THESIS [)ate 0-7639 This is to certify that the thesis entitled CHILDREN'S SENTENCE COMPREHENSION AT DIFFERENT RATES OF SPEECH presented by Alkistis Charalambous has been accepted towards fulfillment of the requirements for Audiology and Mvo degree in SpeeCh SCienceS Mug/1%,? Major professor February 21, 1986 MS U is an Affirmative Action/Equal Opportunity Institution 3 U—sAfi I ”SCHE‘ ,lz‘L‘e“fi1 (*aig. n ‘III~IQ> $5333; . lelwtga.4¢. MSU LlBRARlES —_ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. CHILDREN'S SENTENCE COMPREHENSION AT DIFFERENT RATES OF SPEECH BY Alkistis Charalambous A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Audiology and Speech Sciences 1986 Copyright by Alkistis Charalambous 1986 ' ABSTRACT CHILDREN'S SENTENCE COMPREHENSION AT DIFFERENT RATES OF SPEECH BY Alkistis Charalambous This study investigated the effects of altered rates of speech on young children's sentence comprehension. Thirteen normal language learning children between five and nine years of age served as subjects. Individual children were presented with taped sentences under three temporal conditions: normal rate, expanded rate, and compressed rate. Results indicated the following: 1) the mean difference of children's performances between the normal condition and the expanded condition was not significant; 2) comprehension decreased for all groups under the compressed speech condition; however, the differences were significant only for the seven—year-old children; 3) comprehension was influenced by Speaking rate as a function of age; the older children performed significantly better than the younger children in the normal and expanded conditions; and 4) a strong positive correlation between accuracy and response latency was noted for the expanded and compressed rates for the six—year-old children, whereas for the seven-year—olds there was a strong negative correlation for the normal rate. ACKNOWLEDGEMENTS ”A picture is worth a thousand words." A tribute to those people who have had the greatest impact on my work and in my life, worth more than many thousands of words. I would like to eXpress my sincere appreciation to my thesis advisor, Dr. Michael W. Casby for his influence and encouragement in my attempts to develop research interests. His support, enthusiasm, knowledge, and friendship have been important and invaluable for the completion of this study. I would also like to express my sincere appreciation to the other members of my thesis committee, Dr. Leo V. Deal and Dr. John B. Eulenberg, for their friendship and support as well as for their comments on this research. A special word of thanks to Dr. Deal, my academic advisor, for his invaluable contribution, concerns, and encouragement over the extent of my graduate studies. Many thanks to Dr. Robert Muth and Mrs. Pat Muth, my ”great American" friends. Finally, my thanks to my parents and family for their continued love, support, and confidence in me. ii TABLE OF LIST OF TABLES . . . . . LIST OF FIGURES . . . . . INTRODUCTION . . . . . . REVIEW OF THE LITERATURE Experimental studies of children. . . . . . . Temporal manipulation METHOD . . . . . . . . . Subjects . . . . . . Procedures . . . . . Stimuli. . . . . . Rate condition . . Pre-test. . . . . . . Experimental task . . Scoring of responses. Reliabiliy C O O O 0 RESULTS 0 O O O O O O O 0 Accuracy of response. Latency of response . Data Analysis . . . . of speech 111 CONTENTS stimuli language disordered page vi 1? 23 23 28 28 29 29 30 31 32 33 33 34 36 Scores within Chronological Age Groups across Rate Conditions. . . . . . . . . . . . . . . . . . . . Time within Chronological Age Groups across Rate Conditions. . . . . . . . . . . . . . . . . . . . Scores across Chronological Age Groups for Rate Conditions. . . . . . . . . . . . . . . . . . . . Time across Chronological Age Groups for Rate conditions. 0 O O O O O O O O O O O O O O O O O 0 Relationship between Accuracy and Timing. . . . . LANGUAGE IMPAIRED CHILDREN. . . . . . . . . . . . ReSUI-ts O O I O O O O O O O O O O O O 0 O O 0 Accuracy of Response. . . . . . . . . . . . . Latensy of Response . . . . . . . . . . . . . DISCUSION O O O O O O O O O O O O O O O O O O O 0 REFERENCES 0 O O O O O O O O O O O O O O O O O O i 0 iv 36 36 37 37 37 41 42 42 42 44 48 LIST OF TABLES Table page 1. Individual data, means and standard deviations for chronological age (CA), Columbia Mental Marurity Scale (CMMS), Token Test for Children (TTC), Clinical Evaluation for Language Functions (Cl), Normal Score (NS), Expanded Score (ES), Compressed Score (CS), Normal Time (NT), Expanded Time (ET), Compressed Time (CT). . 25 2. Summary table of the groups' mean time performance of the various measures. . . . . . . 35 LIST OF FIGURES Figure page 1. Cell means for the Accuracy of performance. 2. Cell means for the Latency of response. . . vie 39 40 INTRODUCTION An increased interest in auditory processing and its relation to speech and language has been the focus of research in the last decade. This interest in research has advanced our knowledge with regard to the perception of speech and language and the variables influencing auditory perception. As a result of these investigations, three theoretical models have been proposed: 1) whether auditory processing provides the foundation of language; 2) whether auditory processing is dependent upon language in order to function efficiently; 3) whether there is an interaction between auditory processing and linguistic knowledge. This third model proposes that there is an interaction between ”data-driven” process (i.e., based upon acoustic-phonetic properties of the input) and "knowledge-driven” process (i.e., the linguistic processes of semantics, syntax, and pragmatics) that is involved in language and auditory processing (Marslen - Wilson and Welsh, 1978; Tyler and Marslen - Wilson, 1981; Duchan and Katz, 1983). One of the variables which is considered to be important to auditory processing and, thereby, possibly important to comprehending and learning language is the rate of the auditory message. Presentation rate and its relationship to comprehension in both normal children and specific language impaired children have been investigated in experimental studies. Auditory perceptual deficits have been hypothesized by some investigators as the cause of language and learning impairment in language impaired children (Lowe and Campell, 1965; Tallal, 1975; Tallal and Piercy, 1973 and 1974). Research is needed that will compare the comprehension of time-altered Speech between normal children and language-impaired children; however, before any meaningful comparison can be made, further data must be collected on the impact of altered rate of speech on normal children's comprehension. The following review of the literature presents information in four areas: 1. research into the nature of general language impairment in children, 2. research which has argued that the basis of language impairment rests in the child's inability to process rapid temporal auditory Signals, 3. research which has investigated the temporal manipulation of speech stimuli and the effect of that manipulation upon speech perception, and 4. research concerning the impact of rate-altered speech on normal children's auditory comprehension. Review of the Literature -o---------------------- Several bases of language disorders in children have been hypothesized. Contemporary thought regarding impaired language development in children suggests that there are three tenable hypotheses for research regarding