EFFECT OF ACOUSTIC COUPLER ON AIDED SPEECH RECEPTION THRESHOLDS AND SPEECH DISCRIMINATION SCORES USING A CROS HEARING AID Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY ALBERT J. JETTY 1968 THE-TRIS This is to certify that the thesis entitled Effect of Acoustic Coupler on Aided Speech Reception Thresholds and Speech Discrimination Scores using a CROS Hearing Aid. presented by l Albert J. Jetty has been accepted towards fulfillment of the requirements for PhoD. degree in AUlelogy and Speech Sciences <5 (/7 T 2’ it an 9' I 3Mxé-gtw Major prdfessor Date July 25, I968 I 0-169 ABSTRACT EFFECT OF ACOUSTIC COUPLER ON AIDED SPEECH RECEPTION THRESHOLDS AND SPEECH DISCRIMINATION SCORES USING A CROS HEARING AID by Albert J. Jetty The major purpose of this study was to investigate the effects of four types of acoustic complers (conventional. vented, Open, and crimped polyethylene tubing) on aided Speech reception thresholds and Speech discrimination scores of three groups of hard-of-hearing adults. Group I was composed of ten subjects with conductive hearing impairments. Group II consisted of ten subjects having a sensorineural type hearing impairment with normal hearing in the low frequencies and a preCiPitOus drOp for frequencies higher than 500 to 1000 Hertz. Group III was composed of ten subjects having a sensorineural type hearing imp airment with a gradually sloping (5 to 10 decibels per octave) CO“figuration with the low frequencies also being affe cted, Pure —conduction thresholds were ob- ‘tone air and bone tamed prior to the Speech audiometric tests. Speech re- cepthm thresholds and speech discrimination scores were Albert J. Jetty obtained in the sound-field under unaided and aided condi- tions, while the nontest ear was occluded by a wax impreg— nated ear plug. All subjects were tested with the same CROS hearing aid at a gain setting of 35 dB, and all Speech discrimination scores were obtained at a 26 dB sensation level. The data were examined by means of two-way analyses of variance for both Speech reception thresholds and Speech discrimination scores. Significant differences were further investigated by employing Duncan's New Multiple Range Test. Results Showed that for the conductive hearing impaired subjects, the mean Speech reception threshold obtained with the conventional earmold was significantly lower than the mean thresholds obtained with the acoustic modifier and crimped tubing. The mean aided Speech discrimination scores showed no Significant differences among the four acoustic COUplers in the aided condition. For the group having a sensorineural hearing impair- ment with a precipitous drop, there were essentially no inter- coupler differences in the mean Speech reception thresholds obtained with the various acoustic couplers. The mean Speech discrimination scores were Significantly improved under the aided conditions. The mean unaided Speech discrimination Score was 69.4 percent in comparison to a mean aided score Of 76.8 percent utilizing the conventional earmold. Thus, a Albert J. Jetty gain of 7.4 percent was achieved. The mean Speech discrimi— nation Score of the modified couplers combined was 87.5 percent, a gain of 17.9 percent over the mean unaided sound— field score and 10.5 percent over the mean score obtained with the conventional earmold. There were no significant intercoupler differences in the mean Speech discrimination scores obtained with the modified acoustic couplers. For the group having a gradually sloping sensorineural hearing impairment, the mean Speech reception threshold, utilizing the conventional earmold, was 25.6 dB while the mean combined Speech reception threshold of the modified COUplers was 28.0 dB. This group obtained a mean unaided Speech discrimination score of 79.6 percent and a mean aided score utilizing a conventional earmold of 75.0 percent. The mean Speech discrimination score of the modified acoustic couplers combined was 85.5 percent and thus was an improve- ment of 8.5 percent over the conventional earmold. The following conclusions were drawn: Persons with con- ductive hearing impairments obtain better aided Speech reception thresholds with the conventional earmold than with the modified acoustic couplers, whereas Speech discrimination scores Show essentially no differences among coupling condi- tions. The aided Speech reception thresholds of persons having a sensorineural hearing impairment with a precipitous drOp are essentially the same under all COUpling conditions, Albert J. Jetty whereas Speech discrimination is markedly improved with the modified acoustic couplers as opposed to the conventional earmold. Persons with a gradually SlOping sensorineural hearing loss obtain better aided Speech reception thresholds with the conventional earmold, whereas Speech discrimination is improved with the modified acoustic couplers. EFFECT OF ACOUSTIC COUPLER ON AIDED SPEECH RECEPTION THRESHOLDS AND SPEECH DISCRIMINATION SCORES USING A CROS HEARING AID BY Albert J. Jetty A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Audiology and Speech Sciences 1968 ACKNOWLEDGMENTS This work was supported (in part) by a Neurological and Sensory Disease Traineeship Award from the Public Health Service, U. S. Department of Health, Education, and Welfare. The writer wishes to express appreciation to Dr. William F. Rintelmann for his help and guidance as thesis advisor. Thanks is also due to Drs. Herbert J. Oyer, Edward J. Hardick, Leo V. Deal, and Martha E. Dale for serving as guidance committee members. Special gratitude is extended to Dr. Edward J. Hardick for his assistance with the statistical design of the study, and to Mr. Donald E. Riggs for his assistance in making the necessary electro-acoustical measurements. Grateful appreciation is also extended to my wife, Judy, for her understanding and encouragement and to Lisa and Michael for doing without an often too busy father. ii Chapter II. III 0 IV. TABLE OF CONTENTS LIST OF TABLES . . . . . . . LIST OF FIGURES. . . . . . . LIST OF APPENDICES . . . . . INTRODUCTION . . . . . . . . Purpose of the Study. . . Significance of the Study Definitions . . . . . . . REVIEW OF THE LITERATURE . . History . . . . . . . . . Acoustic Phenomena. . . . Modified Earpieces. . . . Summary . . . . . . . . . EXPERIMENTAL PROCEDURES. . . Subjects. . . . . . . . . Equipment . . . . . . . . Test Environment. . . . . Test Materials. . . . . . Test Procedures . . . . . RESULTS AND DISCUSSION . . . Descriptive Summary of Results. Reliability . . . . . . . Speech Reception Thresholds . Speech Discrimination Scores. Clinical Significance . . iii Page viii ix I-P-UIN P \J 11 20 41 45 45 47 58 59 61 70 70 72 77 86 94 TABLE OF CONTENTS - Continued Chapter Page V. SUMMARY AND CONCLUSIONS . . . . . . . . . . . 107 Summary. . . . . . . . . . . . . . . . . . 107 Conclusions. . . . . . . . . . . . . . . . 110 Recommendations for Further Research . . . 112 BIBLIOGRAPHY. . . . . . . . . . . . . . . . . 115 APPENDICES. . . . . . . . . . . . . . . . . . 120 iv Table II. III. IV. VI. VII 0 VIII. IX. LIST OF TABLES Percent Harmonic Distortion of Radioear, Model 950 Hearing Aid by A.S.A. Method. Input 75 dB SPL re 0.0002 microbar. . . . . . . . . . . . . Means (M), Medians (Med), Standard Deviations (SD), and Ranges of Speech Reception Thresholds Obtained for Three Groups of Subjects Under One Unaided and Four Aided Listening Conditions . . Means (M), Medians (Med), Standard Deviations (SD), and Ranges of Speech Discrimination Scores Obtained for Three Groups of Subjects Under One Unaided and Four Aided Listening Con- ditions . . . . . . . . . . . . . . . . . . . . Coefficients of Correlation (Spearman Rank- Order) Between Test and Retest for Each of the Four Acoustic Couplers Across All Three Subject GrOUps. . . . . . . . . . . . . . . . . . . . . Summary of Two-Way Analysis of Variance Compar- ing the Effects of Differences in Kind of Hear- ing Loss and Type of Acoustic COUpler on Speech Reception Thresholds. . . . . . . . . . . . . . Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for SRTS Within the Conductive Group . . . . . . . . . . Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for SRTS Within the Sensorineural Precipitous DrOp Group Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for SRTS Within the Sensorineural Gradually SIOping Group . . . . . . . . . . . . . . . . . . . . . Summary of Two-Way Analysis of Variance Compar- ing the Effects of Differences in Kind of Hear- ing Loss and Type of Acoustic Coupler on Speech Discrimination Scores . . . . . . . . . . . . . Page 58 71 75 77 79 81 85 85 88 LIST OF TABLES - Continued Table X. XI. XII. XIII. XIV. XVII. Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for Speech Discrimination Scores Within the Con- ductive Group. . . . . . . . . . . . . . . . . Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for Speech Discrimination Scores Within the Sen- sorineural Precipitous Drop GrOUp. . . . . . . Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for Speech Discrimination Scores Within the Sen- sorineural Gradually Sloping Group . . . . . . Mean Speech Reception Thresholds Obtained for Three Groups of Subjects Under one Unaided and Four Aided Listening Conditions. . . . . . . . Mean Speech Discrimination Scores Obtained for Three Groups of Subjects Under One Unaided and Four Aided Listening Conditions. . . . . . . . Median (Med), Mean (M), and Standard Devia- tions (SD) of Speech Discrimination Scores Ob- tained with N.U. Auditory Test No. 6 at a 24 dB Sensation Level for Ten Subjects with Norm- al Hearing During the First Test Session. (Scores Represent Percent of Items Correctly Repeated). . . . . . . . . . . . . . . . . . . Median (Med), Mean (M), and Standard Devia- tions (SD) of Speech Discrimination Scores Ob- tained with N.U. Auditory Test No. 6 at a 24 dB Sensation Level for Ten Subjects with Normal Hearing During Retest Session. (Scores Represent Percent of Items Correctly Repeated) Difference Between Mean Discrimination Scores from Test to Retest at a 24 dB Sensation Level for Ten Normal Hearing Subjects on N.U. Audi- tory Test No. 6 (Negative Difference Indicates Higher Score in Retest than in Test Session) . vi Page 89 90 95 97 100 122 125 125 LIST OF TABLES - Continued Table XVIII. XIX. Page Coefficients of Correlation (Pearson r) and Standard Error of Measurement (Se) Between Test and Retest for N.U. Auditory Test No. 6 Administered to Ten Subjects with Normal Hearing at a 24 dB Sensation Level. . . . . . . 125 Unoccluded and Occluded Speech Reception Thres- holds Showing Attenuation Provided by Wax Impregnated Ear Plugs (Flents) for Spondee Words. (Thresholds in dB re Speech Audio- metric Zero). . . . . . . . . . . . . . . . . . 127 vii LIST OF FIGURES Figure 1. Audiogram Showing mean air and bone conduction thresholds for Group I (Conductive Impairment), N = 10 . . . . . . . . . . . . . . . . . . . . Audiogram showing mean air conduction thres- holds for Group II (Sensorineural Impairment with a Precipitous Drop) N = 10- . . . . . . . Audiogram Showing mean air conduction thres- holds for Group III (Sensorineural Impairment with a Gradually SIOping Loss). N = 10. . . . The acoustic Spectrum of a broad band white noise through the Grason-Stadler LoudSpeaker, Model 162-4. . . . . . . . . . . . . . . . . . Frequency reSponse of the Radioear, Model 950 CROS type hearing aid with an input Signal of 60 dB re 0.0002 microbar . . . . . . . . . . . Schematic diagram of the test environment Show- ing placement of equipment and location of the subject during testing . . . . . . . . . . . . Mean discrimination scores yielded by ten normal hearing subjects for Lists I, II, III, and IV of N.U. Auditory Test No. 6 during both test and retest sessions . . . . . . . . . . . Mean audiogram of three subjects with normal hearing showing attenuation provided by wax impregnated ear plugs (Flents) for pure tones. (Thresholds ISO-1964 Standards). . . . . . . . viii Page 46 48 49 54 57 65 124 126 LIST OF APPENDICES Appendix Page A. Results of Research Project with N.U. Audi- tory Test No. 6. . . . . . . . . . . . . . . 121 Attenuation Provided by Wax Impregnated Ear Plugs (Flents) for Pure Tones and for Spondee Words. . . . . . . . . . . . . . . . 125 Summary Showing Data for Each Subject. . . . 128 ix CHAPTER I INTRODUCTION The recommendation of a hearing aid in the rehabilita— tion process of a hard-of-hearing person has become an increasingly important aSpect of clinical audiology. A re- cent National Conference concerning hearing aid evaluation procedures has focused attention on this problem.1 However, the manner in which the hearing aid is connected to the ear has been neglected even though it is this final acoustic COUpling which ultimately determines the reSponse of the instrument. The standard procedure has been to utilize a conventional type earmold irreSpective of type of hearing loss or pure-tone audiometric configuration. The only cri- teria for the earmold was that it provide a tight seal at the ear, so that there was no leakage of the amplified sound to create acoustic feedback. The problem of acoustic feedback has been eSpecially pronounced in ear-level aids because of the close proximity of the micrOphone and receiver, and little could be done to modify the earmold without creat- ing an even greater problem. However, with the advent of luA Conference on Hearing Aid Evaluation Procedures," ASHA Reports, Number 2, 1967. the CROS1 type hearing aid, it has become possible to use modified earpieces, since the problem of acoustic feedback is reduced. Research is now needed with various types of acoustic couplers to determine their effects on different types of hearing losses and pure-tone audiometric configura— tions. Purpose of the Study The major purpose of this research was to determine whether variations in the way a hearing aid is acoustically COUpled to the ear affects the Speech reception thresholds and Speech discrimination scores obtained by subjects having different types of hearing losses and audiometric configura- tions. Specifically, this investigation was concerned with the effects that a conventional (stock) earpiece, a vented earpiece (the Zenith acoustic modifier), an Open earpiece, and crimped polyethylene tubing have on the aided Speech reception thresholds and Speech discrimination scores of sub- jects having either a conductive hearing loss or a sensori- neural type hearing loss. Subjects with a sensorineural type hearing loss had one of two types of audiometric configura- tions: (1) a precipitous drOp for frequencies higher than 500 or 1000 Hertz or (2) a gradually sloping loss with the lower frequencies being affected. lEarl Harford and Joseph Barry, "A Rehabilitative Ap- proach to the Problem of Unilateral Hearing Impairment: The Contralateral Routing of Signals (CROS)," Journal of Speech and Hearing Disorders, 50 (1965), pp. 121-158. In addition to the above purpose, the subjective quality of the amplified sound presented through the various acoustic couplers was investigated by obtaining quality judgments from the subjects. Significance of the Study In the 1940's investigators such as Schierl and Grossman2 pointed out that the way a hearing aid receiver was coupled to the ear could drastically alter the reSponse character- istics of that aid. Even though this has been an accepted fact for many years, relatively little research has been done in this area. The present study is significant by virtue of the fact that it is a controlled investigation of the Speech reception threshold as a function of the type of acoustic coupler em- ployed. The determination that modified earpieces have a significant effect on the Speech reception threshold of hard- of-hearing subjects has ramifications for changing the hear— ing aid evaluation procedures now employed in many clinics. Perhaps, the subtle differences among various hearing aids might be better indexed by evaluating their performance with different types of earmolds. That is, where differences 1Mayer B. A. Schier, "The Earpiece-~In Testing for and Fitting Hearing Aids," Laryngoscgpe, 51 (1941), pp. 52-60. 2Frederick M. Grossman, "Acoustic Sound Filtration and Hearing Aids," Archives of Otolaryngology. 58 (1945). pp. between two aids were not obvious in the past, differences in performance may Show up when they are coupled to the ear in different ways. Hard-of~hearing persons with certain types of sensori— neural hearing losses often have a great deal of difficulty in discriminating acoustic stimuli. Their problem may be such that a hearing aid is of no benefit to them when a con- ventional earpiece is employed. Some of these people might be helped by employing a different type of coupling. Dis- crimination scores might be improved by employing an ear- piece that takes advantage of the natural resonance character- istics of the external auditory canal and that does not change the impedance characteristic from the normal state by completely closing the ear canal. Considerable controversy exists regarding the clinical fitting of hearing aids. In view of this, research which contributes information on this aSpect of clinical audiology will make a contribution to this expanding body of knowledge. Definitions The following definitions of terms were employed in this investigation: Acoustic Modifierl--With this particular type of earpiece, the portion entering the external auditory canal is almost lZenith Radio Corporation trade name for their patented, vented earmold. entirely removed and the remaining portion is enlarged. Usually two small vents are cut in the flat portion of the mold and communicate with the larger inner Opening. The vents have thin discs of wax impregnated lamb's wool inserted in them. Conventional Earmold or Earpiece--There are actually many different types, but the concern in this study was the fact that the ear canal was completely sealed by the earmold. The earmold is solid, and there are no vents. Open Earpiece--This type of earpiece is designed so that the contour of the concha is outlined by a plastic rim which has a projection extending into the schaphoid fossa by means of which it is held in position. The lower part of the plastic rim has a small Opening into which is inserted the poly- ethylene tube that delivers the amplified signal. This ear- piece doeS not occlude the ear canal, a fact which is its important feature. Crimped Tubingr-This consisted of a stock piece of polyethylene tubing measuring 2% inches in length and 0.077 inches in diameter bent in such a manner (approximately 90 degrees) that it remained in the ear canal during the testing of a hearing aid. Conductive Hearing Impairment--For purposes of this study, a conductive hearing loss was defined as one where bone con- duction thresholds were within the normal range (no greater than 25 dB ISO—1964 Standards) and where there was an air- bone gap of at least 20 dB for the test frequencies 500, 1000, and 2000 Hertz. High Frequency Sensorineural Hearing Loss-—Normal hearing (25 dB or better ISO-1964 Standards) for the low frequencies with a precipitous drop Of at least 20 dB for the first octave beginning at 500 or 1000 Hertz. Air and bone conduc- tion thresholds were interweaving. Gradually SlOping Sensorineural Hearing_Loss-—This was defined as a progressively greater loss for higher frequencies at a Slope of 5 to 10 dB per octave with the loss beginning in the low frequencies. Air and bone conduction thresholds were interweaving. CHAPTER II REVIEW OF THE LITERATURE A review of the pertinent literature includes a brief history of the development of earmolds. The development of hearing aids is also reviewed since they are used in conjunction with earmolds. The second section of the literature review is con- cerned with studies of the acoustic prOperties Of the head and external ear and with the principles behind the develop- ment Of modified earmolds. Finally, studies concerned with applying the fore- going acoustic principles in experiments with earmolds are reviewed. History The development Of acoustic couplers to connect the hearing aid receiver to the ear, Of necessity, closely parallels the development Of hearing aids themselves. Before the turn Of the century, hearing aids consisted of sound collecting devices such as the ear trumpet, which collected sound and funneled it into the external ear canal.1 lLeland A. Watson and Thomas Tolan, Hearing Tests and Hearing Instruments (Baltimore: Williams & Wilkins Company, 1949). p. 268. They not only collected sound, but also were resonators which amplified certain frequencies within the Speech range and often yielded a 10 to 15 dB gain in acoustic energy reaching the ear.1 These sound collecting devices were acoustic couplers, which actually extended the external ear canal and, in so doing, could be eXpected to change the overall resonance and impedance characteristics of the ear giving rise to changes in the subjective quality of sound in addition to a small amount of amplification. This is supported by the Observa- tions Of Schier2 in the early 1920's. He experimented with small ear trumpets, which he made by taking a modeling com- pound impression Of the ear and making a vulcanite repro- duction. Various canopies Of different sizes, shapes, and Openings were vulcanized to it. He found that the varia- tions of Size, Shape, and cavities gave a different subjec- tive quality to sound. 3 Kranz, in describing the ear trumpet or "ear horn" as he termed it, stated, "The Size Of the horn will of course lHallowell Davis and S. Richard Silverman, Hearing and Deafness (New York: Holt, Rinehart, & Winston, 1960, p. 266. 2Mayer B. A. Schier, "Clinical Phenomena in Conductive Media: The Individual Earpiece," Journal of the Acoustical Society of America, 17 (1945). PP- 77-82. 