v __.A A STUDY OF THE ELECTROACOUSTICAL RESPONSE-CURVE CONSISTENCY OF rwsmsromm HEARING AIDS The“: for flu Dunn of. M. A. MICHIGAH STATE UNIVERSITY Ronald M. Rogers 1965 “ A L [BR A R Y Michigan State Universxty THESIS ABSTRACT A STUDY OF THE ELECTROACOUSTICAL RESPONSE-CURVE CONSISTENCY OF TRANSISTORIZED HEARING AIDS By Ronald M. Rogers The purpose of this study was to analyze and com— pare the results obtained from the electroacoustical re- sponse measurements of current transistorized hearing aids so as to determine their reliability on repeated trials. The criteria employed for the desired measurements were the American Standards Association's Basic Frequency Re- sponses and Saturation Output Responses. Eight current clinically-used hearing aids were selected from the stock of a clinical facility under the auspices of the Michigan State University Speech and Hear— ing Clinic. Eight aids which matched the first selection were selected from another clinical facility under the same auspices. The measuring system employed consisted of the following: one hearing aid test box, one audio frequency spectrometer, two condenser microphones, one audio oscil- lator, one 2 cc coupler, one cathode follower, one cathode follower cable, and one amplifier. Bach aid was evaluated by the stated criteria on three distinct trials for the frequencies of 250, 500, 1000, 2000, 3500, and 4000 cps. The response measurements of the aids, in dB coupler SPL, Ronald M. Rogers were recorded for statistical analysis. The following null hypotheses were tested: 1. There is no difference in the mean coupler SPL between Hearing Aids I and II of the eight models as measured by the following criteria: (a) ASA Basic Frequency Responses (b) ASA Saturation Output Responses 2. The difference in coupler SPL between the first and second hearing aids of the same model do not vary as a function of auditory frequency (There is no aids—by-frequency interaction.) as measured by the following criteria: (a) ASA Basic Frequency Responses (b) ASA Saturation Output Responses 3. The difference in coupler SPL--taken over all frequencies--between the first and second hear- ing aids is the same for all models investigated (There is no aids-by-models interaction.) as measured by the following criteria: (a) ASA Basic Frequency Responses (b) ASA Saturation Output Responses The results of the analyses demonstrated that the main effect and the two interactions under consideration for the basic frequency responses were statistically sig- nificant at better than the .01 level of confidence. It further demonstrated that of the one main effect and the two interactions under consideration for the saturation Ronald M. Rogers output responses, only one interaction (aids x frequencies) did not reveal a statistically significant F statistic below the .01 level of confidence. The conclusions drawn from this study were that the recurrent statistically-significant results were not necessarily clinically significant. A STUDY OF THE ELECTROACOUSTICAL RESPONSE-CURVE CONSISTENCY OF TRANSISTORIZED HEARING AIDS By Ronald M. Rogers A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Speech 1965 LIST OF .LIST OF LIST OF Chapter TABLE OF CONTENTS TABLES O O O O O O O O O O O O O O FIGURES O O O O O O 0 O O 0 O O O APPENDICES O 0 O O O O O O O O O 0 STATEMENT OF THE PROBLEM. . . . . Introduction. . . . . . . Definition of Terms . . . Purpose of Study. . . . . Importance of the Study . Organization of Study . . . . . REVIEW OF THE LITERATURE. . . . . Trend in Hearing Aid Sales. . . Response Measurements of Vacuum Indirectly Related Studies. . . EQUIPMENT AND TESTING PROCEDURES. Equipment . . . . . . . . . . . Procedures. . . . . . . . . . . Summary . . . . . . . . . . . . RESULTS AND DISCUSSION. . . . . . Results . . . . . . . . . . . . Discussion. . . . . . . . . . . SUMMARY AND CONCLUSIONS . . . . . Summary . . . . . . . . . . . . Conclusions . . . . . . . . . . BIBLIOGRAPHY O O O O O O 0 0 O O 0 O O O O APPENDICES. O O O O O O O O O O 0 O O O 0 ii Page iii iv l-l 0000me 10 10 ll 12 l6 l6 19 23 24 24 29 4O 4O 42 44 46 LIST OF TABLES Table Page I. Summary of Analysis of Variance for Basic Frequency Responses . . . . . . . . . . . . . 27 II. Summary of Analysis of Variance for Saturation Output Responses. . . . . . . . . . . . . . . 28 iii Figure l. 2. LIST OF FIGURES Block Diagram of Equipment Arrangement for Aid Response Evaluation . . . . . Mean Coupler SPL of Aids I and II Frequency Responses . . . . . Mean Coupler SPL of Aids I and II Output Responses. . . . . . . Mean Frequency Responses for Aids Basic Frequency Responses . . Mean Frequency Responses for Aids Saturation Output Responses . Mean Coupler SPL of Aids x Models for Basic Frequency Responses Mean Coupler SPL of Aids x Models for Basic for Saturation I I Interaction Interaction for Saturation Output Responses iv and II for and II for Page 19 31 32 34 36 37 38 LIST OF APPENDICES Appendix Page A. Raw Data for Basic Frequency Responses . . . 47 B. Raw Data for Saturation Output Responses. . . 49 C. List of Hearing Aids Employed in Study. . . . 51 CHAPTER I STATEMENT OF THE PROBLEM Introduction Many clinical audiologists are engaged in the daily process of hearing aid evaluations and recommendations. The audiologist selects several instruments for evaluation on a client. The criteria for the selection of the aids for evaluation are generally the frequency response char- acteristics of the instrument. Information regarding the frequency response characteristics of the aid is presented in the manufacturers' specification brochures. The fre-3 quency response characteristics are based on standards es- tablished by the Hearing Aid Industry Conference (HAIC)l and the American Standards Association (ASA), Subcommittee SB-W-34.2 The basic information provided in the manufaC- turers' specification brochures is usually adequate for the evaluation procedures. However, information regarding the reliability of the frequency responses of the aids is not published in the specification brochures. Some ls. F. Lybarger, "Standardized Hearing Aid Measure- ments," Audecibel, 10 (2), 1961, 8. 