i PERFORMNCE Q? CHELDREN 0F VARYING AGES Ui‘ébER €ONDEYION3 Q? DELAYES SPEECH giifl-BACK Thesis {or $510 Degree 0? M. A MECIHGAN STATE WEVERSETY John Eamea Gawronskfi 1958 TH EGIB J z; . . J '. t "v E? ‘3 .' ”’Mtxf PERFORMANCE OF CHILDREN OF VARYING AGES UNDER CONDITIONS OF DELAYED spasca FEEDBACK by JOHN JAnzs GAWRONSKI AN ABSTRACT Submitted to the College of Science and Arts Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1958 m «in ‘ { \ IHESIS ABSTRACT ‘Delayed speech feedback refers to an experimental condition of thich the time component of auditory feedback or side-tone is delayed beyond .03 seconds . This modifica- tion in the monitoring system that constantly informs the speaker of his continuous speech performance , leads to cer- tain speech disturbances. Research with adults using delayed speech feedbacldDSP) had indicated that these disturbances are generally in the form of increased vocal intensity, re- tarded reading rate and increased number of articulatory errors(th‘e most prominent being syllable repetition character- istic of stuttering} . The present study was designed to inves- tigate the behavior of children under DSP, with the considera- tion that the degree of disturbance effected in these children under US? might reflect the developmental level of training in linguistic skills. Fifty Ss (ages: 6.9 - 13.6) were used in this study forming I; groups of male 33, 10 8s per group and one female group of 10 3s. The instrumentation used to induce and amplin the (1er signal was: (a) a tape recorder modified to produce a; delay of 0.63 seconds, which pro-tests had Shown to be consid- erably disturbing to children, (b) a pro-amplifier and amplifier system used to amplify the delayed signal which was not limited at the headset and (c) a microphone and headset used to transmit and receive the delayed signal. Each 3 us presented with 6 9‘ "“nrbr‘”01 1e db1ck trials (35 F) fOLlOEJed by 6 delave m ‘U o L“ w bixk trialsI The 35 were instructed to recite a simule 6 word sentence which was identical for all groups, The initial encounter with D3? eventuated in defrenens tel changes in the speech performance of all Fronps in terms of increases in sound pressurea syllable duration and articu~ latory errors. During this first delay trial significant positive correlations were found between sound pres dz3 and articulatory errors and syllable duration and articulntcrv errors, As DEF trials progressed, all groups reaponded with continuous increases in Sound pressure. The only groun stow— t‘ in? snee ch centrol during 33? n3 ,ne eldest 553U?(4T&2:1?.h~ C!) 1356). whom mi Ltained the same le el OE syllable duration and decreased in articulatory errors. This finding seems to ngest a dichotomy between earlier and later Speech d2'alOnm UT. 1 terms of those factors which mediate stable syeezh benaviwr when the Speech situation becomes far from idzzi. It The female group és fiohn3.rd j eli other spougg g: 1-;. 17'? ‘ ‘-' ' 5 e133eat levels of sound pressure during bqu When e;i~ 1 c' , . «I»; n‘ 1:. ‘ad with a comparable male,age group ;n r t? e1» n3 sitnzficant differences were fourw.£kn-zxur measure. -. f‘ m ‘ \u. i . ' “v 2-. '. -.-' .'. a ’ p“1"e-’” ’» - .n " ‘e ,s-~ «2 1 : 9-” . ' -. »' - ». ' ’ a v-' C 1. 0v 3 a 75.1 .; .‘TY..- 1:; l I". K- Ct; to tilt: 'C I. dc: 5': t. €2.23); 9 L‘. f; 11’ :LI; a!“ 1...: _, i -133- -. ’ .., " ,. -.“‘ - - - ‘ ' - , V . , I. -3 a g _ ,‘z _ , "I ~~ , I.“ intranzuwt fig the reletienshin between behavior under OxF rd ’ .-., T »'h ': r- . . fl 1*; -=. . ,; . ‘- ‘ __ . 1 I: r ' - - . ,._ - _. I y _ . tn as l Fa1:.tzf u<1lis XECUlLlEe can iderobie concentratzcra 3%;h as fluent si ent reading and comnrehwnsien and these ennetiznees requiring a considerable amount of speech tfifitfolfi sgeh as formal classroom speaking, PERFORMANCE OF CHILDREN OF VBRYIHG AGES UNDER CONDITIONS OF DELAYED SPEECH FEEDBACK by JOHN JAMES GKWROHSKI A THESIS Submitted to the College of Science and Arts Michigan State university of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MISTER OF ARTS Department of Psychology 1958 ACKNOWLEDGEMENTS The writer wishes to express his sincere gratitude to: Dr. Stanley C. Ratner, his graduate advisor, under whose direction this study was designed and carried through: Drs. M. Ray Denny and Charles Hanley, for their constructive criticism of the manu- script: Sister, Matthew Ann, Mother Superior of Resurrection Catholic Parochial School, under whose cooperation students were asseMbled for this study. TXBLE OF CONTENTS INTROI)UCTION o o o o O o o o o o a o o o o o o o O PrOblem . O O O O O O O O 'O O O O O O O O O 0 METHOD . . . . . . . . . . . . . . . . . . . . . . Subjects . . . . . . . . .-. . . . . . . . . . Instrumentation . . . . . . . . . . . . . . . Procedure . . . . . . . . . . . . y . . . . . Measurement . . . . . . . . . . . . . . . . . RESULTS . . . . . . . . . . . . . . . . . . . . . . Performance During synchronous Feedback. . . . Effects of‘Initiel Encounter with Delay . . . Age Group Differences Hanifested in ReSponse to Delay . . . . . . . . . . . . . Changes in Performance Accompanying Subsequent Encounters with Delay . . 9;. . . . . . . . Age Group Differences Manifested During the Final Encounter with Delay_. . . . . . . . . Relationship Between the Measures . . Performance of the Female Speakers With Delay DISCUSS IO}; 0 g o o a o o 0 o 0 O O O O I O O O O 0 Performance During Synchronous Speech Feedback Performance During Delayed Sneech FeedbaCk . . 11 17 17 18 18 20 PAGE Changes in Performance During Subsequent Encounters with Delay . . . . . . . . . . . . 31 RelationShip Between Measures . . . . . . . . . 36 Performance of Female Speakers With Delay . . . 36 WHY O O C O O O O O O O O O O O O O I O O O O 0 - 3 8 REFEREIICES O O O O O O O O O O O O O O O O O O O O O 61 APPENDIX A O O O O O O O O O O O O O O 0 O O O 0 O 44 APPENDIX B O O O O O O O O O O O I O O O O O O O O 47 TABLE I. II. III. IV. A LIST OF TABLES PAGE Statistical Analysis of Variance to Determine Whether the Varying Age Groups Differ Significantly in Sound Pressure During the Final Synchronous Trial . . . ; 48 Statistical Analysis of Variance to Determine Whether flhe Varying Age Groups Differ Significantly in Syllable Duration During the Final Synchronous Trial . . . . . 48 Statistical Analysis of Variance to Determine Whether the Varying Age Groups Differ Significantly in Sound Pressure During the Initial Delayed Trial . . . . . . 49 Statistical Analysis of Variance to Determine Whether the Varying Age Groups Differ Significantly in Svllable Duration During the Initial Delayed Trial , . . . . 69 Statistical Analysis of Covariance to Determine Whether the Varying Age Groups Differ Significantly in Syllahle Drration During the Initial Delayed Trial Unon Taking into Account the Age Differences Indicated for Syllable Duration During fine Final Synchronous Trial . . . . . . . . . . . . . 50 TABLE VI. VII. VIII. IX. PAGE Statistical Analysis of Variance to Determine Whether file Varying Age Groups Differ Significantly in Number of Articulatory Errors Made During die Initial Delayed Trial . . . . . . . . . . o o o o o . 50 Statistical Analysis of Variance to Determine Whether the Varying Age Groups Differ Significantly in the Amount of Change in Sound Pressure Occurring Between the Initial andthe Final Delayed Trials . . . . 51 Statistical Analysis of Variance to Determine Whether the Varying Age Groups Differ Significantly in the Amount of Change in Syllable Duration Occurring Between the Initial and the Final Delayed Trials . . . 