Recognition of Time-Compressed and Natural Speech With Selective Temporal Enhancements by Young and Elderly Listeners

Peter J. Fitzgibbons Gallaudet University, Washington, DC Sarah A. Friedman University of Maryland, College Park Contact author: Sandra Gordon-Salant, Department of Hearing and Speech Sciences, 100 LeFrak Hall, University of Maryland, College Park, MD 20742. E-mail: sgordon{at}hesp.umd.edu . Purpose: The goal of this experiment was to determine whether selective slowing of speech segments improves recognition performance by young and elderly listeners. The hypotheses were (a) the benefits of time expansion occur for rapid speech but not for natural-rate speech, (b) selective time expansion of consonants produces greater score increments than other forms of selective time expansion, and (c) older listeners benefit from time expansion of speech Method: Participants ( n = 10–16 per group) were younger and older adults with normal hearing or with hearing loss. A repeated-measures design was used to assess recognition of sentence-length stimuli presented in 2 baseline speech rates: natural and 50% time compression. Selective time expansion of consonants, vowels, or pauses was applied to the natural-rate and time-compressed sentence-length stimuli. Results: Listeners showed excellent performance for natural-rate speech, regardless of time-expansion method. Recognition was significantly poorer for the time-compressed sentences, but performance by elderly listeners and listeners with hearing loss improved with selective time expansion, particularly when applied to consonant segments. Conclusion: The findings support the hypothesis that older listeners and listeners with hearing impairment benefit from selective time expansion of consonants applied to rapid speech, without a corresponding decrement when applied to normal-rate speech. KEY WORDS: geriatrics, sensorineural hearing loss, speech perception CiteULike Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?