2"Methods for Measurement of Electroacoustical Characteristics of Hearing Aids," American Standards Asso- ciation Bulletin, 33.3 (1960), 7-15. audiologists would like to have more information regarding the frequency response characteristics of the instruments they wish to evaluate on their clients. The question that confronts the practicing audiolo- gist is the reliability of the frequency response charac- teristics of current hearing aids. Will a particular hear— ing aid respond significantly different on different trials for the electroacoustical response characteristics of Basic Frequency Responses and Saturation Output Responses? Will the response characteristics of two different aids of the same make and model differ significantly? Are the electro— acoustical characteristics of current hearing aids, with the increasing interest in miniaturization, alike enough to justify the recommendation of the purchase of a like instrument to the client? That is, will the aid the client purchases differ significantly from the instrument which performed successfully during the hearing aid evaluation? Some practicing audiologists have advocated the procedure of having the client purchase the actual instru- ment that performed successfully during the aid evaluation, thereby recognizing and attempting to eliminate the prob- lems with which this study is concerned. A review of the literature has revealed an absence of research with current transistorized hearing aids related to the questions and problems with which this study is con- cerned. Therefore, there appears to be a distinct need for research findings which will attempt to answer or shed light upon the major questions of the reliability of the frequency response characteristics of current hearing aids. Definition of Terms For the purposes of this study the terms used are defined in the following manner. Hearing aid evaluation.--A series of audiological tests for the purpose of determining the relative effective- ness of the available hearing aids for recommendation to a client. Acoustic gain.—-"The amount in decibels, by which the sound pressure level developed by the hearing-aid ear- phone in a specific coupler exceeds the sound pressure level in the free-field into which the hearing aid is introduced."1 Hearing Aid Industry Conference (HAIC).--A confer- ence of hearing aid manufacturers who meet for the purposes of establishing ethics, manufacturing standards, public relations, etc., for the hearing aid industry. American Standards Association (ASA).--"A voluntary association of manufacturers and consumers which has written standards for many branches of American industry including the field of acoustics."2 1Ibid., p. 7. 2Hallowell Davis and S. Richard Silverman, Hearing and Deafness (New York: Holt, Rinehart and Winston, Inc., 1962), p. 177. Frequency response characteristics.--The pattern of the amplification of acoustic energy by a particular hearing aid for the stated ASA criteria. Current hearing aids.--Hearing aid models which have been manufactured within the last three years. ASA Basic Frequency Response.—-"The frequency re- sponse for a specified input sound pressure level, main— tained constant over the specified frequency range, and a specified output sound pressure level at 1000 cps that is chosen as a reference response for purposes of descrip- tion."1 ASA Saturation Output Response.--The maximum SPL obtainable in the coupler from the earphone of the aid with the gain control of the aid at the full-on position. Clinically-used hearing aids.--Hearing aids which have been used for the purposes of clinical evaluation on clients in the daily routine of a clinical audiology facil- ity. These instruments have had neither excessive abuse or care. Clinicallyegood operating condition.--A condition whereby the audiologist would not hesitate to use this in- strument for fear of less than adequate functioning. Randomly-selected hearing aids.--"Hearing aids with different frequency—response characteristics selected 1"Methods for Measurement of Electroacoustical Characteristics of Hearing Aids," loc. cit. randomly from the clinic stock without regard to specific response characteristics, acoustic gain, maximum output, etc."1 Body-worn hearing aids.--Instruments which have been designed to be worn on the body below the shoulder line. Ear-level hearing aids.——Instruments which are de— signed to be worn at the level of the ear either in eye- glasses or curving over and behind the ear. Coupler.—-"A coupler is a device for the acoustic loading of earphones. It has a specified arrangement of acoustic elements and is provided with a microphone for the measurement of the sound pressure developed in a spec- ified portion of the device."2 Sound Pressure Level (SPL).--"The sound pressure level (SPL) of any sound is the ratio between its pressure and a standard reference pressure, usually 0.0002 dyne/cm2."3 Coupler SPL.--The sound pressure level measured in the coupler. Clinically significant difference.—-A difference lConstance Rae Walton, "Discrimination by Normally Hearing Subjects for Filtered Speech Under Conditions of Hearing Aid Amplification" (unpublished Master's thesis, College of Communication Arts, Dept. of Speech, Michigan State University, 1964), p. 7. 2"Methods for Measurement of Electroacoustical Characteristics of Hearing Aids, op. cit., p. 7. 3Hayes A. Newby, Audiology (New York: Appleton- Century—Crofts, Inc., 1964), p. 11. between Aids I and II that is greater than 5 decibels. Purpose of the Study, The purpose of this study is to analyze and compare the results obtained from the electroacoustical response measurements of current clinically—used hearing aids. The aids evaluated in this study have been used for daily rou— tine clinical evaluations and were considered to be in clinically-good operating condition. The electroacoustical response characteristics of Basic Frequency Responses and Saturation Output Responses were obtained according to pro- cedures recommended and outlined in the American Standards Association Bulletin 83.3 (1960).1 One of the purposes of the Basic Frequency Response measurement is to aid the clinical audiologist in his se- lection of hearing aids for evaluation on the client. This measurement provides the response characteristics of the various frequencies to a fixed input and a fixed gain ad- justment on the aid. The purpose of the Saturation Output Response is as follows: To determine the maximum rms coupler sound pressure level that the hearing aid is capable of producing with gain control and maximum, using as much input sound pressure level as is needed to produce maximum output at each test frequency. This test gives information which is of great value when considering whether the maximum 1"Methods for Measurement of Electroacoustical Characteristics of Hearing Aids," loc. cit. intensities available from the hearing aid may be dangerous to the ear.1 From this analysis it was hoped that information regarding the reliability of the electroacoustical responses of current hearing aids could be evaluated. With this goal in mind, the following null hypotheses have been formulated: 1. There is no difference in the mean coupler SPL between hearing aids I and II of the eight models as measured by the following criteria: (a) ASA Basic Frequency Responses (b) ASA Saturation Output Responses The differences in coupler SPL between the first and second hearing aids of the same model do not vary as a function of auditory frequency (There is no aids-by-frequency interaction.) as measured by the following criteria: (a) ASA Basic Frequency Responses (b) ASA Saturation Output Responses The