USE OF A CORRELATIONAL METHOD TO ESTIMATE A LISTENERS WEIGHTING FUNCTION FOR SPEECH

The purpose of this study was to determine if it is feasible to use th e correlational method (Lutfi, 1995; Richards and Zhu, 1994) to estima te how listeners use or weight the information contained within variou s frequency bands of speech. Three naturally spoken vowel-consonant-vo wel (VCV) syllables (/aba/, /aga/, and /ada/) were presented monaurall y to listeners. Each of the VCV waveforms were filtered into three sep arate frequency bands (i.e., low, mid, and high). Each band was then i ndependently and randomly degraded at various signal-to-noise (S/N) le vels (-7, -5, -3, -1, or +1). On each trial, listeners were asked to i dentify the VCV that was presented to them. For each trial, the S/N le vel of each frequency band, the stimulus that was presented, and the l istener's responses were all recorded and stored in a file. From this trial-by-trial data, a point biserial correlation was computed between the listener's response (correct or incorrect identification) and the degradation within each frequency band. The stronger the correlation, the greater influence that given frequency band had on the listener's performance on the task. From these relations it was shown that it is possible to obtain a listener's weighting function for speech. Result s showed that although most listeners weighted the mid-frequency band the greatest, several of the listeners used different weighting strate gies to perform the task. Several methodological issues are discussed in regard to improving the future application of the correlational met hod to speech. (C) 1996 Acoustical Society of America.