51 Statistical Analysis of Means to Determine Which Groups Differ Significantly From Each :weragzz Other in the Amount of Change in Sound owsAflwd' Brassere Occuring Between the Initial and the Final Delayed Trials . . . . . . . . . 52 Statistical Analysis of Variance to Detenmine Whether the Varying Age GrOUps Differ Significantly in the Amount of Change in Articulatory Errors Occurring Between the Initial and the Final Delayed Trials 53' 00.000000000000000 \ TABLE. XI. XII. XIII. XIV. XV. PAGE Statistical Analysis of Means to Determine Whidh Groups Differ Significantly From Each Other in the Amount of Change in Articulatory Errors Occuring Between the Initial and the Final Delayed Trials . . . . . . . . . . ..g 53 Statistical Analysis of Variance to Determine Whether the Varying Age Groups Differ Significantly in Sound Pressure During the Final Delayed Trial . . o . . . . . . . . . ._53 Statistical Analysis of Variance to Determine Whether the Varying Age Groups Differ Significantly in Syllable Duration During the Final Delayed Trial . . o . . . . . . . . . 54 Statistical Analysis of Variance to Determine Whether the Varying Age Groups Differ Significantly in the Number of Articulatory Errors Made During the Final Delayed Trial.‘ 54 Statistical Analysis of Means to Determine Which Groups Differ Significantly From Each Other in Number of Articulatory Errors Made During the Final Delayed Tri'al . . . . . . 55 Statistical Analysis of Variance to Determine Whether the Female and Male Group Differ Significantly in Sound Pressure During the Initial Delayed Trial 0 o o o o o o o O 55 TABLE XVII . XVIII PAGE A Statistical Analysis of Variance to Determine Whether the Female 31d Hale Group Differ Significantly in Syllable Duration During the Initial Delayed Trial . . . . . 56 Statistical Analysis of Variance to Determine Whether the Female and Male Group Differ Significantly in the Number of Articulatory Errors Mada. During the Initial 56 Delayedmalooooooooooooooo LIST OF FIGURES FIGURE PAGE 1. A Close-up Photograph of the Instrumentation EmployedinthisStudy. . . . . . . . . . . . 7a 2. A Photograph Indicating the Position of the Experimenter and the Subject During the Experimentation . . . . . . . . . . . . . . . 7b 3. Means of Male Groups for Three Peak Readings of Sound Pressure Level During the Final - . Synchronous Trial and Six Delayed Trials . . . 18a 4. Means of Male Groups for Syllable Duration, a Measure Obtained by Dividing Actual Phoneting Time Per Trial by the Number of Syllables Produced Per Tria1(repetitions included), During the Final Synchronous Trial and Six . Delayed Trials . e . . . . . . . . . . . . 0 18b 5. means of Male GrOUps for the Number of Articulatory Errors Made.During Six Delayed Trials.......o............IBC INTRODUCTIOI The speaker is constantly being informed of his continuous speech performance and guided accordingly by a system composed of several feedback channels, the most prominent of these being feedbacks via bone conduction, muscular movements of the speech organs, and the vocalized aspect of this system referred to as auditory feedback or side- tone. Interference with any of these feedbacks should eventuate in certain speech changes. Recently researdh on the auditory feedback or side-tone has been implemented by use of the Audio~ Signal Delaying unit which delays the side-tone at various time delays. Invading this system of synchronous air-conducted speech by delaying the side-tone leads to certain speedh dis— turbances. Characteristic of these experimentally induced speech disturbances are the following: (a) retarded speaking rate (1) (b) increased phonetion time (2) (c) increased sound pressure (3) (d) increases in articulatory errors, the most prominent being syllable repetition. Characteristic of stuttering (4) (e) physiological changes. such as increased heart and blink rate and decreased 68R (5) (f) qualitative reactions such as blocking. facial contortions and trembling (6) .lost of the research employing delayed speech feed- back has been conducted with adults who are instructed to read or recite a given passage. A.number of independent and dependent variables have been investigated in this way: (a) (b) (o) (d) (e) (f) (9) (h) (1) varying the time interval of delay (7) varying the intensity of delay (8) varying the verbal material employed (9) assessing the effects of practice on per- formance under delayed conditions (10) employment of delayed speeCh feedback as a a stress producing condition.(ll) using delayed speech feedback for detecting auditory malingering and psychogenic (functional) deafness (12) (13) 'using delayed speedh feedback as a method to investigate certain personality characteristics (14) (15) using delayed speech feedbaCk to investigate the physiological correlates of disrupted speech (16) using this device to assess the quasi-permanent effects of disrupted speedh (l7) and the adaptation to these effects (18) At the time this study was undertaken no known research had employed children as subjects for the study of reactions ,to delayed speech feedback.1 Developmental changes in lan- guage are knosn to occur from birth through the grade school period, and may occur throughout the entire life span of an individual.“ Beginning from the time the child hears his own voice for the first time (supposedly during the "birth cry”) typical changes follow (20): from the random vocal behavior occurring during the first few weeks of life, the first aspect of speech to be controlled by the organism seems to be volme, later control extends to the vocal cords thus over pitch. Thereafter, at about five months, the complicated muscles nediating speech sounds are regulated by the infant]... During-later development, after the production of sev- eral words, the child learns to express himself in a meaning- ful we; to others in his imediate environment. As the child reaches the school period, emphasis on primarily the vocal- iscd aspect of speech is reduced as he begins reading, in- creasing the complexity of his verbal constructions, accumula- ting a. wider vocabulary, attending to meanings implied in written and spoken discourse, and attaining a broader frame of reference. During the closing stages of the school period ‘ and throughout adult life , language development continues , 1 feedback wit? ghefidgetgdzu? £31331“ ifzhzfi ii???“ ”9““ 4 enhanced by the various experiences afforded an individual.' Thus. not only does the type of language experience probably change from childhood to adulthood,‘but also the absolute amountof practice with language. speaking and listening, increases.with age. Therefore. the degree of interference in speaking performance caused by delayed speech feedback nay‘be related in some simple fashion to language development. With the above consideration in View. this study was conducted to investigate the characteristic reactions of children of varying ages under conditions of delayed speech feedback. Jonson Subjects Prom.a total of 72 elementary students subjected to conditions of delayed speech feedback, 40 male and 10 female subjects were selected to comprise four groups of male sub- jects. 