differences in coupler SPL--taken over all frequencies--between the first and second hear- ing aids is the same for all models investigated (There is no aids-by-models interaction.) as measured by the following criteria: (a) ASA Basic Frequency Responses (b) ASA Saturation Output Responses 1Ibid., p. 12. Importance of the Study This study is considered to be important in that it will provide the practicing clinical audiologist with information that will help to guide his daily decisions concerning the recommendations of hearing aids to his cli— ents. This study will also hopefully add to the badly needed research regarding the reliability of current transistorized hearing aids. The need for clinically practical research is supported by the lack of informa- tion in the literature as evidenced in the literature review of this study. If a client is to gain the value of the clinical findings of a hearing aid evaluation, it is important that the instrument he purchases does not differ significantly from the instrument which was determined to be helpful during the evaluation. The primary importance of this study is to con- tribute up-to-date research findings which are of prac- tical value to the practicing audiologist in the field. This study intends to provide information which would be useful to the clinical audiologist concerning the relia- bility of the transistorized hearing aid. Organization of Study_ Chapter One contains the statement of the problem that led to this study. An introduction to the topic and an outline of the purposes of the study are included. The terms to be used are defined. Hypotheses are stated and the importance of the study is discussed. Chapter Two contains a review of the literature pertinent to this topic, and Chapter Three consists of a discussion of the equipment and the procedures utilized in this study. Chapter Four discusses the results of the study and Chapter Five contains the summary and conclusions. CHAPTER II REVIEW OF THE LITERATURE Trend in Hearing Aid Sales A review of the literature has revealed an ever increasing trend on the part of the consumer to purchase fewer and fewer body-worn aids in favor of ear-level instru— ments. With this demand by the consumer, and the interest of the manufacturer to please the consumer market, the im- provements in the miniaturization of hearing aids are clearly reflected by the increased percentage of ear-level hearing aids being purchased each year. The trend toward the smaller ear—level instruments was reported at the 1961 HAIC meeting: In 1959, conventional aids accounted for 27% of total sales. In 1960 this figure fell to 26% and in 1961 to approximately 23%. Eye-glass aids have dropped in this three year period from 20% of sales in 1959 to 41% in 1960 and 40% in 1961. Behind-the-ear aids have gained from 23% of sales in 1959 to 33% in 1960, and an estimated 35% in 1961.1 The further increase in the popularity of the smaller ear-level instruments was reported in 1962 by Bolstenin: A percentage comparison of types of hearing aids sold during 1961 and 1962 is given. 1962 l"HAIC Meeting Notes Progress," Hearing Dealer, 11 (November, 1961), 15. 10 11 sales included conventional aids, 22%; eyeglass aids, 34.67% behind-the-ear aids, 39.33% and in- the-ear aids, 4%.1 In 1963 the National Hearing Aid Journal reported an estimate of the annual figures for the hearing aid in- dustry for 1964. A still greater increase in the sales of ear-level hearing aids, and particularly behind-the-ear instruments is indicated: "It is estimated that of the aids sold, 34% will be eyeglasses, 45% behind—the-ear, 19% conventional, and 2% in-the-ear."2 Response Measurements of Vacuum Tube Aids The literature has revealed a number of studies prior to 1953 which are directly concerned with the elec- troacoustical responses of hearing aids. Hudgins3 and Hector et al.4 reported studies which were concerned with the electroacoustical responses of vacuum tube hearing aids. However, these studies are obsolete as a result of the cur— rent use of transistors in hearing aids. "Transistors took the place of bulky and fragile vacuum tubes in hearing aids lM. Bolstenin, "From the Editors Desk," National Hearing Aid Journal, 16 (February, 1962), 3. 2"Ninth Annual Facts and Figures," National Hear- ing Aid Journal, 17 (January, 1963), 6. 3C. V. Hudgins, "Testing the Performance of Hear- ing Aids," Volta Review, 49 (1947), l. 4G. L. Hector et a1., "Recent Advances in Hearing Aids,“ The Journal of the Acoustical Society_of America, 12 in January of 1953."1 Indirectly Related Studies A review of the literature has clearly indicated an absence of studies which are directly concerned with the electroacoustical response characteristics of current transistorized hearing aids. In 1962 Goldberg reported a new electronic device which evaluates the acoustic properties of a hearing aid. The acoustic properties which can be evaluated with this device are frequency responses, maximum acoustic gain, maximum output, and harmonic distortion. The Electronic Ear is considerably different from the many components which were combined to obtain the measurements for this study. The complete device is about the size of a port- able audiometer. The Electronic Ear consists basically of a ceramic microphone set into a six cubic centimeter coupler, and a calibrated amplifier. The coupler may be converted to two cc by inserting a miniature receiver adapter plug. Additional adapters are provided to make measurements of sub-miniature 2 receivers and those receivers using plastic tubes. Goldberg did not report the results of any measurements with this instrument. The following study by Shore, Bilger and Hirsh is l"Short Shorts About Hearing Aids," The Hearing Dealer, 6 (December, 1956), 18. 2Hyman Goldberg, "The Electronic Ear," Audecibel, 11 (May, 1962), 21. 13 not directly concerned with the reliability of the electro- acoustical responses of hearing aids; however, they did recognize the possibility of electroacoustical fluctuation. They employed fifteen clinical patients and four popular hearing aids which they set at a variety of adjustments. Their purpose was to test the reliability of repeated meas— urements of gain and speech discrimination in noise and in quiet. These tests were performed on four different days. They concluded that the reliability of these measures is not sufficient to justify the investment of a large amount of clinical time in selecting hearing aids. The following statement is evidence of their awareness of possible fluc— tuation in the electroacoustical responses of the aids on different trials: This conclusion does not imply that there are no differences among conventional monaural aids; but rather it suggests that whatever differences there might be are not detectable by these three usual measures of speech audiometry. The following study by McConnell, Silber and McDonald in 1960 is again not directly concerned with the reliability of the electroacoustical responses of hearing aids. How- ever, the results rendered might well be interpreted in the interest of this present study. They conducted a study to determine the test-retest reliability of speech audiometry 1Irvin Shore, Robert C. Bilger, and Ira J. Hirsh, "Hearing Aid Evaluations: Reliability of Repeated Measure— ments," Journal of Speech and Hearing Disorders, 25 (1960), 112. 