the bases of language development in children: a neurolinguistic hypothesis, a conceptual/ representational hypothesis, and an auditory processing hypothesis. The neurolinguistic hypothesis regarding the nature of language disorders in children makes the argument that language impaired children have difficulty in learning language because of some deficit in the neurological substrata of language competence. The following conditions have been implicated: prenatal or postnatal lesion in the language related areas of the brain (i.e., left hemisphere); lack of anatomical or functional cerebral asymmetries (i.e., diminished left hemisphere size); and bilateral lesion of the temporal lobe association cortex (cf. Ludlow, 1980). The conceptual/representational hypothesis is based on the view that children's disordered language is the result of delayed deve10pment of the concepts which underlie language, a difficulty in the linguistic mapping of these concepts and/or impoverished use of language in context. The third hypothesis regarding the nature of language impairment in children has disordered auditory processing as its core. This hypothesis states that there might be a causal relationship between impaired auditory processing abilities and language disorders in children. Various studies (Lowe and Campell, 1965; Tallal and Piercy 1973, 1973a, 1973b; 1974, 1975, 1978, Tallal, 1976) have implicated a deficit in the ability to make temporal perceptual judgments as a prime contributor to language impairment in children. Tallal and her colleagues have reported a series of experiments which demonstrated that language impaired children have difficulty temporally ordering rapid sequences of sounds. It has been reported that in order to respond correctly the language impaired children require significantly longer interstimulus intervals than do chronological age matched normal language children. The research detailed in this study concerns the auditory perceptual hypothesis. Eresrisestsl-§§9§is§-9§-99999a99-91§9§§s§e§-99319599 Tallal and Piercy (19733) studied developmentally aphasic children, aged six to nine years, and CA matched contol normal children for their ability to perceive nonverbal stimuli in the auditory and visual modality. The children were required to discriminate two complex tones of differing fundamental frequency (100 Hz and 305 Hz) at durations of 75 to 250 msec. with inter-stimulus-intervals (181) of 15 to 350 msec. Their performance was studied as a function of the duration of stimulus elements and the number of elements in a sequence. Results indicated that aphasic children performed significantly poorer than their age matched normals in discriminating nonverbal auditory stimuli when the duration of the stimulus elements and the interstimulus (ISI) interval were decreased (e.g., with tone durations of 175 msec. and 151 less than 15 msec.). With tone duration of 250 msec. they performed as well as normals for all 1813. It was also found that the disordered children performed as well as the controls on the visual tasks. The authors hypothesized that aphasic children were unable to perceive auditory information with short 1815 as well as normal children and suggested that this difficulty in auditory processing may underlie their language impairment. The same dysphasic children were examined for their ability to discriminate two synthesized steady state vowels (/E/ and Az/) and consonant-vowel syllables (/ba/, /da/). No impairment in discriminating the two vowels was found when they were of 250 msec. duration. However, the dysphasic children showed significantly impaired discrimination of the two consonant-vowel syllables of 250 msec. duration with a discriminable transitional component of only 43 msec. The authors suggested that the brief duration of the formant transition resulted in the impairment in discriminating consonant stimuli. Still, another possibility was suggested that involved the inability of dysphasics to process transitional information irrespective of its duration. To differentiate these possibilities, Tallal and Piercy (1975) studied the perception of V-V syllables and C-V syllables. For both vowels and consonants, dySphasic children performed significantly poorer than normals when the discriminable information of the two stimuli was brief (43 msec.). When the transition was expanded to 95 msec. or larger, the dySphasics did not differ from normals in their ability to discriminate the stimuli. It was concluded that the dysphasics have no difficulty in discriminating transitional auditory information but that their impaired perception of stop consonants was attributable to the brief duration of the discriminable components. The authors stated that the results support their original hypothesis that the language defect of these children is not specifically linguistic but is secondary to an impaired rate of processing auditory information. i In another study, Tallal (1976) investigated the development of rapid auditory processing in dysphasics six to nine years of age and normal children (4;6 to 8:6 years) matched for CA and nonverbal intelligence. The performance of each group was compared with that of an adult control group under three stimulus conditions: short ISI (8 to 305 msec.); long ISI (more than 428 msec.); and constant ISI at 428 msec. Two 75-msec. complex tones were used as stimuli. It was found that all normal groups younger than 8:6 responded significantly poorer than adults to rapid auditory stimuli presented in 8-305 msec. When the interval between the two tones was increased, normal children 6:6-year-old or older could respond correctly to the same stimuli, whereas the 4:6 and 5:6 year olds continue to show significantly inferior performance compared to that of adults. The pattern of performance of dysphasic children was not similar to that of any age group of normal language development. They performed significantly poorer than even the 4:6 year old normal group on rapidly presented auditory sequences but not significantly different from normal children their own age on auditory stimuli presented more slowly. By the age of 6:6 the normal subjects performed as well as adults on the long ISI. However, these normal children did not perform as adults on the short ISI until the age of eight years and six months. Based on the above results, the author concluded that the ability to respond correctly to auditory information, presented rapidly, develops progressively with age. She also suggested that the dysphasic children's develOpment of auditory processing is so delayed that it reflects that of even younger children. The picture of the comparison between dysphasic and normal children could be expected, since the two groups were matched for CA and not for linguistic performance. An interesting finding is that the normal Children between four and seven and a half years-old had still not reached adult levels of discriminating, though they were developing language normally. One may question how "normal” was their language development, since we do not know their language level. If we accept that these (normal) children have normal language development in the presence of a less than perfect rate processing, then we may question the possibility of seeing auditory processing related to the rate of speech as a necessary prerequisite in language development. These series of studies suggested, as Tallal, Stark and Curtis (1976) stated, that the specific auditory perceptual deficit might be sufficient to explain the language