g... -mm..a_.-.—.~_~, . . . H ION, MANUAL F TOTAL. COM'MUNICAT 0 THE EFFECTS COMMUNICATION, ORAL COMMUNICATION AND READIN 7 RNING 0F FACTUAL INFORMATION, ' ON THE L19 G IN RESIDENTIAL SCHOOL DEAF CHILDREN Thesi for the Degree of Ph. D.. MICHIGAN STATE UNIVERSITY ALFRED HENRY WH ITE, JR. 1972 .:. , .. . ...,..4.o¢ . 4 . 3:... yfi? 3. .Iatfi . GI... I ... IIIIII IIIIZIIIII II “431w. mu... 3 29310 5143 tummy ' " in» This is to certify that the 'ps‘ ,' .," thesis entitled The Effects of Total Communication, Manual Communication, Oral Conmmnication and Reading on the Learning of Factual Information in Residential School Deaf Children presented by Alfred Henry White, Jr. has been accepted towards fulfillment of the requirements for Ph 0 D o deg-Cc in Educat ion 4 79%. 3"“ ' . Major professor ,‘34" ‘ ‘ Date La; ‘17,: 1972 0-7639 ABSTRACT THE EFFECTS OF TOTAL COMMUNICATION, MANUAL COMMUNICATION, ORAL COMMUNICATION AND READING ON THE LEARNING OF FACTUAL INFORMATION IN RESIDENTIAL SCHOOL DEAF STUDENTS by Alfred H. White. Jr. The polemic between proponents of oral and combined methods of communication for deaf children has been an active controversy for many years and it is not likely that it will be resolved very soon. The issue cannot be overlooked because it holds a central place in the philosophy of deaf education. Rather than trying to ignore the controversy, or trying to make ubiquitous claims regarding what is best for all deaf children, this study attempted to focus upon two pepula- tions of hearing impaired children at two different residential schools, and discover the method of communi- cation under which those students assimilate more factual information. A stratified random sample of 45 Se was drawn from the Maryland School for the Deaf. Ss ranged in age from 11.0 to 18.7 years and in IQ from 60 to 140. These Se were presented factual information through four methods of communications (1) oral communication, (2) total communication, (3) manual communication, and (4) reading. The independent variables in the study were; method of Alfred H. White, Jr. communication, age, and intelligence. The dependent variable was the amount of information assimilated. It was hypothesized that hearing impaired children would assimilate more factual information when it was presented to them through total communication, manual communication and reading than they would through oral communication. It was hypothesized that there would be an interaction between method of communication and intelligence; the lower IQ Ss being able to assimilate more through manual communication (speech deleted) and the average and bright 83 being able to assimilate more through total communication. It was also hypothesized that hearing impaired children would assimilate more information through reading than they would through oral, total, or manual communication. An experimental design was used to eliminate criticism directed against the use of ex post facto designs which employ matching techniques. A 3 x 3 x 4 factorial repeat- ed measures design, fixed effects model was used. 83 were presented four passages of factual information through each of the four methods of communication: each 8 was compared to himself across the four methods of communica- tion. The use of a repeated measures design eliminated the need to match or randomly assign heterogeneous $3 to Alfred H. White, Jr. treatment groups. Four different passages were used with each of the four methods of communication; two of the four passages were at the 2nd grade level of diffi- culty, and two were at the 4th grade level. EQuivalent passages were randomly assigned for use with the four methods of communication. The same certified interpreter for the deaf was employed to present the material to all Ss under all methods of communication. Information was presented over a three day period, at three different periods each day. ”Time of day", "day", and “order of presentation” were eliminated as confounding variables through system- atic scheduling. The results of the analysis of the data suggest the following conclusions: (1) hearing impaired children assimilate more factual information through reading than they do through oral or total communication; (2) hearing impaired children assimilate more factual information through total and manual communication than they do through oral communication; (3) all categorical sub-groups of hearing impaired children assimilate more information through total communication and manual communication than they do through oral communication; (A) the speech com- ponent in total communication does not increase the amount of information assimilated over that assimilated Alfred H. White, Jr. through pure-manual communication; (5) bright, average and low functioning hearing impaired children do not differ in their ability to assimilate information through oral communication; however, average and bright children do significantly better than low functioning children through total communication, manual communication and reading. A replication of the Maryland study was conducted at the Michigan School for the Deaf. In general, the results of the replication supported the findings of the Maryland study. THE EFFECTS OF TOTAL COMMUNICATION, MANUAL COMMUNICATION, ORAL COMMUNICATION AND READING ON THE LEARNING OF FACTUAL INFORMATION IN RESIDENTIAL SCHOOL DEAF CHILDREN By Alfred Henry White, Jr. A THESIS Submitted to Michigan State University . in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Special