Audibility equalization reduces but cannot eliminate hearing-loss effects on speech-in-speech

Evidence

In the first experiment, 4FAHL was negatively correlated with SRM in both the low SL (r = ‑0.33, p = 0.05) and high SL (r = ‑0.39, p = 0.02) conditions. In the second experiment, 4FAHL was not significantly correlated with SRM (r = ‑0.10, p = 0.57). In summary, increasing SL improved performance and increased SRM for all listeners. Furthermore, careful equalization of audibility across listeners reduced the effects of hearing loss on SRM. On the other hand, no manipulation was able to guarantee equal performance across listeners with various degrees of hearing loss. This suggests that while audibility is clearly an important factor, other factors may impact speech-in-speech intelligibility (e.g. aging, auditory filter width, or comorbidities associated with cognition, working memory and attention). — Dijk, Pim van, Physiology, Psychoacoustics and…, 1726

Categorization

• DEEPTHINKING: The finding that equalizing audibility removes the SRM-hearing-loss correlation yet leaves group threshold differences intact is a carefully delimited, non-obvious result pointing at non-audibility factors (aging, filter width, cognition) — a paradigm-shaping boundary for the ‘it’s all audibility’ hypothesis.
• QUESTION: Which specific non-audibility factors — cochlear nonlinearity, efferent function, binaural processing, cognitive capacity — drive the residual hearing-loss effect after equalization? This demands follow-up.
• SEED: An essay on the limits of ‘making speech audible’ as a hearing-rehab end-goal: audibility is necessary but not sufficient, and the talker-perceptual boundary lies in supra-threshold processing.