14 measures with randomly selected hearing aid users. Speech discrimination scores were found to have a markedly high degree of test-retest reliability even when obtained by different clinicians. Aided speech reception thresholds were less consistent on repeated measures. No significant difference was found between test results from both types of measures when the tests were administered by dif- ferent clinicians.l It is interesting to note that the results of the aided tests were less consistent than the unaided results. This might possibly indicate variation as a function of the electroacoustical characteristics of the hearing aids. It is apparent from the high test-retest reliabil- ity found in this study that none of the many variables operating was sufficient to affect the results of the re- peated measures. One of the variables of this study was, of course, the electroacoustical responses of the hearing aids. However, there was apparently no significant differ- ence in the reliability of the electroacoustical responses of the instruments used in this study. If a significant difference had occurred, the test-retest reliability would have been affected. While the "Harvard Study," reported by Davis in 1947,2 is not primarily concetned with the reliability of 1F. McConnell, s. F. Silber, and D. McDonald, "Test- Retest Consistency of Clinical Hearing Aid Tests," Journal of Speech and Hearing Disorders, 25 (1960), 112. 2Hallowell Davis et a1., "Hearing Aids: An Experi- mental Study of Design Objectives" (Cambridge, Massachusetts: Harvard University Press, 1947). 15 the electroacoustical responses of hearing aids, it repre- sents an awareness of the need to provide the practicing clinicians with more information than is presented by the manufacturers. It also represents an attempt to help the practicing clinician more effectively evaluate his testing procedures in light of the effectiveness of the equipment he is evaluating. This review of the literature is evidence of the need for current research directly concerned with the elec- troacoustical characteristics of transistorized hearing aids. This writer searched the literature extensively and was unable to find any such study. CHAPTER III EQUIPMENT AND TESTING PROCEDURES Equipment Sixteen current clinically-used hearing aids were randomly selected for this study from the hearing aid stock of two clinical audiology facilities which are under the auspices of the Michigan State University Speech and Hear- ing Clinic. A sample of eight aids was randomly selected from the population of aids at one of the clinical facil- ities, then matched with eight aids of the same make and model from the other clinical facility. The sample for this study is, therefore, composed of two aids of each make and model selected for a total of sixteen aids. The instruments selected can be classified into two major categories with one of the categories subdivided into two types. The two major categories and two types are: I. Body-worn instruments II. Ear-level instruments A. Behind-the-ear models B. Eye-glass models The following hearing aids and equipment were em- ployed for the purposes of this investigation. 16 l7 Hearing Aids: I. II. Body-worn aids: Zenith Super Extended Range, serial number 5103113, External adjustments -- "C" setting, Receiver Y—S. Zenith Super Extended Range, serial number 5103108, External adjustments -- "C" setting, Receiver Y-S. Sonotone 300, Serial number 302852, Internal adjust— ments -- all left ”Normal LO—2," "Selector Switch" setting -- "Normal," Receiver 41.21. Sonotone 300, Serial number 319065, Internal adjust- ments —- all left "Normal LO-2," "Selector Switch" setting -- "Normal," Receiver 41.21. Ear-level instruments: A. Behind—the-ear instruments: Beltone Jubilee, Serial number 809556, Internal adjustments -- "clear dot setting," Acoustic tubing #16 gauge. Beltone Jubilee, Serial number 812567, Internal adjustments —— "clear dot setting," Acoustic tubing #16 gauge. Zenith Delegate B, Serial number B4170, In- ternal adjustments -- None, Acoustic tubing #16 gauge. Zenith Delegate B, Serial number B4164, In- ternal adjustments —- None, Acoustic tubing #16 gauge. Radioear 891, Serial number IRA98, Internal adjustments -- Normal, Acoustic tubing #16 gauge. Radioear 891, Serial number IVD94, Internal adjustments -— Normal, Acoustic tubing #16 gauge. Sonotone 55, Serial number 77351, Internal adjustments -- None, Acoustic tubing #16 gauge. Sonotone 55, Serial number 89752, Internal adjustments -- None, Acoustic tubing #16 gauge. 18 B. Eye-glass instruments: Sonotone 75, Serial number 704212, Internal adjustments -- None, Acoustic tubing #16 gauge. Sonotone 75, Serial number 704194, Internal adjustments -— None, Acoustic tubing #16 gauge. Zenith Z-20, Serial number 7203217, Internal adjustments -- None, Acoustic tubing #16 gauge. Zenith Z-20, Serial number 7202928, Internal adjustments -- None, Acoustic tubing #16 gauge. Equipment: Hearing aid test box, Bruel and Kjaer (type 4214). Precision sound level meter, Bruel and Kjaer (type 2203). Audio frequency spectrometer, Bruel and Kjaer (type 2112). Pistonphone, Bruel and Kjaer (type 4220). Low frequency audio oscillator, Hewlett-Packard (type 202C). Condenser microphone, one inch/pressure, Bruel and Kjaer (type 4132). Condenser microphone, one inch/pressure, Bruel and Kjaer (type 4132). Cathode follower, Bruel and Kjaer (type 2613). Cathode follower cable, Bruel and Kjaer (type AO—0033). Acoustic coupler, 2cc., Bruel and Kjaer (type DB-0138). Amplifier, Ampex (model 620). See Figure 1 for a block diagram of the equipment arrange- ment. 19 xom heme ee< mceummm .oa AMODHCOE UHmHMIUCDOmV MCOSQOHUHE ummcmpcou umHQSou 00 N maozaouuaz Homcmpcou Hmzoaaom moonumu mabmu umzoaaom mpocpmu umbmeoubumam mucmdqmum OHUS< hobo: Hm>malpczom umemeans< uoumHHHumo MucmsquMIzoq .m .m .h .m .m .v .m .N .H oCOHflMDHMNrM mmcommmm UH< Mom #cmEmocmuum EMHOMHU xuonII.H mudmflm .w .h .m OOH 20 Procedures Response Test Procedures: The procedures followed for obtaining the Basic Frequency Responses and Saturation Output Responses for hearing aids are those outlined by the American Standards . . 