disorder of these children. It was also hypothesized that if developmental dysphasia stems from failure to perceive certain speech sounds (e.g., stop consonants in CV syllables which require auditory processing of rapidly changing formant structure), then it would seem that these sounds would be produced incorrectly or omitted. This hypothesis was investigated in their 1976 study (Tallal, P., Stark, R., and Curtis, 8., 1976). Dysphasic children between 7:1 and 9:5 years old, matched with normal controls for CA, were tested for their ability to perceive and produce speech. They were required to imitate 1) isolated steady-state vowels (/E/, £17): 2) stop-consonants in CV nonsense syllables, in which rapid acoustic changes occur; 3) stop-consonants in CVC nonsense syllables; 4) stop-consonants in clusters. The children were also required to produce the names of pictures of objects shown on cards representing CV and consonant cluster syllable structure. The results showed that the dysphasic children made a significantly greater number of errors in all speech production categories (p<.001) than normals. Comparison between the dysphasic's responses to the receptive task demonstrated that five of the twelve subjects were "unimpaired" in their discrimination of CV syllables comprising 43 msec. formant transition. However, their performance in consonant clusters was significantly impaired. The remaining seven children did not reach criterion on their performance on 43 msec.: therefore, they were called perceptually ”impaired." The performance of the perceptually "impaired" children was significantly worse than that of the perceptually "unimpaired” on all production tasks. Their ‘pattern of impairment corresponded to their perceptual impairment: i.e., the production of isolated vowels and nasals was significantly less impaired than that of stop-consonants, particular in clusters. Later in her review on ”recent research pertaining to auditory processing disorders in children," Tallal (1980) referring to the same study, concluded that speech sounds involving rapid spectral changes critical for their perception are most difficult for dysphasic children to perceive and are also most often misproduced by these children. She also suggested that ”this result adds further 10 support to the hypothesis that some developmental language delays result, at least in part, from a primary nonverbal auditory perceptual deficit. This deficit precludes normal speech perception which, in turn, disrupts the normal development of speech and language" (p. 96). There are, however, some serious problems with the Tallal, Stark and Curtis (1976) investigation. For example, the 40% agreement of the two listeners for the transcription of subject's speech demonstrates weak interjudge reliability of the production test. Though all dysphasic subjects reached criterion on perception on test of perception of steady-state vowels, they did make errors in their production of isolated vowels. This is contradictive of the statement that the speech production deficits of dySphasic children ”mirror their defects of speech perception” (p. 75). Even if it could be shown that the dysphasic children's articulatory problems corresponded to CV speech perception deficits, it is not clear how these problems could affect other aspects of language development that were defective in dySphasic children. Moreover, the author do not adequately describe the children's level of language skills. Tallal, Stark, Kallman and Mellits (1981) reexamined the nonverbal perceptual abilities of language impaired and normal children as a function of age and modality stimulation. Nonverbal and memory tests were presented to 11 language impaired and normal children between five and nine years old, matched for CA and nonverbal intelligence. The stimuli were presented in three modalities: auditory, visual and cross-modal. Results indicated that the language impaired group made significantly more errors than the controls, regardless of modality stimulation, when sequencing pairs of stimuli were presented at various rates. This lack of significant difference in the performance of language impaired in the two modalities did not replicate previous reported data (Tallal and Piercy 1973b) that 7 to 9 year old language impaired children's temporal processing deficit was specific to the auditory modality. Further analysis for investigation of the effect of age on language impaired and normal children's performance showed a dissociation between modality in perceptual functioning and age in language impaired but not in normal children. Whereas the younger (5 and 6 year-olds) language impaired children's skills in the auditory and visual modality were not significantly different, the older language impaired children's visual skills were significantly better than their auditory skills (p < .005). No significant difference was found between the performance of auditory and visual modality for either the younger or older normal subjects. This developmental difference did not occur for the control group. In summary, the (5 and 6 yr old) language impaired 12 were impaired in processing rapidly presented stimuli regardless of modality. However, older (7 and 8 yr old) language impaired children were impaired only in processing rapidly presented auditory stimuli (the same result found by Tallal and Piercy with children of similar age). In general, the research of Tallal and her colleagues provides a strong rationale for the continued investigation of language disordered children's auditory temporal processing skills. In particular it calls for research investigating the functional relationship between children's language comprehension and the temporal characteristics of the speech stimuli. Normal children seem to benefit from additional processing time. Berry and Erickson (1973) investigated the effect of speaking rate on children's comprehension of sentences. Kindergarten children, five to six-years of age , and second grade children from seven to eight years of age were required to listen to a sentence at one or another of the rates of 2.6, 3.4, 4.7, 5.3 or 6.3 syllables per second and then to choose the one picture of four that best corresponded to the verbal stimuli. The Ngrthwesterp Syptax Spreepipg_Te§t stimulus sentences were spoken by a normal speaker who varied her speaking rate, and each sentence was checked for accuracy using a hand-held stopwatch. Results indicated that the performance of both groups of children was significantly better at the two slower rates than at the three faster ones. 13 Studies have also been conducted on learning disabled children's responding to rate altered Speech. McCroskey and Thompson (1973) investigated whether comprehension-of a Spoken message by children with specific learning disabilities was affected by altering the rate at which the message was presented. Twenty learning disabled children between five and seventeen years of age participated in this study. The subjects had normal hearing and an IQ of 75, or better. The stimuli were 50 simple declarative 'sentences presented at five predetermined rates. There were two conditions of expansion (2.9 and 2,3 syllables per second (Sps)), a normal rate (3.6 sps) and two conditions of compression (6.8 and 5.0 sps). The stimulus sentences were mechanically time-altered using a Rate changer. The linguistic structure of the stimulus sentences was consistent throughout the test: and the degree of difficulty was kept at the level appropriate to the youngest