Education 1972 1‘0‘)‘ ,9 (9 5A ACKNOWLEDGMENTS The author would like to take this opportunity to express sincere appreciation to several people for their meaningful and timely contributions to this endeavor. To Dr. Edwin J. Keller, chairman, appreciation must be expressed for the many hours he has spent in guidance throughout the authorfs doctoral program; as a person, sensitive to the needs of others, and as a scholar of unique acumen, Dr. Keller is most worthy of commendation. The author's gratitude is also extended to Mrs. Vivian Stevenson, Dr. Mary Ellen McSweeney, and Dr. Julia Falk for their sensitivity, meaningful suggestions and general interest during his association with them, and particular- ly during the completion of this work. Deep appreciation is also eXpressed to the adminis- trators and staff at the maryland and Michigan Schools for the Deaf. Without their cooperation this study would have been impossible. Finally, my parents, children and dear wife, Susanne, must be thanked for their continual support. Each has been a unique source of strength in encouraging the completion of graduate studies and this dissertation; without their unselfishness and sacrifices, efforts to complete this work would have been futile. ii TABLE OF CONTENTS LIST OF TABLES O O O O O O O O O O O O O 0 LIST OF FIGURES O O O O O O O O O O O O 0 CHAPTER I. II. III. THE PROBLEM . . . . . . . . . . . Purpose of this Study . . . . . Definition of Terms . . . . . . Relevant Literature . . . . . . Research Supporting Ora Communicat on Research Supporting Combined Methods Of communication 0 o o o a o Sensory Overloading . . . . . . Reading as a Method of Communication HyPOtheSeS o o o o o o o o e 0 DESIGN or THE STUDY . . . . . . . Sample a o o o o o c o a o o 0 History of Sample and School . Design and EXperimental Methods Definition of Methods . . . . . Scheduling of $8 for Testing . Materials Used 0 e o o o o o 0 Procedures for Testing . . . . RESULTS OF THE ANALYSIS . . . . Main Analysis . . . . Exact Hypostheses . . Ancillary Findings . Summary . . . . . . . iii Page viii CHAPTER IV. REPLICATION OF THE MARYLAND STUDY Procedures . . Sample...... Results . . . . Main Analysis . . Summary of Michigan St udy V. SUMMARY AND DISCUSSION . . Discussion . . . . . maryland Study . . Michigan Study . . Major Conclusions . Implications and Limitations . Suggestions for Future Research LIST OF REFERENCES . . . . . . . . . . . APPENDIX A. Total Communication . . . . . . . Bo Manual comunication o o o o o 0 iv Page 63 64 66 66 69 78 80 8A 84 91 93 99 101 103 109 8. 9. 10. 11. LIST OF TABLES Formal Definitions of Methods of Communication . Stratification of the Population of Hearing Impaired Children . . . . Means for Age, IQ, Hearing Loss and Reading Level for the Stratified Sample . . . . Means and Standard Deviations for Each Level of Age and IQ Means and Standard Deviations for the Entire Sample on Hearing Loss, IQ, and Reading Level . Analysis of Variance Table on Hearing Loss for the Factors of Age and IQ Mean Length of Time in Which Material was Presented . Composition of Testing Groups Schedule of the "Day” and Order of Method of Presentation Schedule Indicating "Day”, Method Received . Schedule Indicating the Insertion of Reading Treatment Over the Three Day Testing Period and the Passages Presented Under Each Method of Communication . "Time of Day", Testing Groups Were Tested and the Page 33 34 3h 35 35 #0 #2 43 “3 #6 TABLE 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 2h. 25. 26. Analysis of Variance Table for Rate of Presentation of Factual Information . . Matrix of Correlation Coefficients Between Methods of Communication . . . Mean Number of Correct Responses and Standard Deviations for Methods of Communication............. Grand Mean Number of Correct Responses for Levels Of Age and IQ o a a a a a a Summary Table for 3 x 3 x A Analysis Of varianc e O O C O O O O O O O O O O O Contrasts on the Main Factor of Methods of Communication . . . . . . . Mean Scores Across Low, Middle and High IQ Levels for Methods of Communication............. Contrast Comparing Means of Low IQ $8 on Total and Manual Communication Contrasts on the Factor of Age . . . . . . Contrasts on the Factor of IQ . . . . . . Means for: Age; IQ: Hearing Loss; and Reading Level for the Sample . . . . . Means and Standard Deviations for Each Level of Age and.IQ . . . . . . . . . . Means and Standard Deviations for the Entire Sample on Hearing Loss, IQ. andReadingLeveleo ooeaooooo Mean Length of Time in Which Passages Were Presented for Methods of Communication . . . . . . . . . . . . . Mean Number of Correct Responses and Standard Deviations for Method of Communication . . . . . . . . . . . vi Page 49 52 53 53 5h 57 58 59 6O 61 67 67 68 68 69 TABLE 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. Grand Mean Number of Correct Responses for Levels on Age and IQ Summary Table for 3 x 3 x h Analysis of Variance o o o o o o One-Way Analysis of Variance Table Comparing IQ Groups under Oral Communication . . . . Contrasts on the Factor of IQ Under Oral Communication . Contrasts on the main Factor of Methods of Communication . . Contrasts on the Factor of Age . Contrasts on the Factor of IQ Means for Passages Presented Orally and for Questions on Which Subjects Guessed . . . . . . . Analysis of Variance Table Comparing Means on Passages 17-B, 21-B and 22-3 by Planned Comparisons Analysis of Variance Table Comparing Means on Passages 17-D, 19-D and 24-D