3Fred W. Kranz, Hearing Aids (Elmsford, New York: Sonotone Corporation, 1941), p. 9. influence its effectiveness, while the shape of the horn will influence the quality Of the sound received through it." Thus, for more than twenty-five years there has been recog- nition that the physical attributes of the acoustic coupler had an important influence on the amplification and quality of the perceived sound. In 1900, Dr. Ferdinand Alt of the Politzer Clinic in Vienna conceived of and produced the first amplified elec- trical hearing aid.l'2 In 1902, the "Oriphone" was produced by C. W. Harper and the "AkOUphone" was produced by Miller Reese Hutchinson. These early carbon type hearing aids employed a flat, over-the-ear magnetic receiver kept in place by means of a headband.3 With this type of receiver there was no need for any coupler to the ear, Since the receiver itself covered the ear much like modern earphones. Although the literature is not clear as to the exact date, the small, button type receiver was develOped for use with some of the earlier carbon type hearing aids and later with some of the vacuum tube type hearing aids.4 The hearing aid companies developed stock connectors or earpieces in order to hold the receiver in the ear. Their acoustic lChevalier Jackson, Diseases of the NoseL Throat, and Ear (Philadelphia: W. B. Saunders, 1945), p. 278. 2Watson and TOlan, Hearing Tests and Hearing Instru- ments, p. 270. 31bid., p. 275. 4Ibid., p. 276. 10 importance was not recognized, and was Simply a means of retaining the receiver in the ear. The individually molded earpiece was used only in cases where a person had a prob- lem keeping the receiver in his ear.1 Gradually individually fitted earpieces began to be accepted. These were usually made of hard rubber and proved to be inadequate, since it was difficult to control the de- gree and thoroughness Of vulcanization of hard or soft rubbers resulting in inconsistent acoustic qualities from one mold to another. Another problem was the fact that the acoustic prOperties of rubber tended to change over time due to deterioration, so in the late 1920's and early 1950's the possibility of using other materials was eXplored. It was 2 who claimed to have develOped the first acrylic Schier earpiece. His eXperiments with miniature hearing trumpets led him to develOp a variety of earmolds which would enhance the amplified sound from a hearing aid. The interest in various types of earmolds to improve sound reception can be understood in light of the fact that early hearing aids were usually Of the carbon type. These were bulky and noisy, with a limited frequency range, and almost no tone adjustments.3 Thus, the develOpment Of 1Schier, "Clinical Phenomena," p. 78. 21bid., p. 79. 3Kranz, Hearing Aids, p. 11. 11 various types of modified earpieces was an attempt to counterbalance the deficiencies of the carbon type amplifier hearing aid. Acoustic Phenomena In order to appreciate efforts to improve sound recep- tion through modified earmolds and the principles behind these efforts, it is necessary to review the acoustic prOper- ties of the head and external auditory canal, and the acous— tic pathway between the transducer and the tympanic membrane. One of the principle investigators Of this area was Bekesy.l In 1952 he demonstrated that the sound pressure develOped at the surface of the head was quite different from that in the undisturbed sound field and that it increases at the head as the frequency gets higher. He also found a difference in the amount of sound pressure develOped at the entrance of the external canal from that develOped at the tympanic membrane. His data indicate that the sound pressure at the eardrum may be as much as three times greater in the important Speech frequency range between 2000 and 5000 Hertz. The resonance curve determined by Bekesy is, of course, dependent on the medium through which the sound passes, the length and cross—sectional area of the external canal, and the impedance of the tympanic membrane. Any changes in the dimensions Of these factors, such as the insertion of an lGeorge Von Bekesy Cited by Stanley Smith Stevens and Hallowell Davis, Hearing: Its Psychology and Physiology (New York: John Wiley & Sons, Inc., 1958), p. 55. 12 earmold into the external canal, would result in a change of the resonance curve. In 1955, Sivian and White1 carried out an extensive study of minimum audible pressures (MAP) and minimum audible fields (MAF) using pure tones to measure threshold. The MAP threshold is measured under earphones while the MAF was defined as the intensity Of the free field measured prior to the insertion of the Observer. The sound field into which the Observer was placed facing the source was substantially that of a plane progressive wave. The data represent monaural MAF thresholds on 14 ears over the frequency range from 100 tO 1500 Hertz and binaural hearing on 15 Observers over the frequency range from 60 to 15,000 Hertz. Their findings indicate the MAP thresholds are from 5 to 12 dB higher (poorer) than MAF, with a maximum difference in the frequency range between 2000 and 4000 Hertz. The average difference was on the order of 6 dB. A number of hypotheses have been offered as an eXplanation Of the dis- crepency noted between the two kinds of measurements. Probably the most generally accepted explanation is that thresholds become higher as a function Of the amount of air enclosed between the transducer and the tympanic membrane.2 1L. J. Sivian and S. D. White, "On Minimum Audible . Sound Fields," Journal of the Acoustical Society of America, 4 (1933), pp. 288-521. 2Tom W. Tillman, Robert M. Johnson, and Wayne 0. Olsen, "Earphone versus Sound Field Threshold Sound-Pressure Levels for Spondee Words," Journal of the Acoustical Society Of America, 59 (1966), pp. 125-155. 15 Because Of the trapped volume of air, the natural resonance of the external canal and the impedance characteristics Of the ear are quite different from the natural state. The maximum difference noted between MAP and MAF in the frequency range between 2000 and 4000 Hertz should be an important consideration in the fitting of amplification to a sensorineural hearing loss, since it is precisely this important Speech frequency range which is most likely af- fected. It would appear that the acoustic coupler between the hearing aid receiver and the ear should be designed to take advantage of the natural resonance and impedance characteristics of the ear. Sabinel in 1942, studied the resonance characteristics of small cavities from 2.0 to 8.45 cubic centimeters in volume. The acoustic pressure develOped in the cavities was compared to that developed at the face of an unapertured baffle. The following results were found: (a) Cavities of the order of magnitude here considered Show marked resonance characteristics, with pres- sure level amplifications at resonance as great as 20 dB. (b) For frequencies well below resonance, pressures within the cavity do not differ markedly from those at the face of the unperforated baffle. (c) For frequencies well above resonance, there is a marked pressure attenuation within the cavity. (d) The resonant frequency decreases with increasing 1Paul E. Sabine, “On the Acoustic PrOperties of Small Cavities," Journal of the Acoustical Society of America, 15 (1942), pp. 74-78. 14 cavity volume and increasing hole depth, and in- creases with increasing hole diameter.1 The above findings indicate that the differences between MAP and MAP noted by Sivian and White2 can be expected be- cause of the cavity resonance reSponse Of the external ear. This reSponse terminated by the compliance of the tympanic membrane and the attached ossicles is Similar to that of the reSponse Of the apparatus used by Sabine. In that case the reSponse was terminated by the compliance of the cavities. This investigation Of Sabine's grew out Of his attempts to quantify hearing aid performance objectively. The dif- ference in pressure level was determined between a hearing aid microphone placed on the chest Of a "dummy" and the diaphragm of a condensor micrOphone terminating an artificial ear employing a 2 cc coupler mounted in the head of the dummy. Sabine stated: "A very marked difference in results following even slight changes in the coupler dimensions sug- gested a more thorough going investigation of coupler ef— 3 It seems plaus- fects on the results of such measurements." ible to assume that these effects take place when the dimensions of the acoustic coupler between the hearing aid receiver and ear of a hard-of—hearing person are changed. lIbid., p. 77. aSivian and White, "Minimum Audible Fields," pp. 288- 521. 3Sabine, "Acoustic PrOperties Of Small Cavities," p. 78. 15 In 1946, Weiner and Ross1 measured the sound pressure at different points along the auditory canal of a number of male and female subjects placed in a sound field which was essentially that of a plane progressive wave. The measurements were made by means of a small, flexible probe micrOphone (Western Electric Type 640-AA condenser micro- phone). The probe was placed at various locations along the auditory canal. Their data Showed that the sound pres- sure at the tympanic membrane is greater than the free field pressure and reaches a maximum Of about 17 to 22 dB near 5000 Hertz. The ear canal, then, acts as an acoustic ampli- fier over most of the important Speech frequency range. The authors attribute the increase in sound pressure at the eardrum over that of a free-field to the combination effect of diffraction by the head and pinna and resonance in the external auditory canal. A few years later, in 1950, Munson and Weiner2 studied the variability among methods for determining threshold for pure tones. Included in their study were measurements of MAP and MAF, and their data indicated a discrepancy in which MAF thresholds were lower in sound pressure level by an 1Francis M. Weiner and Douglas A. Ross, "The Pressure Distribution in the Auditory Canal in a Progressive Sound Field," Journal of the Acoustical Society of America, 18 (1946), pp. 401—408. 2W. A. Munson and Francis M. Weiner, "Sound Measure- ments for Psychophysical Tests," Journal of the Acoustical Society Of America, 22 (1950), pp. 582-586. 16 average of 6 dB. In 1952, they conducted a more thorough investigation of MAP and MAF in the low frequencies.1 Using a pair of Western Electric 711A receivers and a large dynamic loudSpeaker coupled to a folded horn, measurements were made at 60, 120 and 240 Hertz. The average differences of the MAP/MAF ratio over ten subjects were 15.5 dB, 9.6 dB, and 5.2 dB reSpectively with higher thresholds by MAP. It is interesting to note that the authors found that a slight air leak caused by a poor fit of the receiver cap over the ear resulted in a drOp of sound pressure at low frequencies. This is, of course, the reason why vented earmolds were recommended for high frequency hearing losses, since, in effect, with prOper venting the earmold can become a high pass acoustic filter. Rudmose2 eXplained the difference between MAP and MAF at the low frequencies as a function of the mechanical "isolation“ between receiver and ear, the quality Of the seal, and the volume enclosed between receiver and ear. The studies concerning MAP and MAF thus far reviewed have all employed sinusoidal Stimuli. The next area of concern is the effects of Speech on thresholds obtained under 1W. A. Munson and Francis M. Weiner, "In Search of the Missing 6 dB," Journal of the Acoustical Society of America, 24 (1952), pp. 498-501. 2Wayne Rudmose, “Free-Field Thresholds vs. Pressure Thresholds at Low Frequencies," Journal of the Acoustical Society_of America, 22 (1950), p. 674. 17 1 made an extensive these two conditions. Breakey and Davis investigation of the difference between MAP and MAF for Speech stimuli. In their eXperiments they used Psycho- Acoustic Laboratory Test No. 9, which is comprised of Spondee words, and Test NO. 12, which consists Of Simple sentences. Ten subjects with normal hearing and ten who were hard-of- hearing were administered the above tests monaurally and binaurally through headphones. In addition, the normal group listened binaurally in a sound field. Combining all listening conditions for the normal- hearing group, they found that mean thresholds were about 5 dB lower undertflmesound-field conditions than under FDR-10 earphones. The authors stated: The difference of about 5 dB (average of all tests) between field and receiver listening is real, although not so large as would be expected from the classical data on minimum audible pressure and minimum audible field thresholds for pure tones. The smaller difference in the present series may be due in part to the fact that our field is not a "free" field. It is also due in part to the unusually low average threshold found for Test NO 9 by receiver listening. Because of the uncertainties expressed by the authors of the foregoing study, Tillman, Johnson, and Olsen3 1M. R. Breakey and Hallowell Davis, "Comparisons of Thresholds for Speech: Word and Sentence Tests; Receiver vs Field and Monaural vs Binaural Listening," Laryngoscqpe, 59 (1949), pp. 256-250. 21bid., p. 241. 3Tillman, Johnson, and Olsen, "Earphone versus Sound Field," pp. 125-155. 18 undertook an investigation in an attempt to define the dif- ference between MAP and MAF threshold sound-pressure levels for Spondee words. A secondary goal was to determine the effect of earphone type (conventional versus insert type) on MAP and MAF differences. Two groups of subjects were utilized in this study. The first group was composed Of 12 individuals with normal bilateral hearing, and the second grOUp was composed of 10 persons with mild to moderate bilaterally symmetrical hear- ing losses Of the sensorineural type. Monaural thresholds were measured using tape-recorded Spondaic words used in the construction of CID Auditory Tests W-1 and W-2. The re- ceiver used was a TDH—59-1OZ earphone housed in an MX 41/AR cushion, which enclosed approximately a 6 cc volume of air between its diaphragm and the eardrum, and a Radioear M75 insert type receiver coupled to the subject's ear via a stock earmold, which enclosed approximately a 2 cc volume of air. The results indicated that the differences between MAP and MAF were essentially the same for both groups Of sub- jects. The average difference between MAF and MAP for the conventional earphones was 7.5 dB, and for the insert ear- phones this difference increased to 12.5 dB with MAF being lower in both instances. The authors offered the following conclusions from this eXperiment: 19 First, the "missing 6 dB" initially described by Sivian and White is indeed a real phenomenon and can be demonstrated utilizing Speech as well as pure- tone stimuli. Second, the magnitude of the difference between MAP and MAF increases if the pressure thres- holds are measured using an insert-type receiver rather than the more conventional earphone. Assuming that, insofar as a Spondee test signal is concerned, the volume of air trapped between the earphone dia- phragm and the tympanic membrane represents the major difference between the two pressure transducers used in this investigation, one could restate this latter conclusion as follows. The difference between MAP and MAF increases in magnitude as the volume of air enclosed between the pressure transducer and the ear- drum decreases in magnitude.1 The authors further stated that the difference between MAP and MAF is caused, in part, by diffraction effects and, in part, by impedance mismatches resulting from enclosure of the ear canal by the transducer. From the studies discussed in this section, eSpecially the last two, the following question was raised for the present research project: "Will subjects Obtain lower Speech reception thresholds using either an Open earmold or crimped tubing rather than the conventional earmold?" It would ap— pear plausible to hypothesize that lower Speech reception thresholds can be obtained with the Open earmold or crimped tubing, Since Tillman, Johnson, and Olsen2 found a 12.5 dB difference between MAP and MAF using a transducer coupled by a stock earmold enclosing a volume Of air of approximately two cubic centimeters. If, in fact, the volume of air lIbid., p. 131. 21bid., p. 150. 20 enclosed between the transducer and the tympanic membrane is reSponsible for the difference, then the difference should decrease when an open earmold or crimped tubing are used, since with these couplers a sound-field condition is more closely approached. The absolute magnitude of the difference would not be eXpected to be as great as that found by Tillman et al., Since this eXperiment was conducted in a sound-field rather than a free-field. By the same token, the eXpected difference should be greater than the 5 dB found by Breakey and Davis,1 since the FDR-10 receivers used by them enclose approximately a 6 cc volume of air, and this larger volume would tend to decrease the difference. In other words, the difference in SRT between the standard earmold and the Open earmold should be somewhere between 5 dB and 12.5 dB under sound—field conditions. Modified Earpieces The foregoing section discussed some Of the important principles that must be considered in the transmission of sound to the human ear. A discussion of how these principles were applied to the develOpment of, and eXperimentation with various types of earmolds is now in order. In 1956, Littler2 carried out and discussed a number Of 1Breakey and Davis, "Comparison of Thresholds for Speech," p. 242. 2T. S. Littler, "Hearing Aids for the Deaf," Journal Of Scientific Instruments, 15 (1956), pp. 144-155. 21 experiments involving the design, use, and testing of appara- tus used for hearing aid purposes with hard-of—hearing sub- jects. He pointed out that a large number of cases have a high frequency hearing loss, such that amplification which has a reduced high-frequency reSponse seriously affects Speech intelligibility for these peOple. From his studies he concluded: "There is a need for improvement in the manner of applying the sounds to the ear, as it seems that the present design of earpiece causes a serious loss in the upper frequencies."l By 1941, Schier2 had become vehement in his criticism of the quality of custom fitted earmolds. He felt that they fell short of the claims made for them and that they followed commmercial dictates rather than the needs Of the individual. He pointed out that a great deal of energy was invested in all things pertaining to the hearing aid itself but that not much thought was given to the earpiece, which could readily change the reSponse of the instrument. His Specifications for an earmold, in addition to being small, inconSpicuouS, and light in weight, included this statement: "That portion known as the actual tip should be as long as comfortable "3 depth Of entry into the canal will permit. Further on he lIbid., p. 155. 2Schier, "The Earpiece," p. 55. 31bid., p. 55. 22 stated: "The longer the earpiece tip, the greater and truer the sound conduction."l Other researchers would disagree with Schier's Specifi— cations for tip length of an earmold, since lengthening the tip would tend to attenuate the high frequencies and such a mold would not be suited for a high-frequency hearing loss.2 3 did feel, however, that filtration in the acoustic Schier path had demonstrated advantages. His filtration consisted Of inserting a small device composed of miniature acoustic chambers between the receiver nib and sound channel. He found that "The frequency-response curve of an instrument can be so affected as to modify the relativity of the low, medium, and high frequencies as emitted from receiver."4 His find— ings were confirmed by tests conducted at the Sonotone Corporation Laboratories at his request. These tests re- vealed reSponse curves quite different from one another using various chambered devices in the line between receiver and sound channel. The effects of these devices could be seen as shifts in the peaks Of the frequency-reSponse curves from lower to higher frequencies and also by a change in peak intensities. lIbid., p. 59. 2Thomas H. Halsted and Frederick M. Grossman, "Modern ASpectS of the Hearing Aid Problem," New York State Journal of Medicine, 42 (1942), pp. 1944-1950. 3Schier, "Clinical Phenomena," p. 80. 41bid. 25 In 1941, Halsted and Grossmanl reviewed the different classes of hearing impairment and discussed the various types of amplification best suited for each. At that time they advocated the individually fitted earmold but did not go into detail as to its Specifications. In a follow-up article published the following year, they made more Specific recommendations as to the type of earmold best suited for each of four broad classes Of hear- ing impairment.2 They stated: "The small air volume between the receiver and the drum and the shape of it has an influ- ence on the final characteristics of the acoustical per- formance. The same applies to the size, width, and length of the sound-conveying canal of the ear mold."3 For Class I, or conductive losses, they recommended an earmold with a long tip and a sound canal of approximately 5 mm. in diameter. The reasoning behind this is that the larger surface area Offered by the longer tip makes contact with the walls of the ear canal and hearing is improved by bone conduction, which is normal or near normal. Class II is a mixed type Of loss for which they do not give any recommendations, since they had not observed a 1Thomas H. Halsted and Frederick M. Grossman, "Some Problems Involved in the Fitting Of Hearing Aids," New York State Journal Of Medicine, 41 (1941), pp. 552-558. 2Halsted and Grossman, "Modern ASpects of the Hearing Aid Problem," pp. 1944-1950. 31bid., p. 1947. 