Normative Data and Repeatability of a 4-kHz Pure-Tone Psychoacoustic Tuning Curve Using a Single Clinical Audiometer
Article information
Abstract
Background and Objectives
This study aimed to establish normative data and evaluate the test-retest variability of 4 kHz pure-tone psychoacoustic tuning curves (PTCs) measured using a conventional two-channel clinical audiometer.
Materials and Methods
Thirty-four adults with normal hearing (≤25 dB HL) underwent two PTC testing sessions separated by a 2-week interval. A 4,000-Hz signal tone presented at 10 dB above threshold and seven pure-tone maskers (3,150-4,896 Hz) were administered using an Interacoustics AC40 Audiometer through the 3-down/1-up approach. Participants were randomly assigned to either low-to-high or high-to-low masker-frequency-order groups. Test-retest variability was analyzed across masker frequencies, and correlations with participant demographics were determined. Q10 values were evaluated as indices of frequency selectivity.
Results
All participants completed the testing without auditory discomfort. Approximately 80% had a test-retest variability of ≤8 dB, indicating high reliability. The largest intersession differences were observed at 4,896 Hz (9.80±7.35 dB) and 3,150 Hz (5.77±6.95 dB). Test-retest variability was positively correlated with age (r= 0.50, p=0.006). The mean Q10 was 5.76±1.72, and decreased significantly with age (r=-0.49, p=0.014). The average tuning curve exhibited a V-shaped pattern with a tip at 4,000 Hz, consistent with classical findings.
Conclusions
The simplified pure-tone PTC protocol using a standard clinical audiometer provided reproducible and relevant estimates of auditory frequency selectivity. These findings support its clinical feasibility for assessing frequency selectivity and age-related changes in the auditory filter.
Introduction
Psychoacoustic tuning curves (PTCs) are used to assess the frequency selectivity and auditory filter characteristics of the human auditory system [1]. To obtain the PTC, two tones are presented simultaneously. One is a pure tone with a fixed frequency and intensity, which serves as the signal tone, and the other is a masking tone with a variable frequency. The minimum level of the masker that can mask the signal tone (the masking threshold) is determined for each masker frequency. The collection of these threshold values across frequencies constitutes the PTC. The PTC has a V-shaped pattern centered on the signal-tone frequency, and the sharpness of the V shape indicates the degree of frequency selectivity.
Previous and current methods for evaluating PTCs have often required specialized software or complex stimulus generation systems, which limits their applicability in routine clinical settings [2-5]. Obtaining a PTC using a single clinical audiometer may provide a more accessible approach for assessing frequency selectivity in routine otologic practice. However, the reliability and test-retest variability of data obtained through such simplified methods and the extent to which they correspond to established experimental data have not been fully investigated.
Pure tones or narrowband noise (NBN) can be used as PTC maskers [1]. NBN is preferred because it helps eliminate beat phenomena, which facilitates the production of more stable responses near the signal frequency and enables more accurate estimation of auditory filter characteristics [6]. However, pure-tone masking may reveal the intrinsic frequency selectivity of the auditory filter more accurately by minimizing spectral spread and focusing the masking interaction on a single frequency component. The beat phenomenon refers to the periodic amplitude fluctuation that occurs when two tones of slightly different frequencies are presented simultaneously [6]. This phenomenon becomes prominent during pure-tone PTC measurements when the masker frequency approaches the signal frequency [6]. Previous studies have reported minimal beat interference in PTC experiments at 4 kHz compared with lower frequencies such as 500 Hz, 1 kHz, and 2 kHz [6].
The present study aimed to develop and evaluate a simplified sinusoidal PTC protocol using a single clinical audiometer at 4 kHz. The objectives were to 1) examine the test–retest reliability of the sinusoidal PTC and its association with age, hearing threshold, and masker frequency and 2) establish normative data for comparison with previously published findings.
Materials and Methods
Participants
The study protocol was reviewed and approved by the Institutional Review Board of Busan St. Mary’s Hospital (BSM 2025-07). Thirty-four volunteers provided informed consent and participated in the study. All the participants underwent pure-tone audiometry of both ears before the PTC assessment. Only those with normal hearing, based on hearing thresholds of ≤25 dB HL at all frequencies between 500 and 8,000 Hz for both ears, were included. Participants with a history of sudden sensorineural hearing loss but had recovered or those with tinnitus symptoms (Tinnitus Handicap Inventory score of 38 or higher) were excluded. Their demographics are provided in Table 1.
