Modeling the Relationship Between Listener Factors and Envelope Fidelity: A Pooled Analysis Spanning a Decade
Bibliographic record
Abstract
OBJECTIVES: There is a large variability in speech intelligibility with hearing aids. This variability remains despite the current clinical approaches that provide individualized frequency-specific adjustments to gain in hearing aids. Much of this variability documented in the literature may also be due to differences across studies in terms of outcome measures, test conditions, etc. The objective of this study was to model sources of individual variability in speech intelligibility with hearing aids, in a pooled analysis across four distinct studies that used common methodologies and outcome measures. DESIGN: Deidentified data from 80 unique listeners with bilateral mild to moderately severe sensorineural hearing loss and aged 49 to 92 years were pooled from four published studies. A hierarchical Beta-Binomial (generalized linear mixed-effects) model was implemented to estimate the probability of correct word recognition in the pooled data using a Bayesian framework. Across studies, word recognition was measured for low-context sentences, in multi-talker babble, for a range of signal to noise ratios. Signals were processed through a hearing aid simulator or a wearable device and were customized to the listener's audiogram. Individual studies involved systematic manipulations of wide dynamic range compression, frequency lowering, or microphone directionality. Individual working memory ability was measured using the reading span test. A well-established auditory metric was used to quantify cumulative envelope fidelity (cepstral correlation) from background noise and the hearing aid processing for each listener. RESULTS: The model showed a strong relationship between speech intelligibility and envelope fidelity, confirming previous research findings that higher envelope fidelity was associated with better speech intelligibility. Among the sources of individual variability, working memory had a significant effect on the relationship between speech intelligibility and envelope fidelity. Listeners with higher working memory had significantly better word recognition than those with lower working memory, especially when envelope fidelity was worse. In addition, listeners with lower working memory had better word recognition as envelope fidelity increased. Age and degree of hearing loss (four-frequency pure-tone average) did not have a significant effect on the relationship between speech intelligibility and envelope fidelity. CONCLUSIONS: The analysis of the pooled dataset identified sources of individual variability in aided speech intelligibility, while also overcoming limitations of smaller sample sizes in prior research. The model supported the hypothesis that speech intelligibility is affected by the cumulative envelope fidelity arising from a combination of background noise and hearing aid processing. The study findings indicate that individual variability in speech intelligibility with hearing aid processing is related to working memory after accounting for age and degree of hearing loss. The study highlights the need for individualized treatment of hearing loss beyond the pure tone audiogram. Auditory metrics such as the envelope fidelity metric used in the study may be useful tools in clinical decision-making.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.104 | 0.157 |
| Meta-epidemiology (narrow) | 0.003 | 0.002 |
| Meta-epidemiology (broad) | 0.004 | 0.025 |
| Bibliometrics | 0.005 | 0.003 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.003 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".