24 sufficient number of cases. Class III is an abrupt loss of the high frequencies which is sometimes called "boilermaker's deafness," and Class IV is a more gradual loss of the high frequencies. Both losses are of the sensorineural type. They stated that "The ear mold for Class III and IV should have a short tip, and the sound-conveying canal should be as wide as possible. A small acoustic high-pass filter between receiver and mold improves results."l Grossman2 experimented with three individual earmolds each having a different diameter sound-conveying canal. Mold One had a conventional canal diameter of 5 mm, Mold Two 1.5 mm and Mold Three a diameter of 5 mm. The length of each was 22 mm. Each of the molds was connected to a vacuum tube hearing aid in succession, and the examiner Spoke into the microphone of the aid from a distance of 10 feet. Grossman stated the following about his eXperiments: The conditions of eXperimentation were chosen in such a way that the hearing aid connected with mold 1 gave good intelligibility. As pointed out, the im- pression was that the high partial tones were weak. It was quite a strain to make out the consonants, the recognition of which depends primarily on Upper partial frequencies. The results with mold 2 were rather startling. Speech sounded less loud than with mold 1, but after adjusting the volume to a comfortable loudness it was almost impossible to understand a single word. The words sounded dull, and the impression was that the lIbid., p. 1949. 2Frederick H. Grossman, "Acoustic Sound Filtration and Hearing Aids," Archives of OtolaryngolOQY. 58 (1945), pp. 101-112. 25 higher frequencies were cut off entirely. On the other hand, when mold 5 was used, Speech sounded brighter than when mold 1 was used, and more natural. It is be- lieved that in this eXperiment the "naturalness" depends on the smallest interference Of all three molds with the normal dimension of the aural canal and also on the lack of filter action in mold 5.1 Another eXperiment was performed in which a tube, 5 mm long with a diameter of 4 mm, and having a side branch ori- fice which had a diameter of 1.2 mm and a length of 1 mm, was placed between the receiver and mold 5. It was found that with this arrangement the loudness was reduced by a consider- able amount as compared tO when the tube was not used.2 From his eXperimentS, Grossman concluded: "The ear mold of present design renders the acoustic line between the re- ceiver and the drum a finite low pass filter. The longer the inserted sound canal of the mold and the smaller its "3 He reiterated diameter, the stronger is the filter action. his earlier conviction that high frequency losses Should be fitted with an earmold that has a Short tip and a large, straight sound-conveying canal. He also recommended the use of an acoustic high pass filter with this type of loss. Grossman and Molloy‘ carried out further studies aimed at investigating eXperimentally variations in the acoustic lIbido, pp. 105-104. 21bid., p. 104. 31bid. 4Frederick M. Grossman and Charles T. Molloy, "Acoustic Sound Filtration and Hearing Aids," Journal of the Acoustical Society of America, 16 (1944), pp. 52-59. 26 pathway between the hearing aid receiver and the tympanic membrane and to analyze, mathematically, the phenomena in- volved. Four different earmolds were used in the study. Two were conventional with canal diameters Of 2 to 5 mm. The third had a canal diameter of 15 mm, and a fourth had a diameter of 5 mm. Three had a canal 20 mm in length and the fourth had a canal 18fi'mm in length. A brass tube 5 mm long and having a diameter Of 4 mm served as a high pass filter between the receiver and earmold. It had a side branch hole 1 mm long with a diameter of 1.2 mm. Two differ- ent receivers were used, which were connected to an oscil- lator. The output of the oscillator was varied continuously by a motor drive. The receiver was connected to the earmold which in turn was coupled to a dynamic micrOphone. The re— ceiver output was picked up by the microphone and fed to an amplifier and then to a level recorder. Results of these eXperiments showed that the narrow mold yielded a broader peak frequency reSponse than the others. For all the earmolds, the low frequencies were reduced with the filter in place. An interesting comparison was that be- tween the conventional earmold with and without the filter. The filtered reSponse showed a rather uniform weaker output of 15 to 18 dB up to 1000 Hertz. With the filter in place the output of the receiver was slightly higher than without the filter between 5000 and 4000 Hertz. 27 Nichols et al.1 made a thorough investigation of the effect of leakage between the earpiece and the ear canal by means of individually molded earpieces having micrphone probe—tubes mounted in them. They measured the sound pres— sure develOped in the ear canal at the tympanic membrane as a function of frequency by the earphone under two conditions: (1) with the earpiece sealed to the ear canal by means of beeswax and lanolin and (2) with the earpiece worn normally. Three subjects evidencing a very snug fitting earpiece, a moderately snug fit, and a loose fitting earpiece were tested using a number Of different earphones. Their results Showed that the reSponse of the various earphones was increasingly affected as the earmold fit be- came less snug. The low-frequency reSponses were weakened. There were no striking differences in the behavior of the various earphones on any particular earmold-ear combination. The authors stated: "The results of these tests indicate clearly that the effects on the reSponse characteristics Of a hearing aid due to the relative snugness of fit of an ear- piece tO the wearer's ear may range in magnitude from prac- tically zero to as much as 15 or 20 dB at low frequencies."2 1R. H. Nichols Jr., R. J. Marquis, W. G. Wiklund, A. S. Filler, D. B. Feer, and P. S. Veneklasen, "Electro-Acoustical Characteristics Of Hearing Aids," Hearing and Hearing Aids, Sec. I, U. S. Office Of Scientific Research ETDevelopment Report No. 4666 (Cambridge, Mass.: Harvard University, 1945), pp. 44-68. 2Ibid., p. 60. 28 All of the early studies, cited in this section, indi- cate that the acoustic stimuli reaching the tympanic membrane of the ear can be drastically changed by modifying the acoustic coupler. Unfortunately, after the initial surge Of eXperimentation with modified earpieces during the early 1940's, interest waned. One of the factors contributing to this loss of interest was the shift in attention to the development of the new vacuum tube hearing aid with the empha- sis of modifying the frequency reSponse of the hearing aid itself or the receiver rather than the earpiece. In 1954, an English firm, the Thomson Houston Company, started to manufacture small, battery-Operated vacuum tubes; and in 1957, the first wearable vacuum tube hearing aid in America was develOped by Arthur Wengel and marketed under the name Stanleyphone.l Single unit vacuum tube hearing aids, however, were not put on the market until 1945.2 Since the vacuum tube instrument had much better fidelity than the old carbon type, interest was turned toward modifying the reSponse characteristics of the instrument electronically rather than acoustically. Modified earpieces were used in- frequently, and the conventional closed earmold became standard equipment with most hearing aids. The trend toward "mirroring the audiogram" by electron- ically modifying the output of the hearing aid might have 1Watson and TOlan, Hearing_Tests, pp. 280-281. 21bid., p. 512. 29 continued without controversy had it not been for an impor- tant study published in 1947. This research by Davis et a1.1 became known as the "Harvard Study" and had a profound effect on the hearing aid industry. One of the purposes of this study was to determine what type of hearing aid fre- quency reSponse could be used most satisfactorily by patients with various types and degrees of hearing loss. As the result of their very intensive study, the Harvard group concluded that an instrument with a flat frequency re- Sponse, or a rising 6 dB per octave frequency output, was the most suitable for the majority of hard-Of-hearing peOple. In addition, they outlined certain other Specifications to which a hearing aid Should conform. With regard to these Specifications they wrote: It is anticipated that when instruments conform- ing to the above Specifications are produced the problem of individual selection of "fitting" will almost disappear. It will be necessary only to: (a) Provide a well-fitting earpiece that is comfortable and at the same time provides adequate acoustic seal, and (b) Select a model with adequate acoustic gain and make the apprOpriate semipermanent adjustment to provide the prOper limitation of maximum power out— put. From the findings Of this study, it is easy to under— stand why manufacturers tended toward producing hearing aids 1Hallowell Davis, S. S. Stevens, R. H. Nichols, Jr., C. V. Hudgins, R. J. Marquis, G. E. Peterson, and D. A. Ross, Hearing Aids: An EXperimental Study of Design Objectives, (Cambridge, Massachusetts: Harvard University Press, 1947), 197 pp. 2151a., p. 115. 50 with rather flat frequency reSponses or with a slight high- frequency tilt. Another factor which contributed to the manufacture of hearing aids having much the same type of frequency reSponse was the utilization of the transistor in hearing aids in the early 1950's. With the advent of the transistor, hearing aids could be made much smaller and still retain considerable power. However, with miniaturization it is more difficult to maintain good fidelity, Since the small- er components are incapable Of giving the same reSponse as the larger ones. Also, various tone controls and circuitry must be eliminated in order for the aid to be made smaller. The modern, ear-level hearing aid tends to be an instrument equipped with little more than a gain control, thus electrical modification of the frequency reSponse is often impossible. The earmold as a means of acoustically modifying the output of hearing aids once again became an important con- sideration; and in 1958, Lybargerl Offered a thorough dis- cussion Of how the earmold's hole diameter, tip length, leakage, and venting affect hearing aid reSponse. In the discussion on venting, he pointed out that indiscriminate venting may reduce the extreme low frequencies, but that the important lows for Speech may actually be increased. He also stated that the larger the cavity between the earmold tip and eardrum, the weaker will be the sound pressure 1S. F. Lybarger, "The Earmold as a Part of the Receiver Acoustic System," (Canonsburg, Pennsylvania: Booklet pub- lished by Radioear Corporation, 1958), pp. 12. 51 develOped at low frequencies. This last principle is one on which the "acoustic modifier" is based. Also, the length and diameter of the tubing used to connect the receiver of an ear—suSpended or glasses type hearing aid to the earpiece can considerably modify the instrument's output. According to Lybarger,1 by Shortening the tube, with diameter held con- stant, the primary and secondary frequency peaks will be Shifted toward the high frequencies. Increasing the length Of the tube, with diameter held constant, has the converse effect. Since in the ear-suSpended or glasses type hearing aid the micrOphone and receiver are in close proximity to one another, there has always been a problem of feedback from any type of leakage. This is a factor which led Lybarger to con- clude: "Except to provide a good fit with comfort, the actual earmold part of the receiver-earmold system used in an eyeglass type aid does not offer much possibility of acoustic control."2 This last statement, however, is no longer true since the develOpment of the CROS (Contralateral Routing of Sig— nals) hearing aid by Harford and Barry.3 Lybarger,4 in a lIbid., p. 11. 21bid. 3Earl Harford and Joseph Barry, "A Rehabilitative Ap- proach to the Problem of Unilateral Hearing Impairment: The Contralateral Routing of Signals (CROS)," Journal Of Speech and Hearing Disorders, 50 (1965). PP. 121-158. 4S. F. Lybarger, "Earmold Acoustics," Audecibel, Winter, 1967. 52 later publication in which he reiterated the basic principles of earmold acoustics, recognized the CROS as a means of acoustically modifying the amplified signal from an ear-level hearing aid while reducing the problem of feedback. With this particular type Of hearing aid the microphone and receiver are mounted on Opposite sides Of the head, thus providing effective isolation against feedback. This hearing aid was primarily aimed at helping persons with unilateral hearing losses by electrically transferring a signal from a micrOphone mounted on the impaired ear to a receiver mounted on the good ear. From the receiver the acoustic signal is carried to an Open earpiece by means of a polyethylene tube. The Open earpiece is necessary, since the normal or near normal ear has to be left unoccluded to allow reception Of sound on that side without the attenuation of an earpiece. The full import of the CROS hearing aid for helping other than unilateral hearing losses and the use of the Open ear- piece were not realized by Harford and Barry at that time. Concerning the Open earpiece, they reported: "The hearing aids used in this study offered a relatively flat frequency reSponse as reported by the manufacturer. However, it Should be stressed that the polyethylene tubing terminated in an Open ear canal undoubtedly altered the reported frequency ill reSponse to some degree. It had not occurred to them that lIbid., pp. 129-150. 55 the Open earpiece might be effectively used with certain types of hearing loss to improve the Speech reception threshold, discrimination score, or quality Of the acoustic signal by not disrupting the natural resonance and impedance characteristics of the ear. Although many claims have been made as to the effective- ness of modified earpieces, few have been subjected to ex— perimentation. During recent years a few studies have been carried out, but these have been primarily clinical in nature with many variables uncontrolled and yielding con- flicting results. However, a review Of these studies will shed some light on the effectiveness of modified earpieces. In 1962, Lewis and Plotkin1 reported on a study concern— ing the effects Of a vented earmold on Speech discrimination scores and tolerance for amplification with a group of 15 subjects with high frequency hearing losses. Each of the subjects had normal hearing out to 500 or 1000 Hertz with a precipitous drOp in the higher frequencies. All subjects were tested with the same conventional, body-type hearing aid with an HAIC frequency range of 550-5500 Hertz and an HAIC average gain of 65 dB. Each subject was tested with a standard and a vented earmold, and the resulting Speech lErnest Lewis and William H. Plotkin, "The Role of the Acoustic Coupler in Hearing Aid Fitting," Unpublished paper presented at the Annual Convention of the American Speech and Hearing Association, Chicago, Illinois, 1962. 54 reception thresholds and Speech discrimination scores were compared. Results of their investigation showed that Speech re- ception thresholds under the two aided conditions were not significantly different, but PB scores did differ signifi- cantly. For the entire group the mean unaided SRT was 24 dB and the mean PB score was 65 percent. Utilizing the con- ventional earmold, the mean SRT was 5 dB, a gain of 19 decibels, but there was a loss of three percentage points in the mean PB score of 62 percent. With the vented earmold there was an 18 decibel gain in SRT, and a Speech discrimi- nation score increase of 10 percent over the mean unaided PB score. For the purpose of further data analysis, the subjects in the study were divided into two groups: those with PB scores Of 70 percent and better, and another grOUp with PB scores below 70 percent. The mean unaided PB score was 77 percent for the sub-group of 70 percent or better. With the conventional earmold this group eXperienced a 15 percent loss in discrimination from the unaided score, whereas with the vented earmold there was only a 2 percent loss when compared to the unaided score. The mean unaided PB score was 56 percent for the sub- group with poorer than 70 percent discrimination. This group experienced a four percent gain in discrimination with the conventional earmold and a 19 percent gain utilizing the 55 vented earmold compared to the unaided mean PB score. Thus, it would appear that peOple with poorer than 70 percent discrimination received the most benefit from vented ear- molds. Also, the subjects were able to tolerate greater levels of amplification with the vented earmolds. McClellan1 compared the discrimination scores of five subjects utilizing a conventional earmold and the Zenith acoustic modifier in a background Of noise. All subjects had normal hearing for frequencies lower than 1000 Hertz with a precipitous drOp at 2000 Hertz, sound-field Speech reception thresholds of less (better) than 10 dB, and Speech discrimination scores of 82 percent or better. The loss at 2000 Hertz had to be 55 dB or more in both ears. The mean unaided discrimination score in quiet for the entire group was 87.6 percent. However, in a background of Speech noise (+10 dB S/N) which does not interfere with the discrimination Of normal listeners, they Obtained a mean unaided PB score of 70.8 percent. This is a decrease in discrimination of 16.8 percent from the quiet situation. All subjects were tested with a conventional earmold and an acoustic modifier COUpled to the same moderate-gain, ear-SUSpended hearing aid in the background of noise. The results showed that a mean discrimination score Of 70.0 J‘Max E. McClellan, "Aided Speech Discrimination in Noise with Vented and Unvented Earmolds," Journal of Audi- torngesearch, 7 (1967), pp. 95-99. 56 percent was obtained with the conventional earpiece indi- cating no improvement. A mean discrimination score of 86 percent was Obtained with the vented earmold in noise. This was a gain of 15.2 percent over the mean unaided dis— crimination score in noise. The author concluded: "This shows that with the vented earmold subjects could achieve a discrimination score in noise equal to that obtained un- aided in quiet: thus the vented earmold essentially overcomes the discrimination loss caused by the background noise."1 In a clinical study, Dodds and Harford2 compared the discrimination ability of 55 subjects with high-frequency, precipitous, sensorineural hearing losses employing a con- ventional earmold, the Zenith acoustic modifier, and an Open earpiece. Sixteen different hearing aids were used, includ- ing the CROS aid employed with an Open earpiece. The same hearing aid was used when comparing the test results between the conventional and vented earmolds for a given subject. Not all subjects were tested with the three types of ear— pieces, and sub-groups were formed depending on the type of earmolds used during testing. Persons tested with more than two kinds Of earmolds were included in more than one group. 11515., p. 97. 2Elizabeth Dodds and Earl Harford, "Modified Earpieces and CROS for High Frequency Hearing Losses," Journal of Speech and Hearing Research, 11 (1968), pp. 204-218. 57 Group I consisted of 18 cases tested with both conven- tional and vented earmolds. For this group the statistical analysis revealed no significant differences in discrimi- nation scores between the vented and conventional earmolds. The scores were 78.1 percent for the conventional and 77.8 percent for the vented mold. Group II consisted of 14 cases evaluated with both con— ventional and Open molds. At the .01 level Of confidence, statistical analysis revealed a significant improvement in discrimination using an Open earmold. These scores were 71.4 percent with the conventional and 81.4 percent with the Open earmold, an improvement of 10 percentage points. Group III consisted of 12 cases tested with both vented and open earpieces. At the .05 level of confidence, statis- tical analysis revealed that performance with the Open ear- piece was Significantly better. Mean PB score with the vented mold was 74.7 percent, and a mean Of 79.8 percent with the Open mold was achieved, an improvement of 5.1 percentage points. Another interesting finding in this study was the fact that the subjects were almost unanimous in their preference of the Open earpiece because Speech sounded much more "natural" to them. In an unpublished paper, Harrisonl has reported on her 1Anne Harrison, "Some Clinical Uses Of the Modified Ear Insert in Supplying More Acceptable Amplification for Select- ed Sensorineural Hearing Impairments," Unpublished paper pre- sented at the Annual Convention of the American Speech and Hearing Association, Chicago, Illinois, 1967. 58 utilization of modified earpieces in the clinical situation. She found that for individuals with high-frequency hearing impairments above 500 or 1000 Hertz, the conventional ear- mold Often did not prove satisfactory even when coupled with a hearing aid having a high-frequency emphasis. With the use of vented earmolds, however, there were not as many complaints of irritation; and people who had previously re- jected the use of a hearing aid were able to benefit from amplification. Harrison also found that a vented earmold could Often be utilized by persons with more extensive cochlear involve- ment with the low frequencies being affected as well as the high. She stated that "We have found that the modified mold has value for a variety of cochlear impairments along the "1 She cited three clinical entire hearing loss continuum. cases indicating the successful use of modified earmolds. The first case, a thirteen-year-Old girl with a high fre- quency, bilateral hearing loss involving a precipitous drOp at 2000 Hertz, had previously been informed that she could not use a hearing aid. However, with a prOperly vented ear- mold coupled to an ear-level aid, the girl accepted ampli- fication and her articulation improved markedly within a period Of three months. The second case was a man 59 years of age with a moder- ately severe, bilateral hearing loss, which had gradually 11bid., p. 5. 