child, which means not above a 5-year level. The results indicated that rate did not influence comprehension when data from all subjects were pooled. However, the analysis revealed that there were Significant differences among the subjects as a whole; this may mean that certain effects were obscured by the performance of the older children. Analysis of data from the 10 youngest children (5:1 to 10:3) Showed Significant differences in comprehension between various rates of Speech. These young 14 children were able to comprehend better at expanded rates (2.9 sps) than at compressed rates of speech (5,0 sps or faster). However, no significant differences were found between normal (3,6 sps) and compressed (5,0 sps) speech or between normal and expanded speech. This finding seems to be important, since normal rate is critical for processing auditory verbal information. Therefore, one may question the necessity of compressing speech, since there ware no significant differences between normal and compressed or expanded rate of speech. The fact that rate did not have any Significant effect upon comprehension when all subjects were pooled might suggest that the stimulus material (high probability and high frequency words were used -- in terms of linguistic structure and vocabulary) -- was too easy for the older children. Because of the ceiling effect of the polder children, the auditory processing abilities were not taxed by the experimental task. The effect of time compression on the listening comprehension and retention of 20 Educable mentally retarded students was studied by S.H. Zucker and Br.J. D'Alonzo (1981). The students were assigned to one of two groups, either normal rate (1.0) or compressed rate (1.5). The normal rate group had mean 1.0. 67.70 and mean CA = 10:80 (range of 9 to 13). The compressed rate group had 1.0. 63.90 and mean CA a 10 (range of 8 to 12). No other selection criteria were mentioned, such as language level 15 or hearing level. Each group was required to listen to five stories and were tested for comprehension after each story. The stories were played at normal rate with a mean rate of 125 words per minute (wpm) and were 9.00, 15.25, 7.50, 5.00 and 6.75 minutes long. The compressed stories were 6.03, 10.22, 5.03, 3.35, and 4.52 minutes long and had a mean rate of 184.50 me. The results showed no significant difference in retention between the groups listening at the normal and the compressed rates. The 1.0 group scored 3.56, whereas the 1.5 group scored 3.40. Each group achieved approximately 70% of the total score. The authors concluded that Since one group learned the material in one-third less time than the other group, with proper training and practice, comprehension and retention could be expected to be maintained at greater rates. Liles et-al. (1978) investigated the effect of pause on the auditory comprehension of nine children who had been diagnosed as language disordered. The subjects, ranging in age from 3:10 to 6:7, scored within one standard deviation of the mean for their chronological age on the Eeabgdy F39§9§§-Y9999919§Y-T9§§ estimate of IQ and below the norms for their age for spontaneous language. They also had normal hearing, though two of the subjects had a history of mild low frequency hearing loss. The children were presented with two types of commands in a sentence comprehension task. One 3-sec. pause 16 was inserted mechanically at the syntactic boundary in one half of the sentences in a set I and set II, whereas the other half of the sentences did not include any inserted pause.‘In set I, the 3-sec. pause was inserted after the first noun phrase before the conjuction in the pause condition sentences, e.g., "Put on the clown's face (pause) and the red ball." In set II, the 3-sec. pause was inserted after the first NP before the prepositional phrase, e.g., "Put the man's face (pause) on the red vest." The children were required to respond to recorded stimulus sentences by manipulating plastic pieces of a ”colorform' game. Results indicated that subjects' performances improved significantly as a function of the pause time in sentence set II but not Significantly in sentence set I. Comparison between the total number of errors made in pause conditions in set II and those made in set I revealed a Significant difference. Children younger then 5 years made fewer errors in pause condition. In other words, it appears that these "language disordered” children performed better on certain sentence comprehension tasks when processing time was increased and that this increase in time facilitated the performance of the younger group more than it did the older group. However, one might question whether these children were really language disordered, since their receptive language performance on the ?eab9§¥-219§9§e Y9?§P91§§¥-Te§§ was within one standard deviation of the mean for their chronological age. 17 Temperal-venippletipn-pf-§peesn-Stinpli The temporal manipulation of speech and its effect upon Speech perception has been the subject of investigation for some time (cf. Beasley and Maki, 1976). For the most part this research has investigated the effect of time compressed and time expanded speech upon giggle Y9§9-39§§1139193139Y for normal populations and sensorineural hearing impaired populations. These studies generally report that as time compression is increased, Speech intelligibility decreases. There appears to be a gradual decrement in subjects' word intelligibility at approximately 40% compression, followed by a more pronounced decrement in ability around 60% compression (Beasley and Maki, 1976: Beasley, Schwimmer and Rintelmann, 1972). There are a few reported studies which investigated the influence of time compressed Speech on the word intelligibility abilities of young normal children with auditory disorders. Beasley, Maki and Orchik (1976) studied the effects of time compression of speech on children's auditory perception. They presented the W9r§-!ntel;isipility-p¥ Eistsre Identifisetipn (WIPI> and the 993:59 word list at 0%, 30% and 60% compression at the intensities of 16 dB SL and 32 dB SL to 60 normal children between the ages of 3:6 to 8:6. The stimuli were mechanically time-compressed using 18 the Lexicon Varispeech I (Lee, 1971). They reported that the average intelligibility scores increased as a function of increasing age and sensation level, but it decreased with increasing amounts of time compression. The PB§:59 was found to be more difficult than the WIPI for each group under all conditions of time compression and sensation level. On the 995799 measure at 32 dB SL, the decrease in intelligibility for the three age groups was similar as a function of compression, with the 8 year-olds showing the highest and the 4 year-olds exibiting the lowest intelligibility scores (the 8 year-olds performed nearly as well at 30% time compression as they did at 0% time compression; at 60% they performed only 10% below their scores on 0% condition. On the WIPI, the children performed nearly as well as the adults reported on by Beasley et al. (1972b). Procedures similar to those used by Beasley, Maki and Orchik‘(1976) have been applied to various clinical groups including adults with sensorineural hearing losses (Kurdziel, Rintelmann, and Beasley, 1975), persons with hemispheric lesions (Snow, Rintelmann and Miller, 1977), children with reading problems (Freeman and Beasley, 1978), children with acquired aphasia (Oelschlaeger and Orchik, 1977), and children with auditory perceptual problems (Manning, Johnston, and