peelle, Jonathan / Wingfield, Arthur (2005): "Effects of adult ageing oo adaptation to time- and frequency-compressed speech", In PSP2005, 239.