10 per group and one group of 10 female subjects. The four male groups range in age as follows: (a) Group 1. ages 6.9 - 7.9 (6.9 meaning here 6 years, 9 months etc.); (b) Group 2. ages 8.6-- 9.11; (c) Group 3, ages 10.7 - 11.9: (d) Group 4, Ages 12.6 - 13.6. The single female group ranged in age from 8.2 - 9.1.. These subjects were selected from the larger group of 72 subjects in terms of age and grade requirements propose for this study. normal speedh and hearing and compliance with the instructions. These subjects represented grades 1 through 8 from a summer session at Resuarection ParoChial School. a center for summer school students for the various Catholic sdhools in Lansing. Muchigan. It is of significance to note, that, in general, these students were attending summer school because they were either slow learners or had a considerable number of absences during the regular school year. Consider- ation of sex differences with regard to performance under delayed speedh feedback were incidental to the main purpose of this experiment, therefore the female group was omitted from the major statistical analysis. Appendix Ashows the classification of these subjects in terms of ages, grades, schools attended, 10 scores. and number of other siblings in family, when reported. W The apparatus used for inducing and recording delayed speech feedback is shown in Figure l and consisted of the following Components: (a) A Revere Magnetic Recorder modified to deliver the delayed signal. The chief modification con- sisted of the addition of two playback heads to the standard model. The delay interval was determined by the tape speed in conjunction with the established distances between play- back and pick-up heads. By pairing each of the two tape speeds 3.75 ips and 7.5 ips with each playback head, four increments of delay were obtainable: 0.31 sec., 0.48 sec.. 0.63 sec. and 0.67 sec. delay. In this research the 0.63 sec. delay was employed as pre-tests had indicated that this interval was most conducive for maximum interference in children's speech. (b) A Bell System Hi-Fi Phone-Radio Amplifier. Model 2122 A, was used in conjunction with a Viking PB 60 Magnetic Playback Pre-Amplifier Full Fidelity. Model P, 7 watt 117 velt. 50/60 cycles. (d) An Astatic Corporation Microphone, Model JT-30, # 8 852550 with adjustable neck ydke, (a) A standard model Air Force headset. The speech signal was transmitted as follows: through 7a Figure 1. Instrumentation used to. transmit enmlified delayed speech siaal and record speech responses. 7b Figure 2. Experimental set-up indicating position of Subject and the Experimenter. the Astatic microphone to the input of the Magnetic Recorder. This was the maximum distance that the speech signal traveled during the syndhronous speech trials. During the delayed feedback, the Duo-Amplification System was activated by turning the amplifier gain up to 6. The speech signal was then delivered first through the 0.63 .sec. delaying head. through the DuanAmplification System and out via the headset. Experimental findings (21) indicate that when the speech signal at the headset is too weak it may be masked by the 3's increasing the intensity of his speaking. Conversely, if the speech signal at the headset is too intense. it is painful to 3. To avoid either extreme, pre~ tests were carried out with other Sarto determine an intensity of feedback whiCh reduced the effects of bone conduction on speech reception but did not lead to S's discomfort. roced e The procedure was identical for all 3s and consisted of three phases.. The first was an instruction and apparatus adjustment phase. The second was the~synohronous feedback speaking phase in whidh s repeated the sentence “American boys go to school everyday“ six times with a 5 second interval between repetitions (trials). Ten seconds after this phase the delayed feedback phase began during which 3 repeated the sentence about American boys for six more trials with a delay of the delayed speech feedback of 0.63 sec. A 5 second intertrial interval was used for these _trials. Specifically, the 8s were taken individually from their classroom into the school auditorium and seated in front of the table which supported the apparatus. The S faced away from the apparatus‘but was not screened from 3 who was located to the left and adjacent to 8 (refer to Figure 2). -This position of E was considered necessary for instructional pur- poses. especially with reference to the younger 88 who in certain instances temporarily avoided speaking as they en- countered delayed speech feedback for the first time. The microphone with adjustable ydke was fitted over S's head so that the yoke rested upon his shoulders. The microphone was adjusted so that it was 5 inches away from S's mouth. thus approximating the distance from one's mouth to his ear. Each 3 was instructed not to manipulate the microphone but to keep his hands clasped in front.of him. lext the earphones were fitted and were worn throughout both conditions, synchronous 10 speech feedback (SSF). and delayed speedh feedback (08?). In cases where female Ss were used. the hair had to. be sufficiently parted to insure the proper fit of the ear- phones. Following this initial seating and adjustment. S received the following set of instructions with additional comments in those cases warranting them: "We want to see hOW’Well you can speak. try not to make too many mistakes. lbw'here is what you have to say, 'American boys g-- go to sdhool-r- everyday'. This is all you have to say, just one sentence. And you will repeat this sentence every time I raise my hand like this (demonstrated). Later on. it may get a little harder to talk, be- cause the tape recorder will be-on, but try to finish each sentence: Remember to finish each sentence that is the important thing. You will say this sentence every time I raise my hand: 'American boys --- go to school --- everyday‘." For the younger 8s (generally the let grade group) extra- instructional precautions were necessary when the S: (a) anticipated the hand signal, (b) forgot the sentence. (c) was not speaking into the microphone or (d) began to laugh. Complications, when occuring during 58?. generally occured during the first few trials. If necessary, further corrections were made during the 10 sec. interval between conditions. In the event that these complications. such as prolonged laughing. had pronounced effects upon S's speaking 11 this data was discarded from the statistical analysis. upon encountering DSF for the first time if the 8s hesitated, they were prompted by E. The Ss had been told in the standard instructions. that their speaking may‘become more difficult later but were not told when. There were 4 possible cues for 53' anticipating the onset of DSF: .