Loudness judgments at a fixed level vary probabilistically across categories
Evidence
We propose to instead model the probabilistic nature of listener responses; these probabilistic representations can be used to improve loudness measurements and further our understanding of the mechanisms underlying this perception. At each stimulus level, there is a probability distribution across categorical loudness responses; when plotted as a function of stimulus level, these distributions form a multi-category psychometric function (MCPF), modeling the statistics of the listeners’ categorical perception. — Dijk, Pim van, Physiology, Psychoacoustics and…, 2512
The results demonstrate that loudness perception of NH and HL listeners is a random variable at each stimulus level, i.e., each level has a statistical distribution across loudness categories. — Dijk, Pim van, Physiology, Psychoacoustics and…, 2566
A psychometric function describes the probability of a particular response as a function of an experimental variable. We introduce the concept of a multi-category psychometric function, which represents the probability distribution across multiple response categories as a function of an experimental variable. In a MCPF, a family of curves demarcates the probabilities of multiple categories as a function of stimulus level. As an example, we demonstrate the construction of a hypothetical CLS MCPF for a NH listener, for an 11-category CLS scale, in Fig. 1 . Figure 1a shows the listener’s probability distribution across categories at 60 dB SPL. One can see that the loudness judgments are not constrained to one category, but, instead, form a unimodal distribution. — Dijk, Pim van, Physiology, Psychoacoustics and…, 2518
A psychometric function describes the probability of a particular response as a function of an experimental variable. We introduce the concept of a multi-category psychometric function, which represents the probability distribution across multiple response categories as a function of an experimental variable. … Figure 1a shows the listener’s probability distribution across categories at 60 dB SPL. One can see that the loudness judgments are not constrained to one category, but, instead, form a unimodal distribution. — Dijk, Pim van, Physiology, Psychoacoustics and…, 2519
A listener’s CLS function, generally computed from the median stimulus level for each response category, is the measure typically used in analysis. This standard approach treats the variability in the response data as something to be removed. We propose to instead model the probabilistic nature of listener responses; these probabilistic representations can be used to improve loudness measurements and further our understanding of the mechanisms underlying this perception. At each stimulus level, there is a probability distribution across categorical loudness responses; when plotted as a function of stimulus level, these distributions form a multi-category psychometric function (MCPF), modeling the statistics of the listeners’ categorical perception. — Dijk, Pim van, Physiology, Psychoacoustics and…, 2515
We show how the MCPF can be used to improve CLS estimates, by combining listener models with maximum-likelihood (ML) estimation. We also describe how the MCPF could be used in an entropy-based stimulus-selection technique. These techniques utilize the probabilistic nature of categorical perception, a novel usage of this dimension of loudness information, to improve the quality of loudness measurements. — Dijk, Pim van, Physiology, Psychoacoustics and…, 2509
Categorization
- SEED: The idea that perceptual judgments are inherently probabilistic distributions, not points, could anchor an article on how measurement methodology shapes what we know about hearing.
- XREF: Connects directly to this book’s core theme of individual variability in hearing and renders loudness as a distribution rather than a single point — analogous to signal-detection theory’s treatment of thresholds as probabilistic. — connects to: Psychometric functions map continuous stimulus scales onto probabilistic detection judgements — Jennifer Lentz, Psychoacoustics- Perception of Normal…, 284; Equal-loudness contours are a fixed property of human hearing — Toole, Floyd E, Sound reproduction- the acoustics and…, 909; Roughly 75 percent of people have sufficiently normal hearing to make reliable loudspeaker judgments (+5 more)
- DEFINE: The multi-category psychometric function (MCPF) is introduced as a family of curves mapping category probabilities across stimulus level.
- QUESTION: Worth chasing: what physiological variability — neural firing stochasticity, adaptation, criterion shifts — actually generates the response distributions the MCPF models?
- corroborated: 9 independent passages support this claim — Dijk, Pim van, Physiology, Psychoacoustics and…, 2512; Dijk, Pim van, Physiology, Psychoacoustics and…, 2566; Dijk, Pim van, Physiology, Psychoacoustics and…, 2518 (+3 more)