Demographics of the participants
Audiometric settings for the PTC test
All tests were performed using a clinical audiometer (AC40, Interacoustics) with TDH-39 headphones. The two independent channels of the audiometer were used simultaneously to generate the stimuli. Channel 1 was used to deliver the masker tone, while Channel 2 was used to deliver the signal tone. Channel 2 was maintained at 4 kHz using the reference mode of the AC40 audiometer.
The AC40 audiometer allows the presentation of a reference tone through Channel 2 for calibration. The reference tone was first set to 4 kHz on the “Set up → Tone” screen. Continuous presentation of the reference tone via Channel 2 was enabled by pressing the F9 key on the “Audio” screen. The continuous tone was disabled by toggling the Main/Reserve switch, after which the Tone Presentation Key was used to elicit the 4 kHz tone. This tone was used as the signal and was presented at an intensity 10 dB above the 4 kHz threshold of the individual. The pure tones for the masking stimuli were presented via Channel 1 at frequencies of 3,150, 3,364, 3,668, 4,000, 4,362, 4,622, and 4,896 Hz.
An NBN with a bandwidth (BW) of 320 Hz was used at 4 kHz.
Stimulus presentation
The forced-choice method described by Moore, et al. [4] was used. Two maskers of equal duration were presented consecutively, and the signal tone was randomly embedded in one of them.
The examiner manually triggered the stimuli using the tone presentation keys for Channels 1 and 2. The examiner was trained to deliver the stimuli as follows to maintain consistent timing and presentation across trials. The duration of each masker was 400 ms, and the inter-stimulus interval was 0.5 seconds. The embedded signal tone had a duration of approximately 20 ms and was positioned at the temporal midpoint of the masker.
Threshold determination
A 3-down/1-up decision rule was used: the masker level was increased after three consecutive correct responses and decreased after a single incorrect response. This tracks the 79.4% correct point on the psychometric function [7]. The descending method was used to determine the first reversal point, which was defined as the masker level at which the participant failed to detect the signal for the first time. The ascending method was applied in 5-dB increments to determine the second reversal point. The masker level was varied in 2-dB increments from the third reversal point until six reversals were obtained (Fig. 1)
Examples of reversal patterns in one representative participant. A: Examples of ascending and descending tracks that yielded six reversals for each masker frequency. The open circles indicate the first two reversals, which were excluded from the calculation of the mean and SD. The filled circles represent the last four reversals, which were used to calculate the mean and SD. The numbers beside the markers represent the masker levels (dB HL) at the reversal steps. B: Table summarizing the mean and SD of the reversal levels for the seven masker frequencies (3,150–4,896 Hz) for the same participant. The lowest mean masking threshold was observed at 4,000 Hz, corresponding to the psychoacoustic tuning curve tip. The higher values at 3,150 Hz and 4,896 Hz reflect the low- and high-frequency tails of the tuning curve, respectively. SD, standard deviation.
The first two reversals were discarded, and the mean of the remaining reversals was used as the PTC threshold. Retesting was performed for frequencies with a standard deviation (SD) of reversal levels exceeding 4 dB [8]. The measurement was extended to 10 reversals if the SD of the reversal levels exceeded 3 dB.
Order of masker frequencies
The participants were divided into two groups based on the order of frequency presentation: Group A (n=18): 3,150, 3,364, 3,668, 4,896, 4,622, 4,362, and 4,000 Hz; and Group B (n=16): 4,896, 4,622, 4,362, 3,150, 3,364, 3,668, and 4,000 Hz.
Test–retest procedure
Each participant underwent two test sessions separated by a two-week interval. Their pure-tone thresholds at 4,000 Hz were re-measured during the second session. For participants whose 4,000 Hz thresholds shifted between sessions, the signal tone intensity was set to 10 dB above the second session’s 4,000 Hz threshold. Because individual hearing thresholds at 4 kHz varied among participants, it was necessary to normalize the baseline hearing levels to enable comparison of masker threshold intensities across participants. To achieve this normalization, the reference threshold at 4 kHz was standardized to 10 dB HL. The masker threshold intensity across frequencies was adjusted based on the deviation from this value for each participant. For example, the masking threshold of a participant with a 4 kHz hearing threshold of 5 dB was 45 dB at 3,150 Hz. The difference between 10 dB and 5 dB (5 dB) was added to the measured value to obtain the corrected masking threshold of 50 dB at 3,150 Hz.