(a) length of time between experimental conditions. (b) E's movements to turn on the Amplification System, (c) the sudden burst of energy at the headset after the amplifier was activated. (d) the feedback of respiratory sounds which were themselves transmitted dysynthronously just prior to pro- duction of speech at the first trial of DSF. After completion of these trials 3 was asked several brief questions concerning his experiences during D8? and then was instructed not to reveal information of the experiment to fellow students. Following the dismissal of the attending 8, the amplifier gain was returned to the 0 reference point and the next 3 was summoned. The entire procedure for a given S todk approximately 20 minutes. Mgasurement' Three measurements typically employed in DSF research were used in the analysis of the speaking performance of the 12 subjects in each age group, sound pressure level (loudness), syllable duration and number of articulatory errors.2 All measurements on each trial were taken from the recordings in the following ways: Mean SOund Pressure - The means of three peak loudness levels were taken from the recordings of each trial for each 8 with a Bell System Research sound Level Meter, set approximately 3 inches away from the recorder speaker. The amplification gain of the recorder was set at an arbitrary reference point. which was sensitive to the ranges of intensities employed by all 38. In the event a given subject responded with a sus— tained intensity level, this single displacement was recorded and regarded equivalent to a mean of three peak displacements on the level meter. In stances where the meter indicator rapidly vacillated between a narrow interval of two calibrated units the higher reading between this oscillation was recorded. Mean Syllable Duration - This measurement comprised two _— v—v—v: “w While the time required to analyze the measures was not recorded, the following rankings give a fair estimate of the relative time taken to make the three measurements, rank 1 indicating the least amount of time taken: (1) syllable LIJ “'66 DJ 2% (064 U) LIJ {62 E? :> 5C’ O . m, 258 ES 256 18s I ~ /' AGE enoups J/ / e-us I 6.9-7.9 / one = 86-9." o—e *6 IO.7-ll.9 o—-o = I2.6-l3.6 DELAY o s 7 s 9 IO n12 TRIALS Figure 3. Mean Sound Pressure Level: lean of three sk readings of sound pressure TIE-feared for groups for the f nal synchronous trial and six delayed feedback trials. 18b AAGE GROUPS 157 . ' ...--. = 6.9'7.9 .6l 0,-.. = 8.6-9. n [X 455’ ,__. = "174L9 / \ o——o l2.6 43.6 1b. (D ()4 N OJ MEAN .SYLLABLE DURATION N A: on o: DELAY (D O 6 7 8 9 IO ll l2 TRIALS . Figure lo. Mean Syllable Duration: obtained . by dividing number of syllables producedi'utotime in seconds per trial, indicated for the final synchronous feedback trial and six delayed feedback trials. 18c m B DELAY ,I" a: 7 \ ,r P 8. ' \ ’ x, a: ' \ ,F-‘Ac Lu 6 I \ ,I 1‘ / >- '\ \ I,’ g 5 , .22; I ‘r' _I 4 I I :3 ' I 0 I] - I E 3 IAGE GROUPS I <1 I 2 s--s : 6.9 7.9 2 4 o—-o : 8.6 9.” g | H = '007 ".9 o—-o - l2.6 '35 o 6 7 8 9 IO II I2 TRIALS Figure 5. Mean Articulatogz Errors: mean" of total number of articulatory errors produced 5; groups during each of six delayed feedback trials. 19 are positioned fairly close together on trial 7, with respect to the sound pressure variable. The analysis of variance yielded a smaller Mean Square Between Groups than the Mean Square Within Groups (refer to Table III. Appen- dix 3) therefore the null hypothesis was not rejected. It was thereby concluded that the level of sound pressure level under conditions of’DSP does not manifest an age difference with respect to speaking performance. The grsph shown in rigure 4 with reference to trial 7 indicates that the groups are even more closely clustered graphically in terns of syllable duration than for sound pressure with the excsption of Group 3 (age range, 10.7- ll.9). The analysis of variance with respect to this measure indicated an.§ ratio of 1.47 (refer to Table IV; Appendix). which was not significant. However. since the groups dif- fered during the SS? condition with respect to syllable dura- tion, it was considered necessary to control for these initial differences in performance level by statistical methods. There— fore an analysis of covariance was applied to the syllable duration measure of the last 88? trial (trial 6) and the initial trial of DSP (trial 7). This analysis yielded an'g ratio of 1.91. (refer to Table v; Appendix) which was also not significant. Hence. even when taking into account certain 20 age differences in performance during normal speaking conditions, syllable duration does not discriminate be- tween groups first encountering DSF. The final analysis for this section (referring to trial 7) was applied to the measure of total articulatory errors. Figure 5 indicates that the groups are positioned closely together with respect to this variable, with the exception that Group 2 (age range, 8.6 - 9.11) is positioned higher than the other three groups. The analysis of variance (refer to Table VI. Appendix) resulted in an‘g ratio of 1.02 which was not significant. It was there- fore concluded that the varying age groups do not signifi- cantly display a differential reaction in terms of total articulatory errors when presented with DSF for the first time. In summary, it was concluded that none of the measures employed in this study discriminated significantly between groups of varying ages during an initial encounter with DSP. ghgpggs ip ggrfoggance aggggpapyigg subseggent Encounters with Delay In contrast to performance during 58?. speech per- formance during DSF led to certain changes for all groups on particular variables. Refer to Figure 3 where it is shown that as trials progress during DSF, the groups reSponded more loudly. An analysis of variance for sound pressure was applied between the initial (trial 7) and the final (trial 12) DSF trials, resulting in a smaller Mean Square Between Groups than die Mean Square Within Groups (refer to Table VII, Appendix). The null hypothesis was not rejected, which led to the conclusion that this measure of sound pressure does not significantly discriminate between the age groups, with reapect to the amount of ctange occurring during DSF. Refer to Figure 4. where it is indicated that as the trials progressed during DSF, increases in syllable duration were observed with a single exception. The group representing the eldest Ss (GrOUp 4, ages: 12.6 » 13.6) showed decreases in syllable duration following the initial encounter'with DSF. The level of syllable duration to which Group 4 regressed remained constant throughout most of the DSF trials, except on the last trial of this series (trial 12), where a slight increase in syllable duration was shown, This increase in syllable duration during trial 12 was the first increase _in this variable for Group 4 since the initial delayed trial. The analysis of variance testing for differences in syllable duration 22 between trials 7 and 12, yielded an 3 ratio of 4.72 (refer to Table VIII. Appendix) which was significant at the 1% level of confidence. This led to the conclusion that with respect to syllable duration, significant increments of change occured during DSF trials and these changes were significantly different between the age groups. Therefore, a series of‘t tests were made between each pair of groups on the different amount of change that occurred between trials 7 and 12 in terms of syllable duration. The 5; ratios for each group compared with each other group are entered in Table IX of the Appendix. The groups which differed significantly at beyond the 5% level were: (a) Group 1 (Mean: 42.4) with Group 3 (Mean: 31.6); (b) Group 1 with Group 4 (Mean: 20.1); (CJGroupa (Mean: 36.4) with Group 4, and (d) Group 3 with Group 4. The conclusions which were made from these comparisons were that the eldest group. Group 4 had shown a significantly lesser amount of change in syllable duration between trials 7 and 12 than any other group. The final analysis in this section was made for the articulatory error variable. Refer to Figure 5 which indicates increases in articulatory errors during subse- quent encounters with DSF for all groups with one exception. 