59 progressed over a period of 20 to 25 years and was suSpected of being noise induced. He had been unable to adjust to a hearing aid in the past but was going to purchase binaural behind-the—ear aids in another attempt at wearing amplifi- cation. However, he was still dissatisfied with the quality of the amplification. A recommendation was made for him to utilize vented earmolds. AS a result, he reported improve- ment in the quality of amplified Speech and minimal annoyance from background noise and was able to wear the hearing aids in his daily activities. The third case was a woman 54 years Old with a severe high frequency, bilateral hearing loss involving a precipi— tous drOp at 500 Hertz. Using a body type hearing aid with a conventional earmold, the patient was able to derive some benefit from amplification but complained Of "sensations of impact at the eardrum while perceiving both Speech and cer- 1 After further hearing aid tain environmental sounds." evaluations, a vented earmold coupled to a high gain hearing aid with a high frequency emphasis was recommended. This combination appeared to help with gross environmental and Speech discriminations, and overall communication ability was much better with the aid than without it. The effects Of conventional, vented, and Open earmolds on high-frequency hearing losses have been investigated by 1Ibid., p. 5. 40 one or more of the studies mentioned above. Recently, however, some hearing aid dealers and audiologists have been fitting high-frequency hearing losses with CROS aids coupled to the ear by just polyethylene tubing. Schafer1 has been advocating Open ear canal amplification for some time, and during the past four years he and his associates have been using tubing exclusively in conjunction with CROS aids. They report having case files on over 750 peOple who have been using this arrangement, some of them for more than three years. These peOple include unilaterals with varying degrees of hearing sensitivity on the better side and many symmetrical bilateral sensorineural losses. Although polyethylene tubing is being used as a coupler without an earmold, there is no published research indicat- ing its effects on Speech reception thresholds and Speech discrimination scores Of subjects having various types of hearing losses. One purpose of the present research was to Obtain Objective evidence as to how polyethylene tubing, used as a coupler, compares to the conventional, vented, and Open earmolds in its effects on Speech reception threshold and Speech disCrimination. 1Personal communication with Donald W. Schaefer (D. W. Schaefer and Associates, Inc., 25 West Main Street, Madison, Wisconsin), April 16, 1968. 41 Summary The review of the literature has shown how earmolds were developed along with hearing aids from the very early ear trumpets to the modern, transistorized aids. The re- search on auditory phenomena associated with the transmis- sion of an acoustic stimulus to the human ear revealed the principles on which certain types of modified earmolds were based. The early research on modified earpieces arose out of efforts to enhance the amplification of the inefficient carbon amplifier hearing aids. These studies showed that the frequency reSponse Of the acoustic stimulus reaching the tympanic membrane could be modified by changing certain dimensions Of the acoustic COUpler. Also, the amount of amplification is modified. It was pointed out that during recent years there has been very little research concerned with modified earpieces. This may, in part, be due to the fact that the use of the CROS aid with other than unilateral hearing losses is, at the time of this writing, a relatively new concept. This type of amplification now makes possible the use of various types of acoustic COUplers which could not be utilized in the past due to the problem of acoustic feedback. Further, the few existing studies either need to be eXpanded and the results verified or subjected to more rigid research controls. A closer study of the more recent research reveals various 42 methodological problems. The study by Lewis and Plotkinl did employ fairly good controls, but only two types Of acoustic couplers were utilized and only 15 sensorineural subjects with one type Of pure-tone, audiometric configura- tion were investigated. The present study utilized four different acoustic couplers and investigated three types Of pure-tone audiometric configurations. These included con- ductives, sensorineurals with a precipitous drOp, and sen- sorineurals with a gradually sloping configuration. Also, the Lewis and Plotkin study revealed a significant improve- ment in Speech discrimination scores when utilizing the acoustic modifier as Opposed to a conventional earmold, whereas the study by Dodds and Harforda was a descriptive study of clinical cases and was lacking in experimental de- sign controls for this reason. Another shortcoming is that not all the cases were tested with the three kinds of ear- pieces: conventional, acoustic modifier, and Open earpiece. It is possible that those cases not showing a significant gain in the PB score utilizing an acoustic modifier might have done so with an Open earpiece. By the same token, those cases who Showed an improved PB score with the Open earpiece may also have shown an improved PB score with the acoustic modifier had one been evaluated. The study by 1Lewis and Plotkin, "Role of Acoustic Coupler,“ p. 9. 2Dodds and Harford, "Modified Earpieces," p. 12. 45 Dodds and Harford1 does not give any data which compare all three types Of acoustic couplers on the same group of sub— jects. However, the study is an important clinical investi— gation and clearly reveals the trend for Speech discrimina- tion scores to improve when an acoustic coupler is employed which more closely approximates a natural listening Situa— tion by leaving the ear canal unoccluded. Their study needs to be subjected to a more standardized procedure, and all of the various types of acoustic couplers should be employed with the same group of subjects. The present investigation was designed with more rigid controls, and all subjects were tested under exactly the same eXperimental conditions. Further, Dodds and Harford used 16 different hearing aids. The present study utilized a single CROS type hearing aid, thus eliminating any differential effects caused by the use of many types of hearing aids. Harrison's2 report of a clinical application of modified earpieces to various types of pure-tone audiometric config— urations contained only the subjective impressions Of the clinician and the subject as to the improvement in amplified sound Offered by vented earmolds. The report does not con- tain any Objective measures Of improved Speech reception thresholds or Speech discrimination scores. The present study presents measures of both Speech reception thresholds lIbid., pp. 204-218. 2Harrison, "Some Clinical Uses of the Modified Ear Insert,“ pp. 1-6. 44 and Speech discrimination scores as well as a judgment on the part Of the subject as to the quality Of the amplified sound through the various types Of acoustic couplers. CHAPTER III EXPERIMENTAL PROCEDURES Subjects Three groups of hard-Of-hearing adults served as sub- jects in this study. Group I consisted of ten persons having a conductive type hearing impairment. This group was com- posed of six females and four males ranging in age from 17 to 60 years with a mean age of 57.1 years. Each case ful- filled the criteria of having bone conduction thresholds within the normal range (no greater than 25 dB ISO-1964 Standards) and air conduction thresholds which showed an air- bone gap of at least 20 dB for the test frequencies 500, 1000 and 2000 Hertz. Figure 1 shows the mean air and bone conduction thresholds for this group. Group II consisted of ten persons having a sensorineural type hearing impairment with an audiometric configuration showing a precipitous high frequency drop. This group was composed Of ten males ranging in age from 40 to 67 years with a mean age of 51.5 years. Each case fulfilled the cri- teria of having normal hearing (25 dB or better ISO-1964 Standards) for the low frequencies with a precipitous drop of at least 20 dB for the first octave beginning at 500 or 1000 45 Hearing Level in dB (180 1964) 46 Figure 1.——Audiogram Showing mean air and bone conduction thresholds for Group I (Conductive Impairment). N = 10 Frequency in Hertz 125 250 500 1K 2K 5K 4K 8K _A.____ I \\\L f ‘\ ,,/ i L 20 40 X \>\>.< OX \1 i 4 60 80 I 100 Key to Audiogram Ear Right Left Air Conduction O >< Bone Conduction ‘ ' (Masked) 47 Hertz and bone conduction thresholds which interwove with the air conduction thresholds. Figure 2 shows the mean air conduction thresholds for this group. Group III consisted Of ten persons having a sensori- neural type hearing impairment with an audiometric configur- ation Showing a gradually SlOping loss with the low frequen- cies also being affected. This group was composed Of one female and nine males ranging in age from 20 to 62 years with a mean age of 51.2 years. Each case fulfilled the criteria of having a progressively greater hearing loss for higher frequencies at a lepe of 5 to 10 dB per octave and bone conduction thresholds which interwove with the air conduction thresholds. Figure 5 Shows the mean air conduc- tion thresholds for this grOUp. Equipment The following is a list Of the equipment utilized in this investigation: Test Equipment Pure-tone audiometer (Beltone, Model 15C) Speech audiometer (Grason-Stadler, Model 162) LoudSpeaker (Grason-Stadler, Model 162-4) Earphones (Telephonics, Model TDH-59-1OZ) Earphone cushions (Model MX 41/AR) Bone Vibrator (Radioear, Model B70-A) Tape recorder (Ampex, Model 601) Narrow band masking unit (Beltone, Model NB—102) 20 decibel attenuation pad Commercial test room (Industrial Acoustic Company, Inc. 1200 series) Earmolds (Conventional, Acoustic Modifier, Open, and Crimped Polyethylene Tubing) Hearing Level in dB (ISO 1964) 48 Figure 2.--Audiogram showing mean air conduction thresholds for Group II (Sensorineural Impairment with a Precipitous Drop). N = 10 125 250 500 1K 2K 5K 4K 8K 20 Xx 40 Vi‘ \ \\\ 5 6° \ \p x \R\\ \e\ .e—fi-K 80 0 Right ear XLeft ear Hearing Level in dB (ISO 1964) 49 Figure 5.--Audiogram showing mean air conduction thresholds for Group III (Sensorineural Impairment with a Gradually SlOping Loss). N = 10 Frequency in Hertz 125 250 500 l5 21-( 55 4K K 0 20 40 F 60 K <__>< /'\ \ll 80 100 Key 0 Right ear )K Left ear 50 Calibration quipment Sound level meter (Bruel & Kjaer, Type 2205) Octave band filter network (Bruel & Kjaer, Type 1615) Artificial ear (Bruel & Kjaer, Type 4152) Condenser microphone (Bruel & Kjaer, Type 4152, used in conjunction with the artificial ear) Condenser microphone (Bruel & Kjaer, Type 4151, used for sound-field measurements) Artificial mastoid (Beltone, Model M5A) Volt meter contained as an integral part of the audio frequency Spectrometer (Bruel & Kjaer, Type 2112) Pistonphone (Bruel & Kjaer, Type 4220) Equipment Used for Measuring Frequency ReSponse and Distortion Characteristics of Hearing Aid Hearing aid test box (Bruel & Kjaer, Type 4212) Frequency analyzer (Bruel & Kjaer, Type 2107) Audio frequency Spectrometer (Bruel & Kjaer, Type 2112) Sine-Random generator (Bruel& Kjaer, Type 1024) Condenser microphone (Bruel & Kjaer, Type 4152) Level Recorder (Bruel & Kjaer, Type 2505) For the pure-tone testing necessary in the experiment, a commercially available pure-tone audiometer (Beltone, model 15C) was used to drive TDH—59—1OZ transducers housed in MX 41/AR biscuit-type cushions. For the necessary Speech testing, a commercially avail- able Speech audiometer (Grason-Stadler, model 162) was used to amplify and attenuate the electrical output of the tape recorder (Ampex, model 601) used to present the tape re— corded tests described later under test materials. For a given test condition, the output of the Speech audiometer drove one Of two transducers: (1) a TDH-59-1OZ earphone housed in an MX 41/AR cushion, or (2) a loudSpeaker (Grason- Stadler, model 162-4) furnished as an integral component of the Speech audiometer. 51 The Speech audiometer used in this research was cali- brated so that audiometric zero is defined as being 20 dB above 0.0002 dyne/cma. Instead Of using the usual 1000 Hertz tone for calibration, "Speech Noise" was used for calibration in the sound-field according to the procedure described by 1 Their rationale for using Tillman, Johnson, and Olsen. Speech noise in lieu Of a 1000 Hertz signal was that the Spectral configuration of the noise closely approximates the Spectrum Of continuous Speech produced by male Speakers. This Spectrum was drawn as the average of two curves reported in graphic form by Licklider and Miller.2 A description Of the procedure follows: In order to calibrate the TDH-59 earphone, it is coupled to the condenser micrOphone (Bruel & Kjaer, Type 4152) of the sound level meter (Bruel & Kjaer, Type 2205) with its associated octave band filter network (Bruel & Kjaer, Type 1615) by means of a standard 6-cc coupler (Bruel and Kjaer, Type 4152). The level of the noise signal at a given attenuator setting is adjusted until it produces a deflection to zero on the Speech audio- meter VU meter. The resulting acoustic output of the system is measured, and this value is accepted as the intensity Of the Spondee words at the same attenuator setting under the 1Tillman, Johnson, and Olsen, "Earphone Versus Sound- Field," pp. 128-129. 2J. C. R. Licklider, "The Perception of Speech," in Handbook of Experimental Psychology, S. S. Stevens, Ed. (New York: John Wiley & Sons, Inc., 1951), p. 1042. 52 condition in which the peaks of the words also produced a deflection to zero on the VU meter of the Speech audiometer. For example, with the Speech audiometer attenuator set at 60 dB, the output of the artificial ear would be 80 dB SPL re 0.0002 microbar. For calibration of the loudSpeaker, the condenser micro- phone (Bruel & Kjaer, Type 4151) was placed four feet from the face Of the loudspeaker at a height of 42 inches. The condenser microphone was oriented so that its diaphragm was perpendicular to the floor and ceiling of the rest chamber at a zero degree angle of incidence from the loudSpeaker. The intensity of the Speech Spectrum noise generated by the Speech audiometer at a given attenuator setting was then recorded. All measurements were made without the presence of an observer in the field. However, the location of the condenser microphone was approximately where the center of the subject's head would be when a subject was in the test chamber. A pistonphone (Bruel & Kjaer, Type 4220) was used to set the meter needle of the audio frequency Spectrometer (Bruel & Kjaer, Model 2112) from which the intensity of the sound-field was read directly in decibels re 0.0002 microbar. Measurements of the overall SPL of the Speech noise were made on all days that subjects were tested, and the readings were found to be within plus or minus one decibel of 20 dB re 0.0002 microbar throughout the eXperimentation. Attenu- ator linearity was also checked and found to be stable throughout the study. 55 The acoustic Spectrum of broad band white noise through the loudSpeaker at an 80 dB SPL re 0.0002 microbar is shown in Figure 4. From this figure it can be seen that the loud- Speaker has a fairly flat frequency reSponse from 500 through 8000 Hertz. This reSponse remained constant from the begin- ning to the end of the study. Calibration Of the pure tone air and bone conduction systems was also checked on all days on which subjects were tested. The Bruel and Kjaer sound level meter and its associated filter network and the artificial ear were used for calibration Of the air conduction system. The artificial mastoid was used to calibrate the bone conduction system. Attenuator linearity was checked periodically in all systems and any necessary corrections were applied to the data. The 20 dB attenuation pad was checked in our laboratory prior to beginning this study and was accurate plus or minus one dB. In addition to the equipment listed above, the Radioear, model 950 CROS type hearing aid was employed. Utilization of the CROS principle was necessary in order to eliminate acoustic feedback when testing with the Open earpiece and crimped tubing. The Radioear, model 950 has interchangeable bows; thus, the same hearing aid could be used with all sub- jects for all conditions of acoustic coupling.‘ All that was necessary was to place the bow containing the pick-up micro— phone on the ear contralateral to the one which was to receive the amplification. 54 Om owl Ofill all OHVI on... .. on- ONII 1.0NI OT] 1 o? o _ L o Mm Kw XN Kfi 00m 0mm mNH uuuom CH hocmsqwum .enmoa Hoooz .uoxmmmmosoq unaccumicommuo one cmsoucu omHoc ouH£3 pawn Umoun m mo Eouuowmm oaumsoom oceil.w musmflm iueuodmoo issBUOJQS e: gp u: AntsuequI eAtqueu 55 The frequency reSponse Of the Radioear, model 950 CROS type hearing aid was Obtained in the following manner: The hearing aid was placed in a hearing aid test chamber (Bruel & Kjaer, Type 4212) consisting of an external artificial ear, a regulating microphone, and built—in loudSpeaker. The regulating microphone was connected to the micrOphone ampli— fier portion of the frequency analyzer (Bruel & Kjaer, Type 2107), which amplified the signal and applied it to the com- pressor input of the Sine—random generator (Bruel & Kjaer, Type 1024). The generator supplied a Sine-wave signal to the loudSpeaker in the chamber and a condenser micrOphone (Bruel & Kjaer, Type 4152) was connected to the 2 cc coupler of the artificial ear. The output of the hearing aid was then connected to the 2 cc coupler. The micrOphone in turn was connected to the micrOphone amplifier portion of the audio frequency Spectrometer (Bruel & Kjaer, Type 2112), and the amplified voltage was led to the input of the level recorder (Bruel & Kjaer, Type 2505) which automatically re- corded the frequency reSponse of the hearing aid. The level of the input signal to the hearing aid was 60 dB re 0.0002 microbar. These measurements were made in accordance with the procedures Specified by the American Standards Associ- ation.l 1"American Standard Methods for Measurement of Electro- acoustical Characteristics of Hearing Aids," American Standards Association, Incorpgrated, NO. 85-5-1960 (1960), p. 12. 56 The frequency reSponse characteristics Of the hearing aid are shown in Figure 5. In addition, the following characteristics of the hear- ing aid was determined by the HAIC method.l Maximum Gain. . . . . . . . . . 48 dB Maximum Output. . . . . . . . .124.1 dB Frequency Range . . . . . . . .500-5200 Hz A comparison Of the above measurements with the manu- facturer's Specifications indicated a frequency reSponse curve nearly identical to that Specified, except for slightly more gain between 2000 and 4000 Hertz. Other Specifications for this hearing aid model by the HAIC method of computation listed a maximum gain of 52 dB compared to the 48 dB actually measured. The Specified maximum output of 124 dB was identi— cal to that measured, and the Specified frequency range of 460-4800 Hertz was comparable to the 500-5200 Hertz measured. The harmonic distortion of the Model 950 was measured in the following manner. The output of the aid was led to the amplifier input of the frequency analyzer (Bruel & Kjaer, Type 2107). The distortion factor was then measured directly by switching the analyzer to "frequency rejection." In this manner, the fundamental was rejected and the remaining total harmonic distortion was read directly in percent from the instrument meter with the switch set in the R.M.S. position. 1S. F. Lybarger, "A New HAIC Standard Method of Express- ing Hearing Aid Performance," The Hearing Dealer, February 1961. pp. 16-17. 57 women SH wocosqoum x0fi>m¢mm vfluhmdmm 00d Oh Om Om ooa OHH ONH .nmppppse mooo.o mp mp om up Hmcmsm uzmcfl cm rugs esp mcflummr mesh momo omm Hwooz .ummoflomm ecu mo wmcommmu >ocosqmumii.m musmam Jeqoxorm gooo°o a: gp ur qndqno Tenor 58 Table I gives the harmonic distortion as determined by the American Standards Association (A.S.A.) method of computa- . 1 tion. Table I.--Percent Harmonic Distortion of A.S.A. Method. Input 75 dB SPL re 0.0002 microbar. SPL* at Frequency_in Hertz Coupler 500 700 900 Average 80 4.2 5.4 2.2 5.5 90 9.4 2.4 1.2 4.5 100 24.0 5.2 1.0 9.4 Saturation 40.0 5.4 2.0 15.8 *In dB re 0.0002 microbar. Test Environment The test room and all audiometric equipment were located in the basement Of the Michigan State University Auditorium Building. The loudSpeaker and earphones used in the testing of subjects were Situated in the test room (IAC, 1200 Series); and during all conditions Of this eXperiment, the subjects were seated in this room. Previous measurements of the ambient noise in the sound— treated room using the sound level meter on the C scale have shown the noise tO be 42 decibels SPL re 0.0002 microbar. 1"American Standard Methods for Measurement of Electro- acoustical Characteristics Of Hearing Aids," p. 15. 