Audiometric Symbols

1989 / II - 127Guidelines • Audiometric Symbols

Committee on Audiologic Evaluation

The revised Guidelines for Audiometric Symbols wereprepared by the American Speech-Language-Hearing Asso-ciation (ASHA) Committee on Audiologic Evaluation, andadopted by the ASHA Legislative Council (LC 19-89) inNovember 1989. Current and past members of the commit-tee responsible for the development of the guidelines includeSandra Gordon-Salant, chair, 1986-1989; Martin Robinette,chair, 1984-1986; Carmen Brewer; Margaret F. Carlin;Thomas A. Frank; Thomas Folkes; Gregg D. Givens; MichaelP. Gorga; Sharon A. Lesner; Robert H. Margolis; John D.Durrant; Laura A. Wilber; Carol Kamara, former ex officio;and Evelyn Cherow, current ex officio. The monitoring vicepresidents included Gilbert R. Herer, past president andformer vice president for clinical affairs, and Teris K. Schery,current vice president for clinical affairs.

IntroductionThese guidelines were developed by the Commit-

tee on Audiologic Evaluation under the direction of theVice President for Clinical Affairs of the AmericanSpeech-Language-Hearing Association. These guide-lines present a recommended set of symbols based onlisting clinical practice. The spirit of these guidelinesis not to mandate a single way of accomplishing theclinical process; rather, the intent is to suggest stan-dard procedures that, in the final analysis, may ben-efit the clients we serve. The intention is to allow forefficient and uniform transfer of information. Much ofwhat appears in this document was taken fromASHA’s previous guidelines covering the same topic(ASHA, 1974). In addition, the opinions of practicing

clinicians were sought and many of their suggestionsare incorporated into this revision of the Guidelinesfor Audiometric Symbols.

Audiometric symbols used to record the results ofconventional pure tone threshold audiometry weresuggested by ASHA (1974), following a review by theCommittee on Audiologic Evaluation. Jerger (1976)has described further options in the way audiometricdata might be reported in scholarly publications. Manyaudiometric symbols are used universally, while oth-ers are unique to particular facilities and clinics. Sucha situation could cause misinterpretation when datain graphic form are shared among clinics. This is ofparticular concern because graphic representation isprobably the most common form for reporting puretone, audiometric results.

The purpose of these guidelines is to set forth a setof symbols and procedures for use in the graphic rep-resentation of audiometric findings for frequency spe-cific stimuli. There are clinical situations in whichgraphic depictions are not ideal, such as serial hear-ing tests in an industrial setting or in the schools; how-ever, the recommendations to follow should beappropriate to other clinical situations where graphicrepresentations are used.

The AudiogramAs recommended in the ANSI S3.21-1978 (R.I 986)

“Methods for Manual Pure-Tone Threshold Audiom-etry, the audiogram shall be shown as a grid with fre-quency, in Hertz (Hz), represented logarithmically onthe abscissa and hearing level (HL), in decibels (dB),represented nearly on the ordinate. One octave on thefrequency scale shall be equivalent in span to 20 dBon the HL scale. The abscissa shall be labeled “Fre-quency in Hertz (Hz)” and the ordinate shall be labeledHearing Level in Decibels (dB).” Different standardscurrently exist for reference equivalent threshold lev-els for air conduction and bone conduction (these ref-erences may be consolidated in a future standard), andthere are many reference levels that are used for data

Guidelines

Audiometric Symbols

Reference this material as: American Speech-Language-Hearing Association (1990). Guidelines for audiometricsymbols. Asha, 32 (Suppl. 2) 25-30.