The masker threshold differences between the first and second measurements were examined for each frequency to identify the frequency with the greatest discrepancy. For each participant, the test–retest variability was quantified by calculating the mean absolute difference across seven masker frequencies. The participants were classified into four groups based on this variability, and the number of participants in each group was determined. The groups were as follows: Group 1, <5 dB; Group 2, 5–8 dB; Group 3, 8–10 dB; and Group 4, ≥10 dB. The correlations between the test–retest variability and demographic variables of the participants were also determined.
Normative data
The mean and SD of the masking thresholds for each frequency were determined for the participants in Groups 1 and 2 who demonstrated good test–retest variability. Those in Groups 3 and 4 were excluded.
Calculation of Q10 (auditory filter sharpness)
The sharpness of frequency tuning was quantified using the Q10 metric, which is the center frequency (CF; 4,000 Hz) divided by the BW. The BW was determined by identifying the masker frequencies corresponding to points 10 dB above the tip on both sides of the curve. In the BW analysis, data obtained using the 4 kHz NBN masker were excluded, and only the six pure-tone masker frequencies (3,150, 3,364, 3,668, 4,362, 4,622, and 4,896 Hz) were used.
Straight lines were drawn through two adjacent data points on each side of the tuning curve to represent the slopes. The intersection of the two fitted lines, which typically occurs between 3,668 Hz and 4,362 Hz, was defined as the tip. This corresponds to the frequency at which the lowest masker level was obtained. The masking level at this tip was denoted Ltip. The masker frequencies at which the masking levels were 10 dB higher than Ltip were determined for the low- and high-frequency sides and denoted fl and fh, respectively. The BW was calculated as the frequency difference between these two points (fh–fl), and the sharpness of tuning (Q10) was then obtained using the formula:
The lowest measured masker level exceeded the levels typically used for BW estimation in some participants. This resulted in (1) the tip not being located between 3,668 and 4,362 Hz or (2) the low or high cutoff frequencies for BW estimation being outside the expected range (fl >3,150 Hz or fh <4,896 Hz).
The procedure described by Florentine, et al. [2] was implemented for these cases. In these cases, the test tone level at the CF (20 dB HL) served as the tip reference point. A straight line was drawn connecting this tip to the lowest measured masker levels on each side of the curve. The frequencies at which this line intersected masker levels 10 dB points above the tip were designated fh (high-frequency cutoff) and fl (low-frequency cutoff).
Statistical analysis
All correlation analyses were assessed using Pearson and Spearman correlation methods. Group means were compared using Welch’s t-test and the Mann–Whitney U test. Statistical significance was set at p<0.05.
Results
The PTC assessment was completed for all participants without any auditory discomfort. The mean duration of each session (first or second measurement) was 25 minutes.
The mean±SD of the reversal values for the seven test frequencies (3,150–4,896 Hz) ranged from 2.01 to 2.44 dB. This corresponded to a narrow total spread of 0.43 dB and an overall coefficient of variation of approximately 7.8%. These findings indicate minimal variability between the frequencies and suggest that the participants maintained consistent attention across all test frequencies (Fig. 2).
Standard deviation (SD) of the reversal levels across the masker frequencies. Bar graph showing the mean SD of the reversal levels at each masker frequency (3,150–4,896 Hz) obtained during the psychoacoustic tuning curve measurements. Error bars represent the standard error.
Test–retest variability and associated factors
Of the 34 participants, 17 (50.0%) and 10 (29.4%) had highly stable repeated measurements (Group 1) and acceptable reproducibility (Group 2), respectively. Only 3 (8.8%) and 4 (11.8%) had greater variability (>8 dB) and were assigned to Groups 3 and 4, respectively. Approximately 80% of the participants had test–retest variability of ≤8 dB, indicating high overall consistency of the present PTC measurements (Fig. 3).
Distribution of the participants based on the test–retest variability groups. Pie chart illustrating the proportions of the participants in the variability groups based on the inter-session differences in the masking thresholds at the seven test frequencies. Group 1 (<5 dB) had high stability and test–retest variability, Group 2 (5–8 dB) had acceptable reproducibility, Group 3 (8–10 dB) had moderate variability, and Group 4 (≥10 dB) had poor test–retest variability. The numbers and percentages of the participants are provided for each group.