Beasley, 1977). Oelschaeger and Orchik (1977) examined the effect of time compressed speech on the speech discrimination of an 19 11 year-old with an acquired left hemisphere lesion. They presented the WIPI at 0 and 60% time compression. Results indicated a decrease in discrimination in both ears at 60% time-compression and, moreover, Significantly poorer speech discrimination at 60% compression in the ear contralateral to the Site of lesion. The finding was regarded clinically significant and was consistent with previous research findings with adults (Kurdziel, Noffsinger, and Olson, 1976). It appeared that the use of time-compressed speech may be a valuable procedure in the assessment of central auditory function and as a diagnostic tool in the determination of Site of lesion. Using tests of compressed speech and sentential approximations and monosyllabic words as stimuli, Freeman and Beasley (1978) studied the auditory processing abilities of children with reading problems. The 'performance of reading impaired children between 8 to 11 years old who performed at least two years behind on grade equivalent on the 939999994 bpéitery-§9npept9e;izetipn-Test was compared to that of normal children. T9§-W9§9 Enteltigipilitv-hy-Eistsre Teentifipetien test m Hmoficfiao .AUBBV couoafino coo same cmxop .xmzsuc «Hmom sunusuaz Hausa: auneuaoo .1263 won HmofimoHocOuco uOu mcofiumfi>oc oumocmuw 0cm memos .mumo Hmsofi>wocH ” H mamme .‘I---’--------------'--------------------------'----------‘----------. mm.~ mH.N HH.N o.ma «.ma ma m.mm om mud mm.hh .m H>.H vh.a om.a o.HH o.NH o.eH me ac «AH mu .m mv.¢ o~.m hm.m o.hH o.ma o.>H Nv mm mad mm .m mm.H «m.a om.a o.NH o.mH c.5H mm ¢m ova m5 .5 mm.a mo.~ Nv.~ o.NH o.va o.ma mm we ANA Mb .o .290 ..... Hm ....... Hz ..... mu ..... mutiwz----mu-au&a--wma----mui”3mm Emedmumwwfim-Nwm»-omm-n-o3m-mmdmmmm Acm5:wucoov H manta 27 Hm.o oa.o mh.o «.N N.H H¢.H m.m hm.~ h.NH N.OH om mm.H ov.a vo.a m.ma h.ma o.ma av mm.vm mad mm x a~.~ am.H ae.~ o.MH o.sa o.kH . as am sea as .ma mH.H AN.H -.H o.¢H o.aH o.aH em am ems ma .NH AH.H mm.a ma.o o.as o.aH o.o~ me mm was can .HH ma.a ha.s as.H o.ms o.o~ o.c~ me am ass as as .‘---'-‘-----'---------‘D--‘----------‘---‘-------------'--'-‘--. so em 92 mo mm mz . Ho pope mzzo aw Amos o.mammc Amp»-m~a u omNm-mmm-msouo Avmacflucoov H mammfi 28 Procedures Parts III and IV of the T9399-T9§§-§9§ Children were presented to the individual children. The stimulus. sentences were administered under three conditions of speech rate: one unaltered normal rate, one temporally expanded, and one temporally compressed. The Lexicon Varispeech II recorder was used to temporally alter the stimuli. The tape-recorded items were spoken by an adult male native speaker of American English. A two dimension color display representing the two standard arrangements of the tokens for parts III and IV of the Tgkeg Test was placed in front of each child. Arrangement A included five color squares and five circles each one of different color (blue, red, yellow, white, green). Arrangement B included five large and five small squares and also five large and five small circles. The test response required the subjects to follow the command as Specified by the tape-recorded stimulus. The child was instructed to listen carefully and to respond to each recorded command by touching the stimuli display. It was stressed that the child should wait until each command was finished, and then respond. This was ensured by having an auditory Signal recorded on the second channel of the stimulus tape immediately following each command. The children were instructed not to respond until they heard the ”beep.” Each youngster was given several practice 29 trials with non experimental sentence stimuli prior to the actual presentation of the experimental stimuli to ensure they understood the task. B§§e_Cgpditigps. The stimulus items were administered to each subject under three Speech rate conditions: 1) normal rate of speech (120 w.p.m.) (0% time compression), 2) 50% expanded rate of speech (60 w.p.m.), 3) 50% compressed rate of speech (240 w.p.m.). In presenting the experimental task to the subjects, the temporal conditions were counter-balanced to control for ordering effects; however, the unaltered rate (0% compression) was always presented first. E§e:Se§S Before administering the experimental task, each child was given a pre-test to assess his knowledge and recognition of the Shapes, Sizes and the five colours of the tokens ' utilized in the test. The purpose of the pre-test was to insure that the children 1) comprehended the content of the experimental stimuli and 2) understood and could do the experimental task. For the pre-test condition, ten stimuli were constructed containing vocabulary items of the experimental stimuli and representing the concepts of ”circle," "large,” “small,” and the colours ”red,“ “green,” ”yellow,” "blue,“ and I'white.‘I The concept-words were preceded by the words ”Show me,“ e.g., ”Show me small,” “Show me circle." The pre-test stimuli were presented with live voice at 30 a normal rate of speech. Prior to administration of the pre-test, the pre-test stimuli were placed in front of the child, followed by these instructions: I am going to ask you to do different things with these (the tokens). Listen carefully. I want you to do what I say. Listen carefully because I can say each one only once. The examiner asked the subject to point to one token described both by colour and Shape (e.g., ”Point to the red circle," ”Point to the large blue circle and the large green circle"). A criterion of 100% accuracy on the pre-test was required for participation in the study. Erosrinentel-tss5 The eXperimental stimuli were presented to individual children of the three groups. The testing took place in a prefabricated single-walled test chamber IAC 1200 series at the Michigan State University Speech-Language—Hearing Clinic. The stimulus items were presented at 60 dB Hearing Level binaurally, via speakers using a tape recorder(Sony model TC 80 L) coupled to a speech audiometer Grason-Stadler, Model 162 located outside the IAC booth. Three of the youngest children were seen in a quiet room located in their pre-school. The stimulus items for these children were presented at a loudness level judged adequate for hearing. For these children the stimuli were presented in free field using a Fisher PH 463 High fidelity tape recorder. This alternate procedure was necessary because of 31 difficulties in the transportation of the pre-school children to the Michigan State University Speech-Language-Hearing Clinic. During each testing session, the subject was seated at a table facing the two speakers, with the experimenter seated next to the child. After the stimuli had been placed in front of the child and followed by the experimenter's instructions, the stimulus items were administered. A second examiner was seated at the other Side and slightly behind the child. Both of the experimenters kept time and accuracy measures on the child's response. A third research assistant controlled the presentation of the stimuli. Scoring-9§-resppnses The children's individual comprehension responses to each test item at the different rates was scored and recorded by two experimenters immediately following each response. Latency of the subject's response to each experimental item was recorded using electronic stop watches; latency of response was considered the period of time occurring between the end of a presented stimulus item and the subject's final