. . .s... 11.1.4: ,.‘..¢..4::.1.t. .. {.1va k1. 7:1,-.. 4. .1. 3. .A wwnruhngfgflkrig u 1: y..n.l\.i..¢!nruusiaw WE... ~ 74.... p . 41474264.; .. .15.. . .r? y .92.... 21".... r . nu): ... . . qu.w.ax.xw$..zwn . .N..;,...,,........,,, 4 Hwy/4..- v. . .1 . :r to .147 D." ‘Yvr . a... .fimmwfi : v 7.412.: O’clv}. p o. {.11. ('4... r . . , . $4.... I“ .z. :. 32:3. . . 25.3.... MW. .... ......L..;u...r. .. if. 1 1.4.4.15. ., €737? . u. [arm/3.. r. {Iran’s .42. by”, Ladydfiuw a...» .?.:x....... E: . 9.: .r .3» .11. .(.x..... v I .fvr .A\ «r. 1.1.... . lo. 1.71 11 V . . .s . .114... . .tnr: 5.: n. 833... 4.. n. 4...... f... 595 I; r J v. 9’. X aflqJI-vaulr J! '7 >.’.'.- '1 I. . . tril— r rhuJallnroiml p. 7- a “'1’! nfll 1 if. O .V- .r . 1 . 15.. J. 1.: tar“! . 'a I '1‘), . I!" Y. a) —\t 1.01.7.11-(7101 >0 ,, (r.1.rq.v!..<$«..flu yo h .b):. :1...‘ c u 1 . I, ‘1'!!!" . — fr.- 7""! 1.04. r I. u .irdplr 1’ ti. .3. v .2 2.}; . S "—J 1' 151.3... hg-IrV'f'luv .r u. a .. , guru: (I 21:! ‘ .33. v. y .> ~11) . . , ‘ anailuftnrr... , . . ‘ A . (cunt! {obi Ito-"r.” (.31.. , .{l' .3 I I . .1 a.» 5.. I) 4 . Vuiro.:ffa.| .4. 1' drawn... w.))1....) .a... x. . . . 40‘}?r:!rlrvorolf . . (.973...) (ff , , . , trip 3" r.r!rr1r.l .V.. r .. .1.-- , . 55-..... 3.. .. ,ln ‘ :3: 3.5.. 9 ....:.53..S::;.. , . 6101:1310! I!!! In"?! it}. .. :1 : .7. . l...a.l:.vf1...r¢I-‘vr V ti??? I . .l .t. r , lyfaz‘sirrl ‘11rrlrfrul :r...: a .n a I r l. rVY—fiIAV'I'N’UM Ivrrdl. ti... 2.. . ,....:.1. . . a I : v. . ? a v3.1.3 . titegfrl (or. ... 9!! (.54); .1. . .34.... .. -V4\In 1‘" :5 rr 2.: . . .rr/rr :rrr...‘ .. .5»:.. r, , 3.1.rrtivlr .rl- cryfir £3.”qu 5.5;? 4? r If! I." . . yr 3:41.; i; I}. . .Ifrra... l.’... .2 .. v5...) 4..- , I f .. 255M / +. a: I :23)?" .7 ..:rv.r..{.;.« I :...:.. Ill». ....vlv.\.:.:|\.......knt1€ 1 : . 7. 4.55311» .. ., . , . «I; . , tiny 1,. .1: Riki“ .... q a. é . . V , . ,. , . ‘ . I x, [.2 5. .mwwaumrgxbgu 1 egg“... _ LIBRARY hitchagan.8¢age University This is to certify that the thesis entitled lntelligibility of Time-Altered CNC Monosyllables As A Function of Contralateral Masking presented by Daniel F. Konkle has been accepted towards fulfillment of the requirements for Ph.D. (19mm, inAudiology and Speech Sciences \ NW % \ Major professor 2-26- 6 Date 7 0-7639 ABSTRACT INTELLIGIBILITY OF TIME-ALTERED CNC MONOSYLLABLES AS A FUNCTION OF CONTRALATERAL MASKING BY Dan F. Konkle Previous research has suggested that time compressed versions of the Northwestern University Auditory Test No. 6 (NU-6) of speech discrimination had clinical potential as a measure of central auditory function. Clinical audiometric procedures, however, frequently require the use of contra- lateral masking in order to minimize non-test ear influences on the behavioral responses obtained from the ear under test. Currently, the effect of contralateral masking on the in- telligibility of time compressed NU—6 words is unknown. The purpose of this investigation, therefore, was to examine the relationship between contralateral masking and the intelli- gibility of temporally accelerated NU-6 word lists. Ninety normal hearing young adult listeners were admin- istered Form B of the NU-6 test (time compressed by 0%, 30%, and 60%) under four different types of contralateral maskers (white noise, speech noise, a four chain multitalker passage, and the multitalker passage 60% time compressed). The intensity level of the NU-6 words was held constant at Dan F. Konkle 75 dB SPL; whereas, contralateral maskers were presented at 30 dB, 60 dB, and 90 dB SPL. An equal number of right and left ears were tested for each experimental condition. Subjects' responses were converted to percent correct scores and submitted to a multiple linear regression analysis with percent correct score regressed on masker intensity level, degree of time compression, ear, masker type, and combinations of these basic variables. Results from this investigation revealed that neither white noise nor speech noise had a significant effect upon NU-6 word intelligibility, regardless of the degree of NU-6 time-compression or the intensity level of the masker. Conversely, the multitalker maskers were found to cause a statistically significant decrease in percent correct scores when these stimuli were combined with other variables (i.e., time-compression, masker intensity level, and ear). The findings from this investigation support the clinical utility of white noise and speech noise as contralateral maskers of time compressed speech stimuli. Additional re- search, however, should be conducted in order to generalize such observations to pathological populations. Results observed with the multitalker maskers served to illustrate a speech perceptual process characterized by an input filter. Moreover, the significant reduction in percent correct scores noted when multitalker maskers were combined with other variables suggested that input filtering may be bound by temporal parameters. INTELLIGIBILITY OF TIME-ALTERED CNC MONOSYLLABLES AS A FUNCTION OF CONTRALATERAL MASKING BY I \ DanwFfLKonkle 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 1976 ACKNOWLEDGMENTS The cumulative efforts of numerous individuals have resulted in the successful completion of this dissertation. Unfortunately, it would be impossible to acknowledge everyone who made either a direct or indirect contribution. There are, however, several individuals whose influence warrants special consideration. First, I would like to express my appreciation to Drs. Daniel S. Beasley and William F. Rintelmann, whose interest in the clinical applications of time compressed Speech led to the formulation of this investigation. The committee that guided this dissertation: Dr. Daniel S. Beasley, Chairman; Dr. Fred H. Bess; Dr. Linda L. Smith; and Dr. Joseph Woefel all provided critical evaluation that was both valuable and constructive. Special thanks needs to be extended to Dr. Bess for his encouragements throughout my professional training. Finally, my most sincere acknowledgment to my wife, Pam, and my two boys, Matt and Fred, who sacrificed so much in order that I might complete this endeavor. TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . Chapter I. INTRODUCTION . . . . . . . . . . . . . . . Time-Compressed Speech as a Test of Central Auditory Function . . . . . . Contralateral Masking and Speech Intelligibility . . . . . . . . . . . Summary and Statement of the Problem . . II. EXPERIMENTAL METHODS . . . . . . . . . . . Subjects . . . . . . . . . . . . . . . . Stimuli . . . . . . . . . . . . . . . . Apparatus . . . . . . . . . . . . . . . Procedures . . . . . . . . . . . . . . . Analysis . . . . . . . . . . . . . . . . III 0 RESULTS 0 O 0 O O O O O O O O O O O O O O 0 White Noise and Speech Noise as Contralateral Maskers . . . . . . . Multitalker Stimuli as Contralateral Masker . . . . . . . . . . . Multitalker 60% Time-Compressed Stimuli as Contralateral Masker . . . . . . . Summary . . . . . . . . . . . . . . . . IV. DISCUSSION 0 O O O O O O O O O O O O O O 0 Implications for Clinical Audiology . . Implications for Speech Perception . . . Implications for Further Research . . . 