The inter-session differences in the masking thresholds varied across frequencies (Fig. 4) The largest differences were observed at 4,896 and 3,150 Hz. The mean inter-session differences in the masking thresholds and their SDs and standard errors (SEs) are provided for the different frequencies.
Mean test–retest differences in masking thresholds for the various frequencies. Bar graph showing the mean differences in the psychoacoustic tuning curve measurements at the first and second assessments for each masker frequency (3,150–4,896 Hz). Error bars represent the standard error.
The values are the mean±SD (SE). They are as follows: 9.80±7.35 (1.26) dB for 4,896 Hz, 5.77±6.95 (1.19) dB for 3,150 Hz, 5.69±3.56 (0.61) dB for 4,622 Hz, 5.23±5.87 (1.01) dB for 3,668 Hz, 4.37±3.80 (0.65) dB for 4,362 Hz, 4.11±4.38 (0.75) dB for 3,364 Hz, and 2.56±2.32 (0.40) dB for 4,000 Hz. The highest variability was observed at 4,896 Hz, followed by 3,150 Hz. However, the frequencies close to the signal CF (3,668 and 4,362 Hz), where beat phenomena were expected to occur, did not show large reductions in test–retest variability.
Table 2 shows the correlations between test–retest variability and participant demographic variables. A significant positive Pearson correlation was observed between age and test–retest variability (r=0.501, p=0.006), indicating that older participants tended to have greater variability between sessions. The Spearman rank correlation was not significant (ρ=0.290, p=0.130). None of the other variables showed a significant relationship with test–retest variability based on the correlation analysis.
Correlation between test–retest variability and participant characteristics
Impact of test sequence (low-to-high vs. high-to-low)
Fig. 5 presents the inter-session masking threshold differences and the average thresholds (mean of the first and second measurements) at 3,150 Hz and 4,896 Hz for the low-to-high frequency (Group A) and high-to-low frequency (Group B) testing groups.
Effect of test order on intersession variability and averaged masking thresholds. The left panel illustrates the inter-session variability at 3,150 Hz and 4,896 Hz based on the test order, and the right panel shows the corresponding changes in masking-level thresholds. The inter-session variability and masking-level thresholds did not differ significantly between the test-order groups at 3,150 Hz. However, they were significantly affected by the test order at 4,896 Hz. *p<0.05 between Groups A and B.
The absolute differences between the first and second PTC measurements at 3,150 Hz were 5.69±7.33 dB (SE=1.73) and 5.86±6.72 dB (SE=1.68) for Groups A and B, respectively. They did not differ significantly (Welch t=-0.07, p=0.95; Mann–Whitney U=126, p=0.55).
The inter-session difference at 4,896 Hz was significantly larger for Group A (13.24±7.89 dB; SE, 1.97) than for Group B (6.14±4.61 dB; SE, 1.19) (Welch t=3.08, p=0.005; Mann–Whitney U=183, p=0.013).
The mean masking levels at 3,150 Hz for Groups A (48.19±6.82 dB; SE, 1.71) and B (48.53±5.75 dB; SE, 1.48) did not differ significantly (p=0.88).
The average 4,896 Hz thresholds for Groups A and B were 38.96±9.09 dB (SE=2.27) and 46.04±9.07 dB (SE=2.34), respectively. The thresholds for Group A were significantly lower than those for Group B (t=-2.17, p=0.038).
These findings indicate that the transition to higher frequencies (4,896 Hz) after the test sequence started with lower frequencies was associated with greater test–retest variability and lower average thresholds. However, the transition to lower frequencies (3,150 Hz) after the test sequence started with higher frequencies (Group B) was associated with smaller differences.
Normative data
The mean thresholds across the seven test frequencies for the participants in Groups 1 and 2 (test–retest variability ≤8 dB; n=27) ranged from approximately 17 to 49 dB HL.
The mean±SD thresholds were 49.12±5.88 dB HL at 3,150 Hz, 46.42±6.65 dB HL at 3,364 Hz, 38.77±5.12 dB HL at 3,668 Hz, 17.33±4.46 dB HL at 4,000 Hz, 27.18±4.98 dB HL at 4,362 Hz, 32.72±6.01 dB HL at 4,622 Hz, and 43.33±9.26 dB HL at 4,896 Hz. The average masking thresholds for all participants revealed a representative V-shaped tuning curve with the minimum threshold near 4,000 Hz (Fig. 6).