touching of a response display. Two judges, who had been trained prior the experimental sessions in timing the child's responses, kept the time for purpose of reliability. The final score for latency of response was the average of the child's elapsed times in completing each test item. The recorded ”beep” on the second channel of the stimulus tape which was synchronized 32 with the ending of the verbal stimulus, served to signal the eXperimenters when to start timing and the child when to reSpond to each command. Reliability During the administration of the experimental task, reliability measures were taken for both accuracy and latency of the subjects' individual responses. Interjpége-reliepilityo To determine interjudge reliability measures for scoring the accuracy of the responses on the Tgkep-Tst, a second person independently observed and scored individual experimental trials selected at random. For each subject this judge observed and scored twenty subject's responses during the experimental sessions, for a total 260 responses. A point-to-point comparison was made between the two experimenters' scoring of the children's responses. The interjudge agreement was 100%. Interjudge reliability for the timing the subject's responses was also determined. This was based on the judges' independent timing of five experimental sessions. This measure was determined by calculating a Pearson product—moment correlation between the time measurement of the two judges. The interjudge reliability for the timing ranged from .85 to .99 with a mean of .92, thereby establishing the reliability of the timing procedure employed. 33 resorts Individual children were assigned a score for each of the following variables: number correct for the normal speech rate, number correct for the expanded Speech condition, number correct for the compressed speech rate, response time for the normal rate condition, response time for the expanded condition, and response time for the compressed speech rate. accoracY-9§-Besponse The scores under the three temporal conditions for each age group were as follows: Group I: a) Normal rate of speech (0% time compression): scores ranged from 9.0 to 18.0 (i a 14.60; so = 3.43); b) 50% expanded rate of speech (60 w.p.m.): scores ranged from 9.0 to 17.0 (x =14.60; SD = 3.20): c) 50% compressed Speech (240 w.p.m.): scores ranged from 5.0 to 18.0 (i = 11.19; so = 4.71). Group II: a) Normal rate condition: scores ranged from 14.0 to 17.0 (E's 16.00; so . 1.41); b) 50% expanded rate of speech (60 w.p.m.): scores ranged from 12.0 to 19.0 (i a 15.25; so - 2.98); c) 50% compressed (240 w.p.m.): scores ranged from 11.0 to 17.0 (X = 13.00; SD = 2.70). Group III: a) Normal rate of speech: scores ranged from 17.0 to 20.0 (x a 19.00; so = 1.41); b) 50% expanded rate of Speech (60 w.p.m.): scores ranged from 17.0 to 20.0 (i a 18.75; SD = 1.25): c) 50% compressed speech (240 34 w.p.m.): scores ranged from 13.0 to 18.0 (i =15.50; so = 2.4). tetencY-9§-Besppnse For the latency of response under the three temporal conditions the scores were as follows: Group I: a) Normal rate of speech: scores ranged from 1.65 to 4.27 (i = 3.22; SD = 1.18): b) 50% expanded rate of Speech: scores ranged from 1.38 to 7.07 (X = 3.87; SD = 2.06): c) 50% compressed speech: scores ranged from 2.39 to 6.63 (f = 4.21; SD = 1.67). Group II: a) Normal rate condition: scores ranged from 1.30 to 2.87 (i = 2.11; so = 0.68); b) 50% expanded rate of speech: scores ranged from 1.64 to 3.20 (i = 2.15; SD = 0.71): c) 50% compressed speech: scores ranged from 1.71 to 4.45 (k = 2.53; so = 1.28). Group III: a) Normal rate of speech: scores ranged "from .93 to 2.67 (§'= 1.64: so = 0.76); b) 50% expanded rate of speech: scores ranged from 1.27 to 1.50 (§’= 1.40; SD = 0.10): c) 50% compressed speech: scores ranged from 1.17 to 2.27 (i . 1.53; so = 0.51). Table 1 presents the means and standard deviations for each age group at the listening rate conditions. Table 2 is a summary of the groups mean time performance on the various measures. 35 Hm.o oH.o 05.0 e.~ ~.H e.H m.~ h.oH ~.oa om v HHH mm.H o¢.H vc.a m.ma h.ma o.ma ~.vm o.oaa o.mm .m mN.H H>.o mm.o h.~ m.~ v.H o.m m.HH mo.v om v HH mm.~ ma.~ AH.~ o.ma ~.ma o.ma o.om ~.m~a m~.>h .M no.a mo.~ mH.H an.v ~.m e.m m.m o.mm m.~ mm m H Hm.¢ hm.m ~m.m mH.HH o.¢a w.¢H o.ov m.voa m.mm .M 2 moomw -Hu ...... Hm. ...... an ..... Wu ..... an ..... Wm - - - dam. - - - -mzzo ..... mu .mousmmme msofium> may co oocmELOLHoQ mean come .mQSOMm on» no oanmu aumEEsm .N wands 36 Pets-ensly-is Nonparametric statistics were used for the data analysis. The Friedman Rank ANOVA for k related groups was used to compare the children's accuracy scores across the three rate conditions of normal, expanded, and compressed. The same kind of analysis was used to compare the children's response latency across the three rate conditions. Scores-sithin-Chronological-9ge-Stoops-ssross-Bpte Conoitions No Significant differences were found for the variable of accuracy for the five-year-old children across the three rate conditions (£2: 3.9, p = .14) nor were there Significant differences found for the Six-year-old children (X2= 4.8, p = .08). A significant difference was observed for accuracy of responding across the three rate conditions (X22 6.12, p = .04), for the seven-year-old children, however. Figure 1 presents cell means for accuracy of performance. iime-3ithin-chronological-599-9ropps-scross-Bete-9999itiops No Significant differences were found for the variable of response latency for the five year-old children across the three rate conditions (X2= 1.2, p > .10) nor were there Significant differences found for the six—year-old children (Xzs .000, p > .10). Also, no significant differences were found for response latency for the seven-year-olds across 2 the three rate conditions (X a .500, p > .10). 37 Figure 2 presents cell means for latency of response Scores-99r9ss-§hronol99icel-499-9ronps-for-Bete-9999tions The Kruskal - Wallis nonparametric test for k independent groups was used to compare the accuracy scores across the three age groups of five-year-olds, six-year-olds and seven-year-olds, for each of the three rate conditions. A similar analysis was completed for the variable of response latency. A significant difference in accuracy was noted across the chronological age groups for the 99§T91-¥§§§ condition (H = 6.30, p = .04). Significance was also found for the expgpded-r§t9 condition (H = 3.43, p = .05). No Significant differences were noted across the three age groups for the compressed rate condition (H = 3.43, p = .17). Time-across-§hr999l99ical-999-9ropps-for-Bste-§999itions For the reSponse latency no significant difference was found across chronological age for the normal condition (H = 3.91, p a .14). A Significant difference was noted across the chronological ages for the expanded rate condition (H a 5.76, p a .05). Significant differences also were found across chronological ages for the compressed rate conditon (H a 7.59, p = 0.02). Relationship-petseen-eccnrscy-sné-limins A Pearson product-moment correlation analysis was performed in order to examine the relationship between the children's accuracy and response latency across the three 38 different rates for each of the three groups. Fiye:year:gld_grggp. The correlation analysis showed virtually no relationship between accuracy and timing for each of the three rate conditions. The correlation; coefficient between accuracy and timing was -.37 for the normal rate, -.03 for the expanded rate, and it was .34 for the compressed rate. Sixyyearfgld_grggp. The correlation analysis showed that both accuracy and timing were related to rate of speech. The correlation coefficient between accuracy and timing was +.17 for the normal rate, +.78 for the expanded rate, and +.99 for the compressed rate. This indicates that while a strong positive relationship is found between accuracy and timing for expanded and compressed rate, no such relationship is noted for the normal rate. Spyep:year:pld_grgup. Accuracy and timing were found to be related to the rate condition. The correlation coefficient between accuracy and timing was -.82 for the normal rate, -.30 for the expanded rate, and -.53 for the compressed rate. This indicates that while there is a strong negative relationship between accuracy and timing for the normal rate, no relationship is noted for the expanded rate, though a weak relationship between accuracy and timing is apparent for the compressed rate condition. 