Index terms: Acoustic impedance audiometry, acousticreflex testing, audiology, hearing assessment, hearingloss, masking, pure tone air conduction testing, puretone bone conduction testing

Document type: Standards and guidelines

II - 128 / 1989

ASHA 2002 Desk Reference Volume 2 • Audiology

plotted on an audiogram but for which no standardfists (e.g., thresholds obtained with insert earphonesor to high frequency stimuli, sound field thresholds).Because of this diversity in how audiograms are used,it is recommended that the reference used to establish0 dB HL should be listed on the audiogram form, al-though not necessarily as part of the label on the ordi-nate. The 0 dB threshold level should be shownprominently so that it stands out from other HL gridlines. The range on the abscissa should include fre-quencies from 125 Hz to 8000 Hz while the range onthe ordinate should include levels from -10dB to 120dB HL Wider ranges of level may be used (e.g., -20 to130 dB HL), because many audiometers are capable ofproducing these levels and there may be clinical situ-ations when the audiologist chooses to test levels be-low -10dB HL or above 120 dB HL Similarly, there areinstances when frequencies above 8000 Hz may betested. If one chooses to record thresholds for frequen-cies above 8000 Hz on the audiogram, a logarithmicfrequency scale should be used, such that the intervalbetween 8000 Hz and 16000 Hz is accurately repre-sented as an octave interval, equal in spacing to allother octave intervals on the form. Clinicians attempt-ing to assess hearing sensitivity for frequencies above8000 Hz are referred to the literature that underscoressome of the special calibration difficulties that are en-countered when using these high frequencies (e.g.,

Stelmachowicz, Beauchaine, Kalberer, Larger, &Jesteadt, 1988; Stelmachowicz, Gorga, & Cullen, 1982;Stevens, Berkovitz, Kidd, & Green, 1987; Stinson &Shaw, 1982).

Grid lines of equal darkness and thickness arerecommended at octave frequency intervals and at10dB HL intervals. If grid lines are used for interactivefrequencies, then they should be finer or dashed inorder to distinguish them from these for octave frequen-cies. It should be recognized that while 750,1500,3000,and 6000 Hz; are often represented and used as geo-metrically centered interactive frequencies, these rep-resentations are technically Incorrect. The errors aresmall and probably not of clinical significance; how-ever, clinicians should be aware that the arithmeticaverage between octave frequencies does not representthe true semioctave frequency. The dashed lines for750,1500, 3000, and 6000 Hz in all of the audiogramsin this document are drawn at the appropriate placeon a logarithmic scale; they are not centered betweenthe lines representing octave frequencies.

An example of an audiogram form is illustratedin Figure 1. Specific recommendations on the type andamount of additional information to be included on theaudiogram, such as patient and tester identification,and results of other tests, are not included in theseguidelines since they may be more appropriately speci-


fied by individual clinics. In addition, some clinics maychoose to represent each ear on a separate audiogramwhile others may prefer to report audiometric data forboth ears on the same graph. Such decisions also areconsidered a matter of preference and are left to thediscretion of individual clinics.

The Audiometric SymbolsThe audiometric symbols are shown in Table 1

and were selected using the following criteria:1. Simple in design, easily drawn, and sharply

reproducible by xerography or other reproduc-tion methods;

2. Mutually exclusive from and internally consis-tent with other symbols in the system;

3. Capable of delineating, without recourse tocolor coding1, the following distinctions:

a) left ear from right ear,b) air conduction from bone conduction,c) masked from masked conditions,d) response from no response, ande) the transducer (earphone2, vibrator, loudspeaker) used to present the stimuli; and

4. Designed to permit multiple notation at a singlelevel on the audiogram.

The symbols recommended in this guideline weretaken directly from those specified in appropriate stan-dards (ANSI S23.21-1978, R-1986; ANSI S3.39-1987),with two exceptions. The rationale for recommendingthese two exceptions will be given when their use isdiscussed in this guideline.

Further SpecificationsSymbols typically represent the transducer place-

ment and not necessarily the ear from which the re-sponse is coming. This general point should be keptin mind, especially when considering the choices ofsymbols for unmasked bone conduction, which aredescribed below.