59 Also, an octave band analysis of the ambient noise level had indicated that the greatest amount of ambient noise (40 dB average) is found in the octave bands below 100 Hertz. For the frequencies from 100 to 8000 Hertz, the ambient noise level averages 14 dB. These levels were sufficiently low so as not to interfere with the tests administered. The pure-tone audiometer, Speech audiometer, tape record- er, and narrow band masking generator were situated in an adjoining control room which communicates with the sound- treated room by means of a window and a two-way electronic communications system, which is an integral part of the Speech audiometer. Test Materials The unaided and aided Speech reception thresholds were obtained with tape-recorded Spondee word lists. These were the same words used in CID Auditory Test W-1 and consist of six scramblings of a single list of 56 Spondaic words.l These words were recorded on magnetic tape by a male talker with a General American dialect who monitored the level Of his vocal productions on a VU-meter so that the two syllable peaks of each word drove the meter to the same deflection plus or minus one decibel as that produced by a 1000 Hertz calibration tone on the same tape. lIra J. Hirsh, Hallowell Davis, S. Richard Silverman, Elizabeth G. Reynolds, Elizabeth Eldert, and Robert W. Benson, "Development of Materials for Speech Audiometry," Journal of Speech and Hearing Disorders, 17 (1952), pp. 521-557. 60 The unaided and aided Speech discrimination scores were Obtained with NU Auditory Test No. 6.1 This test con- sists Of four lists of 50 monosyllabic words patterned after the CNC lists develOped by Lehiste and Peterson2 in 1959 and revised by them in 1962.3 These lists were recorded on mag— netic tape by the same male talker mentioned above who monitored his vocal output by means Of a VU—meter. The carrier phrase, "You will say" preceded each test word. The last word of the carrier phrase was monitored and the CNC word was said naturally. Four additional lists, necessary for the eXperimental conditions in the present study, were made from scramblings of the original four. In other words, two forms (A and B) were used. Form B was constructed by re-recording Form A and cutting and Splicing the tape so that the words appeared in a different order than in Form A. The four CNC lists were standardized earlier by Rintel- mann and Jetty4 on ten young adult subjects with normal lTom W. Tillman and Raymond Carhart, "An EXpanded Test for Speech Discrimination Utilizing CNC Monosyllabic Words (NU Auditory Test No. 6)," U. S. School of AerOSppce Medicine—- Technical Research, 66-55, 1-12, June, 1966. 21. Lehiste and G. E. Peterson, "Linguistic Considera- tions in the Study of Speech Intelligibility," Journal of the Acoustical Society of America, 51 (1959), 280-286. 3G. E. Peterson and I. Lehiste, "Revised CNC Lists for Auditory Tests," Journal of Speech and Hearing Disorders, 27 (1962), pp. 62-70. 4William F. Rintelmann and Albert J. Jetty, "Reliability of Speech Discrimination Testing Using CNC Monosyllabic Words," Unpublished Study, Michigan State University, 1968. 61 hearing following the procedures outlined by Tillman and Carhart.1 The results were comparable to those Obtained by Tillman and Carhart and the lists were equivalent plus or minus 4 percent at a 24 dB sensation level. The results of the standardization at a 24 dB sensation level are given in Appendix A. Test Procedures The following tests were administered to each subject: Pure tone air and bone conduction, both ears monaurally with routine masking by bone conduction. Unaided Speech reception threshold, both ears monaurally, and in a sound-field. Unaided Speech discrimination, both ears monaurally, and in a sound-field. Aided Speech reception threshold and Speech discrimi- nation in one ear under each of the following conditions: (1) Conventional earmold (2) Acoustic modifier (5) Open Earpiece (4) Crimped tubing The unaided tests were administered in the order listed above. The aided tests were presented according to a pre- determined rotation procedure. Masking was used routinely during bone conduction test— ing. The masking agent was supplied by a narrow band masking generator (Beltone, Model NB—102). Previous analysis Of this masking generator had indicated that the band widths, 1Tillman and Carhart, "An Expanded Test for Speech Dis- crimination," pp. 4-7. 62 determined at the level 5 dB down from the peak intensity, were all greater than the critical band widths defined by Fletcher.1'2 The effective masking for a zero dB hearing level was determined at each band by applying the critical band data of Fletcher in the manner described by Sanders and 3 Rintelmann. Bone conduction thresholds were determined by the Hood Technique.4 During this preliminary testing, the subjects were seated in the sound-treated room, and sound-field measure— ments were made with the subject seated so that the midline Of his forehead was four feet from the face of the loudSpeaker at a zero degree azimuth. One loudSpeaker was used for all unaided and aided sound-field measurements. A schematic diagram of the test environment is shown in Figure 6. All pure tone air and bone conduction thresholds were determined by the Revised Hughson-Westlake Technique as de- scribed by Carhart and Jerger.S lHarvey Fletcher and W. A. Munson, "Relation Between Loudness and Masking," Journal of the Acoustical Society Of America, 9 (1957), pp. 1-10. 2Harvey Fletcher, "Auditory Patterns," Review of Modern Physics. 12 (1940). pp. 47-65. 3Jay W. Sanders and William F. Rintelmann, “Masking in Audiometry," Archives of Otolapyngology, 80 (1964), pp. 541-556. 4J. D. Hood, "Principles and Practice Of Bone Conduction Audiometry," Laryngoscope, 70 (1960), pp. 1211-1228. 5Raymond Carhart and James F. Jerger, "Preferred Method of Clinical Determination of Pure-Tone Thresholds," Journal of Speech and Hearinngisorders, 24 (1959), pp. 550-545. 65 Figure 6.--Schematic diagram of the test environment showing placement of equipment and location of the sub- ject during testing. ‘ Subject TEST ROOM J t. .774 f Bone Get Oscillator Earphogfs Af‘ MicrOphoneEarpE%fes S k ‘ pea er 0 1» o o R L B R L .1 . . . -yj: r ZOdB —J7 I ,Pad |->- LB I Beltone 1 Be tone NB. {+JL_L 15C Grason-Stadler Masking R 162 _ Unit +_' Power Ampex CONTROL ROOM ““9 ° T23: - _ Recordw er Speaker for moni- toring subject's reSponses 64 The unaided and aided Speech reception thresholds were obtained with the recorded Speech materials described earlier. Threshold was determined by the method described by Tillman and Carhart.l In this method the following pro- cedures are followed: Initially two test words are presented at a level 10 to 20 dB above the estimated SRT. The intensity of the signal is then attenuated by 2 dB and two more words are presented. The initial presentation level is selected so that the sub- ject correctly repeats a minimum of five of the first six test words. The procedure of attenuating in 2 dB steps and presenting two words at each step is continued until the subject either fails to reSpond or reSponds incorrectly to six consecutive test words. Threshold is then computed by subtracting the number of words correctly repeated from the intensity of the Signal at the starting level and adding one decibel to compensate for the fact that the 50 percent cri- terion is not fully met via this procedure. Since the attenuator of the Speech audiometer is calibrated in 2-dB steps, in the case of an Odd-integer Spondee threshold, the reference intensity used was the level 1-dB higher than the actual SRT. It has been demonstrated that the threshold for Spondee words may vary considerably over time as the test items are lTillman and Carhart, "An Expanded Test for Speech Discrimination," pp. 5-6. 65 repeated and is apparently a function of the person's familiarity with the test vocabulary.1 All of the subjects used in this study had received previous Speech audiometric tests. In this reSpect, they all had about the same familiarity with the Spondee words. However, the intervening time between their last tests and this eXperiment varied, so all were exposed to the Spondees immediately prior to unaided threshold measurement. This was accomplished by the examiner reading the words to each subject prior to testing. The following instructions were given: Before actually beginning testing, I am going to read you a list of 56 two-syllable words. They will be presented at a level so that you should be able to hear them comfortably. Please repeat each word aloud. These words are the same ones that will be used during the testing, although they will be in different order. Please pay careful attention to the words SO that you will become familiar with them. Upon completion of the above, the subject was told that the actual testing would now begin and the following instruc- tions were given: You will now hear a man's voice saying the same two-syllable words you have just heard. The words will begin at a loudness level at which you will be able to hear them easily, but they will eventually get very faint. Your task is to repeat as many of the words as you possibly can. Even though the word may be very faint if you think you know what it is, repeat it aloud. Are there any questions? The unaided and aided Speech discrimination scores were Obtained with the recorded monosyllabic words described lTom W. Tillman and James F. Jerger, "Some Factors Affecting the Spondee Threshold in Normal-Hearing Subjects," Journal of Speech and Hearing Research, 2 (1959), pp. 141-146. 66 under "Test Materials." The words were presented at a 26 dB sensation level (SL) re the SRT for all test conditions. The following instructions were given to the subjects: During the next test you will hear a man's voice saying one-syllable words. These are common words, which will be very familiar to you. The words will be presented at a sufficient loudness level, so that you will be able to hear them easily. They will not become fainter as in the previous test. Your task is to repeat aloud as many of the words as you pos- sibly can. If you think you know what a word is, but are not quite sure, go ahead and repeat what you think it might be. Are there any questions? Upon completion of the Speech tests under earphones, they were repeated in the sound-field under the unaided and aided conditions with the various types of acoustic couplers. The subject was seated facing the loudSpeaker at a distance of four feet. Only one ear of each subject was tested in the sound field under the unaided and aided conditions. This was accom- plished by occluding the non-test ear with a Flent ear plug. The details of this procedure are described below. The sub- jects were chosen so that they had a bilateral hearing loss which was fairly symmetrical. The choice Of which ear to test in the Sound-field unaided and aided with the acoustic couplers was based on the following criteria applied to the results Obtained under earphones: (1) The ear yielding the better discrimination score was selected as the test ear for subjects with a discrimination score poorer than 90 percent. A score differing by 4 percent or greater between the ears was considered significantly different. When the discrimination score yielded by the better ear was 90 percent or greater, the poorer ear was chosen for amplification irreSpective of SRT. 67 (2) If the Speech discrimination score was within 4 percent between ears, then the ear with the better SRT was chosen. A 4 dB or greater difference in SRTS was considered as significant. (5) If the discrimination scores and SRTS were equal, then the ear selected was the one which helped balance the number of right and left ears chosen for amplification. Speech reception thresholds and Speech discrimination scores for the sound-field unaided and aided conditions were obtained using the same procedures as outlined previously; however, in order to rule out participation Of the contra- lateral ear, the threshold was raised by occluding the ear canal with a wax impregnated ear plug (Flent). Since the subjects had fairly symmetrical hearing losses, the attenua- tion of the Flent coupled with the already existing hearing loss provided the isolation necessary to rule out participa- tion of the contralateral ear in the test Situation. Previous to their utilization in this study, the amount Of attenuation provided by these ear plugs was investigated using three subjects with normal hearing. Monaural unoc- cluded pure-tone air conduction thresholds and Speech reception thresholds were obtained, and then these same thresholds were obtained with the ears occluded by the ear plugs. The results are given in Appendix B. The attenu- ation provided for Speech was approximately 55 dB. The gain of the hearing aid remained at the same set- ting during all of the experimental conditions for all sub- jects. This was necessary in order to rule out the influence 68 of different gain settings on the Obtained results. A limit- ing factor to the amount of gain that could be utilized was the acoustic feedback produced under the condition of maxi— mum leakage. When the Radioear, Model 950 with crimped tubing is worn, acoustic feedback can be completely elimi- nated by turning the gain control down somewhat from its maximum position. The setting which eliminated acoustic feedback was determined by the investigator placing the aid on himself and turning the gain control down until the feed- back was eliminated. The gain control was then sealed by tape at this setting and was not disturbed throughout the investigation. The HAIC maximum gain determined at this setting was 55.1 dB. The equipment used and the procedure for measuring gain were described earlier in this chapter. In order to prevent acoustic leakage, which could result from using a stock earpiece, the conventional earmold was sealed as well as possible to the ear canal by Glastrip,l a commercially available compound. Glastrip is provided in powder form but becomes a pliable, clay-like substance when mixed with water. Another possible source of leakage was where the nub at the end of the polyethylene tubing snaps into the earmold, so this junction was also sealed with the same compound. In order to eliminate test order effects, the four ac- coustic couplers were tested according to a predetermined lGlastrip is available from Coe Laboratories, Inc., 6055 Wentworth Avenue, Chicago, Illinois. 69 rotation procedure. The Spondee lists and CNC lists were also rotated so that they were not always presented in the same order or with the same coupling condition. Upon completion of testing with the first two acoustic couplers, the subject was asked which he preferred, number one or number two, as far as the subjective quality of the amplification was concerned. This preference was noted. The third acoustic coupler was then tested and the subject ” was asked to state his preference between this one and the previously chosen one. Through this process of elimination, the subject was able to arrive at a decision as to which coupler he preferred. The first coupling condition under which a subject was tested was repeated. This yielded test and retest scores for both Speech reception thresholds and Speech discrimi- nation scores for each subject under one of the coupling conditions. The correlation between these two scores was then used as an estimate Of the reliability Of the measures. CHAPTER IV RESULTS AND DISCUSSION This chapter is divided into five sections. The first section presents a descriptive summary of the results ob- tained for the various experimental conditions. The second section contains a discussion of the reliability of the ob- tained measures. The next two sections contain the statis- tical analysis of the Speech reception thresholds and Speech discrimination scores. The final section is a discussion Of the clinical Significance Of the Obtained results. Descriptive Summary Of Results The means, medians, standard deviations, and ranges of the Speech reception thresholds (SRTS) recorded for the three groups aided utilizing the four acoustic couplers and unaided in the sound-field are shown in Table II. From the table it can be Observed that the threshold variability of the three groups differs markedly. This how- ever, is not unexpected, since from the standpoint of hearing sensitivity, these three groups are not very homogeneous. However, it Should be noted that for all three groups the variability about the mean threshold remains relatively constant from one coupling condition to another. 70 71 _l .oumu oeuumEOHoom commmm on mo CH * mm mm mm Hm ON mmcmm ©.w m.m o.m >.m N.m Om m.mm m.mm 0.0m m.em m.>¢ p02 m.om ®.mm H.mm m.mm >.m¢ z Ammon Hmsomumv amuswcfluomcmm .3 me we oa em mmcmm ¢.w ¢.m m.¢ o.m H.@ mm O.Mfi o.ma m.ea m.efi o.eN U02 >.ea m.NH m.efi m.ea m.mm 2 Amsouflmflomumv amusmcauomcmm mm we mm am on 095m N.@ N.m >.w >.m O.m Om o.md m.ad o.ma m.m m.mm on: H.ed >.ma m.ma o.ad 0.0m z o>Huosocoo OSHQSB momflmumm “OHMHUOS Hmcofluco>coo mm OOQEHHU some oaumsood OOOHOCD ZOHBHQZOU OZHZMBmHA ADOMO mcofluflocoo Oceanumflq owned usom pom confines 0:0 woos: muomflosm mo mmsouo mouse Mow pecamuno *mOHOmeucs coeumwomm sommmm mo mwmcmm pom .AQmV mcoHDOH>mQ pumOCMDm .flowzv mcmapwz .AZV mcmmzii.HH canoe 72 The means, medians, standard deviations, and ranges of the Speech discrimination scores recorded for the three groups aided utilizing the four acoustic couplers and unaided in the sound-field are presented in Table III. Examination of Table III reveals that the Speech dis- crimination Of the three groups is substantially different. Again, this is not an uneXpected outcome, since the three groups are not very homogeneous with reSpect to their ability m to understand Speech. However, note that for all three groups the variability about the mean Speech discrimination score remains relatively constant from one listening condition to another. With reSpect to within grOUp variability, the group with the conductive hearing impairment is much more homogeneous than the other two groups. The group having a sensorineural hearing impairment with a precipitous drOp shows more vari- ability, with reSpect to Speech discrimination, in the un- aided condition than in the aided utilizing the four acoustic couplers. In other words, the scores become much more homogeneous under the aided conditions. This is not true for the other two groups. Reliability Before considering the differences among the various means Obtained in this experiment, it is necessary to examine the reliability of the measures on which these comparisons 75 .uo0uuoo uc0ou0m SH * mm em om om ow 0mcmm m.m ¢.oa w.m m.m m.oa Om mm me mm mm om O02 m.em 0.0m o.mm o.m> m.m> E A0mon Hmopmuwv amus0cflnomc0m ea ea md 56 mm 0mcmm «.9 >.e m.m N.m >.m Om mm mm mm me oh O02 o.mm 0.5m N.>m m.m> «.mw 2 Amsouflmflo0umv amus0cfluomc0m oa oa m oa m omcmm m.m m.m m.N m.m ¢.m mm mm em mm mm mm O02 N.Nm w.mm w.em ©.Nm o.mm 2 0>Huosocoo OSHQSB 000Hmumm H0Hwflooz HmCOHuc0>coo mm U0mefluu c0mo oaumsood U0UH0cD ZOHBHQZOU OZHzmemHA gomw mcoauaocoo O0OH4 Hsom Ocm ©0OHOCD 0:0 u0ocD muo0flnom mo mmoouu 00u£9 uOm U0cfi0uno i.m0soom COHumcHEHuomHQ co00mm wo m0mcmm pom .Aomv mcoAu0H>0Q OHMUSMDm .AO0ZV mc0H00z .AZV mcm0ZiI.HHH 0HQOH 74 are to be based. As indicated in Chapter III, the first coupling condition under which a subject was tested was repeated. Since the total number of subjects was 50 and the four coupling conditions were rotated, each coupler was retested a total of either seven or eight times. This yielded test and retest scores for both Speech reception thresholds (SRTS) and Speech discrimination scores for each subject under one of the coupling conditions. The Spearman rank-order correlation coefficient was used to investigate the reliability of the repeated measure- ments in this investigation. The use of this statistic requires that both scores be measured in at least an ordinal scale so that individuals can be ranked in two ordered 1 This was achieved for both the test and retest series. scores of the SRTS and Speech discrimination scores, and no further assumptions were necessary for the use of this statistic. An overall correlation was obtained for both SRTS and Speech discrimination scores by combining the test and re- test scores for all four couplers across the three groups of subjects. A correlation of .59 was Obtained for the SRTS and a .95 for the Speech discrimination scores. Although both correlations are Significant beyond the .01 confidence level, the correlation for the Speech discrimination scores lSidney Siegel, Nopparametric Statistics for the Behavioral Sciences (New York: McGraw-Hill Book Company, Inc., 1956i? p. 202. 75 is substantially higher than that of the SRTS. This agrees with the findings of McConnell et al.1 in their investi- gation of the test-retest reliability of clinical hearing aid tests. They did an immediate test-retest of the aided Speech reception thresholds and Speech discrimination scores of 40 hearing impaired subjects using the W-1 Spondee words and the W-22 PB words. A Pearson Product Moment Correlation of .67 was obtained for the SRTS and a .85 for the PBS. After a period of two weeks or more aided SRTS and PB scores were again obtained. This time a correlation of .48 was found for the SRTS and .92 for the PBS. McConnell et al. attributed the lower reliability of threshold measurement, in part, to the increased familiarity of the subjects with the test words from test to retest. Tillman and Jerger2 have also reported such an effect. Since the Speed reception threshold measured in the present study was that intensity level at which a subject reSponded to only 50 percent Of the items, it is reasonable to SXpect some variation in reSponses at this point. This also holds true for Speed discrimination scores at very low sensation levels as found by Tillman and Carhart3 in the lFreeman McConnell, Eileen F. Silber, and Douglas McDonald, ”Test-Retest Consistency of Clinical Hearing Aid Tests," Journal of Sppech and Hearing Disorders, 25 (1960). pp. 275-280. 2Tillman and Jerger, "Some Factors Affecting Spondee Threshold," pp. 141-146. 3Tillman and Carhart, "An Expanded Test for Speech Discrimination," pp. 1-12. 