23 Group 4 (ages: 12,6 . {3‘6} showed overeaseS in tié‘l €"‘m1htn~= “f 53‘ H- JTtiCllaLLr? errors followinm the in USP; This direction of change following the firs' delay trial was also shown by this grouP for syllable duration “as mentioned in the previous analysis. The anaiysis cf variance on the changes in number of articulatory errcrs yielded an _I; ratio of Mt? (refer to rabie- h.” Amman), which was significant at the 1% level of confidrnte. It was thereby concluded that significant changes in artzruii- ‘tory errors occurred during DSF and t ese changew indicated group differences, To determine which groove differed in the amount of change in articulatory errors: a series of 5 tests were run on these changes which occurred between trials 7 and 12, All comparisons with their respective 2 ratios have been entered in Table XI of the Appendix” The significalt differences obtained were : (a) Crotp 1 (Man: ‘12) with Group 4 (Mean: 6.2)(bfim‘Up 2 (2“...11-11: (.1 .9) with group 4 and (c) Group 3_(Mean: 9.3) with Croup a: flhe comparisons indicate that the eldest group (Group 4), differed significantly from all other male groups in tart; of tie amount of engage in articulatory errors made bcfiueen NTS‘t}i&IB 7 and 12, The Changes for Group 4 between crisis 7 and 12 were in terms of decreases in articulatory 24 errors, while for Groups172and 3 the changes were with respect to increases on this variable. Age Group Differences Manifested During the Final Encounter with Delay It was considered quite pertinent to this study to investigate the level of performance which the groups had reached following the series of encounters with DSF. The comparison of the groups on trial 12 was considered the valid comparison for this analysis. The graph of Figure 3 indicates that on trial 2, with respect to sound pressure. Groups 2 and 4 are positioned together and higher than Groups 1 and 3 who are also located together. The analysis of variance yielded a smaller Mean Square Between Groups than the Mean Square Within Groups,(refer to Table XII. Appendix) therefore the null hypothesis was not rejected. It was thereby concluded that the level of sound pressure attained after a series of encounters with DSF, does not discriminate between the variousage groups. For syllable duration with respect to trial 12, Figure 4 indicates that all groups, with the exception of Group 4, are clustered together with respect to syllable duration. Adaptation. mentioned earlier for Group 4, was still in effect at trial 12, which segregated.them from the other groups who increased 25 syllable duration from trial to trial. The analysis of variance for this measure on trial 12 yielded anlg ratio of 2.16 (refer to Table XIII. Appendix) which was not significant. The null hypothesis was not rejected. There- fore, it was concluded that the level of syllable duration readhed after a series of encounters with DSF does not distinguish between the various age groups. For the final analysis in this section, refer to Figure 5, with reference to trial 12, where the number of articulatory errors made by each group following a series of encounters with D8? is shown. The two youngest groups, Group 1 and 2 are clustered together while the eldest groups, Group 3 and 4 are separated and away from 1 and 2. Group 4 had made the least number of errors on trial 12 than on any other previous encounter with DSF. The analysis of yvariance for articulatory errors on trial 12 yielded an .g ration of 3.40 (refer to Table XIV} Appendix) which was significant at the 5% level of confidence. It was thereby concluded that the groups significantly differed in the num- ber of articulatory errors made following a series of encounters with DSF. A series of‘g tests were then run to determine which of the groups paired for this analysis differed. Table xv in the Appendixcontains the g ratios 26 of each group compared with each other group. The significant differences Obtained were: (a) Group 1 (Mean: 7.6) with Group 4 (Mean: 2.8) and (b) Group 2 (Mean: 7.0) with Group 4. From these comparisons the conclusion drawn was that with reference to the number of articulatory errors made on the last trial of DSF. the only significant differences occurred between the eldest group, Group 4 and the two youngest groups. Groups 1 and 2, with Group 4 making the least number of articulatory errors. Relationship Between the Measures Since three measures of responses were employed in this study, the relationship between these measures was examined. Because the groups had not displayed a significant differential reaction on the initial encounter with DSF (trial 7) with respect to any of the measures, all 40 33 were combined for the analysis in this section. The Pearson Product - Moment Correlation Coefficient was applied to the data of trial 7 between sound pressure and syllable duration and between each of these measures and the number of articulatory errors. The correlation coefficient between sound pressure and syllable duration yielded an r of .30 which was not significant. The 27 correlation coefficients for sound pressure versus articulatory errors and syllable duration versus articula- tory errors yielded r's of .38 which were significant at the 5% level of confidence. Therefore, it was concluded that on an initial encounter with DSF, significant and positive relationships were manifested between each measure, sound pressure and syllable duration versus the number of articulatory errors. Performance of Female Speakers With Delay Although the female group had been omitted from the major statistical analysis in this study, certain other analyses and comparisons were made. Upon the first con- tact with DSF, following synchronous speech the female group had shown, increased sound pressure, used longer syllable duration and increased in number of articulatory errors. As mentioned previously, the male groups had also responded this way when encountering DSF for the first time. Other comparisons made between the female group and the other four male groups indicated that the female group: (a) along with Group 1 made the least amount of errors during the initial trial of DSF, trial 7, (b) used approximately the same syllable duration during DSF as 28 Groups 1, 2, and 3, (c) employed the highest levels of sound pressure during D3? and (d) made the greatest amount of increases in number of articulatory errors between trials 7 and 12. Because its members were of similar ages, the female group (ages: 8.2 - 9.1) and Group 2 (ages: 8.6 - 9.11) were run in an analysis of variance for the three variables used in this study on trial 7, the first delay trial. The g ratios have been entered in Tables XVI, XVII, XVIII. Appendix. These analyses, none being significant, indicated that the female group and Group 2, finales) did not differ in terms of sound pressure, syllable duration and articulatory errors during the initial trial of DSP. It was concluded that within the restrictions of this study no significant sex differences were indicated during the initial encounter with delay. DISCUSSION 29 Performance During Synchronous Speech Feedback. The analysis of the synchronous speech feedback data at the first trial of this condition indicated that the varying age groups had not differed significantly in terms of sound pressure level of articulatory errors. The groups had differed significantly, however, in terms of syllable duration with the eldest group speaking most rapidly, the next to youngest group speaking most slowly, and the youngest and next to oldest group speaking at a rate between the two extremes. Perhaps due to greater training in linguistic Skills such as speaking and reading, the finding that the eldest group had spoken at the most rapid rate is not unexpected but the ordering of the other three groups does not yield to any simple explanation in regards to existing information. Performance During Delayed Speech Feedback The 83' first