by Planned Comparisons vii Page 69 7O 71 71 74 7# 75 76 77 77 LIST OF FIGURES FIGURE Page 1. Data Matrix of the 3 x 3 x 4 Repeated Measures DeSign a o a a o a a o o a a 38 2. Mean Number of Correct Responses for Low, Middle and High IQ Groups in the Maryland StUdy a a a o a a o o o o o o 56 3. Mean Number of Correct Responses for Low, Middle and High IQ Groups in the MiChigan StUdy a o a a a o o o o a 0 viii CHAPTER I THE PROBLEM Historically, one of the crucial problems in the edu- cation of deaf children has been the inability of profes- sionals to resolve the oral-manual controversy. Until the last few years proponents of both oral and manual methods of communication have been inclined to defend the efficacy of their preferred methods rhetorically rather than empir— ically. Advocates of the oral method still argue that exposing deaf children to any form of manual communication will reduce their ability to speechread and speak, con- sequently resulting in greater estrangement from the main- stream of society. Proponents of manual methods continue to denounce these assertions arguing that there is no evidence to support such propositions and that the use of signs and fingerspelling is necessary if deaf children are to achieve their maximum potential and live full, rich lives. During the past decade proponents of various manual methods have begun to generate a body of research which supports their claim that general educational achievement is enhanced through the combined use of speech, finger- spelling and signs. The result has been a dramatic change in deaf education. Many oral advocates have re-examined their phiIOSOphy in light of this research and thus 1 2 modified their philos0phical position to incorporate signs and fingerspelling. In short, there has been a marked philos0phical shift from the use of oral-only methods to the use of various manual methods. This shift has precipi- tated several questions: (1) What are the proposed advan- tages of manual methods? (2) For whom are manual methods more effective? (3) How does total communication which encourages the use of all avenues of communication compare to oral-only communication in conveying factual information? (4) Can students assimilate more factual information through total communication than they can through a ”pure-manual" mode of communication? Purpose of this Study The trend towards the use of total communication is supported by the general concept that deaf children re- ceive ”a little information“ through residual hearing, "a little information" through speechreading and a 'lot of information“ through signs and fingerspelling. It is often implied by proponents of total communication that deletion of any of these components from the communication process results in a loss of information. If this proposi- tion is true, then there is another powerful reason for educators to speak while communicating manually in addi- tion to the traditional one which asserts that failure to speak deprives the deaf child of critical speechreading 3 practice. Informed persons, familiar with practices at most residential schools, recognize that all teachers are encouraged to use speech: however, careful observation of teachers' practices frequently reveals that many teachers fail to use speech when communicating manually with chil- dren. In this study, the question is being asked: Does a teacher who fails to speak while communicating manually deprive children of information which they would otherwise receive? Thus, a major purpose of this research is to investi- gate the contribution of speech in the assimilation of factual information when that information is presented through total communication. Although a trend towards total communication exists, many respected oral educators have not been persuaded to alter their philosophy despite the results of current research (Miller, 1970: Fellendorf, 1970: Bruce, 1969). They have criticized the research supporting the use of manual methods on several counts. It is argued that: (1) a majority of the studies have employed ex post facto designs which do not allow control of independent variables: (2) most studies have employed matching techniques to ' achieve random equivalence of experimental and control groups, but such a technique does not control for differ- ential regression or for innumerable determinants beyond those few upon which the groups were matched: and (3) few, 4 if any, of the studies have been conducted at schools where contamination of subjects by fingerspelling and signs was not a factor. Although most professionals in deaf education are weary of the polemic between proponents of oral and various manual methods, the issue cannot be dropped since it holds a nuclear position in the educational process. Hence, another primary purpose of this study is to investigate the effects of four methods of communication in the assim- ilation of factual information: namely, oral communication, total communication, manual communication and reading. This study attempts to improve upon previous studies compar- ing methods of communication by using an experimental design rather than an ex post facto one, and by allowing students to act as their own control thus eliminating the need for matching. many general educators as well as educators of the deaf have stressed the need to formulate teaching strategies based upon the unique needs of children. However, little attention has been directed to assessing the effects of ~various modes of