The sensitivity of hearing-impaired adults to acoustic attributes in simulated rooms

Abstract

In previous studies we have shown that older hearing-impaired individuals are relatively insensitive to changes in the apparent width of broadband noises when those width changes were based on differences in interaural coherence [W. Whitmer, B. Seeber and M. Akeroyd, J. Acoust. Soc. Am. 132, 369-379 (2012)]. This insensitivity has been linked to senescent difficulties in resolving binaural fine-structure differences. It is therefore possible that interaural coherence, despite its widespread use, may not be the best acoustic surrogate of spatial perception for the aged and impaired. To test this, we simulated the room impulse responses for various acoustic scenarios with differing coherence and lateral (energy) fraction attributes using room modelling software (ODEON). Bilaterally impaired adult participants were asked to sketch the perceived size of speech tokens and musical excerpts that were convolved with these impulse responses and presented to them in a sound-dampened enclosure through a 24-loudspeaker array. Participants' binaural acuity was also measured using an interaural phase discrimination task. Corroborating our previous findings, the results showed less sensitivity to interaural coherence in the auditory source width judgments of older hearing-impaired individuals, indicating that alternate acoustic measurements in the design of spaces for the elderly may be necessary.

The Association Between the Processing of Binaural Temporal-Fine-Structure Information and Audiometric Threshold and Age: A Meta-Analysis

“…Four studies ( Oberfeld & Klöckner-Nowotny, 2016 ; Rost et al., 2018; Whitmer & Akeroyd, 2013 ; Whitmer, Seeber, & Akeroyd, 2014 ) used a three-down, one-up rule, thereby tracking a higher performance level (i.e., 79% correct) than for the other studies. As described in the Derivation of a Performance Measure section, the raw data produced by the TFS-LF test (thresholds expressed in degrees or percent-correct scores for the constant-stimulus procedure) were transformed into values of d ′.…”

“…The higher d ′ value of 1.14 corresponding to 79% correct was used for the four studies that used the three-down, one-up rule, thus allowing comparison across data sets obtained using different stepping rules. For listeners who did not score above chance in the constant-stimulus procedure, d ′ was set to 0 (this happened only in 13 cases; Whitmer & Akeroyd, 2013 ; Whitmer et al., 2014 ).…”