1%]. “/1V Page Vi vii 12 17 20 20 21 24 26 28 3O 32 42 45 47 50 50 51 56 Page LIST OF REFERENCES . . . . . . . . . . . . . . . . . 58 APPENDICES Appendix A. Method and Procedure Used to Obtain Frequency Spectra of Masking Stimuli . . . 64 B. Standardized Instructions Provided to Subjects Prior to Experimental seSSions O O O O O O O I O O O O O O I O O 66 C. Answer Form . . . . . . . . . . . . . . . . 67 D. Equations Used in Multiple Linear Regression AnaIYSis O O O O O O O O O C O O O O O O O 69 E. Subjects' Raw Scores for Each Experimental condition 0 O O O O O O O O O O O O O O O 7 0 F. Table of Mean Percent Correct Scores for Each Experimental Condition . . . . . 74 LIST OF TABLES Table Page 1. Zero—Order Correlations for Percent Correct Score Regressed on Independent Variables . . . . . . . . . . . . 31 2. Analysis of Variance, Multiple R, and R-squared for Percent Correct Score Regressed on Independent Variables for Equations I and II through Step Eight . . . . 33 3. Zero-Order through Seventh-Order Standardized Beta Weights (B) and F Values of Percent Correct Score Regressed on Independent Variables for Equation I . . . . . . . . . . . . . . . . . 35 4. Zero—Order through Seventh-Order Standardized Beta Weights (B) and F Values of Percent Correct Score Regressed on Independent Variables for Equation II . . . . . . . . . . . . . . . . . 37 5. Average Percent Correct Scores for each Time-Compression Ratio, Masker Type, Masker Intensity Level, and Right and Left Ears . . . . . . . . . . . . . . . . . . 74 vi LIST OF FIGURES Figure Page 1. Frequency spectra of four maskers (WN = white noise, SN = speech noise, MT = multitalker, and MT 60 = multitalker 60% time-compressed) used as contralateral masking stimuli . . . . . . . . . . . . . . . 23 2. Schematic of apparatus used to present experimental stimuli . . . . . . . . . . . . 25 3. Experimental design . . . . . . . . . . . . . . 27 4. Mean percent correct scores for 0% time— compression as a function of masker type and intensity. The Beasley et al. (1972a) data without contralateral masking are shown for comparison . . . . . . . . . . . . 39 5. Mean percent correct scores for 30% time- compression as a function of masker type and intensity. The Beasley et al. (1972a) data without contralateral masking are shown for comparison . . . . . . 40 6. Mean percent correct scores for 60% time-compression as a function of masker type and intensity. The Beasley et al. (1972a) data without contralateral masking are shown for comparison . . . . . . . . . . . . . . . . . 41 7. Mean percent correct scores for right and left ears as a function of the intensity of a white noise contralateral masker . . . . 43 8. Mean percent correct scores for right and left ears as a function of the intensity of a speech noise contralateral masker . . . 44 vii Figure Page 9. Mean percent correct scores for right and left ears as a function of the intensity of a multitalker 0% time- compressed contralateral masker . . . . . . . 46 10. Mean percent correct scores for right and left ears as a function of the intensity of a multitalker 60% time- compressed contralateral masker . . . . . . . 48 11. Filter Model of Speech Processing proposed by Broadbent (1958). . . . . . . . . 52 viii CHAPTER I INTRODUCTION The Northwestern University Auditory Test No. 6 (NU-6), originally developed by Tillman and Carhart (1966) as a measure of phonemic discrimination, has been used in several recent investigations concerned with temporal acceleration (time-c0mpression) and speech intelligibility. These inves- tigations have provided information about the effects of time compression on the intelligibility of NU—6 stimuli for normal hearing young adults (Beasley, Forman, and Rintelmann, 1972a; Beasley, Schwimmer, and Rintelmann, 1972b; Riensche, Konkle, and Beasley, 1975), aging persons who possessed normal speech discrimination for non-accelerated stimuli (Konkle, Beasley, and Bess, 1975), and adults with sensorineural hearing loss (Kurdziel, Rintelmann, and Beasley, 1975). Based on these findings, it has been recommended that time-compressed versions of the NU-6 be used as a clinical test of central auditory function. In clinical populations, however, it is frequently necessary to use contralateral masking in order to avoid the influence of non—test ear participation on audiometric results. Since there have been no published reports concerned with the effects of contralateral masking on the intelligibility of 1 time compressed NU—6 words, the interaction between masking the non-test ear and performance with time-compressed NU-6 stimuli are unknown. It was the purpose of this investigation, therefore, to examine the effects of contralateral masking on the intelligibility of temporally accelerated NU-6 word lists. Time—Compressed Speech as a Test of Central Auditory Function The central auditory system (CAS), defined by Jerger (1973, p. 86) "... as that portion of the total auditory system lying within the central nervous system," is comprised of a complex arrangement of neurological structures. Several investigators (Jerger, 1960, 1964; Katz, 1969; Willeford, 1969; Bergman, 1971) have noted that the complexity of the CAS provided an internal redundancy such that pure tone and conventional speech audiometric stimuli resulted in near normal responses despite neurological damage to the central auditory system. Consequently, investigators have sought to develop measures specifically designed to evaluate central auditory function. Research on the development of central auditory tests has revealed that speech stimuli, modified to reduce external message redundancy, were the most suitable means to evaluate central auditory processing (Bocca, Calearo, and Cassinari, 1954; Bocca, Calearo, Cassinari, and Migliavacca, 1955; Bocca, 1955, 1956, 1958, 1961; Calearo, 1957; Calearo and Lazzaroni, 1957; Calearo and Antonelli, 1963). These inves- tigators observed that when external redundancy was reduced, the normal CAS could still process the stimulus with a high degree of intelligibility. When there was a central auditory lesion, however, central processing appeared to become over— taxed by the more complex material, and as a result, there was a breakdown in intelligibility. This phenomenon led Jerger (1960) to propose that central auditory processing was characterized by a "subtlety principle" whereby more complex (subtle) stimuli were necessary to detect higher central auditory lesions than were required to locate lesions at lower levels. The majority of central auditory tests developed to date have employed some form of distortion to reduce external speech redundancy. Such stimuli have included filtered speech (Jerger, 1960, 1964; Matzker, 1962; Speaks and Jerger, 1965), periodically switched speech (Bocca and Calearo, 1963; Calearo and DiMitri, 1958; Calearo, Teatini, and Pestalozza, 1962), and interrupted speech (Bocca and Calearo, 1963). In addition, it