Performance of new hearing aids using the ANSI S3.22-1976 standard.

One hundred new hearing aids were tested to determine their compliance with ANSI S3.22-1976 specifications. Thirty-four models representing eight manufacturers were included. Estimates of the test equipment's accuracy were utilized as required by the standard to correct the tolerances permitted for the 11 measurements made. Results revealed that 68% of the instruments tested met all specifications, when the accuracy of the test apparatus was accounted for, while 11% fewer hearing aids passed all tests when it was excluded. No greater than a 10% difference was found in the performance of various types of aids, such as automatic gain control, directional, linear, or nondirectional.

Comparison of total harmonic distortion measures among hearing aid test systems.

Different types of hearing aid test equipment use different methods to calculate total harmonic distortion (THD). Comparisons of the THD measured by four commercially available systems were made by evaluating eight hearing aids, three times each. The results obtained were significantly different (p less than 0.001) among three of the test instruments. Furthermore approximately 10% of the hearing aids, which met the ANSI S3.22-1976 specifications for THD when tested with an apparatus that measures true THD, failed this measurement when evaluated on the other three systems.

Revised estimates of hearing aid test system accuracy.

This study describes a refinement and extension of a procedure to estimate hearing aid test system accuracy reported by Townsend. The present results were obtained from 10 repeated measures on one linear and one AGC hearing aid. Test system variability, represented by standard deviations (SDs) for the 10 trials, was found to be less than or equal to the manufacturer's specifications for 7 of 8 ANSI S3.22-1976 measurements. Furthermore, 4 of these measures revealed variability of only .1 dB. The stability of the SDs across trials varied depending on the measurement; some were consistently reliable whereas others fluctuated considerably. Although data based on 3 trials differed substantially from the 10-trial results, those from 5 trials indicated only minimal differences.

Error analysis on the California Consonant Test by manner of articulation.

Normal hearing subjects, listening to 2 conditions of low-pass, filtered speech, and 2 groups of listeners with sensorineural hearing loss, having audiometric roll-off frequencies analogous to those of the low-pass, filter cut-off frequencies (i.e., 900 and 2000 Hz), were tested with the California Consonant Test (CCT). Analysis or errors based on the three manners of articulation in the test stimuli--fricative, stop, and affricate--indicated that the probability of error per manner changes considerably across roll-off frequency. Because the distribution of test items on the CCT by manner of articulation is not consistent with that found in conversational speech, incorrect estimation of word recognition ability may result. It is suggested that the CCT be altered to reflect the distribution of the manner of articulation found in everyday speech.

Hearing aid measurements using the revised standard: reducing variability and determining test system accuracy.

Unlike previous standards, the revised hearing aid standard, (ANSI S3.22-1976), specifies tolerance requirements for the electroacoustic characteristics of hearing aids. It is essential, therefore, that measurements of performance be made as accurately as possible. To assess the precision with which these measures can be made in a clinical environment, five hearing aids of the same model were each tested five times using Bruel and Kjaer instrumentation. When possible, data were recorded directly from the meter of the measuring amplifier as well as from the chart paper. The variability of the results was always less when data were obtained from the meter than from the chart. The ratio of the variability(meter:chart) differed depending on the test being made. The accuracy of the test system was derived (as required by the standard) both theoretically and empirically. When its estimated tolerances wee added to the tolerances permitted for the aids, all hearing aids were found to perform well within the limits of the standard except for one measurement on one aid.

Effects of age and sensorineural hearing loss on word recognition.

Word recognition scores were compared between a younger age group (15 to 35 years) and an older age group (55 to 85 years) exhibiting similar mild sensorineural hearing losses. No differences were found between the two sets of subjects for recognition ability in quiet and only 5% in noise. Contrary to widespread belief, these results revealed no tendency for the older age group to display poorer recognition performance than the younger listeners. The importance of hearing sensitivity as well as age as a factor affecting work recognition is discussed, and comparative data from the literature are presented.

Effects of phase manipulation on speech intelligibility through communication headsets.

A masking level difference (MLD) paradigm was established by rewiring the earphones of a communications headset out-of-phase. Essentially no release from masking could be measured, however, on a crew of listeners in the cabin of light aircraft. The experiment was replicated in the laboratory so that the exact phase of the aircraft noise masker could be controlled. The substantial MLD obtained in this environment led to the conclusion that the noise reaching the ear in the cockpit was of random phase, which almost eliminates the MLD. Therefore, rewiring headsets out-of-phase provides no advantage in intelligibility.

Speech intelligibility through communication headsets for general aviation.

Word discrimination was measured on eight general aviation pilots listening alternately through each of three communication headsets and an aircraft loudspeaker in the presence of light aircraft noise. Each subject listened at the speech intensity designated by him as yielding optimal intelligibility. Performance varied directly with the degree of attenuation provided by the headset. Mean intelligibility scores ranged from 73% through the loudspeaker to 92% through two of the headsets, although retesting through the loudspeaker at a higher signal level improved mean scores to 85%. Articulation functions obtained on four normally hearing listeners tested under the same conditions as the pilots revealed that, for situations when little or no attenuation was available (loudspeaker and one headset), greater signal-to-noise ratios were necessary to allow discrimination equivalent to that obtained under conditions of greater noise attenuation (two headsets). Although good intelligibility could be achieved through the loudspeaker given sufficient signal intensity (greater than 100 dB SPL) some pilots preferred lower levels even though discrimination was reduced.

A comparison of four hearing aid measurement systems.

Measurements of hearing aid performance were made using four different electroacoustic systems. Nine hearing aids were evaluated three times in each of the four systems: a standard Bruel and Kjaer apparatus, a Fonix 5000, and both a Phonic Ear HC 1000 and HC 2000. Tests included frequency response and, when possible, total harmonic distortion, second harmonic distortion, and third harmonic distortion. Results were equivalent for frequency response and second harmonic distortion, whereas significant differences occurred for total harmonic distortion and third harmonic distortion at both the low and high frequencies. The degree to which the sum of second harmonic distortion and third harmonic distortion fails to equal the total harmonic distortion was also examined. In the low frequency region, second harmonic distortion plus third harmonic distortion yielded values considerably less than the total harmonic distortion, although good agreement was found in the midfrequencies.

Word discrimination for listeners with flat sensorineural hearing losses.

This study examined discrimination scores among listeners (742 ears) with flat sensorineural hearing losses. The discrimination data were analyzed in terms of degree of hearing impairment and subject age. Contrary to previous findings, the results showed that the average subject exhibited a discriminative ability of 70% or better for hearing levels less than 60 dB. A 10 to 20% decline in intelligibility, however, was found for losses 60 dB and greater. Listeners' performance decreased somewhat as a function of age especially for subjects with the greatest hearing losses. Finally, a subject's word discrimination was similar whether hearing loss was based on a three- or a five-frequency average.

Hearing sensitivity in rural Michigan.

In a survey of the hearing sensitivity of a rural mid-Michigan population, 1,325 adults over 15 years of age received pure tone air conduction tests in a "mini" IAC room housed in a small trailer. In addition, a detailed questionnaire was administered to each person to obtain information about his exposure to noise vocationally and avocationally, family history of hearing loss, etc. Results indicated that, on the average, individuals displayed hearing losses considerably greater than was predicted on the basis of presbycusis alone. Also, people who had never had any industrial noise exposure had thresholds almost as poor as those who had worked in factories. These findings are discussed in view of the subject's exposure to various types of nonoccupational noise such as snowmobiles and motorcycles.