20 O 18 16 /o 0. ./ 14 O >— < o (I D 10 O 2% 8 6 4‘ D-—o Normal “—0 Expanded 2. 0--o Compressed I I) 'III onoup Figure 1. Cell Means for Accuracy of Performance 5.00 4.50 4.00 3.50 3.00 TIME (sec) 2.50 2.00 1.50 1.00 Figure 2. 4O 0—0 Normal 0—0 Expanded 0—0 Compressed GROUP Cell Means for Latency of Response 41 tansnage-lmpaired-9hildren For the purpose of comparison, two language impaired children were studied to explore the possibility of any differences between their performance and the performance of normal language learning children. Two language impaired male children, 7:1 and 8:1 years old respectively, were presented with the experimental task. The criteria for including these children were as follows: a mean score of at least one year below chronological age on the Tgkep_T§st-fgr Childrgp and 9iinical-§yalpation-o§-iansoase-€999tions subtest 1 (Cl) - Processing word and sentence structure. The other selection criteria with respect to non-verbal intelligence, hearing level, history of emotional disturbances, mental 3 retardation or neurological impairment, and oral motor or sensory deficit were the same as those of normal language children. The score for the T9599-T§§§-§9§-§9339§99 for each of the children was as follows: Child 1: 35.0, age equivalent: 5:0 years - 5:5 years. Child 2: 39.0, age equivalent: 6:6 years - 6:11 years. Both of the children obtained scores within normal limits on non-verbal intelligence. 0n the golnmpia-vental-Matsrity-§cale the age equivalent for child 1 was 9:8 years: and 8:6 years for child 2. The two language impaired children were presented 42 with the original experimental task, including the pre-test, under the same methodological procedures as the normal language children. 39§93E§2 sccprasy-p§-Bespon-e The results for the children's correct comprehending under the three temporal conditions were as follows: Child 1: a) Normal rate of speech: 11.0; b) 50% expanded rate: 10.0: c) 50% compressed Speech: 9.0. Child 2: a) Normal rate of speech: 13.0: b) expanded rate: 15.0: c) compressed speech: 8.0. The mean score for the two children's correct responses was as follows: Normal rate of speech: 12.0: expanded Speech: 12.5: compressed speech: 8.5. This is compared to the following group means of the youngest group of normal children, (i.e., Group I, five-year-olds): Normal Speech: 14.6, eXpanded Speech: 14.6, compressed speech: 11.19. Recall that the total correct was 20 for each rate condition. Latenc¥-9§-Besponse For the latency of response under the three temporal conditions the scores were as follows: Child 1: Normal rate of speech: 4.72: expanded rate: 3.41; compressed speech: 3.65. Child 2: Normal rate of Speech: 1.30; expanded rate: 1.06; compressed speech: 1.36. 43 The mean score for the language impaired children's time responses was as follows: Normal rate: 301.5, expanded rate: 223.75, compressed rate: 250.7. This is compared to the following group means of the normal children, (i.e., Group I to Group II, five to six-year-olds: Normal rate time: 211.2, expanded rate time: 215.6, compressed rate time: 253.0. DISCUSSION While it was thought that temporal expansion might proved beneficial for auditory processing abilities of children, the mean differences between the normal and the expanded rate conditions were not significant. The expanded rate did not improve the listening comprehension of the five-six-or seven-year—old children beyond that demonstrated in the normal rate condition. It should be noted that the compressed rate decreased the listening performance of all children. However, only for the seven-year-old children were the differences statistically Significant. Their performance was significantly better in the normal and the expanded condition than in the compressed condition. With regard the measurement of the children's latency of response to the experimental stimuli, no significant differences were noted across the listening conditions of normal, expanded or compressed for any of the three age groups. None of the groups required significantly more or less time to respond for any altered rate of speech. There were however, differences noted across age groups, with the five-year-old children requiring considerably more time than the six-year-old children, regardless of listening rate condition. 45 The results concerning the effects of expanded rate of speech on these age groups' performance, do not appear to support the claim that oral language comprehension in normal children is improved by slower than normal Speaking rates (Berry and Erickson, 1973; Nelson, 1976) or faster rates (Woodcock and Clark, 1968). These data do, however, demonstrate that children's comprehension decreases under a faster than normal rate of Speech. Nelson (1976) reported that normal children between five and seven-years-old performed Significantly better at glgw-rates of Speech than at the fast§r_rate§ in a comprehension task. Similar differences, as a function of Slow rate, have been reported by McCroskey and Thompson (1973) in studies involving learning disabled and normal children. Like the present research, these studies found no significant difference in the children's performance between pgrmal and slower rate of Speech. Another point to be made in this study concerns the influence of Speaking rate on the children's comprehension as a function'of age. The seven-year-old children always performed better than the five-and six-year-old children. However, their performance was only significantly better in the normal and expanded rate condition. With the influence of the compressed rate the seven-year-old children did not maintain the significant difference in their performance. Looking at the relation between accuracy and time measurements, it was found that although there were no 46 significant differences between tha accuracy of performance for either the five-year—old or the six—and seven-year-old children in the compressed rate condition, there were significant differences as indicated by response latency measurements. This is interpreted to mean that in dealing with a complex task (i.e., compressed rate of speech) the younger children needed more time in decoding the sentences than did the other children. In other words, in order to maintain an equivalent level of comprehending, the younger children needed more response time, whereas the seven-year-old children did not need the extra time to respond accurately. The older children performed more efficiently in the comprehension task than the youngest children. In summary, the temporal expansion of Speech had no beneficial effects on children's processing of verbal language. The compressed verbal message, however, had extremely detrimental effect on their language comprehension. 