has been shown that temporal parameters play a major role in speech perception (Aaronson, 1967; Beasley and Shriner, 1971; Hirsh, 1967). Thus, temporally altered speech in the form of time compression has been suggested as a unique means to assess central auditory problems. Calearo and Lazzaroni (1957) examined the effects of presentation level and syllable rate on the intelligibility of a list of short significant sentences presented to normal and presbycusic subjects. ‘Sentences were recorded at rates of 140, 250, and 350 words per minute (me) and articulation functions obtained on each group of subjects. Results in- dicated that articulation functions for the normal group were shifted by 5 to 10 dB when sentences were accelerated from 140 wpm to 250 wpm and 10 to 15 dB when presented at 350 wpm. For presbycusic subjects, a 30 dB shift was observed at 250 me and thresholds could not be established at maximum intensity levels when sentences were presented at 350 wpm. The three articulation functions for normal subjects remained essentially parallel, indicating that loss of intelligibility associated with speech acceleration could be restored by an increase in intensity. Calearo and Lazzaroni also reported that when accelerated sentences were presented to a group of subjects with temporal lobe damage, articulation scores were poorer in the ear contra— lateral to the lesion when compared to the homolateral ear. The findings of this investigation are difficult to interpret because Calearo and Lazzaroni failed to describe the exact nature of their speech stimuli, qualify their subject samples, or provide a description of the methodology used to time-compress speech materials. Nevertheless, two general conclusions may be drawn from their results. First, presbycusic individuals had a special difficulty processing time—compressed speech stimuli; and secondly, cortical damage resulted in poorer discrimination ability when stimuli were presented to the ear contralateral to the site-of—lesion. deQuiros (1964) conducted an investigation similar to that of Calearo and Lazzaroni. deQuiros presented an accelerated speech task to twenty normal hearing persons, fifteen subjects with peripheral hearing loss (5 conductive and 10 cochlear), seven subjects labelled as presbycusic, six persons with retrocochlear lesions, and twenty-eight subjects with central auditory pathology. Speech stimuli consisted of sentences (ten words in length) presented to subjects at rates of 140, 250, and 350 wpm. Apparently, speech stimuli were presented to each subject via a monitored live voice technique. Resultant data were examined with respect to the shape of articulation functions, speech detec— tion and reception thresholds, and maximum articulation scores. deQuiros noted that normal hearing subjects usually demonstrated a common articulation curve for each presentation rate. An approximate 10 dB shift was observed for successive word rates, with the exception of detection thresholds, which tended to remain constant. Subjects with conductive hearing loss either obtained results similar to normals, or the entire family of curves was shifted by an amount consistent with the degree of conductive involvement. Cochlear impaired subjects occasionally obtained normal results, but more frequently, the maximum score at 350 wpm was lower than that obtained for normal subjects. Results for presbycusics were similar to those reported by Calearo and Lazzaroni. Unlike the other subject groups, however, presbycusics showed a shift in the threshold of detectability for all presentation rates. For subjects with central auditory pathology, it was found that a bilateral shift in threshold occurred when intercranial pressure resulted from an extratemporal tumor. A bilateral threshold shift was also observed when inter- cranial pressure was due to a temporal lobe tumor, but was more severe in the ear contralateral to the lesion. For extratemporal tumors without changes in intercranial pressure, the threshold shift was small and symmetrical. Based on these findings, deQuiros concluded that accelerated speech testing may provide information that "... when correlated with other findings - may aid in pinpointing the sites of brain lesions, especially those within the temporal lobe" (p. 40). The findings reported by Calearo and Lazzaroni and by deQuiros resulted in several investigations designed to study the responses of aged subjects to time-compressed speech. Luterman, Welsh, and Melrose (1966) found that time-compressing and expanding monosyllabic words (CID W—22's) (Hirsh et a1., 1952) by 10% and 20% of normal duration had a slight effect on overall intelligibility, but there were no significant differences between the performance of young normal hearing subjects, young sensorineurally impaired per- sons, and aged subjects with sensorineural hearing loss. Conversely, Sticht and Gray (1969) noted that time-compression had a differential effect upon the perceptual performance of young and aged persons. They observed that when CID W—22 words were presented at compression levels of 36, 46, and 59% to young and aged listeners matched for hearing sensiti- vity, younger listeners performed better than aged persons and that the difference in performance became greater as the amount of time compression was increased. In addition, they reported that while aged subjects experienced difficulty with time-compressed speech, there were no differential effects on performance due to hearing loss. Sticht and Gray concluded that the detrimental effects of time compression on the speech discrimination scores of aging individuals reflected changes within the central rather than peripheral auditory system. Schon (1970) presented 25-word CID W-22 lists that were compressed by 30% and 50% to groups of young subjects with normal hearing, aged subjects with normal hearing for their age, young subjects with sensorineural hearing loss, aged subjects with sensorineural loss, and young subjects with normal hearing up to 3000 Hz, but a hearing loss of 40 dB at 4000 Hz. In general, Schon's findings supported Sticht and Gray in that aged individuals experienced difficulty in the discrimination of time compressed speech. Unlike Sticht and Gray, however, Schon stated that both peripheral and central auditory factors contributed to the decreased per- formance of aged persons. DiCarlo and Taub (1972) measured the word intelligibility of twenty young and twenty aged aphasic subjects on the same word lists (ZS-word CID W-22's) under the same time compression ratios (30% and 50%) employed by Schon. Their results indicated that temporally accelerated speech stimuli were effective in distinguishing between cen- tral (i.e., aphasic subjects) and peripheral auditory problems, but were less effective in differentiating lesions related to the peripheral auditory mechanism. Although the findings concerned with the intelligibility of time compressed speech and aging populations has been equivocal, the afore-going research has illustrated that