communication with different kinds of deaf children. It is important that research be conducted which focuses upon categorical sub-groups nested under the rubic of deafness and which evaluates how they learn--assim- ilate information--when exposed to various methods of communication. It is conceivable that a method highly 5 effective with bright deaf children, for example, may be very ineffective with retarded or lower functioning deaf children. This study proposes to evaluate the ability of bright, average and low functioning hearing impaired chil- dren to assimilate factual information under different modes of input. In summary, there are three purposes for this study. First, an attempt is made to assess the relative contribu- tion of speech in transferring factual information from "teacher“ to ”student" using total communication. Second, the study investigates the ability of different categorical sub-groups of hearing impaired children to assimilate factual information under different modes of input. And finally, the comparisons between methods of communication were made utilizing an experimental rather than an ex post facto design. Definition 2; Tg§m§_ The terms: ”oral method”, ”Rochester method", ”simulta- neous method”, “sign language“, ”Signed English“ and “total communication" are understood by most professionals in the area of deaf education. However, to eliminate any ambiquity which may exist certain standard definitions accepted in the field will be cited. These definitions served as a framework for the formulation of more explicit definitions used within this study. 6 In schools for the deaf in the United States basically three methods of communication are used: Th; Oral Method. In this method, as practiced in its pure form, the deaf child is instructed through speech and writing. He, in turn, communicates through speech, speechreading, writing and reading. This method also is known as the German Method because of its original widespread use in Germany through the work of Samuel Heinicke in the 18th century (Quigley. 1967: p-BI- The Rochester Method. This method also uses speech, speechreading, writing, and reading as a means of communication between students and instruct- or but adds fingerspelling as an additional communi- cation avenue (Quigley, 1967: p.3). Th2 Simultaneous Method. In the Simultaneous Method, communication and instruction are conducted in the same manner as in the Rochester Method with the addition of manual signs. This method also is known as the French Method due to its original use in France through the work of the Abbe Charles Michel de l'Epee in the 18th century (Quigley, 1967: p. 3). Auditory training and amplification are used in vary- ing degrees in all three methods, usually with greatest emphasis associated with the oral method. American Si n Language. Sign Language is a language in wh oh what are commonly called gestures do the usual work of words, or more precisely, in which cheremes are found instead of phonemes. But, most important, it is also a language that has its own morphology, syntax, and semantics (Stokoe, 1970: p- 5). Signed English. What most hearing observers see when watching an interpreter or teacher of the deaf is not Sign' Language, but rather Signed English. Stokoe says: This--Signed English--is a rapid succession of' glossing the content words of an English utterance more or less approximately and glossing some function best 7 words, but not all. It usually includes fingerspelled words as well as signs. Both the signer and the ad- dressee in this mode must know English well, because the signs are put together as if they were English words and not by the rules of Sign Language syntax (1970. Pa 5‘8). Total Communication. Total communication has probably been defined by Denton: By total communication is meant the right of a deaf child to learn to use all forms of communication available to develop language competence at the earliest possible age. This implies introduction to a reliable receptive-eXpressive symbol system in the preschool years between the ages of one and five. Total communication includes the full spectrum of language modes: child devised gestures, formal sign language, speech, speechreading, fingerspelling, reading and writing (1971, p. 3). Although it is not eXplicitly stated, total communi- cation encourages the early use of gestures, speech, formal signs and any other technique to facilitate language acqui- sition. This is a marked departure from the simultaneous method which has traditionally been used with students after they have first received training through the oral method during the primary years. One further distinction is important. "Combined methods” refers to any method of communication which employs any form of manual communication simultaneously with speech. ”Combined systems” refers to educational systems where deaf children are instructed orally during the primary years and through simultaneous communication during intermediate and high school years. Using these definitions as a frame of reference, more 8 parsimonious and explicit definitions have been formalized in quasi-mathematical form and presented in Table 1 to show the relationship between each of the methods of communica- tion. Table 1 Formal Definitions of Methods of Communication Modes of Communication Method SS FS BL S (G) EXi T.C. - SS xx FS x BL x S x (G) x Exi s.c.