1 Assoc1ation. Placement of Aid in Test Box: The reference point for the hearing aids was at the center of the microphone grill area. The hearing aid was placed with the reference point located toward the sound source so that the direction of the incident sound was perpendicular to the surface of the reference point. The instruments were placed in the same location within the test box. The microphone of the aid being tested was on the same plane as the sound-field monitor- ing microphone at a distance of one inch apart. Standardization of Test Procedures: All test procedures were standardized so that the replication of events would occur under the same condi- tions. Each instrument was evaluated on three distinct different trials for the desired data. The different trials occurred on different days. 1"American Standard Methods for Measurement of Electroacoustical Characteristics of Hearing Aids," Amer- ican Standards Association Bulletin, 53.3 (1960), 10—13. 21 Batteries: The batteries were tested to be certain of the de- sired power before each aid was evaluated. The batteries used in the various aids were those recommended by the manufacturer. Acoustic Tubing: The gauge of the acoustic tubing used on all ear- level instruments was that recommended by the manufacturer for maximum gain and frequency response range. In all cases, for the aids used in this study, number sixteen gauge tub- ing was used. The length of the acoustic tubing was stand- ardized to 2 inches. Internal and External Aid Adjustments: On those instruments which had either internal or external adjustments which controlled the range of fre— quencies or gain responses, the adjustments were made that would produce maximum gain and maximum frequency response range according to the manufacturers' specification bro— chures. In all cases the particular adjustments or set- tings have been stated. Basic Frequency Response Procedures: The basic frequency response characteristics for each of the sixteen aids were obtained on three distinct trials for each of the six frequencies. The desired data were obtained in the following manner: 22 The free field sound pressure level in the hearing aid test box was adjusted to read 60 dB 1 1 dB at the fre— quency of 1000 cps. The gain control of the instrument under evaluation was then adjusted so as to give a sound pressure level in the coupler of 100 dB 1 2 dB at 1000 cps. The aids which did not have sufficient gain to permit this adjustment were set at maximum gain with the volume control in the full-on position. All of the ear-level instruments used in this study were set at maximum gain. The frequencies of the sound source used were: 250, 500, 1000, 2000, 3500, and 4000 cps. Following the 100 dB 1 2 dB coupler SPL measurement at 1000 cps., the output in dB coupler SPL of the other five frequencies were also measured and recorded on each of the three dis- tinct trials. Saturation Output Responses: The saturation output responses for each of the sixteen aids were measured in dB coupler SPL on three dis- tinct trials for each of the six frequencies in the fol- lowing manner: The gain control was turned to the full-on posi- tion. The saturation output measurements were recorded for the selected six frequencies of 250, 500, 1000, 2000, 3500, and 4000 cps. This was done by increasing the coup— ler sound pressure level until the instrument reached its maximum pressure, a point at which the maximum reading 23 fell with a further increase of input sound pressure. The maximum reading was recorded as the saturation output re- sponse for that particular frequency. Summary In order to study the Basic Frequency Response and Saturation Output Response Characteristics of sixteen clin— ically-used hearing aids, the following procedures were employed. Eight hearing aids were randomly selected and matched with eight more aids of the same make, model, and setting. Therefore, two aids of each make and model of the original selection made up the sample. The Basic Frequency Responses and Saturation Output Responses were measured in dB coupler SPL on three distinct trials for each of the six following frequencies: 250, 500, 1000, 2000, 3500, and 4000 cps. All procedures were standardized according to standards set forth by the American Standards Association. CHAPTER IV RESULTS AND DISCUSSION Results The data for this study were obtained by measuring the electroacoustical response characteristics of sixteen current hearing aids for Basic Frequency Responses and Saturation Output Responses as prescribed by American Standards Association Bulletin 33.3 (1960).1 The null hypotheses under consideration are: 1. There is no difference in the mean coupler SPL between Hearing Aids I and II of the eight models as measured by the following criteria: (a) ASA Basic Frequency Responses (b) ASA Saturation Output Responses 2. The differences in coupler SPL between the first and second hearing aids of the same model do not vary as a function of auditory frequency (There is no aids—by-frequency inter- action.) as measured by the following criteria: (a) ASA Basic Frequency Responses 1"Methods for Measurement of Electroacoustical Characteristics of Hearing Aids," American Standards As- sociation Bulletin 53.3 (1960), p. 7. 24 25 (b) ASA Saturation Output Responses 3. The differences in coupler SPL-~taken over all frequencies—-between the first and second hear— ing aids is the same for all models investigated (There is no aids-by—models interaction.) as measured by the following criteria: (a) ASA Basic Frequency Responses (b) ASA Saturation Output Responses Appendix A contains the raw data obtained from the Basic Frequency Response measurements of the hearing aids. Appendix B contains the raw data obtained from the Satura- tion Output Response measurements of the hearing aids. Since it was the desire of this investigator to determine if any significant difference and/or interaction existed between the aids, models, and frequencies, a 2 x 8 x 6 factorial analysis of variance was employed. The design employed is described by Edwards.1 The statistical anal— yses of this study were computed by the CDC 3600 computer at the Michigan State University Computer Laboratory. The computer routine employed for the analysis was the Option 2 of FACREP.2 The statistical analysis of the results of the lAllen L. Edwards, Experimental Degign in Psycho- logical Research (New York: Rinehart and Company, Inc., 1960), pp. 201-205. 2D. F. Kiel, A. F. Kenworthy, and W. L. Rubel, "Analysis of Variance Routines” (East Lansing: Michigan State University, September 30, 1963), p. 23. 