temporally accelerated speech provides a unique stimuli for central auditory testing. It is important to note, however, that results from this series of investigations did not provide adequate data for clinical diagnostic purposes. For example, investigators either failed to use suitable time compression procedures (Calearo and Lazzaroni, 1957; deQuiros, 1964), neglected to match subject groups for perceptual ability under normal listening conditions (i.e., undistorted speech) (Schon, 1970; DiCarlo and Taub, 1972), or used small subject samples (Sticht and Gray, 1969). Moreover, the half- list revisions of the CID W-22 word lists (Campbell, 1965) used by Schon and DiCarlo and Taub have not received stan- dardization for reliability and equivalency for time compressed conditions. In order to obtain a clinically useful test, procedures for time-compression must be well defined and standardized test materials should be used as stimuli. Since the CID W-22 speech discrimination test had been criticized as being too easy (Carhart, 1965), several inves- tigators have examined the effects of time compression on the NU-6 test of speech discrimination. Beasley, Schwimmer, and Rintelmann (1972b) obtained intelligibility scores for ninety— six normal hearing young adults on Form B of the NU-6. They presented stimuli at six time compression ratios (0, 30, 40, 50, 60, and 70%) and at four sensation levels (8, 16, 24, and 32 dB) to an equal number of right and left ears. Results indicated that discrimination performance decreased gradually until the 70% time compression condition, at which point a precipitous drop in intelligibility was noted. As sensation level was increased intelligibility scores also increased and there were no significant differences in performance between ears. Beasley, Forman, and Rintelmann (l972a) ex- panded the data of Beasley et al. (1972b) to a 40 dB sensation level and found a slight, non-significant improvement in discrimination scores. Beasley et al. (1972a) concluded that since "... normative speech discrimination data is available at five intensity levels for six time compression conditions, it is now possible to compare meaningfully the performance of patients with auditory pathology to normal Ss" (p. 74). Kurdziel and Noffsinger (1973) used the normative data of Beasley et a1. (l972a,b) to investigate the performance of adults with cortical (temporal lobe) brain damage. They reported scores for NU—6 stimuli time-compressed by 60% to be significantly poorer in the ear contralateral to the brain damage as compared to scores for the ipsilateral ear. These findings lend support to the observations of Calearo and Lazzaroni (1957) and deQuiros (1964) who also noted scores to be poorer in the ear contralateral to cortical damage. In an attempt to resolve the equivocal results reported for time compressed speech in aging populations, Konkle, 10 Beasley, and Bess (1975) administered time compressed NU-6 lists to 118 aging subjects. Their subjects, who had excellent discrimination ability for NU-6 words under normal conditions (i.e., 0% time-compression), were divided into four age groups (54-60 years, 61-67 years, 68-74 years, and 75 years of age and older). Experimental stimuli were presented at O, 20, 40, and 60% time-compression and at sensation levels of 24, 32, and 40 dB to an equal number of right and left ears and male and female subjects. Konkle et al. found that intelligibility scores became poorer as a function of increased age and time compression ratio, but that increased sensation level caused scores to improve. Since their subjects demonstrated normal discrimination ability at 0% time-compression, Konkle et a1. argued that differences in performance between the four aging groups could not be attributed to peripheral auditory function. Consequently, Konkle et al. concluded that the differential performance may be due to changes within the central auditory system associated with the aging process. Only one investigation has been reported that was de- signed to examine the effects of time compressed NU-6 lists on the discrimination ability of individuals with peripheral auditory lesion. Kurdziel, Rintelmann, and Beasley (1975) presented NU-6 monosyllables at six levels of time compression (0, 30, 40, 50, 60, and 70%) under four sensation levels (16, 24, 32, and 40 dB) to a group of nine subjects with bilateral noise-induced hearing impairments. Results indi- 11 cated that the effects of time compression on discrimination ability were similar to those reported for normal hearing subjects by Beasley et al. (1972a,b). That is, performance gradually decreased with increased time-compression up to 60% acceleration, and a dramatic breakdown in scores occurred at 70% time-compression. Maximum discrimination performance was obtained at the 32 dB sensation level and a slight roll- over phenomenon was observed when stimuli were presented at the highest sensation level (40 dB). This latter finding stressed the importance of presenting stimuli at several intensity levels in order to determine maximum performance. The review of literature concerned with time compressed speech as a test of central auditory function has revealed that accelerated speech stimuli may be useful in delineating problems in the CAS from those confined to the peripheral auditory mechanisms. More specifically, it appears that sufficient data exists for time compressed versions of the NU-6 to receive serious consideration as a clinical diagnostic tool. It is important to note, however, that every investigation reported to date has presented time compressed NU-6 word lists monaurally under earphones. Numerous investigators (Konig, 1962a,b; Naunton, 1960; Palva, 1954, 1958, 1962; Zwislocki, 1951, 1953) have demonstrated that an auditory signal presented to one ear may be perceived in the other ear whenever the level of presentation exceeds interaural attenuation. This phenomenon has been commonly termed "cross hearing". For 12 air conducted speech stimuli presented via earphones the minimal interaural attenuation has been found to be approximately 40 dB. Since stimuli presented to one ear by an earphone is transduced to the opposite cochlea pri- marily by bone conduction, Studebaker (1967) has recommended that the non-test ear be masked whenever the presentation level in the test ear exceeds bone conduction sensitivity of the non-test ear by 40 dB. Contralateral masking (i.e., masking the non-test ear) was not used in any of the investigations employing time-compressed NU-6 stimuli. Hence, it is possible that the results from these investi- gations may have reflected non-test ear participation. Contralateral Masking and Speech Intelligibility The primary purpose of contralateral masking has been to elevate the threshold sensitivity in the non-test ear to avoid the influence of "cross hearing" on responses obtained from the test ear. Research has shown, however, that contra— lateral masking may also affect the responses from the test ear. The majority of research in this