encounter with DSF, trial 7, led 85% of them to increase sound pressure and syllable duration and led 100% of them to increase the number of articulatory errors. That is, these children showed a similar mode of reaction to DSF as adults do under conditions of DSF. Fairbanks (22) using various time delays with the speech 30 output amplified by a constant amount found that adult subjects reading a prose text increased sound pressure, employed a higher fundamental frequency, and increased duration of speaking time and number of articulatory errors. Ratner and Gawronski (23) in an unpublished study, observed that all 20 male adult Ss following a series of undelayed trials, reduced speaking rate as DSF was introduced for the first time. Tiffany and Hanley (24) employing DSF as a test for detecting auditory malingering, found that in spite of the fact that the Ss were in- structed as to the properties of DSF and following a five minute period of familiarization with a 100 word passage, responded with significant decreases in reading rate when subjected to DSF for the first time. Statistical analyses of the reactions to DSF as a function of the ages of the children yielded non-significant differences on the initial delayed trial for all measures employed. Even the analysis of covariance, adjusting for the group differences displayed for syllable duration during the last synchronous trial, yielded a non-significant difference. That is, the ages of the children and thus their varying levels of practice with the language, did not lead to significant levels of reaction when the age 31 groups encountered DSF for the first time. Disregarding statistical significance, the means of the groups on each measure do not order themselves in terms of the ages of the groups employed. ‘ghgngesiig Performance During Subsequent Encounters with Belay- Statistical analyses of the changes in performance from the first to the last DSP trials indicated that the groups differed significantly on changes in articulatory errors and syllable duration, but not in sound pressure. To a large extent these significant differences arose from the behavior of group 4, the eldest group, which tended to make less errors while the other groups became poorer (increased in number of errors) and which tended to maintain a constant Level of syllable duration while the other groups increased syllable duration. All groups tended to increase sound pressure between trials 7 and 12 which seems to be a common reaction to any acoustical interference in the ongoing speedh process, eg. noise. Of significance in these findings is that only the eldest group, Group 4 (ages: 12.6 -l3.6) showed any consistency in improvement between trial 7 and 12 such as decreases in the number of articulatory errors. 32 And only this group displayed any constancy in response, such as maintaining the same level of syllable duration throughout most DSF trials. Although the eldest group displayed these trends, in no instance had they regressed to their level of speech performance observed during synchronous speech. This type of speech behavior dis- played by Group 4 under conditions of DSF is in some ways characteristic of the performance of adults employed under DSF, who, under these conditions, had shown considerable improvement in speech performance. Melrose (25) using 20 Se who read a series of 40 matched sentences, found that the number of correct words per sentence for each individual reading increased in the general pattern of a sigmoid learning curve, while duration increased steadily during the first half of the readings and then decreased. Tiffany and Hanley (26). using 20 53 who were subjected to 80 db. of delayed side-tone, found that during 24 readings of a 45 word prose passage 83, in general, tended to avoid omission of words, syllables and sounds, but did not learn to overcome the rate effect. Miller (27) found while reading a 100 word passage over a 200 massed trial period, he reduced the number of articulatory errors to almost zero, however, no significant decreases in duration 33 time of trials were observed. The literature on DSF research with adults had not reported any studies which found any consistent decreases in sound pressure level during DSF, which was also indicated for children in the findings of the present study. Since the speech behavior of adults under con- ditions of DSP reported in the literature, resembled in some ways the behavior of the eldest group in this study' and because no improvement was found for the younger groups used here, it is suggested that perhaps the eldest group (Group 4, ages: 12.6 - 13.6) represents a breaking point between earlier and later language deveIOpment, i.e.. especially in terms of those processes which mediate stable speech behavior as the speaking situation becomes far from ideal. It should also be noted that Group 4 is approaching the age period of puberty during which certain biological and behavioral changes are known to occur, such as increases in strength, deepening of the voice, acceler- ation in growth. increases in sexual sensitivity and increases in number of social contacts. Important in this respect is the report by Hockett (28) who finds that fundamental speech habits are firmly established by the age of puberty and the most important force shaping dialect 34 is the speech of others. As the three younger age groups used in this study showed no significant improvement as their con- tacts with DSF increased, it would seem that these groups are in some way dominated by the effects of the delayed feedback. This would seem to indicate, that as language development is a gradual process, and since the three younger groups performed significantly poorer than Group 4, under DSF, that these younger groups are still in the process of acquiring the necessary skills that call for a considerable amount of control over the speech monitoring system. With regard to the speech performance of the age groups on trial 12, the only significant group differences found were on the measure of articulatory errors. The t tests had indicated that to a large extent this signifi- cant difference found for errors could be attributed to the behavior of Group 4, which although making the first increase in articulatory errors since trial 7, made a significantly smaller number of errors than either Groups 1 and 2 on trial 12. Despite the finding that Groups 3 and 4 did not differ significantly in errors. the groups were systematically ordered in terms of age’ 35 in articulatory errors with Group 1 making the most errors, Group 2 making the next highest number and so forth. It is of significance to note that at trial 12 all groups had some degree of familiarity with DSP, and the ordering of the age groups in terms of performance at trial 12, may be some conservative estimate of the age group level of training in terms of linguistic skills. This may be even more so when articulatory errors are considered more directly related to language than either syllable duration and sound pressure. In summary, it was found that all groups reacted to DSF, in some ways the same, for instance, all groups showed increases in sound pressure, syllable duration and number of articulatory errors when encountering DSF for the first time. And in some certain ways they had differed significantly, that is, significant group differences were found for the amount of change in articulatory errors and syllable duration occurring between trials 7 and 12, and the number of errors made at trial 12. As‘pointed out earlier, these significant differences were attributed to the findings that Group 4 remained the same in syllable duration during D8! while the other groups increased syllable duration and that Group 4 decreased in articulatory errors and other groups increased. 