ABS TRAC T SPEECH DISCRIMINATION AND RESPONSE LATENCY OF NORMAL HEARING AND HEARING-IMPAIRED CHILDREN AS A FUNCTION OF TIME COMPRESSION by Jean Eileen Maki Response accuracy and latency of response were used in conjunction with time compression, first, to assess the use of time-compressed speech in evaluation of children exhibiting peripheral hearing impairment; and second, to compare speech processing skills of normal hearing and hearing-impaired children. Test lists from the Word Intelligibility by Picture Identification speech discrimination measure (Ross and Lerman, 1970) were processed at 0%, 30%, 40%, 50%, and 60% time compression. Test stimuli were presented at a 32 dB sensation level (re speech reception threshold) to 60 normal hearing and 21 hearing-impaired children. Each subject received five test lists in randomized order, each list processed at five time compression levels. Correct, error, and mean reaction time values and percent correct were calculated for each test list. In addition, a vocabulary pretest and a measure of simple reaction time were also administered. '1 III ‘r 1‘ .ulll' T. L c v .. C» Ad Q. AC 4 O N . . rs .Iu Q» A» .3 .3 nu. Jean Eileen Maki Variance analyses performed on speech discrimination and mean choice reaction time data of normal hearing children showed significant main effects for age and time compression for both measures. An analysis of variance performed on simple reaction time data also showed a signi- ficant main effect for age. No main effect was found for sex and no significant interactions occurred. Speech discrimination and reaction time data of hearing- impaired subjects were limited to descriptive analysis because of excessive heterogeneity of subjects. In addition, the number of subjects per age and hearing loss category was neither equal nor of sufficient size to allow statistical analysis. Since the factors of age and degree of hearing loss interacted, interpretation was difficult with the variable age and hearing loss characteristics. Grouping subjects on the basis of performance across time compression conditions indicated that the age factor probably had a greater effect on the performance of this hearing-impaired group than did degree of hearing loss. Results of this investigation were compared with previous studies and discussed relative to speech perception theory and application of time compression in audiological evaluation of hearing—impaired children. SPEECH DISCRIMINATION AND RESPONSE LATENCY OF NORMAL HEARING AND HEARING-IMPAIRED CHILDREN AS A FUNCTION OF TIME COMPRESSION by Jean Eileen Maki A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Audiology and Speech Sciences 1975 Accepted by the faculty of the Department of Audiology and Speech Sciences, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Degree of Doctor of Philosophy. Guidance Committee: ACKNOWLEDGEMENTS I would like to thank all those who contributed their time, talents and suggestions so that I could complete my dissertation within the set time limit. For their help and friendship, I'd like to offer my gratitude to the following people. To Dr. Daniel S. Beasley, Dr. John M. Hutchinson, Dr. Fred H. Bess, and Dr. Oscar I. Tosi, for your criticisms and suggestions concerning the dissertation content and organization, and for expressing, each in your own way, constant support and encouragement during the completion of this study. To Ms. Linda Seestedt and Dr. Fred Bess, for taking the responsibility and time to contact and schedule the children tested at Central Michigan University, and to the faculty and students who assisted us, altered routines and "rearranged the clinic" to meet our needs. Your help was very much appreciated as was the overall interest and enthusiasm which we experienced while at Central. To Jane Shoup, Kris King and Daun Beasley, for assisting in the testing of subjects; to Judy Frankmann, Ron Shoup, Toni Tryon and Ken Stonebrook for help in obtaining neces- sary equipment; to Jane Shoup, for serving as consultant, editor and typist; to Jim Mullin and Ron Shoup, for assistance ii n: m w‘o‘r- 1‘,“ with the data analysis; and to Randy Arman, for willingly accepting this typing responsibility and "seeing me through to the end." The enjoyment of any thesis involving children exists in the testing of subjects. This study was no different and grateful acknowledgement is offered to the children and their parents, many of whom brought with them an interest and willingness to help which made the task con- siderably easier. The "unique" circumstances surrounding this disserta- tion warrant special acknowledgements, therefore, I'd like to thank my committee members for being so incredibly flexible, and also Kris and Larry King, Jane and Ron Shoup and Dan Konkle for providing continuous support during the past year. In addition, I'd like to express special gratitude to my advisor, Daniel S. Beasley, for without his confidence and direction the doctorate never would have been attempted, much less achieved. And finally, during a time when many females were discouraged by their families from continuing in school, I'd like to thank my mother and brother for continually . expressing their pride in me and what I was doing. And the one "thank you" that is left goes to my mother, just for being the person she is. iii TABLE OF CONTENTS Page ACKNOWLEDGEMENTS.................................. ii LIST OF TABLES.................................... vii LIST OF FIGURES................................... ix Chapter I. INTRODUCTION............................... Time Compression in Audiological Assessment A Model of Speech Perception............... \OO\|-'l-’ Reaction Time Measurement.................. Statement of the Problem................... 12 III EXPERINENTAL PROCEDURES. I I I I I I I I I I I I I I I I I I I 15 Subjects................................... 15 Testing Environments....................... 16 Generation of Experimental Tapes........... 1? Instrumentation for Reaction Time Measurement........................... 21 Presentation Procedures.................... 23 Vocabulary Pretest.................... 23 Hearing Screening and SRT Determina- tion............................. 25 Speech Discrimination Testing......... 26 Simple Reaction Time.................. 27 DataAnalySiSIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 27 iv Chapter Page III. RESULTS.................................... 29 Normal Hearing Children.................... 30 Hearing-Impaired Children.................. 33 IV. DISCUSSION............... ...... ............ 39 Normal Hearing Children.................... 40 Speech Discrimination and Time Compression...................... 40 Response Latency and Time Compression. 46 Test-Retest Reliability............... 51 Hearing'lmpaired Childrene e o e e e e e e e e e e e e e e e 52 Speech Discrimination and Time compreSSionIIIIIIIIIIIIIIIIIIIIII 52 Response Latency and Time Compression. 62 Theory of Speech Perception and Memory..... 65 Clinical Application....................... 68 Implications for Future Research........... 69 LIST OF REFERENCES................................ 72 APPENDIX A. INDIVIDUAL SUBJECT INFORMATION FOR THE HEARING-IMPAIRED GROUP B. COMPOSITE AUDIOGRAM FOR HEARING-IMPAIRED SUBJECTS DIVIDED ACCORDING TO DEGREE OF HEARING LOSS C. CALIBRATION PROCEDURES D. SUBJECT INSTRUCTIONS E. TEST FORMS: SUBJECT INFORMATION AND RESPONSE SHEETS F. INDIVIDUAL SUBJECT TEST-RETEST DATA FOR 10 NORMAL HEARING CHILDREN V Appendix G. INDIVIDUAL DATA - NORMAL HEARING SUBJECTS H. INDIVIDUAL DATA - HEARING-IMPAIRED SUBJECTS vi Table 1. LIST OF TABLES Speech discrimination scores on the WIPI in mean percent correct for normal hearing children, ages 4, 6, and 8 years under each condition of time compression. Standard deviations for main effects are also shown................................ Mean total reaction time (RT ) in milli- seconds for normal hearin children, ages 4, 6, and 8 years under each condition of time compression, and mean simple reaction time (RT ) in milliseconds for each age group. Standard deviations for significant main effects are also Shame-cocoa...oeeeeeeeeeeeeeeeeeeeoeeeeeo Mean error reaction time (RT ), mean total reaction time (RTt)' and Dean correct reaction time (RT ) measured in milli- seconds for norma hearing children ages 4, 6, and 8, under each percentage of time compression.......................... Speech discrimination scores on the WIPI in mean percent correct for hearing- impaired children grouped according to degree of hearing loss. Also shown are number of subjects per group, and mean age, 3-frequency pure tone average (M PTA) and speech reception threshold (M SRT) for each hearing loss group............... Mean total reaction time (RTt) and mean simple reaction time (RTS) in milli- seconds for hearing-impaired children grouped according to degree of hearing lOSSIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII vii Page 31 32 3A 36 37 Table 6. Mean error reaction time (RTe), mean total reaction time (RTt), and mean correct reaction time (RTS) for each percentage of time compression for all hearing- impaired subjects combined............... Mean Speech discrimination scores on the WIPI from three investi ations of normal hearing children (ages , 6, and 8) tested under 3 conditions of time com- pression at 32 dB sensation level........ Results of three investigations showing differences in percent between mean speech discrimination scores of normal hearing children, ages 4 and 6 years and 6 and 8 years for 3 conditions of time compression......................... Comparison of hearing-impaired subjects and experimental procedures used in the Ross and Lerman (1970) study and the present investigation............................ viii Page 38 41 44 53 LIST OF FIGURES Perceptual and memory model taken from Norman and Rumelhart (1970)............. Timing of test sequence on track one and tone burst on track two plus 4 second silent interval for subject response.... Location of instrumentation, subject and experimenters for collection of response accuracy and reaction time data......... Percent correct scores (A) and mean reaction time in milliseconds (B) for normal hearing children, ages 4, 6, and 8 years as a function of five conditions of time compression (0%, 30%, 40%, 50%, and 60%)IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII Percent correct scores (A) and mean reaction time in milliseconds (B) for hearing-impaired children with various degrees of hearing loss, as a function of five conditions of time compression (0%, 30%, 40%, 50%, and 60%)............ Percent correct scores (A) and mean reaction time in milliseconds (B) for hearing-impaired children displaying little variability across time compres— sion conditions (Group I) and those displaying comparatively greater varia- bility across time compression condi- tions (Group 2). Mean ages were 10 years-0 months and 8 years-3 months for groups 1 and 2 respectively............. ix Page 22 24 50 57 60 CHAPTER I INTRODUCTION Research concerning the processing of auditory informa- tion has afforded increased knowledge relative to peripheral and central nervous system function. One of the major concerns relative to assessment of the central auditory nervous system is that pure tone and speech signals do not adequately assess central auditory functioning (Jerger, 1960). This fact has been attributed to the existence of two forms of redundancy. The first has been referred to as an intrin- sic redundancy where the complex structure of the human auditory nervous system permits considerable reduction in acoustic information before speech discrimination is affected. Subsequently, speech distortion has been employed to reduce signal redundancy (extrinsic redundancy), thus taxing the disordered auditory nervous system to a point where it cannot process the signal efficiently. One of the frequently used distortion techniques is time compression, a method which reduces temporal redundancy in the signal through deletion of random segments of the signal. Time Compression in Audiological Assessment The time compression technique and many of the practi- cal applications of time compression were based on the l 2 finding that the speech signal is highly redundant, where numerous cues are available for adequate intelligibility (Miller and Licklider, 1950: Garvey, 1953). Thus, removal of random segments of the signal does not affect intelligi- bility until a substantial portion of the signal has been discarded. Fairbanks, Everitt, and Jaeger (1954) described an electromechanical time compressor-expander which was designed to delete fixed segments of the speech signal while retain- ing the general pitch characteristics of the signal. Tech- nological advances since the development of the Fairbanks compressor have resulted in several computer-based compres- sion devices such as the Lexicon Varispeech I (Lee, 1971) time compressor. This unit contains an analog to digital to analog converter which allows for random sampling of the acoustic signal and subsequent removal of information through deletion of sampled time segments. The unsampled speech segments are returned to analog form as time- compressed speech, while retaining frequency characteris- tics of the signal. The Lexicon Varispeech I is capable of processing up to 60% time compression, which means that 60% of the signal is removed and thus, the processed signal requires only 40% of original playback time. The sampling interval duration for the Varispeech I is sufficiently small (25 to 30 msec) so as to minimize the possible removal of entire phonemic units (Fairbanks and Kodman, 1957) - 3 For purposes of investigating perceptual prOCessing differences, time compression has been investigated with aging adults (Konkle, Beasley, and Bess, 1974) and with adults having normal hearing (Beasley, Schwimmer, and Rintel- mann, 1972; Beasley, Forman, and Rintelmann, 1972), sensori- neural impairments (Kurdziel and Rintelmann, 1971) and central auditory disorders (Kurdziel and Noffsinger, 1972). With children, two standard speech discrimination measures have been time—compressed and used with several populations. Beasley, Maki, and Orchik (1975) presented the time-compressed Word Intelligibility by Picture Identification (WIPI) speech discrimination measure (Ross and Lerman, 1970) and the Phonetically Balanced Kindergarten (PB—K-50) word lists (Haskins, 1949) to normal hearing children, ages 4 to 8 years. The time-compressed WIPI has also been administered to articulation-defective children (Orchik and 0eschlager, 1974) and learning-disabled children (Freeman and Beasley, 1975). In general, results of investigations with both adults and children have shown performance differences when comparing speech discrimination scores of pathological and normal hearing subjects under time compression conditions. The results obtained with children have supported the use of time compression in eliciting differential performance relative to speech processing and thus providing additional diagnostic information. Orchik and 0eschlager (1974) admin- istered the WIPI to three groups of children varying in articulation skills. Group I was composed of children with be mn . ‘9‘ .~ ‘ .1 s 1.: Qk zflK NC L, normal articulation, and groups II and III were composed of children with one to three misarticulations and 4 or more misarticulations, respectively. It was reported that there were no differences between the groups at 0% time compression; however, a significant difference at 60% time compression was found for children displaying 4 or more misarticulations when compared to the other two groups. The authors considered that possibly a more subtle listening task was necessary to elicit differences in perceptual processing of speech between the three groups. Further support for use of the time-compressed WIPI in perceptual testing was provided by Freeman and Beasley (1975) who presented the discrimination measure to two groups of children, both containing children enrolled in classrooms for the learning disabled. One group displayed auditory diffi- culties (auditory group) and the other did not (non-auditory group). Preliminary results of this study indicated that speech discrimination of the auditorially impaired learning disabled children was appreciably lower at 60% time compres- sion than the non-auditory group, regardless of the fact that both groups performed similarly under the 0% time compression condition. In addition, the difference between groups was further increased when the intensity level was decreased from 32 dB SL (re SRT) to 16 dB SL, thus further increasing the listening difficulty level. The use of time compression with hearing-impaired children has not, as yet, been investigated. At present, 5 evaluation of hearing-impaired children is based on peripheral tests and administration of speech discrimination measures, the difficulty of which cannot be altered to meet situa- tional needs. Only through differences in the response task, i.e., picture-pointing closed set response vs word repetition open set response, can the difficulty level be increased (Sanderson and Rintelmann, 1971; Beasley, Maki, and Orchik, 1975). This practice, however, does not assess differences in speech discrimination skills, but rather assesses differ- ences in ability to respond to different task requirements. Development of the WIPI was intended, in part, to supply a speech discrimination measure which would not be affected by the child's articulation and language skills (Lerman, Ross, and McLaughlin, 1965). The difficulty level of the measure, however, cannot be altered to allow presentation of a more difficult speech discrimination task. Through the use of time compression, however, the listen- ing difficulty level can be increased without increasing the difficulty of the response task. Evaluation of this proce- dure with hearing-impaired children is necessary since analysis of the performance of hearing-impaired children would reflect the effectiveness of this measure as a means of evaluating the auditory system under progressively more difficult listening conditions. With adults demonstrating central impairments, the response at 60% time compression is considerably different than that of adults with sensorineural impairment. In the ear 6 contralateral to cortical damage, the centrally disordered population showed marked decrease in speech discrimination scores at 60% time compression (Kurdziel and Noffsinger, 1972), whereas sensorineural subjects demonstrated an over- all depression in scores when compared to normal listeners (Kurdziel and Rintelmann, 1971). No dramatic decline in speech discrimination scores at 60% time compression was found, however, for adults with sensorineural impairments. Results with normal hearing children (Beasley, Maki, and Orchik, 1975) were found to parallel those found with adults (Beasley, Schwimmer, and Rintelmann, 1972); therefore, performance of hearing-impaired children with sensorineural impairments should logically parallel that of adults with sensorineural impairments. This would appear to be the requisite to further research with children diagnosed as having both central and peripheral involvement. Whereas time compression has been employed to distin- guish perceptual processing differences among various experi- mental groups, reference to a perceptual and memory model developed by Norman and Rumelhart (1970) offers one theoreti- cal explanation for speech discrimination results obtained with time compression. A Model of Speech Perception Shown in Figure l is the perceptual and memory model of Norman and Rumelhart (1970). The authors provide explana- tion of the perceptual and memory processes at several levels .Aonmav pnmnamazm use smsnoz 809% meMP Hove: macaw: cam HmsPQmopmm .H oHSMam Sukm>m >m02w—2 msfikm>m Jmc vamvsmpm .c0flwmmpasoo weflv we :oHpHGSOo some nous: mane» m can .w .3 mmmm .Cmnuawco wcflpmm: Adeno: now pomnnoo enmohmm cmms ca HmHz one no mmpoom nowpmGHEHnomHv nommmm .H mnm¢e 32 m.eam m.o~w m.mow m.wam e.mmm mm m:.mmda m.dama m.woma m.am:a m.mema o.owma E m.mma a.mme m.mmm o.emoa 0.0HNH m.omHH m.mooa m.mmoa m.:om w m.aem o.mmm m.amm 5.:oma 2.0::H m.amma w.emma m.mmma m.:oaa o m.omm o.aom m.:mm m.emma m.maam H.mHam m.moom :.maaa m.msma : mm domme nae mm m. ace mam! ea: Sam e0 mam new: omme ea msHe 20Hemn unaccepm .msonm 6mm none 90% mucoommwaafls ca Amamv msfip coapommu mflaswm sums cam .SoammmeEoo we“? mo nowpflcSoo comm nouns memo» m ucm .w .: mmwm .Sopuawno wSASan adage: pom mucoommfiaaws SH Apemv oEwp Soavommn Hmpop cams .N mqmmm am .aa .:a .ma .oa .m mm a.mm m.ma m.am m.mm a.wm a.om 0.8m a.me a-oa mamamw -aawmmamgma ma .0a .a mm m.am o.ma m.am m.mm 0.0m m.am m.m: :.ae m-m mammmeos m .a .m .m .a mm a.mm 0.0a w.am o.ma o.mm m.am 0.6m m.em aa-a eaaz E apm. mow Time: mom mo amm E. «am E. wwa_E mmoa mammamm zoammmmmzoo mzae ezmommm mo mmmomn .msoam mmoa wCaHan some how Aamm 2V vaonmmnsp Soapmmoma soommm cam Am mcop mafia Eocmsvmnw m .mmm name use .m:ohm awn mpoonnzm Ho Hones: mam ssonm oma< .wmoa wsaamms Ho mmHMou op mafinmooom cmmzonm Smacaaso voaammeH umSHAan pom pomanoo enmoamn same :a HmHz map so mmuoom SoapmsflEHnomau commam .: mAm¢E 37 o.amm m.mm: m.ome 0.0mm N.oom :.mewa m.mmma m.mema a.mmma o.mm:a o.am:a a.a:ma m.amaa m.aaaa m.mmma o.:m:a m.aama m.oo:a m.mmma o.mo:a 0.:mma m.mmma m.aama m.aama o.moom o.mmma o.oomm m.amma m.omma m.mawa 0.3mma o.mama e.mmoa m.wama m.mm:a E moo Eon Ame: mam mo zoammmmmsoolwzaa sawmmmm msaw+zoae -oamm mamzam E. ma .mH .NH .HH mm mpm>mm Hm .mH .da .MH .OH .0 mm mum>mm [Hampmamvos ma .wa .m mm mpmamvoz a 5 .a .m .a $ UHHE mmog 02Hm¢mm mo mmmwma .wmoH MSEHan mo moHMmc op mSHUAOOOM ummsopm Swmcaano voyamquumSEHmmn Mom mesoomm uaaafis CH Amemv mafia coapommn magsHm same can Apemv meap Cowpomma Hmpop cams .m mamme TABLE 6. RTe RT RT 38 Mean error reaction time (RTe)' mean total reaction time (RTt)’ and mean correct reaction time (RTC) for each percentage of time compression for all hearing-impaired subjects combined. PERCENT TIME COMPRESSION 0% 30% 40% 50% 60% ‘fi 1731.6 2015.9 2219.8 2446.0 2373.1 2166.9 1481.0 1483.0 1668.9 1743.5 1828.8 1642.4 1379.7 1379.5 1437.0 1512.8 1643.3 1470.6 CHAPTER IV DISCUSSION In the present investigation, test