76 develOpment of NU Auditory Test NO. 6 and later by Rintelmann and Jetty.1 In other words, it is not unreasonable to eXpect lower correlations close to the listener's threshold than when Speech audiometric material is presented at high sensa- tion levels. Further in the present study, an examination of SRTS from the first coupling condition and the repetition of the same condition reveals that SRTS tended to be lower for the repetition. Thus, 18 (60%) subjects had SRTS that were lower for the repeated condition, 6 (20%) subjects Obtained the same thresholds, and 6 (20%) had higher thresholds. In other words, a little more than half of the subjects showed im- provement in the retest SRT while the remainder Of the sub- jects did not. These results suggest that the familiarity factor was Operating in the present study even though the subjects had been familiarized with the Spondee words before testing was begun. Another possible explanation is that the subjects became more adept at listening through the CROS hearing aid from test to retest and were thus able to ob- tain lower thresholds during the retest. Since the aid remained at the same gain setting, it is also possible that some of the subjects became more tolerant Of the degree of amplification by the time they had reached the retest condi- tion. At any rate, it appears that some improvement in SRTS occurred in the retest because Of familiarity with the task. lRintelmann and Jetty, "Reliability of Speech Discrimi- nation," Unpublished Study, Michigan State University, 1968. 77 Correlation coefficients were also obtained for each of the four acoustic couplers across all three groups of subjects and are diSplayed in Table IV. Table IV.--Coefficients of Correlation (Spearman Rank-Order) Between Test and Retest for Each of the Four Acoustic COUplers Across All Three Subject Groups ACOUSTIC COUPLERS Acoustic Open Crimped Conventional Modifier Earpiece Tubing (N=8) (N=7) (N=7) (N=8) SRT .96* .70 .71** .95* Speech Discrimination .90* .97* .97* .91* * **Significant at .01 level. Significant at .05 level. Again, it can be seen that the correlations for the SRTS are somewhat lower for two of the coupling conditions. However, from the array of high positive correlation coeffi- cients, it may be concluded that the Speech reception thres- holds and Speech discrimination scores obtained in this study were sufficiently reliable to allow comparisons from one coupling condition to another to be made with confidence. Speech Reception Thresholds In order to test the differences among the mean Speech reception thresholds Obtained in this investigation, the following null hypothesis was postulated: 78 There are no significant differences among the mean aided Speech reception thresholds Obtained with the conventional earmold, acoustic modifier, Open earpiece and crimped polyethylene tubing regardless of the kind of hearing loss or pure-tone, audiometric configuration. In order to determine the significance of differences among the variables in this comparison, a two factor analy- sis Of variance with repeated measures on one factor was employed.1 The repeated measures were on acoustic couplers (Couplers). The second factor was subject groups (Groups): the group of subjects with a conductive hearing impairment (Conductives); the group having a sensorineural impairment with a precipitous drop (Sensorineural Precipitous); and the group having a sensorineural impairment with a gradual- ly SlOping loss (Sensorineural Gradually Sloping). The two factors were arranged in a two dimensional table Of a 4 x 5 design. The four columns represented acoustic coupling conditions, whereas the three rows represented subject groups. The obtained Speech reception thresholds were the criterion measures entered in the resulting twelve cells. The F-ratio was used in testing the statistical signifi- cance of the variance attributable to the two main effects and the two-way interaction. The analysis of variance was conducted on a Control Data Corporation 5600 Digital Computer.2 1B. J. Winer, Statistical Principles in Experimental Design (New York: McGraw-Hill Book Co., 1962)? pp. 519-555. 2Agricultural Experiment Station, "Stat Series Descrip- tion NO. 14: Analysis of Variance with Equal Frequency in Each Cell," (East Lansing, Michigan State University, 1968). 79 The mean Speech reception thresholds used in the com- parison are Shown in Table II, page 71, and a summary of the analysis is given in Table V. Table V.--Summary of Two-Way Analysis of Variance Comparing the Effects of Differences in Kind of Hearing Loss and Type Of Acoustic Coupler on Speech Reception Thresholds . Sum of Mean F Source Of Variance Squares DP Square Statistic Type Of Hearing Loss (A) 4799.217 2 2599.608 14.295* Z=C+AC 4552.250 27 167.861 Couplers (B) 226.967 5 75.655 16.921* A x B 125.585 6 20.897 4.674* Z=ABC+BC 562.150 81 4.471 Total 10045.967 119 * Significant beyond the .01 level. As shown in Table V, the F-statistic was Significant beyond the .01 confidence level for the two main factors and the interaction between these two factors. Thus, the null hypothesis of no difference among the means was rejected. A number of Observations can be made about the analysis Of variance. First, the type Of hearing loss affects the Speech reception thresholds obtained. Second, the type of acoustic coupler affects the obtained SRTS when considered over the different kinds of hearing loss. Third, there is 80 an interaction between kind Of hearing loss and type of acoustic coupler, indicating that the magnitude and di- rection of the effects of coupler conditions on the SRTS differ according to kind Of hearing loss. The presence Of significant interaction indicates pos- sible effects which are due to peculiar combinations Of the two variables under consideration. Thus, caution must be used in making predictions from knowledge Of only one factor. Haysl stated: ”When interaction effects exist, the best estimate one can make of a difference attributable to one factor depends on the particular level Of the other factor." In other words, in order to predict how well a hearing im— paired person will do with a particular acoustic coupler, we must know what kind Of hearing loss or pure-tone audio- metric configuration he has. Since these factors are usually known before a hearing aid evaluation is attempted, the inter- action Observed in the present statistical design is only of passing interest and its implications need not be explored further. The significant F-ratios shown in Table V were investi- gated by employing Duncan's New Multiple Range Test in a comparison Of the treatment means.2 All means were compared. but the interest of this investigation is concerned with the 1William L. Hays, Statistics for Psychologists (New York: Holt, Rinehart and Winston, 1965), p. 590. 2Allen L. Edwards, Expprimental Design in Psychological Rpseappp (New York: Holt, Rinehart and Winston, 1960), pp. 136-141. 81 comparison Of the treatment means within each group of sub- jects. The results of the comparison within the conductive group are shown in Table VI. Table VI.--Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for SRTS# Within the Conductive Group Conventional Acoustic Open Crimped Earmold Modifier Earpiece Tubing Means 11.0 15.9 12.7 14.1 Conventional Earmold 2.9* 1.7 5.1* Acoustic Modifier 1.2 0.2 Open Earpiece 1.5 # In dB re Speech audiometric zero. * Significant at .01 level. From Table VI it can be seen that the mean SRT obtained with the conventional earmold is significantly different statistically from that obtained with the acoustic modifier and the crimped polyethylene tubing. The differences Of about 5 dB are in the direction of a lower mean SRT with the conventional earmold, and although the difference was not statistically significant, the mean SRT with the conventional earmold was also lower than with the open earpiece by about 2 dB. The mean SRTS obtained with the acoustic modifier, 82 Open earpiece, and crimped tubing were not significantly dif- ferent from one another. The tendency for SRTS within the conductive group to be somewhat lower with the conventional earmold can be under- stood in terms of the amount of sound pressure developed in the ear canal. An examination of the mean audiogram (Figure 1, page 46) for this group reveals that the loss of hearing sensitivity is greater for the lower frequencies. The longer tip on the conventional earmold allows greater sound pressure to be develOped in the ear canal at low frequencies, whereas with the modified couplers the sound pressure is reduced at the lower frequencies due to venting and an increase in the size of the cavity between the earmold and the tympanic mem- 1 stated that a reduction in the size of this brane. Lybarger cavity by an earmold tip would produce an improvement in low- frequency output Of about 5 dB. Greater lower frequency out- put could also be expected if the effects Of leakage are re- duced. Thus, in the present study, a combination of the above factors were probably working to enhance the sound pre- sure develOped in the ear canal with the conventional earmold resulting in lower SRTS. The results of Duncan's New Multiple Range Test, employed in investigating significant differences among mean SRTS within the sensorineural precipitous drop group, are shown in Table VII. lLybarger, "Earmold Acoustics," p. 4. 85 Table VII.--Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for SRTS# Within the Sensorineural Precipitous DrOp Group (.01 Significance Level) Conventional Acoustic Open Crimped Earmold Modifier Earpiece Tubing Means 14.5 14.8 12.5 14.7 Conventional Earmold 0.5 2.0 0.2 Acoustic Modifier 2.5 0.1 Open Earpiece 2.2 # In dB re Speech audiometric zero. * Significant at .01 level. From Table VII, it can be seen that there were no statis- tically significant differences at the .01 confidence level among the mean SRTS for this group. In contrast to the con- ductive group, and examination of the mean audiogram (Figure 2. page 48) for this group reveals normal hearing in the low frequencies with a hearing loss for the high frequencies. Since thresholds were already within the normal range, the increase in sound pressure in the ear canal Offered by the conventional earmold was not great enough to lower the thres- hold significantly from those Obtained with the modified couplers. Another possible explanation may be hypothesized from the discrimination scores Obtained from this group. 84 Table III shows that this group Obtained significantly higher mean discrimination scores when the modified earpieces were employed. Since the discrimination function does play a part in determining threshold with Spondee words, it is quite possible that the improved discrimination obtained with the modified couplers was enough to offset any advantage in greater sound pressure at the low frequencies Offered by the conven- tional earmold. The results Of Duncan's New Multiple Range Test applied to the differences among the mean SRTS within the sensorineural gradually sloping group are diSplayed in Table VIII. At the .01 confidence level, Table VIII shows that the mean SRT Obtained with the conventional earmold was signifi- cantly different from the mean SRTS obtained with the acoustic modifier and the crimped tubing. Again, these differences Of 4.5 dB for the acoustic modifier and 6.7 dB for the crimped tubing are in the direction of a lower mean SRT with the con- ventional earmold, and although not significant at the .01 confidence level, the mean SRT for the conventional earmold was 2 dB lower than with the Open earpiece. Table VIII also shows that the mean SRT obtained with the Open earpiece was significantly lower (4.7 dB) than that obtained with the crimped tubing. The largest difference (6.7 dB) was found between the conventional earmold where the ear canal is completely sealed and the crimped tubing where it is Open. The mean audiogram (Figure 5, page 49) for this 85 Table VIII.--Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for SRTS Within the Sensorineural Gradually SlOping Group Conventional Acoustic Open Crimped Earmold Modifier Earpiece Tubing Means 25.6 28.1 25.6 50.0 Conventional Earmold 4.5* 2.0** 6.7* Acoustic Modifier 2.5** 2.2** Open Earpiece 4.7* # In dB re Speech audiometric zero. * Significant at .01 level. *at- Significant at .05 level. group shows a loss Of hearing sensitivity in the low frequen- cies as well as in the higher frequencies. Again, the tendency for SRTS to be lower with the conventional earmold can be explained by the greater sound pressure Obtained in the low frequencies with the conventional earmold. Since with this group there is a loss Of hearing sensitivity for the low fre- quencies, there is greater dependency Of the SRT on the amount of sound pressure in the ear canal at these frequencies than is the case with the sensorineural precipitous drOp group where low frequency thresholds are normal. Although both the conductive group and the sensorineural gradually SlOping group Obtained lower mean SRTS with the conventional earmold 86 as opposed to the modified couplers, the differences are greater for the sensorineural gradually SlOping group. This outcome may be explained in terms of the discrimination scores obtained with these grOUpS. An examination of Table III, shows that the mean Speech discrimination scores for the sensorineural gradually SlOping grOUp are lower under all listening conditions than those of the conductive group. Since the SRT is somewhat dependent on discrimination ability, the poorer Speech discrimination of this group does not allow as much compensation for the loss of sound pressure at the low frequencies as is the case with the conductive group. When the .05 confidence level is considered, then all of the means for the sensorineural gradually sloping group are significantly different from one another. There is appar- ently more variability within this group than within the other two groups. This variability is attested to by the larger standard deviations Obtained with each Of the acoustic COUplers for this group as Shown in Table II. Speech Discrimination Scores In order to test the differences among the mean Speech discrimination scores Obtained in this investigation, the following null hypothesis was postulated: There are no significant differences among the mean Speech discrimination scores obtained under all listening conditions regardless of the kind of hearing loss or pure-tone, audiometric configuration. 87 The analysis of variance for the mean Speech discrimi- nation scores was the same design as for the SRTS except that the mean unaided sound-field Speech discrimination scores were added to the listening condition factor (COUplers). This resulted in a two dimensional table of a 5 x 5 design. The five columns represented the coupling conditions, whereas the three rows represented subject groups. The Obtained Speech discrimination Scores (means) were the criterion measures entered in the resulting 15 cells, and the F-ratio was used in testing the statistical Significance of the variance attributable to the two main effects and the two-way interaction. The mean Speech discrimination scores used in the com- parison are shown in Table III, and a summary of the analysis is given in Table IX. As shown in Table IX, the F-statistic was significant beyond the .01 confidence level for the two main factors and the interaction between these two factors. Thus, the null hypothesis of no differences among the mean Speech discrimi- nation scores was rejected. Again, a number Of observations can be made about the analysis of variance. First, the type Of hearing loss affects the Obtained Speech discrimination scores. Second, the type of acoustic coupler affects the Obtained Speech discrimi- nation scores when considered over the different kinds Of hearing loss. Third, there is an interaction between kind of 88 Table IX.--Summary of Two-Way Analysis of Variance Comparing the Effects of Differences in Kind of Hearing Loss and Type of Acoustic Coupler on Speech Dis- crimination Scores Source of Variance Sum Of Mean F ' Squares DF Square Statistic Type Of Hearing Loss(A) 5075.415 2 2557.707 17.787* Z = C + AC 5852.160 27 142.672 Couplers(B) 1657.440 4 414.560 15.526* A x B 1828.520 8 228.540 8.565* Z = ABC + BC 2882.240 108 26.687 Total 15295.575 149 *- Significant beyond the .01 level. hearing loss and type of acoustic coupler, indicating that the magnitude and direction of the effects of coupling con- ditions on the Speech discrimination scores differ for the different kinds of hearing loss. The Observed interaction is not Of any important concern in this study for the same reasons discussed earlier for the SRTS. The significant F-ratios shown in Table IX were further investigated by employing Duncan‘s New Multiple Range Test in a comparison of the differences between treatment means.1 All means were compared, but the interest Of this investiga- tion was concerned with the comparison of the treatment means lEdwards, Experimental Design in Psychological Research, pp. 156-141. 89 within each group of subjects. The results of the comparison within the conductive group are Shown in Table X. Table X.--Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for Speech Dis- crimination Scores# Within the Conductive Group Unaided Conventional Acoustic Open Crimped SF Earmold Modifier Earpiece Tubing Means 95.0 92.6 94.6 95.6 92.2 Unaided SF 2.4 0.4 1.4 2.8 Conventional Earmold 2.0 1.0 0.4 Acoustic Modifier 1.0 2.4 Open Earpiece 1.4 # In percent correct. * Significant at .01 level. As indicated in Table X, the mean Speech discrimination scores Obtained under all of the listening conditions were not Significantly different statistically from one another at the .01 confidence level. Normal Speech discrimination is usually found with conductive impairments, and evidently the acoustic couplers utilized in this study had no adverse ef- fects on this discrimination ability. The results of Duncan's New Multiple Range Test applied to the differences between the mean Speech discrimination 90 scores within the sensorineural precipitous drop group are diSplayed in Table XI. Table XI.--Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for Speech Discrimination Scores Within the Sensorineural Precipitous Drop Group I Unaided Conventional Acoustic Open Crimped SF Earmold Modifier Earpiece Tubing Means 69.4 76.8 87.2 87.0 88.0 Unaided SF 7.4* 17.8* 17.6* 18.6* Conventional Earmold 10.4* 10.2* 11.2* Acoustic Modifier 0.2 0.8 Open Earpiece 1.0 In percent correct. * Significant at .01 level. Table XI shows that the mean aided Speech discrimination scores were all significantly higher than the mean unaided scores at the .01 confidence level. The 7.4 percent higher discrimination score obtained with the conventional earmold over the unaided sound-field indicates that there may be a certain amount of frequency—response tilting occurring with the combination of the conventional earmold and the CROS hearing aid. Again, referring to the mean audiogram 91 (Figure 2, page 48) for this group, it can be seen that the greatest loss of hearing sensitivity for the important Speech range occurs from 1000 to 4000 Hertz. An examination of Figure 5, page 57 showing the frequency reSponse of the hearing aid used in this study, reveals that it is precisely those frequencies at which the hearing loss is greatest that the aid yields the greatest amount of amplification. Thus, it appears that the pure-tone audiometric configuration of this group in combination with an aid, which gave greater amplification in the mid frequencies yielded a statistically significant increase in the mean Speech discrimination score over that Obtained unaided in a sound-field. From Table XI, it can also be seen that the mean Speech discrimination scores obtained with the modified acoustic couplers were all significantly higher than the mean discrimi- nation Score obtained with the conventional earmold. This outcome is a reasonable expectation when it is remembered that the conventional earmold, much like a low-pass filter, transmits low frequency sounds to the tympanic membrane with- out a reduction in their strength. At the same time the strength of high frequency sounds is reduced by the long narrow channel tip of the conventional earmold. A person witheasensorineural precipitous drop hearing loss eXperiences a discrimination problem because of the sharp difference be- tween his hearing sensitivity for low frequency sounds and his sensitivity for high frequency sounds. When amplification 92 is used in conjunction with a conventional earmold the low frequencies partially mask the high frequencies. However, all Of the modified acoustic couplers utilized in the present study provided a means for the low frequency sounds to escape to the outside atmOSphere, and they had no long channel tips to reduce the high frequency transmission. Thus, when these modified couplers were used in conjunction with a CROS hearing aid, the subjects in this group obtained higher Speech discrimination scores than with both the un- aided condition and the conventional coupler condition. The mean Speech discrimination scores obtained with the acoustic modifier, Open earpiece, and crimped tubing were not significantly different from one another. Evidently the varying degrees to which the ear canal was left unoccluded by these couplers was not great enough from one coupler to another to index any difference in discrimination ability, at least within the limitations of the present SXperiment. The results of Duncan's New Multiple Range Test applied to the differences between the mean Speech discrimination scores within the sensorineural gradually SlOping group are diSplayed in Table XII. As indicated in Table XII, at the .01 confidence level the mean unaided sound-field Speech discrimination score is not Significantly different from any of the mean aided scores. However, the mean Speech discrimination scores obtained with the acoustic modified and crimped tubing are significantly 95 Table XII.