26 Basic Frequency Response measurements is summarized in Table I. The statistical analysis of the results of the Saturation Output Response measurements is summarized in Table II. There is a significant difference in the mean coup- ler SPL between Hearing Aids I and II of the eight models as measured by ASA.Basic Frequency Responses at the .01 level of confidence. The F attained by this analysis was 19.8. This indicates that part (a) of the first hypothesis can be rejected at better than the .01 level of confidence. There is a significant difference in the mean coup- ler SPL between Hearing Aids I and II of the eight models as measured by ASA Saturation Output Responses at the .01 level of confidence. The F attained by this analysis was 36.79. This indicates that part (b) of the first hypoth- esis can be rejected at better than the .01 level of con- fidence. The second null hypothesis states that the differ- ence in coupler SPL between the first and second hearing aids of the same model do not vary as a function of audi— tory frequency (There is no aidseby—frequency interaction.) as measured by ASA Basic Frequency Responses. An examina- tion of Table I indicates that part (a) of the second hy— pothesis can be rejected at better than the .01 level of confidence. The F attained by this analysis was 4.88. An examination of Table II indicates that part (b) of the second hypothesis was not rejected at the .01 level 27 TABLE I SUMMARY OF ANALYSIS OF VARIANCE FOR BASIC FREQUENCY RESPONSES Sources of Variance Sum of Squares df Mean Square F Aids (A) 63.28 1 63.28 19.80‘ Models (B) 6899.41 7 985.63 - A x B 202.52 7 28.93 9.05‘ Frequencies (C) 25686.72 5 5137.34 - A X C 77.94 5 15.58 4.88‘ B x C 11847.19 35 338.49 - A x B x C 1002.07 35 28.63 - Error (Within Treatments) 613.33 192 3.19 Total 46392.49 287 ‘Significant below the .01 level of confidence. 28 TABLE II SUMMARY OF ANALYSIS OF VARIANCE FOR SATURATION OUTPUT RESPONSES Sources of Variance Sum of Squares df Mean Square F Aids (A) 96.83 1 96.83 36.79‘ Models (B) 20492.10 7 2927.44 - A x B 1498.02 7 214.00 81.31’ Frequencies (C) 7239.22 5 1447.84 — A x C 15.05 5 3.01 1.14 B x C 3014.07 35 86.11 - A x B x C 127.57 35 3.64 - Error (Within Treatments) 505.33 192 2.63 Total 32988.24 287 ‘Significant below the .01 level of confidence. 29 of confidence. The F attained by this analysis was 1.14. This indicates that no significant difference in the coup- ler SPL was demonstrated between the first and second hear- ing aids of the same model as a function of auditory fre- quency as measured by ASA Saturation Output Responses. The third hypothesis states that the differences in coupler SPL--taken over all frequencies-~between the first and second hearing aids is the same for all models investigated (There is no aids-by-models interaction.) as measured by (a) ASA Basic Frequency Responses and (b) ASA Saturation Output Responses. An examination of Table I indicates that there is a significant difference in the third hypothesis as meas- ured by part (a) at the .01 level of confidence. The F attained by this analysis was 9.05. This indicates that there is a significant aids-by-models interaction as meas- ured by ASA Basic Frequency Responses. An examination of Table II indicates that there is a significant aids-by-models interaction as measured by ASA Saturation Output Responses. The F attained by this analysis was 81.31. This indicates that part (b) of the third hypothesis can be rejected at better than the .01 level of confidence. Discussion The analysis of the data has shown that parts (a) and (b) of the first hypothesis were rejected. This 30 hypothesis concerned the reliability of the mean coupler SPL between Aids I and II of the eight models as measured by ASA Basic Frequency Responses (part (a) of the first hypothesis). Although the study has shown statistical sig— nificance, as evidenced in Table I, observation of Figure 2 reveals that the mean coupler SPL difference between Aids I and II is only 1 dB. The 1 dB difference, while statistic- ally significant, is not significant in its practical or clinical application. The fact that 1 dB is not significant in its practical or clinical application is supported by ASA Bulletin 53.3,1 a 1 1 dB for error in the free—field sound pressure level, and a i 2 dB error in the gain adjust— ment of hearing aids as stated in the procedures for the elec- troacoustical measurements of hearing aids. Therefore, the ASA allowed error could exceed the 1 dB difference observed in the results of part (a) of the first hypothesis. Part (b) of the first hypothesis was also shown to be statistically significant. This would indicate that there is a significant difference in the reliability of the mean coupler SPL between hearing aids I and II of the eight models as measured by ASA Saturation Output Responses. An observation of Table II will confirm the significant F. However, an observation of Figure 3 reveals a mean coupler SPL difference between Aids I and II of only 1 dB. Although 1"Methods for Measurement of Electroacoustical Characteristics of Hearing Aids," op. cit., p. 12. dB coupler SPL 90 85 80 75 31 IIH lllllllll Aids Figure 2.--Mean coupler SPL of Aids I and II for the frequencies of 250, 500, 1000, 2000, 3500, and 4000 cps, as measured by ASA Basic Frequency Responses. dB coupler SPL 120 —— 115 - 110 - 105 32 4. llllllllllllll] I II Aids Figure 3.-—Mean coupler SPL of Aids I and II for the frequencies of 250, 500, 1000, 2000, 3500, and 4000 cps, as measured by ASA Satura- tion Output Responses. 33 the mean coupler SPL difference between Aids I and II re- veals a statistical significance, there would seem to be no significant difference with regard for the practical or clinical application. The analysis of the data has shown significance in part (a) of the second hypothesis and a lack of signif- icance in part (b) of the second hypothesis. The results of part (a) indicate a significant difference in the mean coupler SPL between the first and second hearing aids of the same model as a function of auditory frequency (an aids- by—frequencies interaction) as measured by ASA Basic Fre- quency Responses. An inspection of Table I will confirm the significant F of 4.88. Figure 4 graphically demonstrates the aids by frequencies interaction of part (a) of this hypothesis. The higher frequency range of 2000 through 4000 cps seems to be more reliable in Basic Frequency Re- sponses than the lower frequency range of 250 through 1000 cps. Although the statistical analysis of those data clearly demonstrates significance, there appears to be no significance in terms of the clinical application of the results. The greatest aids by frequency interaction is at the frequency of 1000 cps. As Figure 4 indicates, this difference is only 2 dB between the mean coupler SPL of Aids I and II. Reference to Table II will indicate a lack of dB coupler SPL 105 100 95 90 85 80 7'5 70 34 IHI IITF lHl [Ill Illl Hlf IIII 250 500 1000 2000 3500 4000 Frequencies Aids x Frequencies Interaction Figure 4.--Mean frequency responses for Aids I and II of each model for Basic Frequency Responses. 