Preliminary comparison of second-formant discrimination thresholds in cochlear implant users and young normal-hearing listeners

Abstract

Formant discrimination thresholds (FDTs) may provide insights regarding factors limiting vowel perception by cochlear implant (CI) users, but have not been systematically studied in this population. In the present study, estimates of second-formant (F2) FDTs obtained in three CI users were compared to FDTs obtained from three young normal-hearing (YNH) listeners. Procedures and stimuli were modeled after Kewley-Port and Watson (1994, JASA 95, 485-96) but employed fewer trials and an expanded F2 frequency range. Stimuli were formant-synthesized versions of three target vowels. FDTs were estimated using an adaptive 3AFC task with feedback and based on six consecutive 80-trial stimulus blocks. FDTs for the three YNH listeners were comparable to previously reported FDTs (2.4% of reference frequency versus 1.5% in Kewley-Port and Watson). FDTs for two of the CI users were about 70% larger than the average for the YNH listeners. FDTs for the third CI user approached YNH average values in one frequency region but were enlarged in another region. Data for this CI user could not be explained by place-pitch thresholds (obtained in a previous study) and suggest that CI users' spectral acuity for complex stimuli may not be directly predictable from measures of spectral acuity for simple stimuli.