36 a o s etwee Measures. The significant correlations obtained in this study were between sound pressure and articulatory errors and syllable duration and articulatory errors. Since these correlations were positive, they suggest that as either increases in sound pressure or syllable duration occur, concomitant increases in articulatory errors are made when the groups encounter DSF for the first time. That is, as the children are responding in one way to DSF, such as increased syllable duration, they are also responding in another way, such as increased number of articulatory errors. Although the relationship between sound pressure and syllable duration was not significant, the obtained 5 of .30 suggested a positive relationship between these two. variables. This is in agreement with the findings of Peters (29) who reported the tendency of adult readers to decrease their reading rate as sound pressure was in- creased during DSF. Reading rate can be considered an estimate of syllable duration. Performance of Female Speakegg With Delay. Upon the initial contact with DSF, trial 7, 70% of 10 female speakers increased sound pressure, 90%.increased 37 syllable duration and 100%.of them increased in terms of articulatory errors. As compared with all of the other male groups, the female group employed the highest levels of sound pressure level throughout the DSF con- dition and displayed the greatest increase in number of errors when progressing from trial 7 to 12, although these were not tested for significance. No significant differences were found, when the female group (ages: 8.2 - 9.1) were compared with Group 2, males, (ages: 8.6 - 9.11), for the three measures herein employed during the initial encounter with delay. -From the extent of these findings, no significant sex differences were shown in this study. 38 SUMMARY The present study using 40 male and 10 female Ss, ranging in age between 6 years, 9 months, and 13 years, 6 months, was designed to investigate the performance of children of varying ages under conditions of delayed speech feedback (DSF). Research with adults using DSF had in- dicated profound disturbances in their ongoing speech behavior, generally reported in terms of stutteringblike responses, retardation in speaking rate and increases in vocal intensity. The apparatus employed in this study were: a tape recorder modified to deliver a delayed signal of 0.63 seconds, an amplification system and the equipment necessary for the transmission and reception of the delayed signal, i.e. a microphone and headset. Each age group consisting of 10 83, comprising 4 male and 1 female group(s). experienced 6 trials of synchronous speech feedback (SSF). followed by 6 trials of DSF during which they recited a simple sentence composed of six words. The initial contact with DSF eventuated in decremental changes in the speech performance of all groups in terms of increases in sound pressure, syllable duration and number of 39 articulatory errors. This finding suggested that speech behavior is not a static process but requires constant changes in the monitoring system as conditions under which a person is speaking change. During further encounters with DSF all groups performed progressively poorer, with the ex- ception of the eldest group (ages: 12.6 - 13.6). For the most part the eldest group displayed no changes in syllable duration during DSF and improved on articulatory errors. This performance displayed by only that group is characteristic of adult performance under D3? which was reported in the liter- ature. It was suggested that the eldest group employed in the present study may represent a dichotomy between early and later stages in language development,,in terms of those factors which mediate stable speech performance when the speech situa- tion becomes far from ideal. During the initial delayed trial, results indicate sig- nificant positive correlations between sound pressure level with articulatory errors and between svllable duration with articulatory errors. Although no significant correlations were found between sound pressure and syllable duration, results show a positive trend. The female group employed in this study reacted in much 40 the same way as the other male groups during the first encounter with DSF. That is, at the first trial of delay the female group increased sound pressure, syllable duration and number of articulatory errors. When compared with a male group whose Ss were similar in age to the female group, no significant differences were found for any of the variables during the initial DSF trial. Comparisons made during sub- sequent DSF trials indicated that the female grOUp employed the highest sound pressure than any of the male groups. The wide individual differences in reacting to DSF displayed by the children used in this study are consistently reported for adults in the DSF literature. That is, some 83 reSpond to DSF in a manner somewhat resembling their normal speech behavior while other Ss seem.for a period of time to be completely captivated by the effects of this condition. It is suggested that further research with DSF might employ 88 similar in age to the eldest grOUp used in this study,investigating the relationship between behavior under DSF and those linguistic skills requiring considerable con- centration, such as fluent silent reading and comprehension and those eXperiences requiring a great deal of Speech con- trol, such as formal classroom speaking. 1. 3. 8. 9. 41 REFERENCES Davidson, G. D. The effect of altered external side- tone transmission time upon oral reading rate, precision of articulation and pitch variability. Abstr. Ph.D. Dissert., Ohio State U., 1955. Rawnsley, A. I. and Harris J. D. A comparative analysis of normal speech and speech delayed side-tone by means of sound spectrograms. Med. Res. Lab. Rep. #248, Bur. of Med. and Surg., Nav. Dept. Proj. NM 003 041.56.03, 1954. Atkinson, C. J. Research study of the psycho-acoustic effects of human and artificial side-tone. First Quarterly Report, Contract No. DA 36-039 sc-45262, Signal Corps Supply Agency and State Univ. of Iowa, l952. Lee, B. S. Effects of delayed Speech feedback. g. Acoust. Soc. Amer., 22,1950, 824-826. Doehring, D. G. Changes in psychophysiological reSponses produced by delayed speech feedback. USN Sch. of Aviat. Med. Res. Rep., NM 001 102 502, Rep. #1, 1956. Lee, B. 8. Some effects of side-tone delay. g. Acoust. Soc. Amer., 22,1950, 639-640. Black, J. W. The effect of delayed side-tone upon vocal rate and intensity. g. Speech and Hear. Disord. 16, 1951, 56-60. “ Young, N. Certain effects on vocal pitch of frequency modulated auditory feedback. Abstr. Dissert., Purdue Univ. . Spilka, B. Some vocal effects of different reading pas- sages and time delays in speech feedback. g, §peech and Hear. Disogg. 19, 1954, 37~47. . 10. Neely,. K. K. The effect of oral practice of different 11. conditions of side-tone upon the rate and the sound pressure of the Speech of a grOUp of'stutterers. Abstr. Ph.D. Dissert., Ohio State U., 1952. Leith, w. R. and Pronko, N. H. Speech under stress: a study of its disintegration. §peech Monographs, 24, 1957, 285-291. 12. 15. 14. “15. 16. 17. 18. 19. 20. 21. 22. 