lists from the Word Intelligibility by Picture Identification (WIPI) speech discrimination measure (Ross and Lerman, 1970) were presented to normal hearing and hearing-impaired children. Word lists were presented under 5 conditions of time compression (0%, 30%, 40%, 50%, and 60%) at 32 dB sensation level (re speech reception threshold). Measures of response accuracy and reaction time were obtained for each experimental group under all time compression conditions. Results showed that for normal hearing children, speech discrimination scores improved as a function of age and decreased as a function of time compression. Response latency was also affected by age and time compression where mean total reaction time values increased as time compres- sion increased and decreased with increasing age. For both measures, no significant main effect was found for sex and no significant interactions occurred. In general, similar trends were found with hearing- impaired subjects, where speech discrimination scores decreased and reaction time values increased with higher percentages of time compression. In addition, as age 39 40 increased, discrimination scores improved and mean total reaction time values decreased. Degree of hearing loss also affected speech discrimination and reaction time performance in that as degree of hearing loss increased, speech discri- mination scores decreased and reaction time increased. Results showed an apparent interaction of age and degree of hearing loss. For the hearing-impaired children, careful interpretation of results is crucial since the number of subjects per age and hearing loss category were small and unequal. Normal Hearing Children Speech Discriminatign and Time Comppession. At present, there are three investigations which have provided speech discrimination results of normal hearing children on the time-compressed WIPI. Results of the Beasley, Maki, and Orchik (1975) and Shoup, Beasley, and Bess (1975) studies and the present investigation can be found in Table 7. When comparing across studies, it can be noted that in general, speech discrimination scores improved with each successive investigation, with lowest scores obtained by Beasley et a1. and successively higher scores obtained in the Shoup et a1. study and the present investigation. These differences can best be explained by procedural differences among the studies. In the Beasley, Maki, and Orchik (1975) investigation, no Practice items were presented before test lists and time compression-sensation level conditions were presented 41 TABLE 7. Mean speech discrimination scores on the WIPI from three investigations of normal hearing children (ages 4, 6, and 8 years) tested under 3 conditions of time compression at 32 dB sensation level. TIME COMPRESSION 0% 30% 60% AGE 4 6 8 u .6 8 4 6 8 Present Study 93.4 96.4 98.4 94.2 97.0 97.8 85.8 88.0 91.8 Shoup et a1. 90.2 95.8 97.7 84.0 93.0 97.6 72.0 88.9 92.5 Beasley et a1. 86.8 94.8 97.6 80.8 92.0 96.4 68.0 82.0 86.8 in randomized order. In the Shoup, Beasley and Bess (1975) study, practice items were used and the order of presentation of time compression-sensation level conditions for each subject was either 0% and 30% time compression at 32 dB SL followed by 0% and 30% time compression at 16 dB SL or 0% and 60% time compression at 32 dB SL followed by 0% and 60% time compression at 16 dB SL. In the present investigation, randomized order of time-compressed conditions was used and practice items were presented before each test list. The primary factor which would probably account for differences between‘muapresent results and those obtained earlier was inclusion of a pretest prior to presentation of test lists. During the pretest, subjects became familiarized with test words which probably eliminated much of the ambiguity of test pictures. Some constions noted during pretesting included 42 both difficulty with identification of words, such as spring, ‘pgpk, pggp and kpgg plus confusions resulting from logical substitutions for correct identifications, such as egg for ,pg§3;‘gipi (wearing a dress) for pkipp; igpm for pgpp;.pgk§ (on a plate) for plate; house (with a door) for gppp; and gppgg (neck portion of the dress) for pgpk. Correction of subject's errors during pretesting probably allowed for better performance during the test session, resulting in higher overall scores. This would seem to imply that the time-compressed WIPI is assessing more than speech discri- mination skills, especially for 4 year olds where at 30% and 60% time compression, score differences between the Shoup et a1. study and the present study were 10.2 and 13.8 percentage points, respectively. At 0% time compression, results for 4 year olds were still higher when compared to the Shoup et a1. study; however, an upper limit is being approached where possibly it is difficult for the 4 year old group to improve much beyond 90% correct. Thus, familiarity with words does not improve results at 0% time compression to the same degree as at 30% and 60% time compression. Another possibility for the overall improvement in scores of the present investigation is that only one sensa- tion level (32 dB SL) and five time compression conditions were used. This situation allowed subjects to listen to four actual conditions of time compression (excluding 0% time compression) whereas only two (30% and 60%) were used in previous investigations under two sensation levels. 43 Possibly, there existed an overall improvement in scores as a result of listening to more time-compressed speech at a relatively high intensity level. When considering the child's adaptation to the listening task, the present investigation required that the child adapt to changes in only one para- meter, rate of presentation, whereas two parameters were altered in the previous investigations, rate of presentation and intensity level. Another discrepancy between the investigations was the slight improvement between the 0% and 30% time compression conditions for 4 and 6 year old children. This finding can probably be attributed to normal subject variability, especially since there is little difference between 0% and 30% time compression in terms of listening difficulty. To explain why this has not been found in the previous investi- gations, the same rationale as discussed above might be applied to this finding. The effect of practice items plus the pretest could have served to improve the 30% condition to equal that of 0% time compression. In addition, the randomized presentation of five time compression conditions might account for the improved performance at 30% time compression. Beasley et a1. randomized time compression- sensation level conditions and Shoup et al. used a clinical procedure where the 30% time compression-32 dB sensation level condition was always presented second, thus no oppor- tunity was available for adaptation to more difficult listening conditions. By nature of the randomized 44 presentation order, each time compression condition (with the exception of 60%) followed more difficult listening conditions a certain percentage of the time. The 30% condi- tion, in particular, has three other more difficult listening conditions (40%, 50%, and 60%) which it could follow, thus making this time compression condition easier, by compari- son, whereas this opportunity did not exist in previous investigations. The effect of age was found to be consistent across the three studies in that as age increased, speech discrimina- tion scores improved. Table 8 shows the differences between age groups as a function of time compression for the three investigations. TABLE 8. Results of three investigations showing differences in percent between mean speech discrimination scores of normal hearing children, ages 4 and 6 years and ages 6 and 8 years for 3 conditions of time compression. AGE GROUPS 4 to 6 6 to 8 % TC 0% 330% 60% 0% 330% 60% Present 3.0 2.8 2.2 2.0 0.8 3.8 Shoup et a1. 5.6 9.0 16.9 ' 1.8 4.6 3.7 Beasley et a1. 8.0 11.2 14.0 1.9 4.4 4.8 Comparison of the difference scores showed that perfor- mance between age groups was considerably different across studies. In particular, between 4 and 6 years of age, the 45 score differences were all fairly small in the present inves- tigation as compared to previous results. Again, this is attributed to the pretest which served to familiarize sub- jects with the test words. These findings are supported, in part, by results of the pretest which showed that eleven 4 year olds missed one to nine words on the first trial of the pretest resulting in a mean of 98.2% correct. Each child's errors were corrected by the experimenter and a subsequent second trial showed an improvement in performance to 99.8% correct. Similarly, results of 6 year olds improved from 99.6% to 100% and 8 year olds improved from 99.8% to 100%. Whether the pretest results show enough improvement between first and second trials to result in a 10 to 15% improvement for 4 year old children is questionable; however, the familiarity factor, with both test words and pictures, does warrant consideration. Another explanation which might have accounted for the improved performance in the present investigation is that the instrumentation of the present investigation was considerably different from that used previously. Each subject was instructed to press the picture on a response panel as soon as he found the correct picture. When pressure was placed on the panel, a new set of pictures appeared and Shortly after, a new test sequence. It is possible that the experimental task was a more interesting and challenging one than simply pointing to pictures in the test manual, and therefore, subjects maintained a higher 46 level of interest during this task. This suggests that use of a response panel might be beneficial in testing reluctant children where interest in the panel might encourage coopera- tion during the testing situation. In summary, comparison of results showed that speech discrimination performance can be affected by altered testing procedures, especially for younger children under conditions of time compression. Results of the present investigation further suggest that given more opportunity to adapt to the listening task, plus considerable familiarity with test items, speech discrimination performance of 4 year old children can be improved considerably. Although it was shown that different experimental procedures served to narrow the gap between 4-year-old and 8-year-old performance, differences in nervous system function were not overcome in that a significant age effect was still found for the present investigation. Thus, regardless of outside influ- ences, the performance range of the 4—year-old nervous system will significantly differ from the more mature nervous system. Response Latency andTipe Compression. Prior to this investigation, there has been only one study which combined the measures of reaction time with speech discrimination of time-compressed speech. For purposes of investigating speech processing skills in children, Shriner and Sprague (1969) administered the PICSI (Seidel, 1963) at four condi- tions