= SSxFSxBLxSx ‘ Exi M.C. 2 SS x PS X BL x (G) x EX1 R.M. = FS x BL x S x EXi 0.0. = BL x S x EXi Where: T.C. = Total communication 3.0. = Simultaneous communication M.C. = Manual communication R.M. = Rochester method 0.0. - Oral communication and Where: SS = Standardized signs FS = Fingerspelling BL = Body language S 8 Speech (Including both audition and speech- reading) (G) = The Optional use of nonstandardized gestures EXi = The sum of all ossible supportive and x = ancillary techn ques and methods which can accompany any presentation of information. interaction 9 Of those five methods of communication listed in Table 1, total communication, manual communication, and oral communication are being focused upon in this study. Relevant Literature Research dealing with methods of communication is useful only insofar as the outcomes provide educators of the deaf with additional information with which to make better decisions regarding methods to be employed in the educational setting. The outcomes of any research study must be examined against a backdrop of specific educational objectives. I It is generally accepted that the development of language and speech are crucial educational objectives in any educational system for deaf children. Inasmuch as academic achievement is so dependent upon the acquisition of language, most studies have focused upon general aca- demic achievement. As a part of some studies researchers have in addition examined speech proficiency. Although speech, as an educational objective, is subordinate to the acquisition of language, there is no question but that speech occupies a more important place in the minds of proponents of oral methods than in the 1minds of proponents of combined methods. It will be observed in the review which follows that the research suggests that only speech and articulation are better among subjects educated orally. The majority of the 10 research focuses upon academic achievement and supports philosophies which employ signs and/or fingerspelling. At the end of this section a discussion is presented regarding the ability of hearing impaired children to assimilate information through two or more modes of input simultaneously. In addition, a brief discussion is pre- sented giving reasons for including reading as a mode of communication. Research Supporting Oral Communication It has been suggested by some people that there is no evidence to support the oral method of communication. One of the major claims made by oral advocates is that by reducing the oralness of the environment, the probabil-' ity of achieving intelligible speech is also reduced. Results from two studies and extrapolation from two others support this claim. Quigley and Frisina (1961) conducted one of the first studies attempting to assess abilities of deaf children. Their primary objective was to assess the effects of instutionalization upon speech, speechreading, fingerspelling, and vocabulary. To do this they compared the performance of day and resident students at five residential schools. From 120 subjects, they conducted a secondary study--of prime interest in this review--by matching 16 students of deaf parents with 16 students of hearing parents: the implication being that deaf children 11 of deaf parents are eXposed to more signs and fingerspell- ing and less Speech than their peers who come from a hear- . ing environment. The results indicated that the day- students had significantly better speech than the resident students. In addition, day-students of hearing parents had significantly better speech than day-students of deaf parents. From these two findings, the researchers con- cluded that the "oralness of the environment“ significantly effects speech deve10pment. As a result of the above findings the researchers hypothesized that there would also be a significant difference between the speech of deaf students in day schools and deaf students in residential schools. Taking his lead from Quigley and Frisina, White (1969) compared the speech of day-students in a day program with a matched group of students from a residential school. Students were matched on age, sex, IQ, hearing loss and age of onset of deafness. In addition, both programs were under the same administration and teachers in both programs had the same general educational phiIOSOphy: in fact, several of the teachers had taught in both the residential and day program. White's findings supported Quigley and Frisina's hypothesis: deaf students in the day program made significantly fewer articulation errors than did the residential school students. Other researCh conducted by Quigley (1967) and Stuckless and Birch (1966) comparing groups which varied 12 in the oralness of their environments showed that speech is slightly superior for the students from the more oral environment although differences were not statisti- cally significant. Closer examination of Quigley's 1967 study, however, suggest that his conclusion of no statis- tical difference may be in error. First he matched subjects from three different schools using the Rochester method with subjects from three schools emphasizing oral methods. He next assessed the speech proficiency of both groups of subjects according to Hudgin's (1949) techniques and then analyzed the difference by a random samples "t" test: no significant difference was found in speech ability between subjects in the three pairs of schools. However, use of Winer's (1962: pp. 43-45) suggestion for combining 't's" leads to a reversal of that decision. Winer reports a '2' statistic computed by summing across ”t's' and