area has been concen- trated on the effect of a contralateral masker on various measures of pure tone threshold. Numerous investigators (Dirks, 1964; Dirks and Malmquist, 1965; Dirks and Norris, 1966; Ingham, 1959; Sherrick and Mangabeira—Albernaz, 1961; Treisman, 1963; Wegel and Lane, 1924; Zwislocki, 1953; 13 Zwislocki, Damianopoulous, Buining, and Glantz, 1967; Zwislocki, Buining, and Glantz, 1968) have reported changes in the pure tone threshold sensitivity Of the test ear when low-intensity masking was delivered to the non-test ear. This effect has been generally referred to as "central masking" and attributed to central rather than peripheral auditory processes. Conversely, a paucity of information has been published directly concerned with the effect of contralateral noise on speech audiometric results. Martin, Bailey, and Pappas (1965) observed that speech reception thresholds measured via a Bekesy tracking procedure shifted 5 to 8 dB when white noise was administered to the non-test ear at a sensation level of 75 dB. In a follow-up study, Martin (1966) examined the effect of a white noise contralateral masker on speech reception thresholds determined by CID W—l spondee words and on discrimination scores obtained with PB word lists. Results of this investigation indicated a shift in speech reception thresholds of approximately 5 dB with essentially no change in discrimination scores, except in cases where "cross hearing" could be expected. More recently, Young and Harbert (1970) obtained dis- crimination scores with CID W-22's in the presence of ipsi— lateral and contralateral white noise maskers in seven normal hearing subjects, sixty-five subjects with total hearing loss in one ear and normal hearing in the opposite ear, and fifteen subjects with bilateral symmetrical hearing loss. Their 14 results indicated that contralateral masking had little effect on discrimination scores until intensity levels were reached that caused the masker to cross to the test ear, thereby reducing the S/N ratio in the test ear and resulting in depressed discrimination scores. Jerger and Jerger (1975) measured performance on synthetic sentence identification materials administered in the presence of a competing message (i.e., a spoken passage) to sixteen patients with intra-axial brain stem lesions. Performance was measured in various stimuli-to- competition ratios from 0 to -40 dB. Jerger and Jerger reported that performance generally remained within the range observed for normal subjects. The findings from the investigations of Martin (1966), Young and Harbert (1970), and Jerger and Jerger (1975) suggest that contralateral masking has little effect on speech discrimination scores. Spencer and Priede (1974), however, provided data that indicated a small percentage of normal hearing individuals could be expected to show a 1% decrease in intelligibility for every 3 dB increase in sensation level of a wide—band noise contralateral masker. They concluded that this finding was caused by central masking phenomenon. Currently, there is no literature directly concerned with the effects of contralateral masking on time compressed speech discrimination tests. Since these tasks have been shown to be sensitive to central auditory lesions, it is 15 possible that the central component associated with contra- lateral masking may interact with time compressed speech stimuli. For example, Beasley et al. (1972b) did not observe a difference in performance between right and left ears when listeners were presented time compressed NU—6 words. Based on these findings they suggested "... that time-compressed speech can be clinically utilized in a monotic listening task without being confounded by ear laterality effects" (p. 348). Previous research with dichotic listening tasks employing speech stimuli, however, have demonstrated a right ear advantage even when listeners were instructed to report only right or left ear stimuli (Berlin, 1972; Kimura, 1967). Masking the non-test ear while time compressed stimuli are simultaneously presented to the test ear creates a dichotic listening condition. It appears possible, therefore, that contralateral masking may result in a perceptual difference between ears. Furthermore, such laterality effects may be enhanced as time compression ratio increases and external redundancy is reduced. If this were the case, existing nor— mative data for time compressed versions of the NU-6 test may confound diagnostic results pertinent to identification of central auditory lesion. With the exception of Jerger and Jerger (1975), each investigation involving the exploration of contralateral masking on speech intelligibility used a white noise masking stimulus. Research with ipsilateral maskers (i.e., masking the test ear) has revealed that spectral and temporal 16 parameters of the masker may have a direct bearing upon masker effectiveness. Miller (1947) reported "... the masking of speech to depend on three characteristics of the masking sound: (1) its intensity relative to the intensity of the speech, (2) its acoustic spectrum, and (3) its temporal continuity" (p. 106). Hawkins and Stevens (1950) compared the masking curves for the threshold of speech intelligibility to the masking functions for pure tones obtained monaurally with a white noise masker. They found the masking function for speech most closely paralleled the average threshold functions for 500, 1000, and 2000 Hz. These results indi- cated that the most efficient maskers for speech were stimuli characterized by a spectrum with energy in the lower third of the range from 100 to 6000 Hz. Regarding temporal parameters, Carhart and his associates (Carhart, Tillman, and Johnson, 1966; Carhart, Tillman, and Greetis, 1969; Carhart, Nicholls, and Kacena, 1972) have conducted several investigations dealing with the effects on spondee thresholds of maskers composed of speech combined with white noise or with other speech signals. Results of these investigations revealed that when white noise was combined with speech there was approximately 3 dB more masking than could be predicted from simple summation of the average intensities of the two maskers. Moreover, when speech was combined with another speech signal, approximately 7 dB more masking was observed than could be attributed to simple addition. This phenomenon has been termed perceptual 17 masking and is felt to be caused when competing signals are semantically meaningful as compared to masker noise that lacks meaning, e.g., white noise. Carhart et a1. (1972) noted that perceptual masking increased up to the combination of four speech signals. Maskers that contained more than four speech signals did not cause a substantial increase in perceptual masking as compared to the four signal maskers. Presently, the relationship between white noise, filtered white noise (i.e., speech noise), or multitalker complexes used as contralateral maskers has not been examined. Con- sequently, any interaction of these stimuli when presented in a contralateral paradigm