of time compression (0%, 30%, 50%, and 70%) to 51 47 children from 5 to 8 years of age. Correct reaction time, total reaction time, and error reaction time were reported fOr all subjects across three percentages of time compression. Results of the present investigation support the results of Shriner and Sprague insofar as the relationship between the three reaction time measurements is concerned. In both studies, it was found that correct and total reaction times were fairly close, with correct reaction time being somewhat faster. Error reaction time, however, was considerably longer than either correct reaction time or total reaction time for all time compression percentages. When considering reaction time as a function of time compression, Shriner and Sprague reported a decrease in reaction time between 30% and 50% time compression and an increase in reaction time at 70% time compression. Results of this investigation do not support those of Shriner and Sprague since a significant main effect for time compression was found where total reaction time scores increased with increasing time compression. Shriner and Sprague offered two explanations for their results. One interpretation was that at 50% time compression signal information was substantially reduced, therefore, less information was available for processing and, subse- quently, less time was required for analysis or perception of the signal. At 70% time compression, however, this did not occur since the intelligibility of the signal was suffi- ciently distorted to interfere with processing and, thus, 48 negatively affect reaction time performance. The major problem with this interpretation is that even though there is less signal information to be analyzed, this does not necessarily indicate an easier perceptual task. In fact, identification of the stimulus must be made with reduced information resulting in a more difficult task. And further, if the perceptual task is more difficult, reaction time durations should increase (Hohle, 1967), not decrease. The second explanation suggested by the authors was that a pacing effect resulted with increasing time compres- sion such that faster stimulus presentation led to faster response output. At 70% time compression, this effect was minimized because of the difficulty level associated with the highest percentage of time compression. Results of the present investigation cannot prove or disprove the existence of a pacing effect; however, the trend across time compression conditions revealed no decrease in reaction time with increasing time compression. The most plausible explanation for the conflicting data involves the reaction time instrumentation. In the Shriner and Sprague study, measurement of reaction time was made from word onset. In this investigation, reaction time was measured from word offset. With reaction time measured from word onset, there is no way to control for word length. Thus, a word of 1000 msec duration at 0% time compression would have a duration of 500 msec at 50% time compression. With reaction time measured from word onset, the listener would 49 have to wait a full 500 msec longer at 0% time compression to receive all necessary information, whereas the system can begin responding 500 msec earlier under the 50% time compression condition. The same holds true for comparisons between the 30% and 50% time compression conditions where shorter reaction times were reported for 50% time compression. Figure 4 further illustrates the finding that longer reaction times reflect increased difficulty in the perceptual task. The figure shows that, in general, the reaction time. and Speech discrimination functions parallel one another, eSpecially under increased time compression levels where listening difficulty was greater. The reaction time axis has been reversed to better illustrate the similarity in functions of the two measures. This figure also shows that 6 and 8 year old children perform most alike, with the 4 year (old group more isolated. This is particularly evident for the reaction time data. Thus, for this experimental task, the perceptual system of the 4 year old did not perform as efficiently as that of older children, a finding which is not surprising in light of developmental norms which reflect maturation of the nervous system. Results also showed reaction time to be more sensitive to age differences than speech discrimination scores since separation of the age groups is considerably greater for the reaction time data. Results of the simple reaction time (RTS) measurement indicate that speech perceptual differences between the age groups are not solely responsible for differential performance. 50 W 8 N N OO_N OOON 00m. com. com.— 000. 000. 00¢. oom— CON. oo. _ 000. com (038W) 3WLI. NOLLOVEIH "N101 NVEI . $86 98 .mom .Eo3 .fiom .Rov Soammmugsoo camp mo macapaccoo m>flm mo Soapocsm m mm whack m can . $00 oxoon oxoo¢ w .3 meme .cmacaasO wsfiana Hugues mom Amv memoommaaaas ca mSap Soapomoa sews use A m u D mm m -4 Oh om om . 00. 1038800 .LNBOHEId NVBW 51 A significant age effect was found for the RTS task where no pictureswere projected onto the panel and the task was merely one of hitting the panel as soon as possible after hearing the final word in the test sequence. These results suggest that there might be an overall slower functioning of the immature nervous system regardless of task requirement. Although the simple reaction time data can be interpreted to explain some of the differences between age groups, the differences in RTS between age groups are not great enough to account for the differential age performance found during the choice reaction time task. Tpgt—Retest Reliability. Analysis of the test-retest data of 10 normal hearing children showed no significant difference between test sessions for intelligibility scores indicating that speech discrimination performance was stable over time. Mean total reaction time data were analyzed with significant differences found for subject, time compression, and test session. Significant interactions were also found for subject x time compression, subject x session, and time compression x test session. The above results are fairly conclusive in suggesting that the reaction time value was not a stable measure and, thus, extensive efforts should be made to carefully control interfering variables when using reaction time as a dependent measure. Comparison of reaction time values tended to decrease between eXperimental sessions. This suggests that possibly the child is more familiar with the task and what is expected. One method of attempting to 52 stabilize the reaction time response would be to provide a longer practice session for subjects. Having the subject practice with one or two non-test lists would allow still greater familiarity with the test pictures and would permit more opportunity for encouraging the child to respond as fast as possible. The length of practice session, however, must be considered relative to length of test session and the possible effect of fatigue. Hearing-Impaired'Chiidren The performance of hearing-impaired children in this investigation revealed lower speech discrimination scores and longer reaction time values than were found with normal hearing children. For the entire hearing-impaired group, speech discrimination scores were found to decrease as time compression increased and reaction time values generally were found to increase with increasing time compression. Trends indicated that speech discrimination and reaction time performance depended on both degree of hearing loss and age, with possibly more dependence on the age factor. Speech Discrimination and Time_Compressipp. In the present investigation, the mean speech discrimination score of hearing-impaired children at 0% time compression was approximately 9 percentage points higher than that found by Ross and Lerman (1970). The difference in performance between the two groups may be attributed to several factors which are summarized in Table 9. .moa mu.mnh m u 2 mo mm op me oa.op .moe m-.mna 33 9mm on mmCavnoo wa n mwcmm m.ma n Owcmm op acupm me» am mo mm -on ooamp mo m.ao u E. mo a.am u E. .662 mlo.Ha 3 ma acomoAm 3 .moa «in; OH H E. q, me om op .moe a-.moa ma mam mu m u mmcmm op smsamq oz Am me 03 moao> 0>aa 1: mu N.mm u E. .moa mu.mna 3 do 3 mmom pmmpmmm . apamsopca [moaaos soap 9mm E. «emu -oma ma EHMHSDMO0> nMpSmmmam .wouwum E. mmmsnmooma aaazmaammaxm maomemmWI .coavmwapmo>:a vcmmmnm one use Evapm Aomma smegma use mmom one Ca vows monsvmooam AMPS88aamaxm use mpomwnsm vmuamaeau Hades Ho nomanmgSoo .m mnm¢9 54 The most probable explanation is that children in the Ross and Lerman study did not receive the vocabulary pretest prior to the test session as did children in this investiga- tion. This lends further support to the vocabulary pretest being the factor which resulted in differential performance between normal hearing children of this study and previous investigations (Beasley, Maki, and Orchik, 1975; Shoup, Beasley, and Bess, 1975). Another possible explanation concerns intensity level of list presentation. Ross and Lerman presented test lists at 40 dB sensation level re two-frequency pure tone average (best two of 3 thresholds obtained at 500, 1000 and 2000 Hz) and in the present investigation, test lists were presented at 32 dB SL re speech reception threshold. The presentation level, therefore, was of higher intensity in the Ross and Lerman study. Since articulation functions have not been obtained on the WIPI through 40 dB SL for either normal hearing or hearing-impaired children, the possibility exists that maximum discrimination scores might be obtained at 32 dB SL for this particular speech discrimination measure. Thus, lower scores might be obtained at higher intensity levels, a possibility worthy of further research with both normal hearing and hearing-impaired children. A third explanation concerns presentation method where live-voice was used in the Ross and Lerman study and taped presentations were used in the present investigation. The possibility exists that the taped presentations, although 55 not as convenient for clinical testing, do have.the advan- tage of precise articulation of test words plus controlled test word presentation across subjects. Results of the vocabulary pretest confirm the evaluation of Ross and Lerman that test words on the WIPI were within the recognition vocabulary of the hearing-impaired children. Mean percent correct on the first trial was 98.6 with performance on the second trial improving to 99.6 percent for hearing—impaired children in the present investigation. Overall, Speech discrimination scores of the hearing- impaired children were found to decrease as a function of time compression (see Table 4). Figure 5 shows that perfor- mance on the time-compressed test lists probably was related to degree of hearing loss. The age factor interacts, however, and performance of groups 3 and 4 are probably higher than what would be expected if age levels approxi- mated those of groups 1 and 2. Thus, the true effect of hearing loss cannot be determined accurately because of the confounding effect of age. Thus, it is apparent that future research with hearing-impaired children grouped according to both age and degree of hearing loss will be more effective in defining the effects of time compression relative to both age and hearing loss. An alternative grouping of subjects became apparent through inspection of individual subject performance. Among the 18 hearing-impaired children, it was found that children performed with differing degrees of variability across time Figure,5. 