--Duncan's New Multiple Range Test Applied to the Differences Between Treatment Means for Speech Discrimination Scores# Within the Sensorineural Gradually SlOping Group Unaided Conventional Acoustic Open Crimped SF Earmold Modifier Earpiece Tubing Means 79.6 75.0 85.0 80.0 84.8 Unaided SF 4.6 5.4** 0.4 5.2** Conventional Earmold 10.0* 5.0 9.8* Acoustic Modifier 5.0** 0.2 Open Earpiece 4.8** # In percent correct. * Significant at .01 level. “K”!- Significant at .05 level. higher than the mean score Obtained with the conventional ear- mold. This occurred because the mean score with the conven- tional earmold was somewhat lower than the mean unaided sound- field score increasing the magnitude Of the distance between the mean score Obtained with the conventional earmold and those Obtained with the acoustic modifier and crimped tubing. When the results are considered at the .05 confidence level, then the mean unaided Speech discrimination score is significantly lower than the mean scores obtained with the acoustic modifier and crimped tubing. Also at this level of 94 confidence, the mean scores obtained with the acoustic modi- fier and crimped tubing are significantly higher than the mean score‘obtained with the Open earpiece. The mean Speech discrimination scores Obtained with the acoustic modifier and crimped tubing are not significantly different from one another. It is evident that the high frequency emphasis provided by the modified acoustic couplers does provide somewhat im- proved Speech discrimination with sensorineural hearing losses having a gradually SlOping loss of five to ten dB per octave. However, there is greater variability within this group than in the other two grOUpS studied, and modified acoustic couplers cannot be used indiscriminately since not all sub- jects Obtained better Speech discrimination scores with them. Clinical Significance In this section the mean differences in the data and their clinical implications are discussed. However, before clinical differences are considered it is necessary to de- termine what constitutes a clinically Significant difference for both the SRTS and Speech discrimination scores. Tillman and Carhartl in investigating the SRTS Of a group of ten subjects with normal hearing found thresholds to be within four to five dB from test to retest. The SRTS lTillman and Carhart, "Some Factors Affecting the Spondee Threshold," p. 145. 95 tended to be lower for the retest, a result which the authors attributed to word familiarity. McConnell, Silber, and McDonald1 studied the reliability of clinical hearing aid tests. The hearing aid worn during the retest was the same model but not the same aid as on the first test. They found that differences in SRTS were six dB or less on the second test for 25 of the 57 subjects studied. Chaiklin and Ventry2 compared 2-dB and 5-dB methods Of Obtaining SRTS and found that for the 2-dB method 27 (95%) subjects out Of 29 had test-retest differences from 0 dB to plus or minus 6 dB. While for the 5-dB method all subjects had test-retest differences within plus or minus 5 dB. Prior to beginning the present study, Rintelmann and Jetty3 Obtained test— retest SRTS while examining the list equivalency of a mag- netic tape recording of N.U. Auditory Test No. 6. Ten young adults with normal hearing served as subjects, and Speech reception thresholds were measured with the same tape re- corded Spondee words used in the present study. At least one week intervened between the first and second tests, and it was found that all subjects had test-retest differences in SRTS no greater than 4 dB. l . . McConnell, Silber, and McDonald, "Test-Retest ConSist- ency," p. 279. 2Joseph B. Chaiklin and Ira M. Ventry, "Spondee Thres- hold Measurement: A Comparison of 2- and 5-dB Methods," Journal of Speech and HearingyDisorders, (1964). PP. 47-59. 3Rintelmann and Jetty, "Reliability Of Speech Discrimi- nation Testing," Unpublished Study. 96 From the foregoing studies it can be concluded that test-retest differences in SRTS will usually range from 0 to 6 dB. Thus, for the present study a difference in SRT of greater than 6 dB was considered to be clinically Sig- nificant. To establish what constituted a clinically Significant difference in Speech discrimination scores, information from the preliminary list equivalency study Of N.U. Auditory Test NO. 6, mentioned above, was used. The results Of this study at a 24 dB sensation level are given in Appendix A. The greatest difference in mean discrimination scores from test to retest was 5 percent for list III. The standard errors of measurement for the four lists ranged from 2.82 percent to 4.51 percent with a mean standard error Of measure- ment across all lists Of 5.52 percent. From this information it was concluded that for the purpose Of this study a dif- ference in a discrimination score Of greater than 4 percent would be considered a clinically significant difference. In order to facilitate the discussion in this section, the mean SRTS obtained under the various listening condi- tions for the three groups are shown in Table XIII. Speech reception thresholds under earphones for the con- ductive group ranged from 14 to 57 dB with a mean SRT of 42 dB for the right ear and 57.1 dB for the left ear. The un- aided sound-field SRTS ranged from 26 to 56 dB with a mean SRT Of 55.6dB. Referring tO Table XIII, a comparison of the 97 .ou0N UHuu0EoHosm £000mm 0H mo SH .x. m.om m.mm a.mm p.mm k.me Ammon Hmspmuoo H0HS0CHuomc0m >.¢a m.md m.efi m.ea m.mm Amsouflmflo0umv amus0cfluomc0m a.¢a >.Na m.m« O.Ha m.mm 0>Huosocou OSHQDB 0O0Hmumm u0amaooz OHOEumm mm U0QEHHU o0mo oaumoood Hmcoauc0>cou U0UH0cD ZOHEHQZOU 073szqu mDOMU mcofluflocoo moac0umaq p0pad usom 0cm p0oflmcD 0co u0ocD muo0flosm mo mmsouu 00HLB MOM O0CHMDDO *moaocm0uca coaum0o0m co00mm cm0zii.HHHX 0HQOB 98 unaided scores to the aided for the conductive group reveals that significantly lower SRTS were Obtained in the aided conditions with all four acoustic couplers. Intercoupler differences in SRTS were not clinically significant, and the small existing differences could probably be overcome by adjusting the gain control of the hearing aid. This is an important outcome, since it reveals that any of the various acoustic couplers can be utilized with persons having con- ductive hearing impairments with comparable results concern- ing SRT. Speech reception thresholds under earphones for the sensorineural precipitous drOp group ranged from 5 to 47 dB with a mean SRT of 24.4 dB for the right ear and 24.6 dB for the left ear. The unaided sound-field SRTS ranged from 8 to 42 dB with a mean SRT of 25.8 dB. Referring to Table XIII again, comparison of these thresholds to the aided thresholds reveals significantly lower SRTS for this group. This is important, since it shows that the SRTS offer no contraindi- cation tO the use of modified acoustic couplers with persons having sensorineural hearing losses with precipitous drOps beyond 500 or 1000 Hertz. For the sensorineural gradually SlOping group, the Speech reception thresholds under earphones ranged from 29 to 66 dB with a mean SRT of 47.1 dB for the right ear and 50.5 dB for the left ear. The unaided sound-field SRTS ranged from 50 to 56 dB with a mean SRT Of 45.7 dB. This 99 group had the greatest loss Of hearing sensitivity and, as Shown in Table XIII although the aided mean SRTS are lower than the unaided, they are not as low as those ob- tained with the other two groups. Intercoupler variability was also greater within this group. The mean SRT obtained with the conventional earmold was 6.7 dB lower than that Obtained with the crimped tubing. This is a clinically significant difference and indicates that caution must be used in recommending the use of crimped tubing with indi- viduals who have a loss of hearing sensitivity in the low frequencies, since the lack of sufficient sound pressure developed in the ear canal at low frequencies seriously affects the SRT. The mean Speech discrimination scores Obtained under the various listening conditions for all three grOUps are shown in Table XIV. Speech discrimination scores under earphones for the conductive group ranged from 86 to 100 percent with a mean score of 95.6 percent for the right ear and 95.6 percent for the left ear. A comparison of these scores with those shown in Table XIV reveals no significant differences between the mean unaided scores under earphones and the mean unaided sound-field. There are also no significant differences between the mean unaided and mean aided scores nor among the mean scores Obtained with the various acoustic couplers. Since there were also no significant differences among the 100 .uo0uuoo uc0ou0m SH * p.4m 0.0m o.mm o.ma m.ma Ameon Happened amus0cfluomc0m o.mm 0.4m N.em m.mk e.mp impengmgomnmo H0HS0CHuomc0m N.mm m.mm m.em O.Nm o.mm 0>Huosocoo OCHQSB 000Hmumm u0flwflooz OHOEHmm mm U0QEHHO c0mo oaumsooe Hmcoauc0>coo O0oflmca ZOHBHDZOU OZHszmHA mbomw mcofluflocoo acac0umflq U0OH< usom pom ©0oflmca 0co u0ocD 0000mnsm mo mmsouo 00uze now O0CHMDQO *m0uoom coeumcaefiuomflo co00mm c0021i.>HX 0H£08 101 mean aided SRTS for this group, it appears that persons with conductive hearing losses can be fitted with any of the types of couplers used in this study, and personal pref- erence or type of hearing aid may dictate the one chosen. Each subject within the conductive group was asked for his subjective preference for the various acoustic couplers. Five subjects preferred the conventional earmold, two preferred the crimped tubing, one preferred the acoustic modifier, one preferred the conventional earmold and Open earpiece equally, and one preferred the Open earpiece and crimped tubing equally. The fact that the conventional ear- mold was preferred by half the subjects may be due to the loss of hearing sensitivity in the low frequencies found in this group. Amplification through the conventional ear- mold may sound more natural, since the low frequencies are not filtered out. However, since the crimped tubing was preferred by some subjects and since it yielded mean SRTS and Speech discrimination scores which were not significant- ly different from the mean scores obtained with the other couplers, it could conceivably be used with some peOple having conductive hearing losses. It may be possible to utilize the crimped tubing in conjunction with a CROS type hearing aid with conductive losses having a chronic drainage problem, since the crimped tubing in no way occludes the ear canal. 102 Speech discrimination scores under earphones for the sensorineural precipitous drOp group ranged from 58 to 78 percent with a mean score of 59.6 percent for the right ear and 57.6 percent for the left ear. A comparison of these scores with those shown in Table XIV reveals that there is a significant improvement in the mean Speech discrimi- nation score obtained unaided in the sound-field from that Obtained under earphones. There is also a significant mean improvement under all of the aided conditions as opposed to the mean unaided sound-field score. With the conventional earmold the mean Speech discrimination score was 7.4 percent higher than the mean unaided sound-field score. The reasons for this difference were discussed above under the statisti- cal results for this group. With the modified acoustic couplers, the increase in the mean discrimination scores from the mean unaided sound-field are even more substantial, 17.8 percent for the acoustic modifier, 17.6 percent for the Open earpiece, and 18.6 percent for the crimped tubing. A comparison of the mean discrimination score Obtained with the modified acoustic couplers to the mean score Obtained with the conventional earmold shows a 10.4 percent increase for the acoustic modifier, 10.2 percent for the Open ear- piece, and 11.2 percent for the crimped tubing. These are all clinically significant differences and indicate that the high frequency emphasis provided by these couplers is prob- ably an important factor in the improved Speech discrimi- nation scores for this group. 105 There were no significant differences among the mean discrimination scores obtained with the modified acoustic couplers employed in this study. Evidently the degree to which the high frequency emphasis was changed from coupler to coupler was not great enough to make significant differ- ences in the mean discrimination scores Obtained with the sensorineural precipitous drOp group. Of interest also is the fact that no subject obtained a better discrimination Score with the conventional earmold as Opposed to any of the modified acoustic couplers. This would seem to under- score the importance of utilizing modified acoustic couplers during hearing aid evaluations with persons showing audio- metric patterns similar tO the subjects in this group. As for subjective coupler preference within the sensori- neural precipitous drop group, five subjects preferred the acoustic modifier, two preferred the crimped tubing, one preferred the Open earpiece, one preferred the conventional earmold and acoustic modifier equally, and one had no pref- erence. Thus, the subjective preference of this group for the modified acoustic couplers was almost unanimous. Summarizing the results for the sensorineural precipi- tous drOp group, there were no Significant differences among the mean SRTS obtained with the various acoustic couplers, mean Speech discrimination scores were signifi- cantly improved with the modified couplers, and they were subjectively preferred over the conventional earmold by 104 9 out of the 10 subjects in this group. These results sug- gest the clinical importance of utilizing the modified COUplers with persons having a high frequency hearing loss with normal hearing in the low frequencies. For the sensorineural gradually SlOping group, Speech discrimination scores under earphones ranged from 56 to 90 percent with a mean score of 75.6 percent for the right ear and 67.4 percent for the left ear. A comparison of these scores with those shown in Table XIV reveals that the mean unaided sound-field discrimination score is somewhat higher than the mean scores Obtained under earphones. Unlike the other two grOUpS, however, the mean Speech discrimination score Obtained aided utilizing the conventional earmold is 4.6 percent lower than the unaided sound-field score. This, in effect, increased the magnitude of the difference between the mean Speech discrimination score Obtained with the con- ventional earmold and those Obtained with the modified acoustic couplers. The mean discrimination score Obtained with the acoustic modifier was 10 percent higher, with the Open earpiece it was 5 percent higher, and with the crimped tubing it was 9.8 percent higher. These differences are all clinically Significant according to the definition employed at the beginning of this section. Comparing the mean Speech discrimination scores Obtain- ed with the modified acoustic couplers to the mean score obtained unaided in the sound-field reveals that there was 105 a 5.4 percent higher score for the acoustic modifier and 5.2 percent for the crimped tubing, both of which are clinically significant differences. There was essentially no difference between the unaided sound-field discrimina- tion score and the one obtained with the Open earpiece. Subjective acoustic coupler preference for the sensori- neural gradually SIOping group Showed that five subjects preferred the crimped tubing, three preferred the acoustic modifier, one preferred the conventional earmold and acoustic modifier equally, and one subject had no preference. It is interesting to note that half the subjects preferred the crimped tubing even though the mean SRT was significantly higher with this coupler than with the conventional earmold. However, this choice is not difficult to understand if it is based on the person's feeling that the words were more clear, since the mean Speech discrimination score obtained with the crimped tubing was significantly higher than that Obtained with the conventional earmold. The Speech dis- crimination task was always presented at a 26 dB sensation level so that the words were at the same level across couplers. Another factor which may have influenced the preference Of acoustic couplers was the fact that the Speech discrimination task was always presented after Obtaining the SRT, and then after both of these tests the person was asked to state his preference. Thus, it is very probable that the preference was based on the discrimination task and not on the Spondee task. 106 A summary of the sensorineural gradually sloping group results indicates that mean SRTS may be Significantly high- er when the acoustic modifier and crimped tubing are utilized as acoustic couplers. However, mean Speech dis- crimination scores obtained with all the modified couplers also tend to be significantly higher than the mean score ()btained with the conventional earmold. Variability within this group was greater than in the other two, and an exami- nation of individual scores within the group reveals that not all subjects received significantly higher SRTS with the modified acoustic couplers and, by the same token, not all received higher Speech discrimination scores with the modified couplers. Thus, it appears that caution must be used in recommending modified couplers to those patients evidencing a gradually SlOping audiometric pattern. CHAPTER V SUMMARY AND CONCLUSIONS The basic purpose of this research was to evaluate the effects of the acoustic coupler on the aided Speech recep- tion thresholds and Speech discrimination scores of hard-of- hearing subjects. The effects of the kind of hearing loss and pure-tone audiometric configuration were also investi- gated. Summary A CROS type hearing aid utilizing a conventional, vented, Open earpiece, and crimped polyethylene tubing was employed in obtaining Speech reception thresholds and Speech discrimination scores for three groups of hard-Of-hearing adults. Group I was composed Of ten subjects with a con- ductive hearing impairment. Group II consisted of ten sub- jects having a sensorineural hearing loss with normal hear- ing in the low frequencies and a precipitous drop for frequencies higher than 500 or 1000 Hertz. Group III was composed of subjects having a gradually SlOping, (5 to 10 dB per octave) sensorineural type hearing loss. Pure-tone air and bone conduction thresholds were Obtained prior to 107 108 the Speech tests. Speech reception thresholds and Speech discrimination scores were obtained in the sound-field under unaided and aided conditions while the nontest ear was occluded by a wax impregnated ear plug. All subjects were tested with the same CROS hearing aid at a gain setting of 55 dB re the HAIC method of determining gain. All Speech discrimination Scores were Obtained at a 26 dB sensation level. Retest discrimination scores and SRTS were Obtained on all subjects for a Single aided condition. Results indicated that for Group I, the mean Speech reception threshold was Significantly lower with the con- ventional earmold than with the modified couplers. The mean aided Speech discrimination scores showed no signifi— cant differences from the mean unaided score nor were there significant differences among the means obtained with the four acoustic COUplerS in the aided condition. For Group II, there were no Significant differences among the mean Speech reception thresholds obtained with the various acoustic couplers. The mean Speech discrimination score was significantly improved under the aided conditions. The mean unaided Speech discrimination score was 69.4 percent in comparison to a mean aided score of 76.8 percent utiliz- ing the conventional earmold. Thus, a gain of 7.4 percent was achieved. The mean Speech discrimination score of the modified couplers combined was 87.5 percent, a gain of 17.9 percentage points over the mean unaided sound-field condi- tion and 10.5 percentage points over the mean score yielded 109 by the conventional earmold. There were essentially no intercoupler differences in the mean Speech discrimination scores Obtained with the three types of modified acoustic couplers. The mean Speech reception threshold for Group III, utilizing the conventional earmold was 4.5 dB lower than that Obtained with the acoustic modifier and 6.7 dB lower than the mean Speech reception threshold obtained with the crimped tubing. Although not statistically significant, the mean Speech reception threshold Obtained with the con- ventional earmold was also 2 dB lower than with the Open earpiece. When the .05 confidence level was considered, then all of the means were Significantly different from one another probably due to the greater variability found within this group. GrOUp III had a mean unaided Speech discrimination score of 79.6 percent and a mean aided score utilizing the conventional earmold of 75.0 percent. Thus, a loss of 4.6 percent was found. The mean discrimination score obtained with the acoustic modifier was 85.0 percent, which was a gain Of 5.4 percent over the mean unaided sound- field score and a 10 percent gain over the mean obtained with the conventional earmold. The mean discrimination score Obtained with the Open earpiece was 80.0 percent, which was not Significantly different from the mean unaided sound-field score but was a 5 percent gain over the mean discrimination score yielded by the conventional earmold. 