with accelerated speech, such as the time compressed versions of the NU-6, is unknown. Summary and Statement of the Problem The afore-going review of the literature has shown that accelerated speech provides a unique stimuli to distinguish central from peripheral auditory problems. The NU-6 test of auditory speech discrimination has been time-compressed using controlled methods, and data has been reported for young and aging persons with normal hearing, as well as persons with temporal lobe damage. The effects of contra- lateral masking on the perception of time compressed speech, however, is unknown. Contralateral masking is a well recognized audiometric procedure commonly employed in clinical situations with suprathreshold speech discrimination tasks. Since the two 18 most common clinical maskers of speech stimuli are white noise and speech noise, the effects of these two stimuli as contralateral maskers of accelerated speech need to be determined before time compressed NU-6 word lists may be routinely used in clinical settings. In addition, because multitalker maskers have been shown to cause perceptual masking when presented ipsilaterally, it appears desirable to ascertain the effect of this type masker in a contra— lateral condition. For example, it may be possible that perceptual masking is a time-based process; or at least dependent upon temporal parameters. Such information may help to elucidate the perceptual process whereby a listener is able to selectively attend to one stimulus in the presence of a competing signal. It was the purpose of this investigation, therefore, to examine the relationship between contralateral masking and the perception of time compressed NU-6 speech discrimi- nation lists. Specifically, the following questions were investigated: 1. Will white noise (WN), speech noise (SN), multi- talker (MT), or multitalker 60% time-compressed (MT60%) stimuli presented as contralateral maskers have similar effects on speech intelligibility measured by the NU-6 presented to the ipsilateral ear? 2. Will the intensity level of WN, SN, MT, or MT60% stimuli presented as contralateral maskers have 19 similar effects on speech intelligibility measured by the NU-6 presented to the ipsilateral ear? Will WN, SN, MT, or MT60% stimuli presented as contra— lateral maskers have similar effects on speech in- telligibility measured by time compressed versions of the NU-6 presented to the ipsilateral ear? Will WN, SN, MT, MT60% stimuli presented as contra— lateral maskers to the right ear have a similar effect on speech intelligibility of NU-6 stimuli presented to the left ear, as when the contralateral maskers are presented to the left ear and NU—6 stimuli presented to the right ear? CHAPTER II EXPERIMENTAL METHODS This investigation was designed to eXamine the effects of contralateral masking on a test of central auditory function. The NU—6 test of speech discrimination (time- compressed by 0%, 30%, and 60%) was presented to ninety normal hearing young adult listeners under four conditions of contralateral masking (WN, SN, MT, and MT60%). Be- havioral responses for each condition were obtained for an equal number of right and left test ears. The experimental design used in the investigation is shown in Figure 3, on page 25. Subjects Ninety normal hearing young adults were selected from a university population to serve as subjects. This sample consisted of 22 males and 68 females. Subjects ranged in age from 19 years to 26 years with a mean age of 23.2 years. In order to qualify for the investigation, each subject met the following criteria: 1) pure tone air conduction thres— holds in each ear of 15 dB HTL (re: ANSI, 1969) or better for the octaves 250 through 4000 Hz, plus the half-octave of 6000 Hz, 2) bone conduction pure tone thresholds within :10 dB 20 21 of air conduction thresholds, except for 6000 Hz where bone conduction responses were not obtained, and 3) speech recep— tion thresholds of 15 dB HTL (re: ANSI, 1969) or better bilaterally. Potential subjects who did not meet these criteria were not included in the investigation. Stimuli Speech reception thresholds were measured with tape re— corded versions of the CID W-l word lists described previously by Rintelmann et a1. (1974). The four lists comprising Form B of the NU-6 were used as time compressed stimuli (0%, 30% and 60%) and consisted of the tape recorded materials used by Beasley et a1. (l972a,b). The source for white and speech noise maskers was a two-channel speech audiometer (Grason-Stadler, Model 162). The frequency spectra for these maskers when transduced by a TDH 39-10Z earphone are presented in Figure l. The white noise masker had a flat frequency spectrum within :1 dB from 100 to 4000 Hz. The speech noise masker was characterized by a 3 dB per octave drop from 250 to 1000 Hz and a 6 dB per octave drop from 1000 to 4000 Hz. Four male talkers simultaneously reading a passage about the general scope of psychology (James, 1973) comprised the multitalker maskers. In order to generate these maskers, each talker individually read the passage at normal conversational speech and effort level into a microphone (Electrovoice, Model 635A) located in an audiometric test room (IAC, Series 22 400). The microphone was connected through the wall of the test room to a four channel tape recorder (Ampex, Model AG 440B, frequency response = 50 to 15000 Hz :2 dB per channel). Each passage was recorded on a single track of a four track magnetic tape. This tape was then re—recorded from the same tape recorder, with each track fed to a four channel microphone mixer (Shure, Model M67) connected to the input on an electrical time-compressor/expander (Lexicon, Model Varispeech I, frequency response = 50 to 15000 Hz I3 dB) where each talker passage was simultaneously recorded on the same track of a master tape. The master tape was then processed following the method recommended by Konkle et al. (1975) through the Lexicon time-compressor/expander at 0% and 60% time—compression, and experimental tapes recorded on the Ampex Model AG 440B tape recorder that was connected to the time-compressor/expander. The Varispeech I was cali- brated to manufacturers specifications before tape recordings were made. The frequency spectra of the multitalker maskers may also be seen in Figure 1. Frequency spectra of both multi— talker maskers approximated the configuration of the speech noise maskers, particularly when the multitalker masker was time-compressed by 60%. Both multitalker maskers, however, contained less low frequency energy (below 100 Hz) and high frequency energy (above 1200 Hz) than the speech noise. When the multitalker masker was at normal duration the major area of energy concentration was at 500 Hz, whereas, for the 23 .HHSEHum mcflxmme Hmuoumamnucoo mm pom: “ommmmumaooloaflu wow Hoxamufiuasfi u owe: pom .Hoxamufluase u as .mmfloc noommm n 2m .omflos moans u 23V muoxmma snow mo muuommm mososvmum .H musmwm N: z. >ozw30mmm xom x0. x0 xm v: 000 CON 00. 00 ON 0 / , 1.1 / z . 5. / .. s." x /. . 1. Jo. / ..x. .