56 Percent correct scores (A) and mean reaction time in milliseconds (B) for hearing-impaired children with various degrees of hearing loss, as a function of five conditions of time compression (0%, 30%, 40%, 50%, and 60%). 57* $00 $00 $0¢ .m ouomaa 20.00mmd200 NEE. Fzmomwn. $00 $0 .. $00 $00 $0? $00 $0 magma .. 4 mmm>mm->._mammooz - O “1.258: .. x . 3.2 -O 00 00 0a. 00 00 00. 1038300 11130838 NVaw 58 compression percentages. Of the five scores obtained for each subject, one for each time compression percentage, a comparison of the individual's highest and lowest scores revealed that some subjects differed by only 12% and others differed by as much as 40%. In order to group subjects according to low (group 1) and high (group 2) within subject variability, subjects were ranked according to the difference scores, obtained by subtracting the individual subjects' highest and lowest scores. The low variable group (9 sub- jects) had difference scores ranging from 12% to 24% with a mean of 16.8% and the high variable group (9 subjects) had difference scores ranging from 28% to 40% with a mean of 31.1%. Figure 6 shows the speech discrimination and reaction time data for these groups. Comparison of the groups revealed that hearing loss for the two groups was essen- tially the same with group 1 having a mean three-frequency pure tone average (PTA) of 60.5 dB (range = 35 to 90 dB) and a mean speech reception threshold (SRT) of 53.1 dB (range = 26 to 82 dB), and group 2 having a mean PTA of 52.8 dB (range = 30 to 73.3 dB) and a mean SRT of 45.1 dB (range = 16 to 72 dB). The major difference between the two groups was age where group 1 contained children aged 6 years-ll months to 14 years-3 months (mean = 10 years-0 months) and group 2 contained children ages 4 years-2 months to 10 years- 9 months (mean = 8 years-3 months). Figure 6. 59 Percent correct scores (A) and mean reaction time in milliseconds (B) for hearing-impaired children displaying little variability across time compression conditions (Group 1) and those displaying comparatively greater variability across time compression condi- tions (Group 2). Mean ages were 10 years- 0 months and 8 years-3 months for groups 1 and 2 respectively. 60 .6 magma... 20.00wmn=200 mic. .rzwomwd $00 $00 $0.... $00 $0 $00 $00 $0¢ $00 $0 If I T m8mm o “—8 comm N / o w 320 o 98.. mooom _ o o om .I a com. T H. 02.. r o o o o o . m 08.1. / / . I U\@ M com. S a com. 03m 0 1 09m . O 3 , o\ v com. 3 om u n W 08. o/ \ 3 00¢. .. D .. 00 o\\0 ( co: m < Loo. ooo. 1038800 INBOUBd NVBW 61 Apparently, age has an affect which was not detected by the non-distorted test condition since speech discrimi- nation was equivalent for the groups at 0% time Compression. Speech discrimination differences with increasing time compression were found where the discrimination functions for the two groups separated as time compression increased. This result is additional support for the use of time com- pression as a more sensitive indication of differences in auditory perceptual skills. Two factors could account for the existent age difference found under conditions of time compression, which was also reflected in the respective reaction time performance, i.e., longer reaction time values for the younger age group (see Figure 6B). First, as a result of the peripheral hearing loss, the hearing-impaired child may require more time than the normal hearing child to reach a maturational level where auditory information is used most effectively, even under adverse listening conditions such as time compression. To compare normal hearing children and hearing-impaired children directly and imply that the hearing-impaired performed at levels below 4 year old normals is misleading since the hearing loss may never allow the hearing-impaired child to achieve discrimination scores of normal hearing children, even when the system has reached full maturation and is functioning at a "normal" level considering the peripheral involvement. _It might be considered more valid to compare the hearing-impaired child with other hearing-impaired 62 children and also with his own performance across various levels of difficult listening conditions.- This might be a more appropriate way of determining if the auditory system is performing at an efficient level for a Specific age, given a specific type and degree of hearing loss. A second explanation for the difference in performance between the 8 and 10 year old hearing-impaired children is one of different exposure times to amplification. The possi- bility exists that the older group ~performed better, not because of maturational differences, but because of longer experience with auditory information through longer use of amplification. To further investigate which of the above factors is the major influence in performance differences, a controlled investigation is required which would pair children of various ages who received amplification at different times. This would indicate if a maturational and/or amplification factor is more influential in speech discrimination performance. This investigation showed that speech discrimination performance of hearing-impaired children varied under condi- tions of time compression where no difference was found under the 0% condition. This suggests that the time-compressed stimuli were more sensitive in identifying differences in functioning of the system. Response Latency and Time Compression. In general, reaction time performance of hearing-impaired children followed the same trend as that found with normal hearing children, 63 that is, as time compression increased, reaction time values increased. The single exception to this was for the severely hearing impaired who showed decreased reaction time between 0% and 30% time compression (see Figure 5B). Again, inter- pretation is difficult since that performance trend is representative of only 4 subjects who differed considerably in age. The variability in reaction time performance was controlled through adequate sample size in the normal hearing group; however, this was not achieved with the hearing-impaired subjects relative to age and degree of hearing loss, therefore, only tentative conclusions should be made based on the present data. The reaction time values for the two hearing-impaired groups containing children with high and low variability revealed consistent trends and can be seen in Figure 6B. The reaction time values separated the two hearing-impaired groups, with increasing separation as time compression increased. These groups were equal in number and had essen- tially the same degree of hearing loss, and reaction time showed a fairly consistent and predictable trend where reac- tion time values increased as time compression increased and was dependent on age. The clinical utility of reaction time measurement for use in conjunction with speech discrimination testing is far from being determined. Considerable research must be con- ducted to determine what variables exist within the test situation and the subject which could affect reaction time 64 performance. At present, familiarity with the task can be considered an important variable since reation time perfor— mance of 10 children during the retest session was faster than during the test session, when considering the mean data. As a research tool, the reaction time measurement pro- vides additional information which the response accuracy might, in some instances, not provide. In this investiga- tion, analysis of speech discrimination and reaction time data showed that both were affected by age and time compres- sion, yet the reaction time measurement showed a considerable difference between the 4 year old group and the other age groups, an aSpect which was not as apparent in the discri- mination results. In this study, statistical analysis showed main effects for time compression and age with both speech discrimination and reaction time results; however, in an investigation where discrimination scores may not be as sensitive in detecting differences, the reaction time measurement may isolate more subtle processing differences between the experimental groups. To further assess the value of reaction time in discri- mination testing, additional research is needed to determine the response patterns of children displaying hearing losses of various etiologies, with and without suspected central auditory system pathology. Unique response patterns as a function of time compression may further delineate differences in processing of auditory information. hi mi: 65 The value of reaction time measurement is particularly evident in attempting to relate the experimental data to a theoretical model of perceptual processing. The data of the present investigation support a model of perception and memory proposed by Norman and Rumelhart (1970), with the reaction time data supplementing the discrimination results to provide stronger support for the theorized function of the perceptual system. Theory of Speech Perception and Memory Both Speech discrimination and reaction time results obtained with normal hearing and hearing-impaired children supported the model of perception and memory proposed by Norman and Rumelhart (1970). Various stages of the informa- tion processing model were supported by the data, including the feature extraction, naming, and memory systems. Normal and Rumelhart maintain that the feature extrac- tion system operates during the presence of a stimulus in the Sensory Information Store. They contend that lengthening stimulus duration would improve perceptual functioning since more time was available for feature extraction. Speech discrimination results of this and other studies (Beasley, Schwimmer, and Rintelmann, 1972; Kurdziel and Noffsinger, 1972; Konkle, Beasley, and Bess, 1974; Beasley, Maki, and Orchik, 1974; Orchik and 0eschlager, 1974; etc.) support their contention in that as time compression increased, i.e., signal duration decreased, significantly poorer speech discri— mination scores were obtained. Since the authors support 66 parallel feature extraction rather than serial feature extrac- tion, the number of features to be extracted per unit of time is critical. Thus, manipulation of the signal duration through time compression altered perception in a manner predicted by the model, i.e., discrimination scores decreased with decreased Signal duration. Reaction time data further supports the model where reaction time values increased as a function of increased time compression. This supports the consideration that as more features are presented per unit of time (i.e., increasing time compression), the extraction process becomes more diffi- cult which slows down the system and thus results in slowed reaction times. The theorized function of the naming system is to match extracted features of the signal with the psychological name associated with the features. This information enters into the memory system prior to entering the decision-making stage. Improvement in efficiency of the naming system may have resulted from inclusion of the vocabulary pretest. Speech discrimination results of the present study, especially for the youngest age group, were shown to improve from two previous investigations (Beasley, Maki, and Orchik, 1975; Shoup, Beasley, and Bess, 1975). Possibly, the inclusion of practice items and the pretest session allowed the perceptual-memory system to obtain information which improved subject performance during the test session. 