Effects of Training on Speech Recognition Performance in Noise Using Lexically Hard Words Purpose This study examined how repeated presentations of lexically difficult words within a background noise affect a listener’s ability to understand both trained (lexically difficult) and untrained (lexically easy) words in isolation and within sentences. Method In the 1st experiment, 9 young listeners with normal hearing completed a ... Research Article
Research Article  |   February 01, 2007
Effects of Training on Speech Recognition Performance in Noise Using Lexically Hard Words
 
Author Affiliations & Notes
  • Matthew H. Burk
    Indiana University—Bloomington
  • Larry E. Humes
    Indiana University—Bloomington
  • Contact author: Matthew H. Burk, Department of Speech and Hearing Sciences, Indiana University, 200 South Jordan Avenue, Bloomington, IN 47405-7002. E-mail: maburk@indiana.edu.
Article Information
Hearing & Speech Perception / Acoustics / Hearing / Research Articles
Research Article   |   February 01, 2007
Effects of Training on Speech Recognition Performance in Noise Using Lexically Hard Words
Journal of Speech, Language, and Hearing Research, February 2007, Vol. 50, 25-40. doi:10.1044/1092-4388(2007/003)
History: Received January 17, 2006 , Accepted June 2, 2006
 
Journal of Speech, Language, and Hearing Research, February 2007, Vol. 50, 25-40. doi:10.1044/1092-4388(2007/003)
History: Received January 17, 2006; Accepted June 2, 2006
Web of Science® Times Cited: 23

Purpose This study examined how repeated presentations of lexically difficult words within a background noise affect a listener’s ability to understand both trained (lexically difficult) and untrained (lexically easy) words in isolation and within sentences.

Method In the 1st experiment, 9 young listeners with normal hearing completed a short-term auditory training protocol (5 hr) while 8 other young listeners with normal hearing completed a similar protocol lasting about 15 hr in the 2nd experiment. All training made use of multiple talkers and was in a closed-set condition. Feedback was provided on a trial-to-trial basis and consisted of either orthographic or orthographic and auditory feedback. Performance on both the trained and untrained words in isolation and within sentences was measured pre- and posttraining.

Results Listeners' performance improved significantly for the trained words in an open and closed-set condition, as well as the untrained words in the closed-set condition. Although there was no mean improvement in the number of keywords identified within sentences posttraining, 50% of the listeners who completed the long-term training showed improvement beyond the 95% critical difference.

Conclusions With enough training on isolated words, individual listeners can generalize knowledge gained through isolated word training to the recognition of lexically similar words in running speech.

Acknowledgments
A portion of this article was presented at the 2005 Acoustical Society of America Meeting in Vancouver, British Columbia, Canada. This work was supported, in part, by National Institutes of Health (NIH) Grant R01 AG08293, awarded to the second author, and by NIH Postdoctoral Training Grant T32 DC00012, awarded to the first author. The authors would like to thank Lauren Strauser for her assistance in data collection and to the participants, especially those in the long-term training protocol used in Experiment 2, for their dedication.
Order a Subscription
Pay Per View
Entire Journal of Speech, Language, and Hearing Research content & archive
24-hour access
This Article
24-hour access