\ / n.... . .\ \ I / /...... \ N / /.... . \ \ m l/ / .\\. IQIU / .s \t al. / /.. ~.. >\ IA :/ .. Vx I On M Inf/ \ . . \ JUMJWHuhwv. mw ...!l.owt2 ...—2 10¢ ...i.. 2m II 23 on 24 speech noise and 60% time-compressed multitalker maskers the major energy concentration centered at 250 Hz. The method and procedures used to obtain the frequency spectra of all four maskers may be found in Appendix A. Apparatus All subjects were tested in an IAC 1200 series double. walled test chamber. Located in the control room of the test suite were a two channel tape recorder (Ampex, Model AG 600-2, frequency response - 50 to 15000 Hz :2 dB) and a clinical audiometer (Beltone, Model 15C). The tape recorder fed speech and multitalker masking stimuli to a two channel speech audiometer (Grason-Stadler, Model 162). Signals from the audiometers were transduced via TDH 39-10Z earphones mounted in Mx 4l/AR cushions located in the attached listening room. Bone conducted pure tone stimuli were delivered by a Radioear B70A white dot vibrator. A talk-back system, com- prised of a microphone (Shure, Model 560) located in the listening room and connected through the speech audiometer to TDH 39-lOZ earphones and associated MX 4l/AR cushions, allowed the examiner to monitor verbal responses. A schematic representation of the apparatus is shown in Figure 2. The ambient noise level in the listening room of the sound suite was measured at 44 dB SPL (re: 0.0002 dyne/cmz) on the C scale of a sound level meter (Bruel and Kjaer, type 4145). This noise level was sufficiently low as not to interfere with the listening tasks. Prior to experimental 25 .HHSEflum Hmucmeuwmxm usmmmum ou poms msumuwmmm mo oapmfimcom .m wusmflm ? mmzozmmzm C11 NOTmm 1.0... £9555 03 ._._mzm._.z_ mmxm<2 cm 00 on 4 a 6.0 «when; 8% @00& w A09. 0 3 N" I. m m 8w .d 1.. m 33 m n omm ... \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ 2.... m \ \w\ \\ \ \ .\ \ \ . \\ \. \. \. \w \. \. \V \x 89 \ \ \ \ \ \ \ \ \ :3 \ K \ 02.: 5 z» z... 02.... .5 z» 23 on»: E z» z; ".9552 .- L 28 condition. Regardless of test ear, the time compressed stimuli were always presented via the right earphone with the masker stimuli presented from the left earphone. Subjects were tested individually in a manner that allowed the four lists of NU-6 Form B to be administered to the test ear at a single time compression ratio. At the same time, each list competed with one of the four maskers presented to the non-test ear at the same intensity level. The word lists and presentation order of the four masker types was counterbalanced across experimental conditions. Finally, a standardized answer form (see Appendix C) was provided prior to the administration of each list for subjects to record their responses. Analysis Subjects' responses on each NU—6 list were tabulated and changed to percent correct scores. A multiple linear regression analysis (Nie, Bent, and Hull, 1970) was performed with the dependent variable (percent correct score) pro- jected on the independent variables (time-compression, masker intensity level, masker type, and ear) with masker type coded as a dummy variable. Two equations were developed (see Appendix D) to determine the amount of the total variance (i.e., predictability) of the dependent variable that was attributed to the independent variables. The equations were submitted to a stepwise computer analysis (Michigan State University, SPSS — Regression, version 6.0) 29 that yielded Pearson rs, multiple correlations (Rs), R-squared, and standardized Beta weights. The percent correct scores obtained by each subject in the various experimental conditions are shown in Appendix E. CHAPTER III RESULTS The results of the stepwise multiple linear regression analysis indicated that neither white noise nor speech noise had a significant effect on the intelligibility of time compressed NU-6 word lists. Conversely, the multitalker maskers were found to react with other variables (i.e., time- compression, masker intensity level, and ear) in a manner that caused a statistically significant alteration in speech in- telligibility. The zero-order correlation coefficients for percent correct score with each variable and variable combination entered into equations I and II are shown in Table l. Fif- teen of the possible thirty-four correlations were statisti- cally significant (p)>0.05). Moreover, all of the significant correlations were negative, thereby demonstrating that in- creasing masker and time compression values were associated with decreasing discrimination performance. Eleven of the fifteen significant correlations were comprised of either the MT or MT60% maskers combined with other variables. These findings suggested that white noise and speech noise presented as contralateral maskers had a negligible effect 30 31 as n 0x Ham u 0x mcofluwsflnEou magmapm> u mvmmx ou mmx Zm H ex cofimmoumEoolmEHB u mx 00092 n 00 23 u mx Hm>ma umxmmz n 0x mo.oAm -l- a 400.0- 000x 00.0 00x 400.0- 000x «00.0- 00000 00.0- 00x 400.0- 00x 400.0- 000x 00.0 00x «00.0- 00x 00.0- 0x 00.0- 00x 00.0- 00x 400.0- 00x 400.0- 0x 400.0- 00x 00.0 00x 00.0- 000x 00.0 0x «00.0- 00x «00.0- 000x 00.0- 0000x 00.0 00 00.0- 0000x 00.0- 00x 00.0- 000x 00.0 0x 00.0- 0000 400.0- 00000 00.0 00x 400.0- 00 00.0- 000x 400.0- 0000 00.0- mmx «00.0- 0x u manmflnm> OHQMHHM> u magmaum> H mHQMHHm> unopsmmmooH so pwmmoummm muoom uomuuoo uswoumm Mom mcoaumamuuoo Hopuonoummll .mwanmwum> .0 mflnme 32 on the intelligibility of time compressed NU-6 stimuli. A more definitive description of the relationship between discrimination performance measured by time compressed NU-6 words and contralateral masking were found from the re— gression equations yielded by the stepwise computer analysis. Table 2 presents the analysis of variance, coefficients of multiple correlation (R), and coefficients of determination (R2) for equations I and II extended through the seventh— order stages improved Rs by only 0.00352 for equation I and 0.00313 for equation II, only the first eight stages (i.e., zero—order through seventh-order) are reported for each equation. The significant F values indicated that differ— ences existed between the various MrL, TC, Ear, and masker type (MrT) experimental conditions. In addition, the R2 values suggested that the percent correct scores obtained by subjects were not completely accounted for by the independent variables. This was not surprising since the purpose of this investigation was to examine the effect of only selected variables (MrL, TC, Ear, and MT), with other important variables such as hearing sensitivity and age held constant. The R2 values indicated that 43.7% of the total variance for discrimination scores was accounted for by the seventh-order equations as compared to 44.8% for total equations. White Noise and Speech Noise as Contralateral Maskers The seventh-order stepwise computer analysis allowed 33 mo.o Am H « ohhmv.o Noamm.o «mH.vm vm.HH mv.ovv Hmm m hm.mhmm mo.wmom Hmoofimmm soammmummm vhhmv.o Noamm.o *mH.vm vm.HH mm.hmm Hmm m hm.mhmm mo.mmom