47 A preliminary investigation of the effects that altered speech might have on the language impaired children's sentential comprehension was also conducted. It was noted that the two language impaired children responded similarly to the normal children's responses across the three temporal alterations of the experimental stimuli. Though thei scores were lower than those of the normal language learning children, no qualitative differences between the responses of normal language impaired children was observed. Given the current interactional models of language comprehension, one could assume that, as with the normal children, interactions between perceptual and cognitive and linguistic processes may be involved in the way language impaired children learn language. 48 Bartenders Arthur: G. (1952). The-Brthpr-ndaptation-9t-the-teiter international-Eeriormance-§cale- Washington. D-C.= Psychological Service Center Press. Baker, H.J., and Leland, B. (1967). petrgit-Test§_gf Leargipg Aptitude. Indianapolis: The Bobbs-Merril Company Inc. Beasley, D., and Maki, J. (1976). Time and frequency-altered speech. In N.J. Lass (Ed.), Contemporary-isspes-in-esperimental-phonetics- New York: Academic Press, Inc. Beasley, D., Maki, J., and Orchik, D. (1976). Children's perception of time-compressed speech on two measures of speech discrimination. Jgprp§l_9f-§peegh_apd Bearing-9i99rdersv 9i: 216-225- Beasley, D., Schwimmer, S., and Rintelmann, W. (1972). Intelligibility of time-compressed CNC monosyllables. 999§993-9§-§P§§99-999-§?§¥399-B§§§§§99r 35: 340'350- Berry, M., and ErickSon, R. (1973). Speaking rate: Effects on children's comprehension of normal Speech. Jgprpgl of-§peech-and-§earing-Besearchv l5: 360-366- Burgemeister, B., Blum, L., and Lorge, I. (1973). Columbia Mept§l_Matp§ity_§g§le. New York: Harcourt Brace Jovanovish, Inc. 49 D'Alonzo, B., and Zucker, S. (1981). Comparison of comprehension scores of learning disabled high school students for content presented aurally at variable rates of speech- Perceptpal-and-hot9r-§tillsv 53, 499-505. Disimoni, F. (1978). T9kep_Te§§_§9§_Childrep. Boston: Teaching Resources Corporation. Duchan, J. and Katz, J. (1983). Language and auditory processing: top down plus bottom up. In E. Lasky, and J- Katz (Eds): central-npditor¥-?rocessipg-disorders Baltimore: University Park Press, 31-45.9 Freeman, B., and Beasley, D. (1978). Discrimination of time altered sentential approximations and monosyllables by children with reading problems. ggugpal_9f-§peegh-apd Fearing-Be§earchv 2i: 497-506- Glasser, A., and Zimmerman, I. (1967). Clinical interpretation-9§-the-9echsler-intelligence-scale-£9! ghildrep (WISC). New York: Grune and Stratton. Kurdziel, 8., Noffsinger D., and Olsen, W. (1976). Performance of cortical lesion patients on 40%, 60% time-compressed speech materials. Jggrpglogf-tbe bmerican-nndiological-§99ietyu 2: 3-7- Kurdziel, S., Rintelmann, W., and Beasley, D. (1975). Performance of noise-induced hearing-impaired listeners on time-compressed CNC monosyllables. qoprnal-9§-the-emerican-bpdi9l99ical-Society: l: 54-60. 50 Liles, B.Z., Cooker, H.S., Kass, M., and Carey, B.J. (1978). The effects of pause time on auditory comprehension of language-disordered children. donrnal-o§-commanication-disordersI ii: 365-374- Lowe, A.D., and Campbell, R.A. (1965). Temporal ' discrimination in aphasoid and normal children. Ludlow, C.L.,(l980). Children's language disorders: Recent research advances. App-Neurgl‘, 7, 497-507. McCroskey, R.L., and Thompson, N.W. (1973). Comprehension of rate controlled Speech by children with specific learning disabilities. Jgg§9§1-9foLearpipg Pisapilities: 9: 621-627- Manning, W.H., Johnston, K.L., and Beasley, D.S. (1977). The performance of children with auditory perceptual disorders on a time-compressed Speech discrimination measure- donrnal-9£-§peech-and-hearing-9isorders. 42, 77-84. Marslen-Wilson, W.D., and Tyler, L.K. (1981). Central process in speech understanding. Psychological mechanisms of Language- rhilgsophical-Transactions 9§-the-Bo¥al-docietv-o§-tgndonv B 295: 317-332; Marslen-Wilson, W.D., and Welsh, A. (1978). Processing interactions and lexical access during word recognition in continuous speech. ngpitiye_Psyghglggy 19, 29-63. 51 Nelson, N. (1976). Comprehension of spoken language by normal children as a function of speaking rate, sentence difficulty and listener age and sex. Child Development: 92: 299-303 Oelschlaeger, M.L., and Orchik, D. (1977). Time-compressed speech discrimination in central auditory disorder: A pediatric case studY- Jeprnel-e§-Speech-and-hearing Pieerderer $2: 483-485- Ross, M., and Lerman (1970). A picture identification test for hearing-impaired children. Jggrpal_gf_§peegh-apd hearing Research: ll: 44-53- Semel, E.M., and Wiig, E.H. (1980). Clipigal-Ey§luatigp-gf Lapguage_nggtipps: Examiner's Manual. Columbus, Ohio: Charles, B., Merril Pub. Co. Snow, J., Rintelmann, W., Miller, J., and Konkle, D. (1977). Central auditory imperception. Lgrypggsggpe, 87, 1450-1471. Tallal, P. (1980). Auditory processing disorders in children. In P. Levinson, and C. Sloan (Eds.), hoditery-preeeeeing and-langeage- New York: Grune and Stratton. Tallal, P. (1975). Perceptual and linguistic factors in the language impairment of developmental dysphasics: An experimental investigation with the Token Test. 99§§9¥r 33' 196'205- 52 Tallal, P. (1976). Rapid auditory processing in normal and disordered language development. Jgurpel-gf_8peegh_epd hearing-8eeearehv l?: 561-571- Tallal, P., and Piercy, M. (1978). Defects of auditory perception in children with developmental dysphasia. In M.A. Wyke (Ed.), Deyelgpmeptel-dy§phe§ie. New York: Academic Press. Tallal, P., and Piercy, M. (1973a). Defects of nonverbal auditory perception in children with developmental aphasia. Neture, 251, 468-499. Tallal, P., and Piercy, M. (1973b). Developmental aphasia: Impaired rate of non-verbal processing as a function of sensory modality. Negggpeyghglggie, 11, 389-398. Tallal, P., and Piercy, M. (1974). Developmental aphasia: Rate of auditory processing and selective impairment of consonant perception. Negggpeyghglpgie, 12, 83-93. Tallal, P., and Piercy, M. (1975). Developmental aphasia: The perception of brief vowels and extended stop consonantS- Feetepeyehologia: l3: 69-74- Tallal, P., Stark, R.E., and Curtis, B. (1976). Relation between speech perception and Speech production impairment in children with developmental dysphasia. erain-and Langeage: 3: 305-317- Tallal, P., Stark, R.E., Kallman, C., and Mellits, E.D. (1981). A reexamination of some nonverbal perceptual abilities of language-impaired and normal children as a function of age and sensory modality. Jgurpel_gf 53 $peeeh-and-hearing-8eeeareh: 29: 351-357- Tyler, L.K., and Marslen-Wilson, W.D. (1981a). Children's processing of spoken language. J99§9§1-9§_Yerbel yearning-and-Yerhel-eehaYior: 29: 400-416- Wechslet: D- <1974l- Weeheler-intelligence-§eale-§or ghild§e9-8eyi§ed (WISC). New York: THe Psychological Corporation. Woodcock, R.W., and Clark, C.R. (1968). Comprehension of a narrative passage by elementary school children as a function of listening rate, retention period, and IQ. Jeernal,ei-Qemmenication: id: 259-271- Zucker, S.H., and D'Alonzo, B.J. (1981). Time compressed speech and the listening comprehension of educable mentally retarded students. Mepte1_8eterdetigp, 19, 177-179.