Air Conduction Symbols

The air conduction threshold symbols should bedrawn on the audiogram so that the midpoint of thesymbol centers on the intersection of the vertical andhorizontal axes at the appropriate frequency and hear-

ing level (for example, see Figures 2, 3, 4, and 6). Thesesymbols should be reserved for those conditions whenstimuli are presented under earphones as opposed toin the sound field (i.e., using a loudspeaker as the trans-ducer).

Bone Conduction Symbols

The specifications for 0 dB HL via bone conduc-tion were determined while masking was presented tothe nontest ear (ANSI S3.26-1981). On the other hand,unoccluded, unmasked bone conduction thresholdsare measured in many clinical situations. In cases ofsymmetrical sensitivity for air conducted signals with-out conductive hearing loss, differences been maskedand unmasked bone conduction thresholds should beclinically insignificant because the central maskingeffect is small (Zwislocki, Buining, & Glantz, 1988).

1 Color coding might still be considered if it is felt that itwould further clarify results.

2 The reader is referred to Zwislocki et al. (1988) for a morecomplete description of the kinds of earphones that havebeen used in audiometry.

II - 130 / 1989


There is disagreement as to how to represent un-masked bone conduction thresholds. Previous guide-lines (ASHA, 1974) proposed using a different symbol,depending upon whether the vibrator is placed on theleft or right mastoid (see Table 1). This convention isconsistent with the idea that symbols only representtransducer placement, and not the source of any mea-sured response. There is no problem with this ap-proach as long as these symbols are not interpreted tomean that the response is coming from the ipsilateralcochlea.

An alternative approach is to plot unmasked boneconduction thresholds in a way that does not suggestthat the response is coming from either ear. This cau-tion results from the fact that an unmasked bone con-duction threshold often cannot be assigned to eithercochlea, regardless of the placement site of the vibra-tor on the head. One suggestion to plot unmasked boneconduction thresholds as a down bracket that does notrefer to either ear (Herer, 1967)3. However, this symbolis comparable to the ones used for masked bone con-duction thresholds. Another option is to plot un-masked bone conduction as an upside down “V” asshown in Table 1.3. This symbol is recommended be-cause it is more consistent with the ear-specific un-

masked bone conduction symbols in the system, withthe exception that it is rotated 90 degrees. When sym-metrical air conduction thresholds are obtained, andthere are no differences between these thresholds andthe unmasked bone conduction threshold (i.e., sym-metrical sensorineural hearing loss), then it is sufficientto record only the air conduction thresholds and theunmasked bone conduction thresholds. An exampleis shown in Figure 3.

Masked bone conduction thresholds, using a mas-toid placement, should be represented as a bracket,with the open side close to but not in contact with thevertical frequency line. The symbol for the left earshould be on the right side of the frequency line whilethe symbol for the right ear should be on the left side ofthese vertical lines. Examples are shown in Figures 2,4,and 6.

Forehead placement of the vibrator may be neces-sary in some cases, such as patients who have hadmastoid surgery. The standards for 0 dB HL at the fore-head and at the mastoid are not identical (ANSI S3.26-1981), requiring the use of a different set of correctionsto derive 0 dB HL. Furthermore, the maximum levelthat can be achieved with a forehead placement is lessthan what is possible at the mastoid. If forehead place-ment is used, unmasked thresholds should be repre-sented as a “V” at the appropriate intersection offrequency and level. For masked thresholds, a “semi-bracket,” such as those shown in Table 1, should beplaced at the appropriate intersection of frequency andlevel, following the convention designated for maskedbone conduction thresholds using a mastoid place-ment.

Multiple Notations

When unmasked or masked air conductionthresholds are identical at a given frequency, the leftand right symbols should overlap. When masked boneconduction thresholds occur at the same HL as airconduction thresholds, the bone conduction symbolsshould be placed adjacent to, but not touching, the airconduction symbols. When unmasked bone conduc-tion thresholds occur at the same HL as an air conduc-tion threshold, the unspecified bone conductionsymbol should be placed slightly above the air conduc-tion symbol. Examples are shown in Figures 2, 3, 4, and6. Alternatively, if one chooses to use the unmaskedbone conduction symbol that reflects transducer place-ment, the symbol should be drawn adjacent to but nottouching the air conduction symbol when these twothresholds are the same.