Hmspwmom GOHmmmumwm owumsvm m 0 60000052 w50m> m mmumowm mo coo: Eooomum mo momumwo mmnmowm mo Sam oOCM0Hm> mo mUHSOm HH cowumovm H coflpmsvm usmouwm How mm90m> omumswm-m pom .m wamfluasz .mosmwum> mo mwmmamswln .unm0m mwum 009000» 00 can H mc00pmsvm How moHQM00m> osmosmmoosH so commoummm ouoom uoouuoo .N magma 34 for eight variables to enter each equation. For each equation, these variables were identical and entered equa— tions in the same order. Tables 3 and 4 present the stan— dardized Beta weights (B) and the F values associated with each variable as a function of successive stages (i.e., step 1 through step 8). Examination of Tables 3 and 4 reveals that white noise and speech noise contralateral maskers lacked sufficient impact on percent correct scores to be included in the seventh-order multiple regression equations. In fact, the WN masker did not enter equation I until the fifteenth stage and only caused an R change of 0.00059. Similarly, the SN masker did not enter equation II until the eleventh stage with an R change of 0.00065. These values were not statis- tically significant (pj>0.05). Graphic illustrations of these findings may be seen in Figures 4 through 8. The greatest differences between conditions for these two maskers occurred when SN was presented to the contralateral ear at an intensity level of 90 dB SPL and the NU-6 words were time-compressed by 30% (see Figure 5). This difference, however, was only 2.2% (see Appendix F). When data obtained in this investigation were compared to the findings of Beasley et a1. (l972a) for the same time compressed NU-6 stimuli without contralateral masking (see Figures 4 to 6), it became readily apparent that contralateral masking with WN or SN stimuli had essen- tially no effect on the intelligibility of NU-6 words regardless 35 va.m *mm.m 0mm.m RON.HN «mm.om {om.va m vH.OI mN.o vm.o ON.OI NM.OI om.OI m .w mv.m *wm.m *Nm.HN 05h.mm awm.mn m mH.OI mN.o HN.OI mm.OI Nv.OI m .m *wo.v «mm.om smN.Nm *mw.mw m HH.o 0N.OI mm.0I vv.ou m .e *mH.om ¥O®.Nm *mw.wm m om.on mN.oI vv.OI m .m *vc.om *hm.mva h mH.OI mm.OI m .N *hv.mma m mm.OI m .H mx NN mx BVMNX hvmx mmmx hmx MNX vam moanw00m> .H COHDmem 00m moanwflum> ucwocmmwosH co owmmmummm whoom uow00oo unmoumm mo mm50m> m cam Amy wunmfloz mpwm owm0o0moswum 00600|£usw>wm flmsounu Hoono-onwmln .0 60000 36 00002 n 0x Hm>mq Hoxmmz u «x s00mmmHQEoo-o50e u 0x ES x 0mm x c00mmmumEoonw809 x 0m>oq uoxmwz n nvmmx a: x 0mm x GOmemumEoo-mEHB n nvmx wows: x c00mmogmfioonoefle x Ho>oq memmz n mmmx 92 x coflmmwumfioo-mawe u nmx cOHmmoumEoo-os0e x 0m>o0 Hoxmmz n mmx 00.020 u .. mm.H mm.m *Nm.m «wh.m «mv.m amm.ma *om.mm mm.o m mo.o HH.oI mm.0I wN.OI mm.o mm.ot mm.0I NH.OI m .m vm.m xom.w «mm.m amo.m «vm.am xmv.om vm.a m HH.OI mN.OI mN.OI vm.o om.OI mm.OI vH.0I m .h mx NX MN hvmmx bwmx mmmx hmx mmx Qwuw mmHQMH0m> oosc0ucOUIl .m magma 37 00.0 000.0 000.0 000.00 000.00 000.00 0 00.0- 00.0- 00.0 00.0- 00.0- 00.0- m .0 00.0 000.0 000.00 000.00 000.00 0 00.0- 00.0 00.0- 00.0- 00.0- m .0 000.0 000.00 000.00 000.00 0 00.0 00.0- 00.0- 00.0- m .0 000.00 000.00 000.00 0 00.0- 00.0- 00.0- m .0 000.00 000.000 0 00.0- 00.0- m .0 000.000 0 00.0- m .0 mx NX MN PWMNX F¢MM mmmx hmx MNX Qwum meQMHHM> .HH c00umsqm Mom meQM00m> unmocwmmocH co Ummmmumom muoom 0000000 ucwommm mo mwoam> m 00m Amy musm0mz wumm tww0tumpcmgm 0mpno-suco>mm stougp uwouo-oummll .0 magma 38 00002 n 02 Hw>wq umxmwz n «x goflwmmHQEOOIoefle n mx 9: x 0mm x :00mmwumEoo-oEHB x 0m>wq umxmmz n 00002 52 x 0mm x qoflwmmumfioo-mfi0e u 00mx 00092 x cofimwwumeoolofi0a x Hm>mq umxmmz u mmmx BS x 200mmwumeoolwefle n 0mx 200mmonE00|wE0B x 0w>mq 002mm: n mmx 00.000 u 0 mN.H mm.m *Nm.w «ww.m *mv.m *mw.wH «Hm.mN mm.o m wo.o HH.OI wN.OI mN.OI mm.o mN.OI NM.OI NH.0I m .m vm.m *om.o *mm.m *wo.m «vN.HN «mv.om vm.H m HH.oI mN.OI mN.OI vm.o HN.0I NM.OI va.ol m .h 02 0x 02 00000 0002 0002 002 000 0000 mmHQM0um> omDCHDcooll .0 mHQme 39 IOO - Me presented at 0% TC \0 95 — '5 uu a: I: CD 090- p. I: n: (J a: UJ “85 _ O—O 90dB SPL Mr III—CI 60dB SPL Mr 80 <>—<> 30dB SPL Mr IL ---- No Mr (Beasleyet al., l972 0) JL 61 I l l I if w~ SN MT M160 MASKER Figure 4. Mean percent correct scores for 0% time- compression as a function of masker type and intensity. The Beasley et al. (1972a) data without contralateral masking are shown for comparison. 40 we? (0 '(fl 1 PERCENT CORRECT “D C) I 00 (fl 39:8 0' 9 \O: 8/8 - O—O 90dB SPL Mr I 4’ Me presented at 30% TC CI—D 60d8 SPL Mr 80 0—0 30dB SPL Mr I -—-- No Mr (Beasley et al., l972 a) ()[m I l l I Figure 5. wn SN MT M160 MASKER Mean percent correct scores for 30% time- compression as a function of masker type and intensity. The Beasley et a1. (l972a) data without contralateral masking are shown for comparison. 41 IOO Me presented at 60% TC 94 T —<) CI PERCENT CORRECT (O (D l 00 (fl O—O 90dB SPL Mr EJ—Cl 60dB SPL Mr <>—-<> 30dB SPL Mr ---- No Mr (Beasley et al., l9720) 01’..._L_r L l 1_J W 13" IflT' IflTéc, MASKER 80 Figure 6. Mean percent correct scores for 60% time- compression as a function of masker type and intensity. The Beasley et al. (1972a) data without contralateral masking are shown for comparison. 42 of the masker intensity level, the amount of NU-6 time- compression, or the ear to which either stimuli were presented (see Figures 7 and 8). Multitalker Stimuli as Contralateral Masker Although the MT masker by itself did not have a signi- ficant relationship with NU-6 intelligibility, the combination of this variable with other variables had a significant effect on discrimination scores. Tables 3 and 4 show that statisti- cally significant F values were obtained when the MT masker was combined with TC, TC by Ear, and MrL by TC by Ear. Furthermore, the MT stimuli was the only contralateral masker that caused an ear laterality effect. In order for this effect to emerge, however, it was necessary to include time-compression as part of the combined variable. This finding suggested that the amount of time—compression served to enhance ear laterality. The same observations may be applied to the variable comprised of MrL by TC by Ear by MT,‘ whereby both the amount of time compression and the intensity level of the MT masker enhanced ear laterality. These results are further depicted in Figures 4 to 7. When NU-6 stimuli were time-compressed by 0%, the effect on intelligibility of MT contralateral masking was negligible (see Figure 4). At 30% time—compression (see Figure 5) discrimination scores decreased approximately 4% when MT masking was presented at 60 and 90 dB SPL. When NU-6 words were time-compressed by 60% (see Figure 6), the MT masker 43 .memmfi Handymamuunoo mmwo: muflnz m «o mufimcmuca 0:“ mo coauossw w mm muwm umma can unmflu How mmuoom uomuuoo unmoumm and: .0 musmwm «Boxes; «000.0 "8 4mm 00 z. >._._mzm._.z_ 00 om on oo 00 on 00 00 xv. _ was,“ A i J ”go! L fl . 5:228 2.: 88* .0 TV. 0. \F0 A a: O 0 1033800 1N3083d .l 00. ¢mq mmwoz nomwmm mmfloz 000:3 . unmouom . mpflmcmucH uoxmmz .muwm #000 was unmwm paw .Hm>wq mufimcmucH umxmmz .0008 memmz .oaumm coflmmmHmEOUImEHB comm How mmuoom uomuuou unmonmm mommm>