In recent years, Internet use in homes has gained increasing popularity. Likewise, with the fast improvement in the quality of the computer audio systems, the Internet has provided the general public with new opportunities in audio applications.

The Online-Hearing Test, developed by Apherma Corp (available at www.apherma.com/web2004/test.htm) is designed to allow individuals to perform self-administrated listening comparisons over the Internet to assess their own hearing. It should be clarified at the outset that the purpose of Online-Hearing Test is not to replace the more accurate traditional assessment performed in the clinic by a qualified hearing care professional. The online test is designed as a screening tool to provide a quick hearing estimate for consumers in the convenience and privacy of their homes. Using the results from the test, the consumer should be able to obtain useful information about the status of their hearing and gain some insight into what their audiogram looks like. Those whom the testing reveals signs of hearing loss are recommended to visit the offices of a local hearing care professional.

f06a.jpg (16426 bytes) Figure 1. The Online-Hearing Test Web page.

Is There a Need for an Online Test?
The reason for developing an online hearing test lies in the fact that, just in the United States, of the 28.6 million people who have a significant hearing loss,1 only a small portion of them have ever gone through any hearing test, and less than 23% have been fitted with hearing aids. There may be various reasons for this unfortunate fact, but one thing is certain: If there is an easily accessible and convincing hearing test available, one can expect that many more of these hearing-impaired people will have a better knowledge of their hearing status—and come to hearing care professionals to seek help.

It should be recognized that there are currently many excellent means for assessing hearing via the Internet. For example, online versions of the Self Assessment of Communication (SAC) test, Hearing Handicap Inventory for the Elderly-Screening (HHIE-S), Significant Other Assessment of Communication (SOAC), and Self Assessment Hearing Tests provide a means to assess hearing via the Internet.

While these and other tests are of great value, they generally represent an indirect means of assessing hearing based on the effects of a possible hearing loss (ie, they don’t actually “test hearing” directly).

How the Online Test Works
The Online-Hearing Test includes two parts. In Part 1 of the test, the hearing threshold of a “normal-hearing” listener is measured to establish the reference for the specific computer audio system used for the tests. This person serves the purpose of “biologically” calibrating the sound system. Needless to say, the reliability of this calibration method is dependent on how close the hearing of the reference listener is to normal.

In Part 2 of the test, the threshold of the individual with suspected hearing loss is measured against the same settings for the “normal-hearing” reference listener that was established in Part 1. Their threshold differences are interpreted as a measure of the amount of hearing loss of the testing subject, or thresholds in dB HL.

Before the test, a software program and sound files are downloaded over the Internet into the consumer’s computer. These sound files represent pure-tones at 5 frequencies (250 Hz, 500 Hz, 1000 Hz, 2000 Hz, and 4000 Hz) with 20 sound levels (at 5 dB intervals) available.

During the test, upon each stimulus presentation, the subject is notified by a screen display that a sound has been played and the system is waiting for his/her response. The subject responds “Yes” or “No” to indicate whether the tone is audible. The sound level is reduced by 5 dB following each “Yes” response, and increased by the same amount with each “No” response. The threshold is defined as the minimum sound level where at least two-of-three responses are positive.

Once hearing thresholds are obtained from both the “normal” reference ear and the testing ear, the “hearing loss” (ie, threshold in dB HL) is obtained by taking the differences between the two sets of values.

Does the Online Test Work?
In a preliminary field test, the authors used 23 ears of 12 subjects. For comparison purposes, the actual audiograms of these ears were also obtained. In processing the results of the Online-Hearing Tests, we have included a correction factor of 7 dB at all frequencies to compensate for the higher ambient noise in typical homes than in a sound-treated booth or a quiet room of typical dispensing offices. This value was based on preliminary measurements of sound levels under the two types of testing environments.

figure Figure 2. One subject’s hearing losses estimated using the online (Blue curve) and conventional (red curve) hearing testing methods.

Figure 2 shows an example of one subject’s hearing losses estimated using the online testing method (blue curve), as well as a conventional audiogram taken in a dispensing office (red curve). The two curves are within about 10 dB of each other, and display similar trends in the frequency pattern of the hearing losses.

To analyze the reliability of the test, we looked at the distribution of the measurements obtained from the Online-Hearing Test versus the audiograms we obtained from the 12 subjects. For this purpose, we calculated the threshold differences for all listeners and frequencies, and present their distribution in Figure 3.