Comparison of the 4,000-Hz psychoacoustic tuning curve (PTC) obtained in this study with the classical data provided by Zwick and Schorn [9]. A: Mean PTC obtained from the current study using a 4-kHz sinusoidal signal and pure-tone maskers (3,150–4,896 Hz). Error bars indicate the standard deviation of the mean. The curve shows a distinct V-shaped pattern with a minimum threshold near 4,000 Hz, representing the tip of the auditory filter. B: Reference PTC redrawn from the data of Zwick and Schorn [9], illustrating a similar sharp tuning pattern centered at 4,000 Hz. The close agreement between the two datasets supports the validity of the present audiometer-based PTC method and its consistency with previously established psychoacoustic findings.
Frequency selectivity (Q10 analysis)
The mean Q10 value for the participants in Groups 1 and 2 was 5.76±1.72 (range: 3.01–9.78; interquartile range: 4.42–6.80). The corresponding Q10 value for participants in Groups 3 and 4 was 5.87±1.79 (range: 2.53–7.71; interquartile range: 5.28–7.04). Comparison of the two sets of groups revealed a negligible difference in Q10, with a minimal mean difference (0.11) and a trivial effect size (Cohen’s d≈0.06).
Table 3 shows the relationships between Q10 and the demographics of the participants. Statistically significant negative correlations between Q10 and age were observed based on Pearson and Spearman coefficients. However, no significant correlations were observed with the 4 kHz hearing threshold or sex.
Correlation between Q10 values and participant characteristics
Discussion
The findings of this study suggest that 4 kHz PTCs can be obtained using a standard two-channel clinical audiometer (AC40) without the need for specialized laboratory equipment. The overall shape and magnitude of the tuning curves were consistent with those reported in previous studies by Vanden Abeele, et al. [3] and Zwick and Schorn [9], demonstrating the physiological plausibility of the tuning patterns obtained through the current clinical implementation. Table 4 presents a comparison between the present study and previous PTC studies that employed either pure tones or NBN as maskers [3,5,6,9-11].
Comparison of methodological characteristics in representative PTC studies
The mean Q10 values (approximately 5–6) were consistent with those reported for normal-hearing listeners at 4 kHz [2,3]. These results indicate that the AC40-based approach yields tuning estimates that do not markedly differ from those reported in previous normative studies. The Q10 values also demonstrated a significant negative correlation with age (r=-0.49, p<0.014), which was consistent with previous reports of age-related widening of auditory filters and reduced cochlear tuning [12,13]. However, given the modest sample size, the age distribution skewed toward younger adults, and the absence of correction for multiple comparisons, the observed age-related association should be interpreted cautiously. Future studies with larger, age-balanced samples and appropriate adjustments for multiple comparisons will be required to validate this finding.
Approximately 80% of participants had test–retest variability of ≤8 dB across the seven masking frequencies, indicating good overall test–retest variability of this simplified clinical PTC protocol. Test–retest differences of up to 5 dB generally denote excellent reliability in conventional pure-tone audiometry, whereas deviations within 8–10 dB are considered clinically acceptable according to Sommers [13] and the ISO 8253-1 (2010) standard [14]. Although the normative dataset was derived after excluding participants with higher test–retest variability, supplementary analyses indicated that this exclusion did not materially affect the core psychoacoustic characteristics of the sample. Specifically, Q10 values were highly comparable between included and excluded participants, with only a trivial effect size. These findings suggest that the exclusion criterion improved measurement reliability without substantially limiting the generalizability of the normative data.
In the present study, participant age showed the strongest correlation with test-retest variability of the PTCs. This finding may reflect age-related instability of the auditory filter, reduced efficacy of the medial olivocochlear efferent suppression, or impaired inhibitory processing at higher levels of the central auditory pathway. Further investigation is needed to elucidate the underlying mechanisms of this age-related increase in variability.
The distribution of test–retest variability did not follow the expected pattern predicted by beat interference theory. Beat phenomena are generally more pronounced when the masker frequencies approach the signal frequency. However, stability did not systematically decline at frequencies near the 4 kHz signal tone based on the current data. This finding suggests that the presence of beats in close frequency regions does not necessarily degrade the test–retest variability of PTC measurements. However, it is important to emphasize that accuracy and precision are distinct concepts. Pure-tone maskers can yield stable results, but NBN maskers may better capture the true shape of auditory filters and are preferable for quantitative assessments of frequency selectivity.
