The discrimination of formant frequency differences as a function of duration by hearing‐impaired listeners

Hearing‐impaired subjects with high‐frequency, noise‐induced hearing loss and normal low‐frequency sensitivity were compared to normal hearing subjects in two discrimination tasks. These tasks were designed to determine formant frequency discrimination thresholds as a function of stimulus duration in steady‐state speech signals. The first task was discrimination of second formant (F2) frequency differences with F2 in the 2200–2840 Hz range, using computer‐synthesized vowels varying in duration in logarithmic steps from 24 to 384 ms. The second task was discriminating first formant (F 1) frequency differences in the 200–360 Hz range for the same durations. Hearing‐impaired subjects took the tests at 35 dB SL (re:SRT) and showed elevated thresholds for discrimination of F2 frequency differences, when compared with the normal‐hearing subjects at both the 35 dB SL and at a 92 dB SPL presentation level. However, these same hearing‐impaired subjects performed similarly to the normal hearing listeners in the F1 discrimination task where hearing sensitivity was normal for both groups. Moreover, the effect of shorter signal duration on formant frequency discrimination appears to be more severe for the impaired listeners.