dividing by the square root of the number of “t's'. The statistic is normally dis- tributed with a mean of zero and a variance of one. Using the information reported by Quigley (1967: p. 42: p. 60) a '2' equal to approximately 2.25 was computed and under the null hypothesis that the mean value for the t-statistic in the population is zero, the null hypthesis can be rejected at the .01 level of significance. Interpreted, this means that if the students in four of the six schools studied by Quigley had equivalent speech, 13 there is less than one chance in a thousand that Quigley would have attained the “t” values he obtained. Thus, it appears that the results of Quigley's work did favor the oral students at a statistically significant level. The purpose for presenting these studies by Quigley and Frisina (1961), White (1969) and the extrapolated interpretation from Quigley's (1967) later study was to make the reader aware that claims by proponents of manual methods that there is no evidence which supports the arguments of oral-only educators are not well founded. It appears that if gains in academic achievement are made possible through the use of combined methods, there is concomitantly a slight loss in speech intelligibility: possibly a result of a reduction in the oralness of the environment. It must be remembered, however, that no causal relationship has been established between the use of signs and fingerspelling and poor speech intelligibility. The studies cited were ex post facto and thus the most that can be said is that there appears to be a slightly negative relationship between the use of manual skills and speech proficiency. Research Supporting Combined Methods pf Communication E; Post Facto Research. Ex post facto research is defined by Kerlinger as: 14 ”...that research in which the independent vari- able or variables have already occurred and in which the researcher starts with the observation of a dependent variable or variables.“ (1967: p. 360). For example: deaf children of deaf parents and deaf chil- dren of hearing parents may be compared to show the effects of a manual versus an oral early environment on later communication skills. The independent variable which has already occurred, is the early environment. But to proceed on the assumption that the groups differ only on that single variable is dangerous. There is a high likelihood that there are other uncontrolled independent variables associated with the independent variable on the basis of which subjects were selected. In this case, for example, deaf parents may be more accepting, or the etiologies of the children of the deaf parents may differ system- atically from those of hearing parents. Thus, from ex post facto studies--as suggested above-- one can only conclude that the independent variable is related to the dependent variable but one cannot assume that the relationship is causal. The crux of the matter is that control of extraneous independent variables cannot be assumed as is the case when subjects are assigned at random to treatments in an experimental design. Six studies are presented in this section which have employed ex post facto designs. These studies are subject to criticism according to the deficiencies 15 just mentioned. Nevertheless, collectively they constitute a formidable argument in favor of combined methods of instruction. In Quigley and Frisina's (1961) study described earlier, they found that students of deaf parents had significantly larger vocabularies than did their matched counterparts. In addition they found a +.87 correlation between vocabulary and academic achievement. This finding suggests that vocabulary and academic achievement have approximately 76 percent of their source of variabil- ity in common. More specifically, the correlation probably reflects greater language competence in the group with deaf parents, and it is possible and probable that language competence constitutes that factor which accounts for both greater vocabulary and academic achieve- ment. It does not establish that signs and fingerspelling are the causal factors effecting language competence even though the researchers make this suggestion. The wide circulation of the results of Quigley and Frisina's research undoubtedly provided impetus which sent other researchers in pursuit of causal factors to account for the differences in academic achievement among deaf children. Shortly after Quigley and Frisina published their results, Stevenson (1964) utilized an ex post facto design and compared 134 deaf graduates of the California School 16 for the Deaf who had deaf parents with a matched group of students who had hearing parents. This group constituted all possible matched pairs of deaf graduates who had attended the school between 1914 and 1961. Only nine percent of the graduates who had hearing parents went to college whereas 38 percent of the deaf students of deaf parents received college training. In addition 90 percent of the graduates who had been exposed to early combined methods of communication through their home experiences reached a higher level of educational achievement. Stevenson’s independent variable was early exposure to and use of signs and fingerspelling. Even though Steven- son's findings conclusively favor deaf graduates of deaf parents, several rival hypothesis exist in addition to his hypothesis that early eXposure to signs and finger- spelling increases significantly academic achievement. Although not reported, probably a