The largest test–retest variability was observed at 4,896 Hz, despite the SD of reversals being comparable to those at other frequencies. This suggests that lapses in attention were unlikely to be the main cause. Stelmachowicz and Jesteadt [13] reported a similar trend, with the greatest variability in their 2 kHz NBN-based PTC measurements occurring at the highest masker frequency (2,600 Hz). This pattern may be explained by a sequential influence, whereby perception during each masking task is affected by the sequence of preceding stimuli rather than being evaluated independently. Most conversational sounds occur at frequencies below 4 kHz. Therefore, distinguishing signal and masker components is easier at lower than at higher frequencies. The relatively easier low-frequency tasks may have induced short-term auditory adaptation or a relaxed decision criterion, leading to slightly elevated thresholds and greater variability in subsequent high-frequency measurements (especially at 4,896 Hz). Conversely, starting with high-frequency trials could help establish a stricter perceptual criterion under more challenging conditions, which may carry over to later, easier tasks and improve overall test-retest variability. To minimize this effect and reduce variability, testing the high-frequency masker conditions first may be preferable. Although our data suggest that initiating testing with higher-frequency masker conditions may help reduce variability, these findings alone are insufficient to establish a standardized testing sequence; confirmation will require prospectively designed studies using refined counterbalanced order assignments and predefined decision rules in larger, age-balanced cohorts.
Several limitations should be acknowledged. First, examiner bias could not be fully eliminated. This is because factors such as the exact temporal alignment of the signal and masker presentation, masker duration, and interstimulus intervals were manually adjusted. The use of manual stimulus presentation introduces the possibility of examiner-dependent variability, particularly with respect to stimulus timing and level control. Such variability could have contributed to increased test–retest differences and minor fluctuations in Q10 estimates.
Second, the number of repeated measurements was limited, primarily due to practical issues related to the cooperation of the participants in undergoing multiple sessions. Consequently, the reliability of the variability estimates may be limited. Participants with excessive variability were excluded from the normative analysis. Additional repeated-measures studies are needed to clarify how intra-individual variability should be characterized and how their data should be appropriately incorporated into future analyses.
Third, the number of masker frequencies employed in this study was limited. Increasing the density of masker frequencies around the signal tone is necessary to achieve more precise estimates of Q10 and better capture the steeper high-frequency slopes reported in previous studies. However, increasing the number of masker frequencies would inevitably prolong the test duration. Future research should explore methods to shorten the duration of measurement (via approaches such as reducing the number of reversals during the adaptive procedure) and verify the validity of these time-efficient modifications for maintaining reliable Q10 estimation.
Finally, the sample size was relatively modest (n=34), and the age distribution was skewed toward younger adults. As a result, the observed correlations between age and both test–retest variability and Q10 values should be interpreted with caution. These analyses were exploratory in nature and were not intended to establish definitive age-related effects on auditory filter characteristics. In addition, multiple correlation analyses were performed without formal correction for multiple comparisons. Given the exploratory purpose of these analyses, no adjustment (e.g., Bonferroni correction) was applied, and the reported p-values should be interpreted as descriptive rather than confirmatory. Future studies with larger and more evenly distributed age cohorts will be required to validate these findings and clarify the role of aging in psychoacoustic tuning.
The normative data in this study were not intended to represent average population values, but rather to define a reference pattern of stable psychoacoustic tuning curves measurable using a standard clinical audiometer. Excluding participants with higher test–retest variability improves internal consistency by minimizing the influence of measurement instability. However, this approach may limit the generalizability of the normative dataset to individuals who can perform the task with high reliability under the present testing conditions. Accordingly, the reported normative values should be interpreted as reference data obtained under stable measurement conditions rather than as population-wide norms.
In conclusion, the findings of this study demonstrate that PTCs measured using a conventional audiometer can provide reproducible and relevant estimates of auditory frequency selectivity in adults with normal hearing. Further well-designed and validated studies are warranted to facilitate the broader application of this method in otologic practice.
Notes
Conflicts of Interest
The author has no financial conflicts of interest.
Funding Statement
None
Acknowledgments
None