35 significance in part (b) of the second hypothesis for the aids by frequencies interaction, as measured by ASA Satura- tion Output Responses, with an F of 1.14. Therefore, on the basis of this statistical analysis, it cannot be said that there is an aids-by-frequency interaction as measured by ASA Saturation Output Responses. The results of these measurements are graphically presented in Figure 5. The analysis of the data for the third hypothesis has shown statistical significance for parts (a) and (b) of this hypothesis; thereby rejecting the hypothesis that there is no aids—by-models interaction as measured by: (a) ASA Basic Frequency Responses and (b) ASA Saturation Output Responses. An inspection of Table I will reveal a significant F of 9.05 for part (a) of this hypothesis. Although statistical significance has been demonstrated, there again appears to be no significant difference with regard to the clinical application of these data. Figure 6 graphiCally presents the aids-by-models interaction. Model H, which shows a 5 dB difference between Aids I and II, is the only model which comes near demonstrating a clinically-significant aids-by-models interaction. An inspection of Table II will reveal a statistically signif- icant F of 81.31 for the aids-by-models interaction as measured by ASA Saturation Output Responses. The graphic presentation of the data in Figure 7 reveals the aids-by- models interaction. It can be seen that Models A, C, D, 125 120 H H U1 dB coupler SPL H H o 105 100 36 *r— 7_ \ ‘_ \ ._ \ \. _ d) i_. __ x = Aid I _ O = Aid II 1.— 250 500 1000 2000 3500 4000 Frequencies Aids x Frequencies Interaction Figure 5.-—Mean frequency responses for Aids I and II of each model for Saturation Output Responses. 37 .mmmcommwm hucmsgmum Uflmmm o HH pom H mpfl< mo 4mm Hmaasoo cmmzll.m musmflm coapumumDCH.mHmpoz x mow< \\ a \ .\ k _\ mm m. mud \ x .\ x x m; \. . . r- \ \ RI... IHI a. .. a m. w v _ . Mil mm \ HH 3a. H 34 IWWI 2: u- m dB coupler SPL 38 .mmmcoommm psauso COHDMHSpmm «m4 ha Umuzmmwe mm mHmUOE unwed on» mo some now mmflucwsgwnm Ham Hm>o HH ocm H mUH< mo 4mm umHQSOU ammZII.e madmflm coapumumDCH mambo: x mUH< mamooz Q U :1: (D m \ _\_r I‘ A A; \\\\\ \ \\\\ \\\ \\ \\‘\ \\ m ‘ A V ‘\\\\\\\\\ \ \ l\\\ HH ea< H ma< \ ) OHH mHH ONH mNH omH ova dB coupler SPL 39 E, and H, while having contributed to the significant sta- tistical interaction, have not individually demonstrated an interaction which has clinical significance. That is, Aids I and II of these five models do not demonstrate a clinically-significant aids4by—models interaction. It can also be seen that Aids I and II of Model B differ by 6 dB. Aids I and II of Model G show the greatest aids-by—models interaction with a mean coupler SPL difference of 10 dB. The 10 dB difference observed, if not compensated for by the audiologist, could conceivably be sufficient to cause undue discomfort, pain, or damage to a client. Therefore, it is important that the practicing clinical audiologist be aware of the possible fluctuation of the electroacoustical response characteristics of hearing aids. It is interesting to note that while statistical significance has been clearly demonstrated for four of the sources of variability tested, only one of the sources has also demonstrated clinical significance. The demonstration of statistical significance without clinical significance can be accounted for by the measuring system that was em- ployed. The measuring system employed was more precise and repeatable than the instruments being measured. The data would seem to support the view that the precision of the instruments being measured was, in the main, adequate for clinical use. CHAPTER V SUMMARY AND CONCLUSIONS Summary During recent years the increased technological and scientific advances in amplification systems, and par~ ticularly the advent of the transistor replacing the vacuum tube, has led to the continuing miniaturization of wearable hearing aids. "Transistors took the place of bulky and fragile vacuum tubes in hearing aids in January of 1953."1 The continuing miniaturization has created questions with regard for the reliability of the electroacoustical responses of the aids. A review of the literature has revealed lit- erally no studies which are directly concerned with the reliability of the electroacoustical response characteris- tics of transistorized hearing aids. Hudgins (1947)2 and Hector g£_§1. (1953)3 reported data they had collected on the electroacoustical response characteristics of vacuum tube hearing aids. However, this information is presently l"Short Shorts About Hearing Aids," The Hearing Dealer, 6 (December, 1956), 18. 2C. V. Hudgins, "Testing the Performance of Hear- ing Aids," Volta Review, 49 (1947), 1. 3G. L. Hector et a1., "Recent Advances in Hearing Aids," The Journal of the Acoustical Society of America, 25 (1953), 1189-94. 40 41 obsolete as a result of the current use of transistors. The following questions are a result of the increas- ing miniaturization of wearable hearing aids. Are these very small amplification systems reliable from model to model and trial to trial? Will the response characteris- tics of two different aids of the same make and model differ significantly? In other words, will the instrument the client purchases differ significantly from the instrument which performed successfully during the hearing aid eval- uation? The purpose of this study was to analyze and com— ‘ pare the results obtained from the electroacoustical re- sponse measurements of sixteen clinically—used hearing aids so as to determine their reliability on repeated trials. The criteria employed for these response measurements was the ASA Basic Frequency Responses, and the ASA Saturation Output Responses. The instrumentation for this study consisted of: one Bruel and Kjaer hearing aid test box, one Bruel and Kjaer sound-level meter, one Bruel and Kjaer audio fre- quency spectrometer, two Bruel and Kjaer one-inch condenser microphones, one Bruel and Kjaer cathode follower, one Bruel and Kjaer 2 cc acoustic coupler, one Hewlett-Packard low-frequency audio oscillator, and one Ampex amplifier. All test procedures, including the placement of the hearing aids in the sound-field, the sound-field pressure 42 level in the test box, the gain adjustment of the aids, etc., were performed in accordance with procedures outlined in the American Standards Association Bulletin $3.3.l Each aid was subjected to six trials at each of the six frequen— cies: three trials for the basic frequency response meas- urements and three trials for saturation output response measurements. The data, in the form of dB coupler SPL, were recorded for statistical analyses. For the analyses of this study a 2 x 8 x 6 factorial analysis of variance procedure was employed. The analyses were computed by the CDC 3600 computer at the Michigan State University Computer Laboratory. The results of the analyses demonstrated that the main effect and the two interactions under considera- tion for the basic frequency responses were statistically significant below the .01 level of confidence. It further demonstrated that of the one main effect and the two inter- actions under consideration for the saturation output re- sponses, only one interaction (aids x frequencies) did not reveal a statistically-significant F statistic below the .01 level of confidence. Conclusions Statistical1y-significant results were obtained for the main effect (aids) and the two interactions (aids x models and aids x frequencies), for the ASA Basic Frequency 1"Methods for Measurements of Electroacoustical Characteristics of Hearing Aids," 0 . cit., p. 7. 