The difference in distribution has a mean of 0.8 dB. Given the size of the standard deviation (9.5 dB), this small deviation from zero is essentially negligible. Therefore, the online test and conventional hearing testing methods appear to measure the same thing (ie, the hearing loss).

figure Figure 3. Distribution of differences between hearing losses estimated using the online and conventional audiogram methods.

The variance of the difference distribution includes two parts: one from the measurement errors for thresholds obtained in the online test, and the other from the measurement errors in the standard audiograms. Standard audiograms are thought to have a standard deviation of about 5 dB, although it can be up to 15 dB in some cases.2,3 The variance of the thresholds obtained from the online test also has two parts: one from the measurement with the “normal-hearing” reference and the other from the measurement with the testing subject. It is reasonable to expect that the two terms are approximately the same, and that they both are close to the variance of the standard audiograms. Therefore, the standard deviation of the difference distribution in Figure 3 is expected to be approximately 9 dB. This is close to the value of 9.5 dB as observed in the present study.

Larger Variances, But Meaningful Results
The above analysis suggests that thresholds from online tests have twice the variance as that of traditional audiograms performed in a hearing care office. In real-world applications, the online test is at a further disadvantage because it:

  1. Is not being administrated by a hearing care professional who has training and experience in assessing the results of a hearing test;
  2. Does not use a carefully calibrated audiometer, and
  3. Is not conducted in sound-treated booth or room.

Compared with standard audiograms, results of Online-Hearing Tests are more susceptible to errors due to variations in the ambient noise level from home to home (or, more specifically, room to room), variations in the hearing level of the “normal-hearing” test references from case to case, and variations of the audio systems from computer to computer.

If the testing subject does have a hearing loss, room noise probably has little effect on the subject’s estimated hearing threshold, because the threshold level is likely to be above the noise level. However, for the truly “normal-hearing” reference, excessive room noise could lead to an overestimated hearing threshold due to masking. This, in turn, could lead to underestimating the amount of hearing loss for the testing subject. Thus, when taking the online-hearing test, it is important that the room is as quiet as possible.

Discussion
Obviously, the Online-Hearing Test is not without its drawbacks. Among the chief concerns is choosing the listener to serve as the “normal-hearing” reference. When the hearing threshold obtained from an impaired ear is used as the reference, the hearing loss of the test subject will be underestimated. Therefore, the consumer should in all cases try to maximize the chance for choosing someone whose hearing is close to normal. For example, all things being equal, ears of younger people and those with less exposure to loud noises are better candidates.

Likewise, computer audio systems come in all shapes, sizes, and quality levels. Those of poor quality can influence testing results due to non-linearity or distortion in the system. However, with the computer industry upgrading its products at a fast pace, the quality of audio systems is undergoing constant improvement. One can expect that, with some exceptions, the quality of the computer audio system will not be a major issue in future considerations.

While not perfect, the Online-Hearing test provides approximate, yet practical, results. When using the Online-Hearing Test, care needs to be taken to:

  1. Reduce the amount of ambient noise in the testing room;
  2. Maximize the chance that the reference listener has a normal hearing, and
  3. Avoid using audio system with poor quality.

However, if these provisions are met, the online test can yield reasonably reliable outcomes for the purpose of hearing screening. When the testing results indicate a hearing loss, hopefully the subject can be motivated to seek out a hearing care professional for further evaluation and advice regarding the potential benefits of amplification. In this way, online testing can serve as a useful and convincing tool.

Ultimately, the true value of any online hearing test resides in creating new opportunities for helping hearing-impaired listeners recognize their hearing loss in the privacy of their homes and convincing them that it’s time to seek help for their hearing loss.

This article was submitted to HR by Zezhang Hou, PhD, and Huanping Dai, PhD, director of research at Apherma Corp. Correspondence can be addressed to HR or Huanping Dai, Apherma Corp, 880 E Arques Ave, Sunnyvale, CA 94085; email: [email protected].

References
1. Kochkin S. The VA and direct mail sales spark growth in hearing aid market. Hearing Review. 2001;8(12):16-24,63-65.
2. Harris JD. Proem to a quantum leap in audiometric data collection and management. J Aud Res. 1978;18:1-29.
3. Harrel RW. Pure-tone evaluation. In: Katz J, ed. Handbook of Clinic Audiology (5th Ed). New York: Thieme; 2002.