239 42 Kline. An experimental study of the nature of hypnotic effects of delayed speech feedback. J, Clinical and EEEEE- Hypnosis. 2, 1954, 145-156. Gibbons, E. H. and Winchester, R. A. A delayed side- tone test for detecting uniaural functional deaf- ness. Argh. Otolaryng., Chicago,66, 1957, 70-78. Spilka, B. Relationships between certain aspects of personality and some vocal effects of delayed speech feedback. g, Speech_and Hear. Disord. 19, 1954, 491-503. fl“ Beaumont, J. T. and Poss, B. M. Individual differences in reacting to delayed auditory feedback. British 5, 2; P32. 48, 1957, 85:89. ' m“"‘ Doehring, D. G. and Harold C. J. The relation between speech disturbance and psydnophysiological Changes resulting from.delayed speech feedback. Nav.'Sch. Aviat. Med., Res. Proj. NM 130299 Sub task 1. Black, J. w. The persistence of the effects of delayed side-tone. g. Speech and Heag. Disord. 20, 1955, 65-68. ' Atkinson, C. J. Adaptation to delayed side-bone. J. Speech and gear. Disord. 18, 1953, 386-391. Chase, R. A. Effect of delayed auditory feedback on the repetition of speech sounds.'g, Speech and Hear. Disord. 25,#5, 1958, 583-590. Osgood, C. E. Method and Theory in Experimental Psychology. New York: Oxford Univ. Press,1953,p.685. Lee, B. S. Artificial stutter. g, Speech and Hear Disord. 16, 1951, 53-55. “' Fairbanks, G. Selective vocal effects of delayed auditory feedback. g. §peech and Hear. Disord. 20, 1955, 333- 345. Ratner, S. C. and Gawronski, J.J. The effect of oral task 'familiarity upon performance under delayed speech feedback. (unpublished study, Michigan State U.,1957). Tiffany W. R. and Ranley C. N. Delayed speech feedback as‘ ggtggt for auditory malingeriflgo Science, 115, 1952, "' c 25. 26. 27. 28. 29. Melrose,J. The temporal course of changes in the amount of vocal disturbance produced by delayed auditory feedback. Abstr. Ph.D. Dissert., Illi- nois Us 9 1954. Tiffany,w. R. and Hanley, C. N. Adaptation to delayed side-tone. g, Speech and Hear. Disord. 21, 1956, 1(34'172. Miller,K (unpublished research investigation, Dept. of Ps1" Michigan Staten ., 1957). Hockett., Age, grading and linguistic change Langkm 25, 449-457. ' Peters, R. W. The effects of change in side- tone delay and level upon rate of Oral reading of normal speak- era. 1. Speech and Hear. Disord. 19, 1954, 483-490. APPENDIX A 45 Classification of Sublgcts by_Age, GradeL SchoolLiggg and Group 1 £95 1 6.9 2 6.9 3 6.10 4 7.0 5 7.0 6 7.2 7 7.3 8 7.4 9 7.8 10 7.9 Group 2 11 8.6 12 8.6 13 8.7 '14 9.1 15 9.3 16 9.5 17 9.8 18 9.8 19 9.9 20 9.11 rou 3 21 10.7 22 10.9 23 11.1 24 11.3 25 . 11.2 26 11.4 27 11.8 28 11.8 29 11.9 30 11.9 GRADE P’NDHFHPJP‘P'Ntdrd whwwuwwwww (flmO‘mUIU'Ul-b-DO‘ Number of Siblings. SCHOOL Resurrection Resurrection St. Mary's Resurrection Resurrection Resurrection Resurrection Resurrection Acquinas Holy Cross Resurrection Resurrection Resurrection St. Casimir Resurrection Resurrection Resurrection Resurrection Resurrection Resurrection St. Mary's Resurrection Resurrection St. Mary’s Resurrection Resurrection St. Mary's St. Mary's St. Casimir Resurrection 1.9 87 121 131 101 110 110 l'" (D H *HOI-I'I'l'fifil-J 0‘ 124 110 95 98 100 90 103 105 77 SIBLINGS p.H P'UIG lloral NHHINV-‘IHN-fi P‘hih-fildrdtnhil h 46 Classification of Subjects by Age, Grade, School, 19, and Group 4 ‘Agg GRADE 31 12.6 6 32 12.10 8 33 12.10 7 34 12.11 6 35 13.1 7 36 13.4 7 37 13.4 8 38 13.5 7 39 13.6 7 40 13.6 7 Female Group 1 8.2 2 2 8.3 2 3 8.4 2 4 8.7 3 5 8.8 3 6 8.8 3 7 8.8 3 8 8.9 3 9 9.1 3 10 9.1 3 i No IQ's Obtained number of Siblings SCHOOL Resurrection St. Therese St. Casimir Acquinas Resurrection St. Casimir St. Therese Holy Cross St. Casimir Resurrection Resurrection Resurrection Resurrection Resurrection Resurrection St. Casimir Resurrection Resurrection Acquinas Resurrection £9 79 116 108 95 79 98 111 104 91 tfil’fifiifittfi SIBLINGS NUTUIU'lF-‘INIH Hrar-Rohowtntohan APPENDIX B TRBLE 1 P. Ratio: Sound Pressure Level. 88?. Trial 6 Sun of Mean Source of Variation Squares _ df Square P Between Groups 62 .9 3 20. 97 . . . Within Groups 1088.6 36 30.24 Total 1151.5 39 TABLE II §111able Duration: 881’. Trial 6 rm 0 ”5635 Source of Variation Squares df Square P Between Groups 686 3 228.7"? 5.53 Within Groups 1490 36 41.38 Total 2176 39 * significant beyond the .05 level of confidence ** Significant beyond the .01 level of. confidence Note: These symbols will hereafter indicate confidence levels. ‘ 49 TRBLB III P. Ratio: pgggpg Presggge Leggl. DSF. Trial 7 Sun of Mean source of variation Squares df ’Square P Between Groups 97.5 3 32-50 ' ... Within Groups 1264.9 36 35.14 Total ‘ 1362.4 39 TABLE IV Syllable‘guration: DSF. Trial 7 Sum of Mean P Source of Variation Squares df Square Between Groups 515.7 3 171.9 1.47 Within Groups 4182.3 6 116.2 Total 4698.0 39 TABLE V 50 P Ratio (Covariance) Szllable Duration. SS? Trial 6 x DSF Trial 7. Source Sum of of Squares variation of Errors Mean of Estimate df Square P Total 4301.0 38 1.91 Within Groups 3695.8 35 105.59 Adjusted Means 605.2 3 201.73 TRBLB VI Articulatogx Errors: DSF Trial 7 Sum of Mean Source of variation Squares df Square P Between Groups 42.5 3 14.17 1.02 Within Groups 498.9 36 13.86 Total 541.4 39 TABLE VII F Ratio: Sound Pressure_Level. Between DSF Trials 7 and 12. so; of Mean Source of variation Squares df Square F Between Groups ,37.40 3 12.46 ... Within Groups 971.00 36 26.97 Total 1008.40 39 TABLE VIII Syllable Duration: Between DSF Trials 7 and 12. Sum of. Mean Source of variation Squares df Square P Between Groups 2676.90 3 892.3 4.72** Within Groups 6812.10 36 189.2 Total 9489.00 39 52 TABLE IX T Ratio: lelable Duration, Between DSF Trials 7 and 12 . ‘5 ‘_‘ _JE; Groups 2 3 4 1 .980 1.76* 3. 63" 2 .782 2.65Mr 3 1.87* Means: Group. 1 (42.4); 2 (36.4): 3 (31.6): 4 (20.1). TABLE X. 1’ Ratio: W m semen DSF Trials-7 and 12. I; i A . Sun of Mean Source of Variation Squares df Square P semen Groups 172.50 3 57.50 4.49 98* Within Groups ' 460.60 36 12.79 Total 633 . 10 39 53 TABLE XI T Ratio: Articulatory Errors, Between DSF Trials 7 and 12. Groups 3 4 1 1.31 1.69 3.62** 2 .375 2.31* 3 1.94* 4‘. Means: Group, 1 (12.0); 2 (9.9); 3(9.3): 4 (6.2) TABLE XII F Ratio: Sound Pressure Level, DSF Trial 12. A _ Sum of Mean Source of variation squares -df Square r Between Groups 50.68 3 16.89 ... Within Groups 2344.10 36 65.11 Total 2394.78 39 54 mm: XIII lelabg Duration: 03?. Trial 12. Sum of . Mean Source of Variation Squares df Square P Between Groups 2442.20 3 814.06 2.16 Within Groups 113557.40 36 376.59 Total 115999.60 39 TABLE XIV Eticglatog Errors: DSF. Trial 12 '7" r fl Stunof Mean source of Variation Squares df Square P Between Groups 138.7 3 46.2 3.40* Within Groups 488.1 36 13.6 TOtal 626. 8 39 55 TABLE XV T Ratio: Articulatory_Errors, DSF Trial 12 Groups .2 3 4 1 .363 1.39 2.91** 2 1.03 2.54** 3 1.52 means: Group, 1 (7.6); 2 (7.0): 3 (5.3): 4 (2.8) TfiBLE XVI ..Female Group 95 Group.2 (M3188) Sound_gres§ure Level:DSF Trial 7. Sum of , Mean Source of variation Squares df Square F Between Groups - l 1 .1 _ ... Within Groups - 748 )8 41.5. Total 749 19 56 TABIE XVII Female Groupgvs Group 2 (Ma1es) Syllable Duration; DSF Trial 7 ~ A Sun of Mean. Source of variation Squares df Square F Between Groups 42.5 1 42.5 ... Within Groups ’ 1340.5 18 74.5 Total 1383.0 19 TBBLE XVIII Female Group vs Group 2 (Males) Articulatory.§;rors: DSF Trial 7. ' Sum of Mean Source of Variation Squares -d£ Square P Between Groups 26 1 26 1.67 Within Groups 281 18 15.6 Total 307 19 «('13: Mo“ 5; 513.55 mg I ‘7! MICHIGAN STATE UNIVERSITY I IN l ”H Ill IIIIIIIIIITIIES O306J2380 3 1293