Modified psychophysical mapping predicts token-specific hearing thresholds far better than the standard.

Evidence

The predicted SRTs were overestimated by 2.8 dB using P and by only 0.4 dB using Pmod . The token-specific SRTs showed a large spread across speech tokens, which was smaller in both model predictions. However, Pmod captured the relative ranking of token-specific SRTs considerably better than P (Spearman’s r of 0.4 and 0.04, respectively). — Dijk, Pim van, Physiology, Psychoacoustics and…, 5607