67 Following processing in the naming system, information concerning the psychological label is stored in the memory store system prior to entering the decision-making stage. At this point, additional information is provided to the system in the form of all possible response alternatives. Results which showed differences between correct reaction time and error reaction time could be applicable to this stage of processing. For words identified correctly, processing of the word may require only one identification attempt. In this case, the system experiences no ambiguity, no confusion, and functions efficiently, indicated by correct identification and fast correct reaction time. For unfamiliar distorted words, however, the input from the memory store may not match any of the response alternatives at the decision- making stage. Thus, return to memory story for reprocessing of information would seem appropriate. At this stage, there is an attempt to regain information and achieve an appropriate match. This strategy may or may not be success- ful. For those instances where it is unsuccessful and errors occur, the reaction time values are longer which reflect additional processing by the system. The only flaw in applying the data to the model is that Norman and Rumelhart provide no way to enter memory storage again to retrieve any information which will aid in correct identification. If the authors assume that information is retained in the "decision-making" stage until a reSponse is 68 determined, regardless of how many comparisons are made between test stimulus and response alternatives, the present data and model are highly compatible. Clinical Appiication Results of this investigation showed that increasing the listening difficulty through time compression provides additional information concerning the hearing-impaired child's use of auditory information. This was apparent where speech discrimination scores for the two hearing- impaired groups were equivalent at 0% time compression, yet separated at higher time compression levels. The clinical application of time compression with hearing-impaired children could be made in both diagnostic and therapeutic aspects. Once sufficient data is collected with hearing-impaired children, individual performance could be compared to the group performance. Advanced auditory training procedures might be suggested if processing of time-compressed speech is poorer than that of a similar group of children. A further consideration is the possibility of using time compression in developing an auditory training program. Use of time-compressed speech would hopefully force the child to develop skills in processing of auditory infor- mation with increasingly less information. This might be warranted with children showing relatively good discrimi- nation under normal conditions. Determining the efficacy of this technique in teaching better processing skills 69 requires additional clinical research utilizing time com- pression in an auditory training program. ippiications fqpinture Research Future research should be conducted to determine the value of using time compression as a means of teaching auditory perceptual skills. The rate of information pre- sentation is an aspect of speech production which is variable in the communication situation as evidenced in different individual speaking rates. The use of time compression might logically be used for advanced stages of auditory training for practice in extracting information with increasingly less information available. A related application to training is the use of time compression in teaching the use of strategies for using available information. The performance of hearing-impaired subjects #8 and #9 demonstrated this need in auditory training. Subject #8 used all information available, auditory and visual, while trying to determine the appro- priate response. For difficult stimuli, it appeared that a decision was made only after comparing each picture label with a stored memory to determine the closest match with the test stimulus. Subject #9, however, began responding in a random manner when the amount of informa- tion was reduced. Research is needed to determine if children like subject #9 can be taught to use all available information as #8 did, thus increasing the amount of infor- mation obtained in that situation. 70 Additional research is also needed using time-compressed stimuli with a larger number of hearing-impaired children. With information collected on a larger sample, correlations could be determined between several variables, including factors such as age, degree of loss, type of loss, and age at which amplification was received. Performance in communication situations and achievement at school, plus performance on other measures of speech perception and production could also be obtained for the same children. Analysis of this data would permit evaluation of many aspects of the child's communication performance. Com- parison to other hearing-impaired children with similar characteristics would possibly allow a closer estimate as to where he stands relative to the best and poorest per— formers in the particular group. From that point, further evaluation and training could be initiated to advance him to a higher level within his group, thus improving his chances of success. The use of reaction time measurement proved valuable in confirming the speech discrimination results and, thus, further supported a theoretical model of speech perception. For such purposes, reaction time appeared to be a valid measure. It could be of greater value, given a situation in which another measure, such as response accuracy, was less sensitive to processing differences. The use of choice reaction time for clinical purposes, however, requires more research for purposes of specifying the variables which 71 affect reaction time performance. Factors such as famili- arity with the task, rate of presentation (i.e., inter- stimulus interval), test instructions, and subject moti- vation need investigation prior to use in the clinical situation. This would be necessary unless response laten- cies of other disordered populations were sufficiently discrepant from normal to such a degree that variables listed above made little difference in differential diag- nosis. Those factors would still require further investi- gation, however, in order to insure a valid and reliable measure of reaction time. LIST OF REFERENCES LIST OF REFERENCES American National Standard Specifications for Audiometers, ANSI S3.6-l969. New York: American National Standards Institute (1969). Beasley, D., Forman, B., and Rintelmann, W., Perception of time-compressed CNC monosyllables by normal listeners. J. Aud. Res., XII, 1, 71-75 (1972). Beasley, D., Maki, J., and Orchik, J., Children's perception of time-compressed speech using two measures of speech discrimination. JSHD (in press). Beasley, D., Schwimmer, S., and Rintelmann, W., Intelligi- bility of time-compressed CNC monosyllables. J. Speech Fairbanks, G., Everitt, W., and Jaeger, R., Method for time or frequency compression-expansion of speech. Trans- lations IIRIEI "" PeGeAe . AU'Z’ 7-12 (1951+). Fairbanks, G. and Kodman, F., Word intelligibility as function of time compression. JASA, 29, 636- 644 (1957). Feldman, R. and Reger, 8., Relations among hearing, reaction time, and age. J. Speech Hear. Res., 10, 479- 495 (1967). Freeman, B. and Beasley, D., Unpublished study. Michigan State University, E. Lansing (1975). Garvey, W., The intelligibility of speeded speech. J. Exp. Psychol., 45, 102- 108 (l953L Goodman, A., Reference zero levels for pure- -tone audio- meters. Maico Audiological Library Series, Vol. IV, #7 (1966) Haskins, H., A phonetically balanced test of speech discri- mination for children. Masters Thesis, Northwestern University, Evanston (1949). Hecker, M., Stevens, K., and Williams, 0., Measurements of reaction time in intelligibility tests. JASA, 39, 1188- 1189 (1966) 72 73 Hohle, R., Component process latencies in reaction times of children and adults. In Lipsitt, L. and Spiker, C. (Eds.) Advances in Child Development and Behavior, Vol. 3, New York: Academic Press (1967). Konkle, D., Beasley, D., and Bess, F., Perception of time- compressed CNC monosyllables in an aged population. Paper presented to the American Speech and Hearing Association, Las Vegas, Nevada, November, 1974. Kurdziel, S. and Noffsin er, D., Performance of cortical lesion patients on 0% and 60% time-compressed speech materials. Paper presented to the American Speech and Hearing Association, Detroit (1973). Kurdziel, S. and Rintelmann, W., Performance of noise- induced hearing-impaired listeners on time-compressed CNC monosyllables. Masters Thesis, Michigan State University, E. Lansing (1971). Lee, F., Varispeech I: Instructional Manual. Lexicon, Inc., Waltham, Massachusetts (1971). Lerman, J., Ross, M., and McLaughlin, R., A picture- identification test for hearing-impaired children. J. Aud. Res., 5, 273-278 (1965 . Miller, G. and Licklider, J., The intelligibility of interrupted speech. JASA, 22, 167-173 (1950). Norman, D. and Rumelhart, D., A system for perception and memory. In D. A. Norman (Ed.), Models of Human Memory. New York: Academic Press (1970). Orchik, D. and 0eschlager, M., Time-compressed speech discrimination in children: its relationship to articulation ability. Paper presented to the American Speech and Hearing Association, Las Vegas, Nevada, November, 1974. . Rapin, I. and Steinherz, P., Reaction time for pediatric audiometry. J. Speech Hear. Res., 13, 203-217 (1970). Ross, M. and Lerman, J., A picture identification test for hearing-impaired children. J. Speech Hear. Res., 13: “4'53 1970)- Sanderson, M. and Rintelmann, W., Performance of normal hearing children on three speech discrimination tests. Paper presented to the American Speech and Hearing Association, Chicago (1971). 