43 Response criteria. Although the F statistics attained clearly indicate rejection of the null hypotheses at better than the .01 level of confidence, there appear to be no clinically-significant differences as defined in this study between Aids I and II for the stated criteria. Statistically~significant differences between Aids I and II were also attained for the main effect (aids) and the interaction (aids x models) for the ASA Saturation Out- put Response criteria. The second interaction (aids x fre- quencies) did not reveal statistically-significant differ- ences at the .01 level of confidence. While the main ef- fect (aids) and the interaction (aids x models) have clearly demonstrated statistically-significant F ratios, only two models (F and G) of the interaction (aids x models) have indicated differences between Aids I and II which have clinical significance as defined in this study. The recurrent demonstration of statistical signif- icance without clinical significance can be accounted for by the measuring system that was employed. The measuring system employed was more precise and repeatable than the instruments being measured. This condition existed despite the fact that the instruments being measured appeared to be highly consistent from a clinical point of view. BIBLIOGRAPHY Books Davis, Hallowell, and Silverman, Richard S. Hearinggand Deafness. New York: Holt, Rinehart and Winston, Inc., 1960. Edwards, Allen L. Experimental Design in Psychological Research. New York: Holt, Rinehart and Winston, Inc., 1960. Newby, Hayes A. Audiology. New York: Appleton~Century~ Crofts, Inc., 1964. Articles and Periodicals Bolstenin, M. "From the Editors Desk," National Hearing Aid Journal, 17 (February, 1962), 3. Goldberg, Hyman. "The Electronic Ear," Audecibel, 11 (May, 1962), 21. "HAIC Meeting Notes Progress," Hearing Dealer, 11 (November, 1961), 15. Hector, Grant L., Pearson, Harry A., Dean, Neal J., and Carlisle, Richard W. "Recent Advances in Hearing Aids," The Journal of the Acoustical Society_of America, 25 (November, 1953), 1189—94. Hudgins, C. V. "Testing the Performance of Hearing Aids," Volta Review, 49 (1947), l. Lybarger, S. F. "Standardized Hearing Aid Measurements," Audecibel, 10 (February, 1961), 7m8. McConnell, F., Silber, E. F., and McDonald, D. "Test- Retest Consistency of Clinical Hearing Aid Tests," Journal of Speech and Hearing Disorders, 25 (1960), 273-80. "Methods for Measurement of Electroacoustical Characteris- tics of Hearing Aids," American Standards Associa- tion Bulletin, 53.3 (1960), 7w15. 44 45 "Ninth Annual Facts and Figures,” National Hearipg Aid Journal, 17 (January, 1963), 6. Shore, Irvin, Bilger, Robert C., and Hirsh, Ira J. "Hear- ing Aid Evaluation: Reliability of Repeated Meas- urements," Journal of Speech and Hearing Disorders, 25 (1960), 152-70. "Short Shorts About Hearing Aids," The Hearing Dealer, 6 (December, 1956), 18. Unpublished Materials Kiel, D. F., Kenworthy, A. F., and Rubel, W. L. "Analysis of Variance Routines" (East Lansing: Michigan State University, September 30, 1963), 32. Walton, Constance Rae. "Discrimination by Normally Hearing Subjects for Filtered Speech Under Conditions of Hearing Aid Amplification." Unpublished Master's Thesis, College of Communication Arts, Department of Speech, Michigan State University, 1964. APPENDICES APPENDIX A 48 RAW DATA: BASIC FREQUENCY RESPONSES m m Frequencies E '3 250 500 1000 2000 3500 4000 g 3 Aids Aids Aids Aids Aids Aids 2 6+ I II I II I II I II I_—II I II A: 1 78 65 86 84 98 100 100 101 93 95 88 89 2 72 67 85 84 101 102 98 101 92 94 88 88 3 68 69 86 84 102 102 99 102 94 95 88 89 B: 1 68 64 88 87 96 94 106 105 100 102 101 104 2 63 65 83 86 93 97 104 110 102 106 92 94 3 76 70 86 86 94 100 106 112 100 109 101 96 c: 1 63 62 72 74 92 94 107 107 89 92 83 90 2 63 60 74 72 92 92 106 103 89 93 86 85 3 63 63 74 73 92 94 106 105 89 92 88 89 0: 1 74 61 78 80 86 89 103 100 84 84 81 88 2 78 62 77 80 84 90 100 100 80 84 84 89 3 76 60 78 82 86 89 102 100 80 82 82 87 E: 1 82 84 105 104 101 101 108 110 106 110 101 101 2 83 84 104 104 100 101 106 109 106 110 101 101 3 78 84 103 103 100 102 108 109 105 109 102 101 F: 1 70 76 78 87 92 100 110 106 101 94 88 86 2 70 67 78 85 92 98 110 104 100 94 88 87 3 70 66 78 86 92 100 110 105 100 94 87 87 G: 1 80 80 98 99 100 100 102 102 90 91 78 78 2 80 82 98 98 100 100 103 100 90 89 80 76 3 81 83 99 98 100 100 103 100 92 88 80 76 H: 1 81 96 88 94 100 102 102 105 72 74 67 70 2 81 96 89 94 100 102 102 105 72 75 69 70 3 83 97 91 96 101 102 102 105 73 76 69 71 The make, model, through H listed above can be found in Appendix C. and serial number of Models A dB coupler SPL APPENDIX B SO RAW DATA: SATURATION OUTPUT RESPONSES m cm Frequencies '3 B 250 500 1000 2000 3500 4000 g '3 Aids Aids Aids Aids Aids Aids 2 E4 I II I II I II I II I II I II A: l 116 115 118 118 119 120 116 117 112 116 107 108 2 115 116 119 120 122 122 116 118 114 116 108 109 3 115 115 120 120 123 122 117 118 114 116 108 109 B: 1 115 117 112 118 114 117 111 115 113 116 107 113 2 109 117 107 116 110 118 108 118 113 120 106 103 3 115 118 112 116 112 121 111 120 112 120 106 103 C: 1 112 112 112 112 116 116 118 118 104 106 102 103 2 111 112 112 112 116 116 118 116 105 109 101 100 3 113 112 112 112 116 116 118 118 104 107 102 103 D: 1 114 116 116 116 113 114 112 112 100 90 116 101 ’2 112 116 116 116 112 114 112 112 99 99 100 101 3 113 117 116 116 113 114 112 112 98 100 100 101 E: 1 120 120 123 123 120 120 116 116 118 120 113 114 2 120 121 123 124 120 121 116 116 117 120 114 114 3 121 120 123 123 120 121 116 116 117 120 113 114 F: 1 125 120 124 120 130 123 129 122 124 120 116 110 2 125 121 124 120 129 124 129 122 125 120 115 110 3 125 120 125 120 130 125 129 122 124 120 115 111 G: 1 147 133 146 133 144 133 144 130 136 126 126 116 2 147 134 146 134 143 134 143 133 136 127 125 116 3 146 134 146 134 144 133 142 132 135 126 125 116 H: l 138 138 138 138 138 138 131 131 120 120 112 112 2 138 138 138 138 138 139 132 131 121 120 113 112 3 138 138 139 138 139 139' 132 131 121 120 113 112 The make, model, and serial number of Models A through H listed above can be found in Appendix Co dB coupler SPL APPENDIX C HEARING AIDS EMPLOYED 52 MODELS AIDS I AIDS II A: Sonotone 75 #704212 #704194 B: Beltone Jubilee #809556 #812567 c: Zenith Delegate B #B4170 #B4164 D: Zenith Z-20 #7203217 #7202928 B: Radioear 891 #IRA98 #IVD94 F: Sonotone 55 #77351 #89752 G: Zenith Super Extended Range #5103113 #5103108 H: Sonotone 300 #302852 #319065 MICHIGAN STATE UNIVERSITY LI BRARIES 7 3 1293 ([3168 839