3 It should be recognized that this symbol has not beenrecommended for use in any published standard.


Acoustic Reflex Thresholds

The American National Standards Institute hasapproved a document that specifies the characteristicsof devices that are used to measure aural acoustic im-mittance (ANSI S3.39-1987). Included in this documentare recommendations for symbols that can be used torepresent ipsilateral or contralateral acoustic reflexthresholds (ART) on the audio-ram. These symbols areshown in Tables 1 and 2, and examples are given inFigures 2, 3, and 6. Implicit in plotting ARTs directlyon an audiogram is the assumption that the elicitingstimulus can be calibrated in dB HL. Although thestandards that are used for routine pure-tone audio-metric testing would be applicable for contralateralARTs (assuming a supra-aural earphone was used), asimilar standard does not exist for ipsilateral stimula-tion. Wilber, Kruger, and Killion (1988) provide “pro-visional reference equivalent threshold levels” forinsert earphones which also may be used for convert-ing the level of ipsilateral stimuli to dB HL.

Sound Field Thresholds

Currently, no standards exist for sound field test-ing and calibration, although the International Stan-dards Organization (ISO) is considering this issue andmay recommend the use of an ear specific symbol if theopposite ear is plugged during testing (L.A. Wilber,personal communication, 1988). These proposed sym-bols are shown for left and right ears in Table 1. TheCommittee on Audiologic Evaluation recommends re-taining the use of the symbol, “S,” to represent thresh-old recorded in the sound field that cannot be assignedto either ear. This symbol should be plotted at the ap-propriate intersection of frequency and level. If onewishes to plot sound field thresholds on an audio-gram, the committee recommends the use of some cor-rection to convert sound field sound pressures toequivalent dB HL (e.g., Morgan, Dirks, & Bower, 1979;Walker, Dillon, & Byrne, 1984). Alternatively, soundfield thresholds can be measured directly in dB SPLwithout attempting to convert these values to equiva-lent dB HL. In these cases, however, the results shouldnot be recorded on the audiogram form.

When functional gain measurements are used inhearing aid evaluations, there is the added concern thatresults with different hearing aids also must be repre-sented. Data for individual aided conditions shouldbe represented by a set of locally determined symbolsthat are exclusive of any other symbols and should bedefined on the audiogram.

II - 132 / 1989


No Response

To indicate “no response” at the maximum out-put of the audiometer, an arrow should be attached tothe lower portion of the appropriate symbol anddrawn downward and at about a 45-degree angle tothe right for the left ear symbols and to the left for theright ear symbols. The brow for sound field and un-masked bone conduction symbols should be attachedat the bottom and drawn straight downward.

The “no response” symbol should be placed on theaudiogram at the HL representing the maximum out-put limit for specific frequencies, transducers, andaudiometers. Each of the “no response” symbols isshown in Table 2. Appropriate usage is shown in Fig-ures 5 and 6.

When a patient has many “no responses,” nota-tion other than by symbol may be used to avoid clut-tering the audiogram. Of course, another clinician maynot be aware of the output limits of the audiometerused to measure thresholds. The rehabilitative impactof knowing exact audiometer limits a function of fre-quency may be relevant to certain clinical decisionssuch as cochlear implant candidacy.

Lines Connecting Symbols

Lines may be used to connect symbols on an au-diogram. When used, a solid line should connect theair conduction threshold. Typically, bone conductionthresholds are not connected by lines, especially whenno air-bone gap is present. In cases of an air-bone gap,a dashed line can be used to connect bone conductionthresholds. Whenever lines are used, it is recom-mended that the lines only approach the symbols; thelines should not touch or go through the symbols. Ex-amples are shown in Figures 2-6.

Symbols representing “no response for either airconduction, bone conduction or sound field testingSHOULD NOT be connected by lines to each other orto any of the response symbols.