74 Seidel, S., The PICSI test: picture identification for children - a standardized index. Masters Thesis, Indiana University (1963). Shoup, J., Beasley, D., and Bess, F., Masters Thesis, Michigan State University, E. Lansing (1975). Shriner, T. and Sprague, R., Effects of time-compressed speech signals on children's identification accuracy and latency measures. J. Exp. Child. Psych., 7, 532-540 (1969) - Utley, J., What's Its Name. Urbana, Illinois: University of Illinois Press (1951). APPENDICES APPENDIX A INDIVIDUAL SUBJECT INFORMATION FOR THE HEARING-IMPAIRED GROUP maaonsm mesmpp appflp manammom maamnsm maampsm thBQMouQ mcahzv CoHPmO :flume vmumpmfisflsom hospos «Hampsm maamnsm czocxss hmoaowpm szonxc: zmoaowpm wnmnponn mum m pad .3 .m mvoow95m : Snowman haaewm czosxcs hMoaome szocxcs Swoaowpm A :oflpwsuomcH Hoseapwuv« no no mo om om ooa OHH no OH OH on mm mm mm om om me me mm om no on on m: om om mm ca om - mz mz ooa mm mm mm m: mm om on mm mm mm m: on om m: mm 0: mm mm o: 0: on ma on m: 0: mm mm om m: on on on 0H 0H mm mm om xn xN xH cow. OWN **muflosmmuaa 8:89-8psm om :m on mm mm 3m 00H m: m: mm mm wm on ma 8mm mmqqmqn mama moan: muoa mum :IOH oaum elm muoa mum muia mu: m:w .mnn Hauma m:w :a ma NH HH 95m 0H% .HO COHPQmOxm mg». SPfiB .mmwwOH MQflkwmfi Hdamgfihomflmm HmePNHfiD UNQ mPomeSw HH< cmppommu you was goPmEOwcsm Ho mpflsfla Pm mmsommmp 02 u mz mEMAMoflosm Pamomh Pmoe Bosh vmcwmppott mwmhamc< mpmo song vmpvfieo mvommnsmt maampsm Pmopohm unsammeo> mcwpzo cmpoc maawxm soapmcweflpomwc nommmm Loom hawfimupxm msoz mmoH mCMMan mo Snowman hawsmm vaflsmmsoo wmoa o>flposcsoo maamnzm mm mm on no ow . on om om . mm mm om m: mm - m: mm 1 mm . m: mm om no mm on 0: mm on mm mm mm mm mz mz mm me do .Ilnmmmmmsuoch ammonpfico< xm gm an: cow, omm :*muaonmmpga ocoanmusm a: flit-3:34:11 a-oa am 0:: omt H-0H ma Hum ma Ha-m AH HH-8 8H muoa ma .umm¢ A .wpsm APPENDIX B COMPOSITE AUDIOGRAM FOR THE HEARING-IMPAIRED GROUP DIVIDED ACCORDING TO DEGREE OF HEARING LOSS HEARING THRESHCLD LEVEL IN dB (ANSI'I969) O) O 3\ <>——<>\ \‘V/ X35 - O-Mild X - Moderate ' O - Moderately- severe - A - Severe | 4 I25 250 500 :000 2000 4000 8000 FREQUENCY (Hz) APPENDIX C CALIBRATION PROCEDURES LEXICON VARISPEECH I TIME COMPRESSOR Calibrated for: compression level Equipment: Beckman 6148 Eput and Timer 1000 Hz calibration tape Procedure: 1. With pitch control off, set compression dial to point where frequency output was 1000 Hz. 2. Activated pitch control and recorded test lists at 0% time compression. 3. Checked frequency count again after processing of tapes. 4. Repeated procedure for higher percentages of time compression. The frequency equivalents were: 30% - 1428 Hz 40% - 1666 Hz 50% - 2000 Hz 60% - 2500 Hz Equation used for determining frequency equivalents: 100 - % time compressipp _ 1000 Hz 100 x where: % time compression = the intended percentage of time compression, and x = the frequency equivalent for that time compression percentage. TAPE RECORDERS: Ampex AG 500 Ampex AG 600-2 Ampex AG 440 B Calibrated for: frequency response Equipment: Ampex calibration tape, 50-15000 Hz Procedure: 1. Used calibration tape as directed (accompanying instructions). 2. Noted intensity output at frequencies from 50 to 15000 Hz. 3. For Ampex Ag 600-2, also checked performance of record head. Recorded calibration tape (taking into consi- deration the playback response of the other tape recorder) and noted playback response of AG 600-2. 4. A11 tape recorders met required specifications (50-15000 Hz :IZ dB @ 7.5 inches per second)- BECKMAN 6148 EPUT AND TIMER Calibrated for: proper operation prior to testing subjects Equipment: no additional equipment needed Procedure: used internal test signal as instructed Note: in the manual Sensitivity was adjusted so that tone burst consistently triggered the counter. Careful observation of triggering was necessary to insure that no other extraneous signal was responsible for onset of timer function. If this occurred, counter was immediately reset. If manual reset was not possible before onset of tone burst, the reaction time measure- ment was omitted. Timer controls were set to trigger with onset of tone burst (i.e., duration of tone burst included in reaction time measurement) and onset of the signal which terminated reaction time measurement (i.e., duration of signal from response panel was not included in the reaction time measurement). TESTING ENVIRONMENT AND AUDIOMETERS Testing environment: Checked for: sound pressure level of ambient noise in test room Equipment: sound level meter (Bruel and Kjaer, .Type 2204) microphone (Bruel and Kjaer, Type 4145) Procedure: sound pressure readings on A and C scales Beltone Model 15 C Pure Tone Audiometer: Calibrated for: accuracy of sound pressure levels of fre uencies 250, 500, 1000, 2000, 000, and 8000 Hz purity of tones Equipment: sound level meter (Bruel and Kjaer, Type 2204) microphone (Bruel and Kjaer, Type 4144) artificial ear (Bruel and Kjaer, Type 4152) accompanying TDH 39-1011earphones audio frequency spectrometer (Bruel Kjaer, Type 2112) Procedure: 1. Measured sound pressure output as specified in ANSI, 1969. 2. Corrected for sound pressure levels not meeting stated require- ments. 3. Purity of tones met stated speci- fications. GRASON STADLER MODEL 162 SPEECH AUDIOMETER Calibrated for: reference threshold sound pressure Equipment: Procedure: level for speech attenuator linearity acoustical fidelity overall distortion electrical noise sound level meter (Bruel and Kjaer, T e 2204) microphone (Bruel and Kjaer, Type 4144) audio frequency spectrometer (Bruel and Kjaer, Type 2112) pure tone oscillator (Hewlett Packard - 4204 A) artificial ear (Bruel and Kjaer, Type 4152) l. Followed procedural requirements as stated in ANSI, 1969. 2. The system met all requirements except that specified for electrical noise. Specifications state 50 dB difference, whereas the system delivered only 42 dB difference between electrical noise and signal sound pressure levels. APPENDIX D SUBJECT INSTRUCTIONS (m "dCTD‘c-O-A 59357 A. tut hea Chi the Do VOCABULARY PRETEST INSTRUCTIONS We're going to look at the pictures in this book. When I say a word, you point to the picture. Let's try a few . . . Show me cat . . . Show me glass . . . Show me rat . . . Good! Now let's go on. 'SUBJECT INSTRUCTIONS FOR SRT DETERMINATION ‘ I'm going to say some words and I want you to repeat each word after me . . . . . . . Now you'll hear a man saying the same words, but this time the words will come from here (earphone). Some of the words will be very soft, so listen very carefully and say the word you think the man said. I'll be sitting in the next room, but I can still hear you. Do you have any questions? INSTRUCTION PROCEDURES FOR SPEECH DISCRIMINATION TESTING Each child was instructed in stages. First, an explanation was given of how the panel worked and the child was instructed to push the panel several times. Second, each child was told to push the picture on the panel which matched the word which was presented by the examiner. Each subject was also instructed to place his hand on the word "go" (a button labelled "go" on the base of the response console) each time after pushing the panel. Practice words (nontest words) were presented until the subject was consistently pushing the panel with sufficient force to advance the projector and placing his hand on "go" following his responses. ' Finally, the child was instructed to push the panel as fast as possible after finding the right picture. Several presentations were given to allow the child to practice while the experimenter repeatedly encouraged the child to go as fast as possible. When the child performed the task to the satisfaction of both experi- menters, the testing procedure was begun. INSTRUCTION FOR SIMPLE REACTION TIME MEASUREMENT This time, as soon as you hear the word, press the panel. When you press the panel, you'll see the light turn off and then come back on. Do this each time you hear the word . . . just like this (demonstrates for the child). Remember, as soon as you hear the word, press the panel, but be sure to wait for the word at the end. Do you have any questions? APPENDIX E TEST FORMS: SUBJECT INFORMATION AND RESPONSE SHEETS DATA FORM Subject Information Name Age: Address ' Summary of Results Vocabulary Pretest: Z correct — lst trial Audiological Data: Normal hearing subjects: Hearing screening: Passed Failed SRT LE RE Hearing—impaired subjects: Pure tone thresholds: RE: AC/BC: 250 / 500 / 1K Date Subject # Years Months DOB Telephone No. 2nd trial LE: AC/BC: 250 / 500 / 1K SRT PTA LE RE Type of Loss: Configuration: Etiology: Primary Means of Communication: Comments: Page 2 Data Form Subject # Speech Discrimination and Reaction Time Data: Presentation order of TC conditions: Raw Score Z Correct MRT CRT ERT 0% TC (List ) 30% TC (List ) 40% TC (List ) 502 TC (List ) 602 TC (List ) 2 TC LIST ONE school ball smoke floor fox hat pan bread neck stair eye knee street wing mouse shirt gun bus train arm chick crib wheel straw pail SAMPLE Z TC LIST TWO broom bowl coat door socks flag fan red desk bear pie tea meat string clown church thumb rug cake barn stick ship seal dog nail RESPONSE SHEET Z TC LIST THREE moon bell coke corn box bag can thread nest chair fly key feet spring crown dirt 811D cup snake car dish bib queen 88W jail Name Z TC LIST FOUR spoon bow # goat horn blocks black man bed dress pear tie bee beet ring mouth skirt gum bug plane star fish lip green frog tail Z TC LIST ONE school ball smoke floor fox hat pan bread neck‘ stair eye knee street wing mouse shirt gun bus train arm chick crib wheel straw pail school broom moon spoon ball bowl bell bow smoke coat coke goat floor door corn horn fox socks box blocks hat flag bag black pan fan can bread red thread bed :3 O. O H O O 7:" I HIII B r: o :3" llllllilllllll Subj.# _ Name Date Vocabulary Pretest lst 2nd lst 2nd neck gun desk thumb nest sun dress gum egg duck leg truck stair bus bear rug chair cup pear bug ear book hair nut eye train pie cake fly snake tie plane kite _____ plate pipe lake knee arm .tea ’____ barn key __~__ car bee star tree farm beans heart street chick meat stick feet dish beet fish teeth ink leaf milk wing crib string ship spring bib ring lip king pig swing___‘ road mouse wheel clown seal crown _____ queen mouth green cow screen;___ house sheep shirt ____ straw church _____ dog dirt _____ saw skirt frog bird __*_. wall girl ball pail nail jail tail lst 2nd llllll B W p. l—' sail Z correct - lst Z correct - 2nd trial: trial: APPENDIX F INDIVIDUAL SUBJECT TEST-RETEST DATA FOR 10 NORMAL HEARING CHILDREN ammpmm u m pmma n a * aam mama amma mma mama mom ooa ma ma ooa ooa ea m 3mm ama amma maoa aooa mma aaa ma ooa ma ooa ooa ca 9 ea: maoa com mma moa asa mm ooa mm ooa ooa ma x mm» mmm asm mmm aam mam ama mm ooa mm mm mm aa a mmm mama mmma smma mmoa mama mm aw mm mm mm m m mmm mama aama omsa maaa mmma mmma om mm ooa ooa ooa m a ma: mama emaa mama soaa mama mm ooa ooa ooa mo ma m ma: smsa maaa mmoa mma mama mm ooa ma ooa ooa ca 9 am* mam mmma maam aama mama mmma mm ma mm am :m m m a» :aea mmmm aamm ommm mmom mamm mm om mm am ooa m a mmm mmma mmma amaa mama omma mm mm mm :m ma 3a m amm amma smma momm sasa mama am ma ma mm mm m 9 am omm mmma mama amaa amaa aama mm mm ooa ooa ooa ma m : mm ama mmaa amam mama mmaa aoma mm mm ooa mm mm! m 9 ¢ mam mom mom mo: mom «0 mom .mom mos mom «0 9mm aomapam :oammmamsoo mafia owMPGoouom nma mmm ama am: mo: 2mm mm mm mm mm ooa o x mmm mmm maoa mmaa mmm mmma mam ooa mm mm mm ooa : a men 3am ama mmm mom mm: mm mm ooa mm ooa a m omm ama :om amm maa amm mm ooa ooa ooa ma ma 9 oma mm: amoa mama moma asm aam mm mm ooa ooa mm m m mmm msaa mmoa mmoa moma aaa ooa ooa ma mm mm m a 33* max mom aom mos mom mo mom mom mos «on «o amm pomapsm soawmougaoo mafia swapsoonom APPENDIX G INDIVIDUAL DATA - NORMAL HEARING SUBJECTS mama mmma aamm mmam mmaa ma ooa am mm ooa ooa mm om a m.: 2 ma mmmm mmmm mmmm amam mamm om mm mm mm mm ooa mm m a mum m ea moam mmaa mmma mama mmaa mm mm mm ma mm mm mm m a_ mu: m ma ammm mmam aamm maam aoma mm mm mm ooa ooa ooa mm ea m a.: m ma mmom oamm mmaa amma. amma mm ma mm mm mm . ooa m m 3.: 2 aa maam aamm mmmm mmom mmsm mm ma mm mm mm - ooa oa a mum m oa mmmm mamm ommm mmom mamm mm om mm mm ooa . ooa m m a-m z m mmmm smam momm mmaa mama om mm mm ooa om - ooa ma m m.: s m mmaa aamm amom amma mmma mm mm mm ooa ooa . ooa ma a on: 2 a amma smma momm mmaa mama am mm ma mm mm ooa mm m a mu: m m smaa saaa mmma amaa mmaa ooa mm mm mm ooa ooa mm oa m mum m m mamm maam mmom seam amma mm mm mm mm mm am am oa m mum m : mmaa omam mama mmaa moma mm mm ooa mm mm - ooa m a m.: m m ommm mmma mmma mamm mmma ma mm om mm :m ooa am m m mu: 2 m amma mmom mama mama aoaa mm mm ooa ooa ooa ooa mm ma m 0:: 2 a New “mom g? RON R0 om fiom ”no: Won ”mo m a S 3 V S S 083. sous-0.9mm H.309 mouoom A M m a.” m m. coapmcasanomam u. w .m. u a w m. m. m. m. # I S R S 1. .A mmoa oamm - mamm mmma - mama mmma mmam omma mmaa ma mmaa ama: mmam mmam omm mam: ammm mmmm mmam mmmm mamm ea mam mmma omsm amaa mmam omm mmaa mam mama mmaa mmaa ma omoa mmmm mma: mamm - . mmmm mmma aamm maam aoma ma om: amm: ammm mmam ma: ommm amaa maam mmma amma aama aa om: ooooa ommm mmma omoa mmma mmmm mamm mmam mmaa mmma oa maaa amom mam: ommm mmmm . mamm mmam mmmm momm .mamm m mmoa mmam mamm mamm - aamm mamm maom ommm mmaa mama m mmm aamm mama mmmm : : aama aamm mama amma mmma a amm maam maam ammm mmma mmam mmma mmma mama amma mmaa m mama - ammm mmma mmma . :maa mmma mmma mmaa maaa m mam mmam mmam amma mmmm ammm mamm mmam mamm mmma aama a mma mmom amm: - amaa amm amaa mmom mama maaa amaa m mma maam ammm moma momm omma aamm maaa aama mmam mmma m om: aom mmmm - - . mmma mama maaa mama aoaa a Iomwmamaflmawwm Rom km? COHWOMwm WWWHN R0 Row imfi’flfiwwmm NWMHHOO E m m .+ a." mmaa ama mmm ama mmm mm ooa mm ooa ooa - ooa _ a m mum m on omaa mmoa mam mmma mmaa mm ooa mm mm mm - ooa m aa a aaum 2 mm mama mmma omam mmma mmaa mm mm mm mm mm - ooa _ oa a mum m mm mmma mmam amma mmma coma mm mm mm ooa ooa . ooa ca 2 oa-m 2 am amaa mmaa maaa moma mmaa ma em ooa ooa ooa ooa am oa a m-m m mm aama mmma maaa mmma mmma om mm ooa ooa ooa ooa mm m 2 mum m mm mmma mmoa ama aoma amm mm mm mm am am - ooa oa a m-m m am aama mmaa moma omaa mam mm mm mm mm mm ooa mm a a m-m 2 mm aooa oam mmaa ama mmm mm ooa ooa ooa ooa . ooa a m oum m mm amaa maaa mmoa mmm mama mm ooa mm ooa ooa . ooa oa a m-m 2 am soma mmma mmma mmaa mmoa mm mm mm mm mm . ooa oa a a.: 2 mm oamm aamm aaam mmam mmom om mm mm mm mm ooa mm ma a oa-m 2 ma mmom omom mmam aama maam em om ma mm om ooa am m a m-m 2 ma mama mmma mama mmm mmm mm mm mm ooa ooa - ooa m m 3-: m aa omma mmma mmaa mmma aama om. ooa ooa ooa mm . ooa m m m.: 2 ma mom 3m mos mom we mom mom mow mom mo m a s a v s s 023. soapomom Hmpoe moaoom A M 3 mm“ m. m. soapmsasaaoman m pm. m... a "w a m. 1. I. m. T. # w m. m S K mam mmma - mma - - moaa ama mmm ama mmm mm as: mmam . mmm mmaa moma maaa mmoa mam mmma mmma mm mam mmaa mmma amm: mmma aaaa mmma mmma ammm amma moma mm mam ammm ooma mmmm . - mmma mmma amaa mmma coma am mom amma mmaa - - . mmma amaa maaa moma mmaa mm mama aama moma - - - mama mmaa maaa mmma mmma mm mmm moam mmoa ama mmoa omm mmma mooa mma mmma amm mm mmm moaa aam mmam scam mmm mama mmma mmma mama oam mm mmm ommm - - - . omm oam mmaa ama mmm mm mas mmm: - mama - - mmoa maaa maoa mmm mmma am amm maaa maom - mama mmm mama omaa mmma mmaa smoa om mmm moaa mmmm osaa - momm mmam aamm ammm mmam maom ma mom ammm ammm mmaa mmam mmm: omaa mmma mama mmma moma ma mmm aomm ammm mmom . - amma amaa mmma .mmm mmm aa mm: mmam - - . amaa mmma mmma mesa mmma mama ma 1 s. Plfilal s 8 mm. gm“... mm: s m 1. an" mmma mama amaa mmoa aama mm mm mm ooa mm - ooa ma m o-m m ma maaa mmoa mmoa moma aam ooa ooa mm mm mm - ooa m a aaua 2 as mmm moa mmm mmm mm: ooa ooa ooa mm ooa ooa mm a a :.m m m: mama mmma amma mmma mmma mm ooa mm ooa mm . ooa a m m-a 2 ms aama mmaa amma amoa amm mm ooa ooa ooa ooa . ooa oa a sum 2 as aama mesa mmma mama mmoa mm mm ooa ooa mm - ooa oa m m-m 2 o: mmaa amma mama mmma mmoa mm mm ooa ooa mm . ooa m m mum 2 mm mmmm mmma mmaa moom mmaa mm mm ooa ooa mm . ooa m m mum 2 mm mmma moaa mmma soma omoa ooa mm mm ooa ooa - ooa m m m-m 2 am mmaa mmaa mmma aaaa mooa mm mm .mm ooa mm - ooa a a aa-m 2 mm mmaa mmma mmma amm mmm mm am mm mm mm ooa mm m m mum m mm amma mmoa aooa mmm aam mm ooa mm ooa ooa . ooa oa a onm m :m aamm omaa mmaa aama smoa mm mm mm mm ooa - ooa ma a a-m m mm aam mmm amm mmm mma mm ooa mm mm mm - ooa 2a a m-m 2 mm mmaa mama mmma mmma omma mm om mm mm mm ooa mm ma 2 oaum m am mom mom $3 Sam 20 mom mom was aam 3 m a s a v s s 959 coapommm ages 3.88 M m m a.“ m m. 203222259039 u. m M... a a. O T: “M. .m. 1. ms 3 av S 1. .A mmm mmmm mmma aama . aamm amma aama maaa mmoa maaa m: mmm . . mmm mmm: mmma mmaa mmoa mmoa oaoa mam a: mma : u : maaa : mmm moa mmm mmm mm: m: mma mmam - mmm - aaoa omaa mmma amma mmma mmma m: mam mmam - - u . aama mmaa amma amoa amm as amm . mmaa - mama omma maaa amaa mmaa omma mama o: om: mmma amma - - mmam mmaa moma mama mmma mmm mm mam mmom mmaa - . aaam soma amma mmaa moom mama mm mmm . mmm soaa - - mmma mmaa mmma mmma omoa am mmm momm aam aamm - aam mmma mmma mmaa aaaa oaoa mm amm maa oaaa omma aaaa mam amaa omma mama mam mmm mm ama maom - amm - : mmaa mmoa mmoa mmm aam am aam amoa moom omam aama . mmom mmma omma maoa smoa mm mmm mma - mmm oam maa mmm mmm mam mom ama mm amm mmaa mmom mam mmmm momm maaa mmaa moma ooma amaa am -ommfimmm as afiwmamf mmma s .8 may 2%... mate 8 m m .+ # mmma aama mmm mmm mom am ooa ooa ooa ooa - ooa m m cum 2 om mmma amoa mmma moma mmma ooa mm mm ooa mm - ooa : m aum 2 mm oama ooaa amaa mmma mmm ooa mm mm mm ooa . ooa m a m-a m mm omma mamm amma omma mmma om am mm mm ooa - ooa m m m-a m am mmoa mmm mmm mma omm mm ooa mm ooa ooa - ooa s a :-m 2 mm ama mma mam mmm mmm mm mm mm ooa ooa - ooa oa m aa-a 2 mm mmma amm amm amm mma mm mm mm mm mm ooa mm a m a-m m :m maoa mmaa mmm mmma mam ooa mm mm mm ooa - ooa : m oaua 2 mm amm mmm ama mmm mmm mm mm ooa mm ooa - ooa s m mnm m mm amaa mam mom mmm mmm mm mm ooa ooa ooa . ooa a a mum m am ama mom amm maa amm mm ooa ooa ooa mm - ooa ma a oa-a 2 om mma moaa mam amm mmm mm ooa mm mm mm . ooa m a aum m m: amma mesa mmaa aama amaa om mm ooa mm ooa ooa mm a a m-a 2 m: maaa aam mmoa smoa mma mm mm ooa mm ooa - ooa : a m-a m as maaa mmoa amm omoa amm ooa mm mm ooa ooa . ooa m m aaua a m: mom mom mob mom mo mom mow mos mom mo m (a s a v s s maam 58.3mm ages 3.83 M w m a m m. sompmsmpfiuomma I Hm we. 1 a a. O I. .1. n 1. B a T. I s m a.” S 1. .A :omom mamsam maaa soapomom hogan mama Compomom poounoo am: omom . - - . :mma mama mmm mmm mom om ama - maa amoa - moma mmma amoa mama moma aama mm mma - ooooa mmm mmm - oama mmma maaa maaa mmm mm mmm amma mmmm maaa moaa - maaa mmam mmma maaa mmma am am: mmmm - aaa - . mmm mmm mmm mma omm mm mm: amm maaa ma: - . mom mma mmm mmm mmm mm ea: mmm: mmm . amm mo: mmaa amm amm maa mom am mmm . mamm mmma ama : maoa mama mmm amma mam mm as: oaaa aam - mma - mmm mmm ama amm mmm mm mam ammm moa - - - mmaa mmm mom mmm mmm am omm mom - - - mm: aaa mom amm maa oam om amm aaaa - oma mom am: maa moaa amm mmm mmm ms aam mmmm amm - aoom : mmma mmaa mmaa mmaa amaa m: amm mmom mmaa - aaam - mmoa amm mmoa omm mma a: mm: - oam mmm . - maaa mmoa aam omoa amm m: nlmmaa,coap mom mom. mos mom mo mom mmm mmm mom mo # ioefqns APPENDIX H INDIVIDUAL DATA - HEARING-IMPAIRED SUBJECTS mmma mmma amma aama mmmm as mm mm mm om - ooa om mm mm m m-ma m ma momm mmma amma amma aam em ma ma mm mm - ooa m.mm m.mm om a m-ma m ma amma aama aoma maaa maaa mm mm ma ma em - ooa m.am a.am mm m m-m m ma maam ammm msom mama mmom mm om om mm mm - ooa om a.am oa a m-m 2 ma aama smoa mmma mmoa mmma om am ma em ma . ooa ma ma mm m suoa m aa amaa mmm mmm aam mmaa mm mm ma am ma . ooa m.am a.am mm a oa-m m oa mama mama mamm aoma ammm m ma ma mm m: - ooa m.aa m.am em a m-m m m: msom mmom mmmm mmmm mom: ma mm mm m: om . ooa m.am m.am ooa m a-oa m m* mmam mmmm ammm mmma mmma mm mm om mm mm - ooa om mm m: a a-m m a maaa omma mmaa aaaa mmoa mm mm mm mm ma - ooa mm am me a m-aa m m aamm ammm ommm ammm mmaa mm mm mm ma ma ooa am m.am 0: mm m m.: 2 m mmma mmma aooa mama mmaa mm mm mm mm mm ooa mm mm mm mm m m-m 2 : moma mmma aama mmaa amoa mm mm mm mm ooa - ooa m.am 0: mm m m-a 2 m mmm mmaa mmm aama amm mm mm om mm mm ooa mm mm mm om m aauma m m ommm ammm momm mmam ammm mm ma om om om mm mm m.aa om ma m m-m m a mom mom mo: Nod mo om Rom mo: mom mo m a z E s a v s s 923 compomom H.308 moaoom M .r .r M m saw m m: comparfismuommn d o J J f. 3 m m m m T.n.+n o 0 4+ B a T. I; m m u. s 1K V V :ommm mamsam maam sompomom aoaaw maaa coHPQMmm Pomupoo mmm mama oamm aamm mmma ammm amaa moma mmaa mmma mama ma mm: mamm maaa mmmm mmm mmm maam moaa mmm mmma mmm ea amm mmma mmmm mmmm mmma maom amma mmma aoma mmaa amaa ma mma aamm mmm: mmam amam aaoa moom maaa mmma mmaa omam ma mom aoma amm mmaa omaa mmom mmm maaa mmma mmm mmm aa mas amaa mmm mmm mmam maaa mmaa omm mam maa amoa oa mmm mmaa mama ammm mama mac: maom mmma amam mmaa aoam m. mmm mmm: mamm aamm osmm mmmm mmam mmmm ammm amam amam m. am: mmom mmam mamm mmmm mmmm ammm mmmm mmam amma amma a mam mmma mama aama mmma mmm maoa mama amoa mmaa oaaa m mmm mmmm maam aamm mamm aamm amma maam osom ooma mmaa m mmm amma maaa mama mmaa mmaa mmma aama oam mmma aama : amm mmmm aama moa mmmm - mmaa maaa mmma maaa amoa m amm ama mmaa mama mam mm: aooa amaa mmm maoa aam m mmaa ammm mmmm omam mmmm moam momm aama mmmm amma omom a maaa soap wmom mom mom mom mo mom mom (mos , jammy mo # 109F908 ammm so: masoo n 920 mammamcm «Pam 80am umppmso mpommpzm * HNmN mmaN mmma mmma mNQH Nm Nm mm mm OOfi : OOH mm m.mm Nu m 0:0H m AN mamm mmaa moma mmam mmmm mm mm ma 0: mm 920 om om mm m m-: 2 om: mmHN aomN ammm mmaa omia mm 00 we om mm : ooa m.Nm m.mm :0 m H:oa m ma maam amma ommm mmmm moma mm mm mm om mm mm om m.mm m.mm m: a a-m 2 ma mmNH mom mNom mooa now :m cm No mm mm : ooa m.Nm m.mm mm H HH:& 2 ma NmNH :ma mmma :dm 0mm Nm mm am No mm 00H mm m.mm m.ad on A HH:@ m om lRom Row R0: R0“ 1R0 Rdm mow “mo: mom R0 N H 2 rt 8 as V S S mama compomom ampoe moaoom m“ us ..a M m mu m m. soapmcmsanoman aaa J 1 n! .mxae a a e . i.ena .b .0 o tg+n . . 1. a.aTL Tque “m "m u” s A v v mam mmmm mamm mmam aaam - maam mmaa mmma mmaa mmma am ama mamm mmaa mama mamm amm: mamm mmma mmma ammm mmmm mm* ma: amam mmam mmma mmaa amma maaa aama mmma mmaa mmma ma mmma ammm ammm mmm: mmam mmma mamm maaa mamm mmmm mmma ma mmm mmmm mam mam mamm mma mmm mmm amma mama mmm aa amm mama amma mmma mmm mmm mmma mmm amma mam mmm ma maam soap mmm mmm (Immm mmm am mmm ramm mmm mmm mm :omom mmmsmm mama Compomom poppm mEHLH. COHPmem PU®.H.HOO # 1oafqns IIIIIIIII Jfilmjlufijfiljfljrilflfytxwm