110 The crimped tubing yielded a mean discrimination score of 84.8 percent, which was an increase of 5.2 percent over the mean unaided sound-field score and an increase of 9.8 percent over the mean discrimination score Obtained with the conventional earmold. Conclusions Within the limitations of the present study, the follow- ing conclusions appear warranted: 1. Persons with conductive hearing impairments tend to Obtain lower aided Speech reception thresholds with the conventional earmold than with the acoustic modifier, Open earpiece, or crimped tubing. This outcome is probably due to greater sound pressure being develOped in the ear canal at the low frequencies with the conventional earmold. 2. There are no significant intercoupler differences in aided Speech reception thresholds obtained with the acoustic modifier, Open earpiece, and crimped tubing for persons with a conductive hearing impairment. 5. There are no significant intercoupler effects on the aided Speech discrimination scores Obtained for persons with a conductive hearing impairment. 4. The modified acoustic couplers can be successfully employed with persons having a conductive hearing impairment provided that sufficient gain (satisfactory SRT) is achieved. 111 5. There are no significant intercoupler differences among the aided Speech reception thresholds of persons hav- ing sensorineural hearing losses with precipitous drOps. 6. Significantly better aided Speech discrimination scores are Obtained with the acoustic modifier, Open ear- piece, and crimped tubing than with the conventional earmold by persons having a sensorineural hearing loss with a pre- cipitous drOp. 7. There are no significant intercoupler differences in aided Speech discrimination scores obtained with the acoustic modifier, Open earpiece, and crimped tubing by persons hav- ing a sensorineural hearing loss with a precipitous drop. 8. Modified acoustic couplers can be successfully used by persons having sensorineural hearing losses with a pre- cipitous drOp. Because of the significant improvement in Speech discrimination scores offered by these couplers, they should be used routinely in hearing aid evaluations with persons Showing this particular pure-tone audiometric con- figuration. 9. Aided Speech reception thresholds Obtained with the conventional earmold are significantly lower than with the acoustic modifier, open earpiece, and crimped tubing for persons having a gradually SlOping sensorineural hearing impairment. 10. Aided Speech discrimination scores tend to be better with the acoustic modifier, Open earpiece, and crimped 112 tubing than with the conventional earmold for persons having a gradually SlOping sensorineural hearing impairment. 11. Modified acoustic couplers can be employed success- fully with sensorineural hearing impairments showing a gradually SlOping audiometric configuration, but caution is indicated, since there appears to be greater variability among persons showing this audiometric pattern. 12. In general, the conventional earmold is subjectively preferred by half of the persons having conductive hearing impairments; whereas, the other half of this group of sub- jects prefers modified earpieces. On the other hand, persons with sensorineural hearing losses almost unanimously prefer modified acoustic couplers. 15. Modified acoustic couplers used in conjunction with the CROS principle Of amplification are valuable in fitting hearing aids to clinical cases with various kinds of hearing losses and pure-tone audiometric configurations. Recommendations for Further Research The present study should be repeated in a background of noise. Various types of noise such as broadband white noise, Speech noise, and Speech babble could be used. This might prove to be a way of indexing any differences that may exist among the various types of modified acoustic couplers that were not evidenced in the present study with the group having a sensorineural hearing loss with a precipitous drOp for frequencies higher than 500 or 1000 Hertz. 115 Further investigation Should be made of the effects of different kinds of hearing aids used in conjunction with the modified acoustic couplers. This should include hearing aids having different frequency reSponse characteristics such as low or high frequency emphasis, since certain combi- nations of frequency reSponse and acoustic couplers may prove to be beneficial as far as improved Speech discrimination is concerned. Methods should be eXplored whereby the reSponse charac- teristics of a hearing aid can be measured when it is termi- nated by various acoustic couplers. When such methods are devised, then the electro-acoustical characteristics of the hearing aid could be related to the clinical results Ob- tained with persons having various kinds of audiometric con- figurations. Perhaps, with the use of a coupler simulating the human ear canal, measurements could be made to determine any changes in frequency reSponse when the closed canal is compared to measurements made with the Open canal. This would give a better understanding of the change in the fre- quency reSponse characteristic as influenced by various acoustic couplers. The Optimum length of crimped polyethylene tubing pene- tration for comfort and maximum benefit needs to be examined if this type of acoustic coupler is to receive further clinical use. 114 Because of the variability found within the sensori- neural gradually SIOping group in the present study, further investigation should be made of the effects of modified acoustic couplers with subjects showing this type of audio- metric configuration. BIBLIOGRAPHY Books Bekesy, George Von Cited by Stevens, Stanley 8., and Davis Hallowell, Hearing: Its Psychology and Physiology. New York: John Wiley & Sons, Inc., 1958. Blalock, Hubert M., Social Statistics. New York: McGraw- Hill Book Company, Inc., 1960. Davis, Hallowell, and Silverman, S. Richard, Hearing and Deafness. New York: Holt, Rinehart 86 Winston, 1964. Davis, Hallowell, Stevens, S. S., Nichols, R. H. Jr., Hudgins, C. V., Marquis, R. J., Peterson, G. E., and Ross, D. A., Hearing Aids: An EXperimental Study of Design Objectives. Cambridge, Massachusetts: Harvard University Press, 1947. Edwards, Allen L., Experimental Design in Psychological Research. New York: Holt, Rinehart & Winston, 1960. Hays, William L. Statistics for Psychologists. New York: Holt, Rinehart and Winston, 1965. Jackson, Chevalier, Diseases of the Nose, Throat and Ear. Philadelphia: W. B. Saunders. 1945. Kranz, Fred W., Hearing Aids. Elmsford, New York: Sonotone Corporation, 1941. Licklider, J. C. R., "The Perception Of Speech," in Handbook of Experimental Psychology, S. S. Stevens, Ed. New York: John Wiley & Sons, Inc., 1951. Gourevitch, Vivian, Statistical Methods: A Problem-Solving Approach. Boston: Allyn & Bacon, Inc., 1965. Siegel, Sidney, Noppgrametric Statistics for the Behavioral Sciences. New York: McGraw-Hill Company, Inc., 1956. Watson, Leland A., and TOlan, Thomas, Hearing Tests and Hear- ing Instruments. Baltimore: Williams & Wilkins Company, 1949. 115 116 Winer, B. J., Statistical Principles in Experimental Design. New York: McGraw-Hill Book Company, Inc., 1962. Periodicals Breakey, M. R., and Davis, Hallowell, "Comparisons of Thres- holds for Speech: Word and Sentence Tests: Receiver vs Field and Monaural vs Binaural Listening," LapyngSCOpe, 59 (1949), 256-250. Carhart, Raymond, and Jerger, James F., "Preferred Method for Clinical Determination Of Pure-tone Thresholds," Journal of Speech and Hearing Disorders, 24 (1959), 550-545. Dodds, Elizabeth, and Harford, Earl, "Modified Earpieces and CROS for High Frequency Hearing Losses," Journal of Speech and Hearipg Research, 11, (1968), 204-218. Fletcher, Harvey and Munson, W. A., "Relation Between Loud- ness and Masking," Journal of the Acoustical Society of America, 9 (1957). 1-10. Fletcher, Harvey, "Auditory Patterns," Review of Modern Physics, 12 (1940), 47-65. Grossman, Frederick M., "Acoustic Sound Filtration and Hear- ing Aids," Archives of Otolaryngology. 58 (1945), 101- 112 O Grossman, Frederick M., and Mollory, Charles T., "Acoustic Sound Filtration and Hearing Aids," Journal of the Acoustical Society of America, 16 (1944), 52-59. Halsted, Thomas H., and Grossman, Frederick M., "Modern ASpects of the Hearing Aid Problem," New York State Journal of Medicine, 42 (1942), 1944-1950. Halsted, Thomas H., and Grossman, Frederick, M., "Some Prob- lems Involved in the Fitting of Hearing Aids," New York State Journal of Medicine, 41 (1941), 552-558. Harford, Earl, and Barry, Joseph, "A Rehabilitative Approach to the Problem Of Unilateral Hearing Impairment: The Contralateral Routing of Signals (CROS)," Journal of Speech and Hearing DiSOrders, 50 (1965). 121-158. 117 Hirsh, Ira J., Davis, Hallowell, Silverman, S, Richard, Reynolds, Elizabeth G., Eldert, Elizabeth, and Benson, Robert W., “DevelOpment of Materials for Speech Audiometry," Journal of Sppech and Hearing Disorders, 17 (1952), 521-557. Hood, J. D., "Principles and Practice of Bone Conduction Audiometry," Laryngoscgpe, 70 (1960), 1211-1228. Jerger, James F., Carhart, Raymond, Tillman, Tom W., and Peterson, John L., "Some Relations Between Normal Hearing for Pure Tones and for Speech," Journal of Speech and Hearing Research, 2 (1959), 126-140. Lehiste, I., and Peterson, G. E., "Linguistic Considerations in the Study of Speech Intelligibility," Journal of the Acoustical Society of America, 51 (1959), 280-286. Littler, T. 5., "Hearing Aids for the Deaf," gournal of Scientific Instruments, 15 (1956), 144-155. Lybarger, S. F., "Earmold Acoustics," Audecibel, Winter, 1967. Lybarger, S. F., ”A New HAIC Standard Method of Expressing Hearing Aid Performance," The Hearinngealer, February 1961, 16-17. McClellan, Max E., "Aided Speech Discrimination in Noise with Vented and Unvented Earmolds," Journal of Auditory Research, 7 (1967), 95-99. McConnell, Freeman, Silber, Eileen F., and McDonald, Douglas, "Test-Retest Consistency of Clinical Hearing Aid Tests," Journal of Speech and HearinngisorderS, 25 (1960). 275-280. Munson, W. A., and Weiner, Francis M., "In Search of the Missing 6 dB," Journal Of the Acoustical Society of America, 24 (1952), 498-501. Munson, W. A., and Weiner, Francis M., "Sound Measurements for PsychOphysical Tests," Journal of the Acoustical Society of America, 22 (1950), 582-586. Peterson, G. E., and Lehiste, I., "Revised CNC Lists for Auditory Tests," Journal of Speech and Hearing Dis- orders, 27 (1962), 62-70. Rudmose, Wayne, "Free-Field Thresholds vs. Pressure Thres- holds at Low Frequencies," Journal of the Acoustical Society of America, 22 (1950), 674. 118 Sabine, Paul E., "On the Acoustic Properties of Small Cavi- ties," Journal of the Acoustical Society of America, 15 (1942). 74-78. Sanders, Jay W., and Rintelmann, William F., "Masking in Audiometry," Archives of OtolagyngIOQY. 80 (1964). 541-556. Schier, Mayer B. A., ”Clinical Phenomena in Conductive Media: The Individual Earpiece," Journal of the Acoustical Society Of America, 17 (1945), 77-82. Schier, Mayer B. A., "The Earpiece--In Testing for and Fitting Hearing Aids," LaryngOSCOpe, 51 (1941), 52-60. Sivian, L. J., and White, S. D., "On Minimum Audible Sound Fields," Journal of the Acoustical Society of America. 4 (1955), 288-521. Tillman, Tom W., and Jerger, James F., "Some Factors Affect- ing the Spondee Threshold in Normal-Hearing Subjects," Journal of Speech and Hearing Research, 2 (1959). 141-146. Tillman, Tom W., Johnson, Robert M., and Olsen, Wayne 0., "Earphone versus Sound Field Threshold Sound-Pressure Levels for Spondee Words," Journal of the Acoustical Society of America, 59 (1966), 125-155. Weiner, Francis M., and Ross, Douglas A., "The Pressure Dis- tribution in the Auditory Canal in a Progressive Sound Field," Journal of the Acoustical Society of America, 18 (1946), 401-408. Unpublished Material Harrison, Anne, "Some Clinical Uses of the Modified Ear Insert in Supplying More Acceptable Amplification for Selected Sensorineural Hearing Impairments," Unpublished paper presented at the Annual Convention Of the American Speech and Hearing Association, Chicago, Illinois, November, 1967. Lewis, Ernest, and Plotkin, William H., “The Role of the Acoustic Coupler in Hearing Aid Fitting," Unpublished paper presented at the Annual Convention Of the American Speech and Hearing Association, Chicago, Illinois, 1962. 119 Rintelmann, William F., and Jetty, Albert J., "Reliability of Speech Discrimination Testing using CNC Monosyl- labic Words," Unpublished Study, Michigan State Univer- sity. 1968. Other Sources "American Standard Methods for Measurement of Electroacoustical Characteristics of Hearing Aids," American Standards Association Incorpprated, NO. S5-5-1960 (1960). Lybarger, S. F., "The Earmold as a Part of the Receiver Acoustic System," Booklet published by Radioear Corporation, Canonsburg, Pennsylvania, 1958. Nichols, R. H. Jr., Marquis, R. J., Wiklund, W. G., Filler, A. S., Feer, D. B., and Veneklasen, P. 8., "Electro- Acoustical Characteristics of Hearing Aids," Hearing and Hearing Aids, Sec. I, U. S. Office of Scientific Research and DeveIOpment, Report NO. 4666, Cambridge, Massachusetts: Harvard University, 1945, 44-68. Schaefer, Donald W. (D. W. Schaefer and Associates, Inc., 25 West Main Street, Madison, Wisconsin), Personal Communication, April 16, 1968. Tillman, Tom W., and Carhart, Raymond, "An EXpanded Test for Speech Discrimination Utilizing CNC Monosyllabic Words (NU Auditory Test NO. 6)," U. S. School Of AerOSpace Medicine--Technical Research, 66-55, 1-12, June, 1966. APPENDICES 120 APPENDIX A RESULTS OF RESEARCH PROJECT WITH N.U. AUDITORY TEST NO. 6 Prior to beginning the present study, the reliability of Speech discrimination testing using CNC monosyllabic words was investigated by Rintelmann and Jetty.l The pur- poses of this study were to confirm the results Obtained by Tillman and Carhart2 and to determine if comparable re- sults would be Obtained when the words were Spoken by a different Speaker. The four lists comprising N.U. Auditory Test NO. 6 were recorded on magnetic tape by a male Speaker with a General American dialect who monitored his vocal output by means Of a VU-meter. The carrier phrase, "You will say" preceded each test word. The last word of the carrier phrase was monitored and the CNC word was said naturally. The Speaker was located in an IAC room and the words were recorded on an Ampex, model AG550-1 console type tape re- corder. lRintelmann and Jetty, "Reliability of Speech Discrimi- nation Testing Using CNC Monosyllabic Words," Unpublished Study, Michigan State University, 1968. 2Tillman and Carhart, "An EXpanded Test for Speech Discrimination," pp. 1-12. 121 122 Ten young adults with normal hearing served as subjects and the test-retest procedures employed were those outlined by Tillman and Carhart. Each of the four lists was pre- sented to the subjects at six sensation levels (-4, 0, 8, 16, 24, and 52 dB) relative to the Speech reception threshold Obtained with recorded Spondee words. At least one week intervened between the test and retest. The results Obtained were comparable to those found by Tillman and Carhart with normal hearing subjects. Only the statistical data at the 24dB sensation level are given in Tables XV through XVIII, since it is the level closest to the 26 dB sensation level used in the present study. Table XV.--Median (Med), Mean (M), and Standard Deviations (SD) of Speech Discrimination Scores Obtained with N.U. Auditory Test No. 6 at a 24 dB Sensa- tion Level for Ten Subjects with Normal Hearing During the First Test Session. (Scores Repre- sent Percent of Items Correctly Repeated) List I List II List III List IV Med 95 94 85 91 M 92.2 95.0 87.4 92.0 SD 5.5 4.1 4.6 5.7 125 Table XVI.--Median (Med), Mean (M), and Standard Deviations (SD) of Speech Discrimination Scores Obtained with N.U. Auditory Test No. 6 at a 24 dB Sensa- tion Level for Ten Subjects with Normal Hearing During the Retest Session. (Scores Represent Percent of Items Correctly Repeated) List I LiSt II LISt III List IV Med 94 95 92 95 L M 95.4 92.6 90.4 94.4 SD 5.6 5.1 4.7 2.8 ' Table XVII.--Difference Between Mean Discrimination Scores From Test to Retest at a 24 dB Sensation Level for Ten Normal Hearing Subjects on N.U. Audi- tory Test No. 6 (Negative Difference Indicates Higher Score in Retest than in Test Session) List I List II List III List IV -1.2 0.4 -5.0 -2.4 Table XVIII.--Coefficients of Correlation (Pearson r) and Standard IError of Measurement (Se) Between Test and Retest for N.U. Auditory Test No. 6 Administered to Ten Subjects With Normal Hearing at a 24 dB Sensation Level List I List II List III List IV r .71 .75 .74 .75 Se 2.96 2.82 4.51 5.79 124 Figure 7.--Mean discrimination scores yielded by ten normal hearing subjects for Lists I, II, III, and IV Of N.UL AuditOry Test No. 6 during both test and retest sessions. 100 90 80 70 60 Percent Correct Discrimination 50 40 50 20. 10 V . ‘ " a V ’4 ‘7‘ - E: V ‘ :3 AV ‘V :3 0 0 List 1 Test 0 List 1 Retest A List 2 Test A List 2 Retest CI List 5 Test - List 5 Retest v List 4 Test ' List 4 Retest a 0 AV L I 4 a 4 i 4 0 8 16 52 40 Sensation Level in dB 48 APPENDIX B ATTENUATION PROVIDED BY WAX IMPREGNATED EAR PLUGS (FLENTS) FOR PURE TONES AND FOR SPONDEE WORDS Prior to their utilization in the present study, the attenuation provided by wax impregnated ear plugs (Flents) for pure tones and Speech was determined. Three young adults with normal hearing served as subjects. The following procedures were employed: Monaural air- conduction thresholds were determined under earphones for each ear, and then Spondee thresholds were determined mon- aurally under earphones for each ear using tape recorded Spondee words. The ear with the better SRT was chosen as the test ear. The unoccluded sound-field SRT for the test ear was determined while the nontest ear was occluded by an ear plug and covered with a Willson Sound-Barrier Earmuff. The non- test ear remained occluded in this fashion throughout the remainder Of the Speech testing. The test ear was then occluded with an ear plug and the SRT was redetermined. Finally, the pure—tone thresholds Of the test ear were again measured. This time, however, the ear was occluded with a Flent which had remained in the ear from the previous test condition. By following the foregoing procedures, once the 125 Figure 8.—-Mean audiogram of three subjects with normal hearing Showing attenuation provided by wax impregnated ear plugs (Flents) for pure tones. Hearing Level in dB (ISO 1964) 126 (Thresholds ISO-1964 Standards) 20 40 #—_$/ 50 6O 70 80 90 100 250 500 1000 2000 4000 Frequency in Hertz 0 Unoccluded 0 Occluded with Flent 8000 127 Flent was placed in either the test or nontest ear it re- mained there until all testing was completed. The amount Of attenuation provided by the ear plugs for pure tones is shown in Figure 8, and the attenuation provided for Spondee words is given in Table XIX. Table XIX.--Unoccluded and Occluded Speech Reception Thres- holds Showing Attenuation Provided by Wax Impregnated Ear Plugs (Flents) for Spondee Words (Thresholds in dB re Audiometric Zero) Subject Unoccluded Occluded with Flent Attenu— SRT SRT ation 1 -6 28 54 2 5 57 54 5 -6 29 55 Mean attenuation for Spondee words = 54.5 dB 128 U0DCHucoo .umm ummu pmpfle .x. em OH om MH Om m mm OH mm m Hm mm mm mm *A AmcHQoB O0QEHHUV mm om m 2 pH OH em m «m NH mm mH Om HH em NH mm mm mm mm *A AHmcoHuc0>couv mm mm m m om m AH0HMHOOS UHDmsoodv mm 4H A em NH Hm OH mm HH mm NH Om m Hm mm mm mm *m m om m A0U0Hmumm c0mov mm Om A Om HH om mH Hm NH Hm NH mm m mm 59 em #9 *m m OH > AmcHQDB U0QEHHOV mm mm A om em mm mm mm mm mm mm mm mm Om mm mm mm *m z OH O mm NH om mH mm NH mm OH mm NH mm we OOH we *4 AHMCOHuc0>coov mm mm m m em m em OH Hm HH om HH em «H mm m mm on mm mm *A AH0HHHOOS UHumsoodv OOH we m 2 ow 5 mm m om HH mm NH mm m mm m mm mm «m mm *4 A0U0Hmumm c0mov mm 05 m m Hm m mm OH mm OH em m mm OH em HH mm Hm em mm *A AmcHQSB p0mEHro mm pm a 2 we N AHMSOHuc0>coov mm mm A «m OH em NH mm m mm MH em HH mm mm Om Hm *m m um H mm 8mm mm Ham mm Bum mm Bum mm 8mm mm 8mm mm BMm Hmm xmm 0m¢ .Oz coHuHocoo U0OH< msHQSB 000Hmumm H0HMHOOZ oHoEumm OH0Hm Ocsom m0conmumu uo0n umuHm mo 000m0m p0meHuo c0mo UHumsoo¢ H0S0Hp¢0>coo O0OHMCD H0ch IQDm A0>HDUSUCOUV H msouw Bomhmbm EUcouv OO HO A Oh 5H OO mH HO OH OO ON H5 OH OH OO On HO *m 2 bO m oxu0HmHOoz UHDOSOUOV OH ON A w“ Hm OH Om NH Om m Om NH OO OH O» OH 05 OH *m 2 OH O NO OH OO OH HO NH OO OH Hm OH OO HH oh OH *A A000Hmnmm c0mov NO HN m 2 OH O Om OH Hm NH Om NH om HH HO NH OO ON OO mm *A ROCHASB OmmEHHOV NO ON m 2 OH H AAMCOHuc0>coov NO NO A On NN HO NN OO mH OO HN Hm ON OH NH On 5H *m 2 OO O AH0HHHUOZ UHDOSOUOV OO OH A NO NH HO OH HO NH HO OH On OH OO ON GO HN *m 2 OH N OO m OO NH OO NH OO HH On HH OO mN HO mN *A A000Hmumm c0mov OO NN m 2 5O H mm 8mm mm 9mm mm 8mm mm 9mm mm 9mm mm 8mm mm 8mm HOW xwm 00¢ .02 COHDHOCOO O0OH< mCHASB 000Hmumm H0HMHOOS OAOEumm OA0Hm Ocsom m0conmumm uo0n umuHm mo um0m0m O0OEHHU ammo UHumsood AmcoHuc0>coo O0OHmcD u0OCD insm . LFII IN Amouo msouHmHo0um H0H30CHuomc0mv AH msouw .nmm ppm» pmeHH * 3835. 668.828 mm OH 4 Om NH Om HH Nm NH om OH NO OH OO OO OO mN *m m ON OH AAMSOHuc0>c00V Om OH A Om HN HO mN Hm HN Om mN Om mN OO OH Nm NO *m 2 OH m OO ON GO OO OO ON HO OO Om ON HO mH NO HO *A AH0HMHOOS OHumsoodv HO OH O 2 HO O A0U0Hmu0m :0mov OO OH A HO NO OO OO HO NO NO OO NO ON OO OH OO OO *m 2 mO m NO OO OO NO Om ON OO OO OO ON OOH OO OO OO *A AOCHADB O0QEHHUV Om HH m 2 OO O m. 28.835288 me E a 1 OO ON HO OO Om OO Om NO NO ON GO HO OO OO *m 2 HO O OO ON Nm OO Om ON Nm ON Nm NN Hm OH OO mH *A AH0HOHOOZ UHumsoodv . OO OO O 2 mH H A000Hmumm c0mov Nm OH A Hm OH Hm OH Om HH Hm HH OO HH OO HO HO OO *m 2 OO O AmcHQSB O0mEHuUV HO OH NO NH OO OO HO OO OO HH OO mH OO HO *m 8 NO N AAmcoHuc0>coov NO OO A OO HO OO OO NO HO om OO Hm mN Om OO OO OO *m z mO H mm Bum mm 8mm mm .55 mm 91mm mm HMO mm 9mm mm .55 HOW xwm 0md .02 COHuHOcoo O0OH< mCHnoB 000Hmumm H0HOHOOZ OAoEumm UH0Hm Ocsom m0conmumm uo0n umuHm mo um0m0m O0QEHHU c0mo UHumsood AmcoHuc0>cou O0OHMCD H0OcD IADO AmchoAm >AAmsOmuO Amus0CHHomc0mv AHH msouw