Color Coding

Color coding is not necessary to transmit informa-tion about sidedness in this symbol system. In prac-tice, color coding becomes meaningless because thiscode will be lost in the copying process.

Masking

While entirely optional, effective masking levelscould be recorded on the audiogram for both air con-duction and bone conduction thresholds, dependingupon individual preference. When this policy is fol-lowed, the maximum effective masking level used toobtain threshold at each frequency should be recorded.This level should be reported for the nontest ear, be-cause this is the ear to which the masking stimulus isbeing delivered. Examples are shown In Figures 2,4,and 6. Oftentimes, both unmasked and masked thresh-olds are recorded on the audiogram. This may be viedby some clinicians as excessive cluttering the audio-gram (see, for example, Figure 4). In cases when mask-ing must be used, it is acceptable to plot only themasked thresholds (see Figure 2).

AcknowledgmentsSpecial thanks are extended to Steve Neely and

Theresa Langer of the Boys Town National Institutefor writing the software and generating the figures andtables that are included in these guidelines.


ReferencesAmerican National Standards Institute. (1978). Methods for

manual pure-tone threshold audiometry (ANSI S3.21-1978,R-1986). New York: ANSI.

American National Standards Institute. (1981). Referenceequivalent threshold force levels for audiometric bone vibra-tors (ANSI S3.26-1981). New York: ANSI.

American National Standards Institute. (1988). Specificationsfor instruments to measure aural acoustic impedance andadmittance (Aural acoustic immittance) (ANSI S3.39-1987).New York: ANSI.

American Speech and Hearing Association. (1974). Guide-lines for Audiometric Symbols. Asha, 16, 260-264.

Herer, G. R. (1967). A suggested symbol for bone conduc-tion testing. Journal of Speech and Hearing Disorders, 32,198.

Jerger, J. (1976). Proposed audiometric symbol system forscholarly publications. Archives of Otolaryngology, 102, 33-36.

Morgan, D. E., Dirks, D. D., & Bower, D. R. (1979). Suggestedthreshold sound pressure levels for frequency-modu-lated (warble) tones in the sound field. Journal of Speechand Hearing Disorders, 44, 37-54.

Stelmachowicz, P. G., Beauchaine, K. A., Kalberer, A.,Langer, T., & Jesteadt, W. (1988). The reliability of audi-tory thresholds in the 8- to 20-kHz range using a proto-type audiometer. Journal of the Acoustical Society ofAmerica, 83, 1528-1535.

Stelmachowicz, P. G., Gorga, M. P., & Cullen, J. K. (1982).A calibration procedure for the assessment of thresh-olds above 8000 Hz. Journal of Speech and Hearing Re-search, 25, 618-623.

Stevens, K., Berkovitz, R., Kidd, G. Jr., & Green, D. (1987).Calibration of ear canals for audiometry at high fre-quencies. Journal of the Acoustical Society of America, 81,470-484.

Stinson, M. R., & Shaw, E. A. G. (1982). Wave effects andpressure distribution in the ear canal near the tympanicmembrane. Journal of the Acoustical Society of America, 71,(Suppl. 1), S88.

Walker, G., Dillon, H., & Byrne, D. (1984). Sound field au-diometry: Recommended stimuli and procedures. Earand Hearing, 5, 13-21.

Wilber, L. A., Kruger, B., & Killion, M. C. (1988). Referencethresholds for the ER-3A insert earphone. Journal of theAcoustical Society of America, 83, 669-676.

Zwislocki, J. J., Buining, E., & Glantz, J. (1968). Frequencydistribution of central masking. Journal of the AcousticalSociety of America, 43, 1267-1271.

Zwislocki, J., Kruger, B., Miller, J. D., Niemoller, A. F., Shaw,E. A., & Studebaker, G. A. (1983). Earphones in audiom-etry. Journal of the Acoustical Society of America, 83, 1688-1689.

Documents

Audiometric Symbols