majority of the deaf parents were graduates of that or other residential schools themselves, and as a consequence were better able to counsel their children regarding the nature of the "educational system”. The wide span of years from which subjects were drawn suggests that the etiological patterns of the children of hearing parents may have been much more heterogenous than those of deaf parents: Certainly the types and causes of deafness had changed between 1914 and 1961. Many children during the early 17 1900's were deafened through meningitis which also can effect brain tissue, but today this type of deafness is much less common. Stevenson's study lacked specificity in describing the exact nature of other differences between the children of deaf and hearing parents. Meadows (1968), however, was much more specific. She matched 59 deaf children of deaf parents (the experi- mental group) with 59 deaf children of hearing parents (the control group), matching on age, sex, and IQ. She made the same tenuous assumption that the only systematic difference between groups was the extent of their exposure to combined methods of communication. The experimental group exhibited an average superiority of 1.25 years in arithmetic, 2.1 years in reading, and 1.28 years in over-all achievement. In addition, scores in oversall achievement indicated that the gap between the two groups increased with age, reaching 2.2 years in senior high school. Furthermore, teachers and counselors rated the experimental subjects as superior in written language, use of fingerspelling, use of signs, absence of communi- cative frustration and willingness to communicate with strangers. Meadows was bold in concluding that her results are a direct reflection of the cummulative effects of manual communication. Stuckless and Birch (1966) also elected to focus upon the early effects of manual communication. They 18 identified 105 deaf children of deaf parents and matched them with 337 deaf children of hearing parents according to the following criteria: sex, school, hearing loss (70 db or greater in the better ear in the speech range, 500 to 2000 cps), and age of onset of deafness (before the age of two). Several deaf children of hearing” parents were matched with each of the children of deaf parents because of the need for further refinement in the sample. In addition every single pair was matched for IQ using the same intelligence test as a basis for com- parison. Finally, all parents filled out a questionaire on which they indicated whether or not manual communication had been used with their children. From the initial groupings 16 matched pairs were finally selected: deaf children of deaf parents who had used combined methods were designated as the experimental group, and deaf children of hearing parents who had used only the oral method were designated as the control group. The researchers found significant differences in favor of the experimental group on reading, lipreading, and written language: however, use of a series of t-tests inflates the possibil- ity of making a Type I error. They also compared speech using a 2 x 2 Chi Square Table. They found no difference in speech although the results slightly favored the control group. As a result of these findings the research- ers concluded that early manual communication: (1) appears 19 to have no influence on the intelligibility of speech: (2) facilitates the acquisition of speechreading skills: (3) facilitates the acquisition of language as mainifested through reading comprehension and written composition, and (4) has no negative influence on the psychological development of deaf children. As cited earlier, the nature of the ex post facto design and matching techniques in no way assures the researchers that their assumption of equivalent, or random equivalence of groups was met: consequently, the conclusions of causality remain suspect. Vernon and Koh (1971) also elected to employ the same basic design as Meadows and Stuckless and Birch in evaluating graduates of the John Tracy Clinic Program. They matched deaf children of deaf parents with the Tracy Graduates who attended the California School for the Deaf at Riverside. Comparing the two groups on speech, speech- reading, academic achievement and reading yielded pre- dictable results: significant differences were found in favor of the deaf children of deaf parents. Because deaf children of deaf parents have much greater exposure to signs and fingerspelling the researchers concluded that these factors caused the observed differences. One of several confounding variables in the four studies reviewed employing ex post facto designs is that of etiology. Not infrequently it is suggested that deaf 20 children of hearing parents are more apt to have multiple prdblems resulting from the non-genetic factors which caused deafness. This being so, one might eXpect to find depressed academic performance in this population of non-genetic deaf children. In an effort to dispose of this criticism Vernon and Koh (1970) conducted another study, this time matching children of hearing parents who were recessively deaf, as evidenced in their case histories, with deaf children of deaf parents. matching resulted in 32 matched pairs. These subjects were then compared on academic achieve- ment, communication skills, and psychological adjustment. Once again the results indicated at a statistically significant level that children exposed to fingerspelling and signs were superior in academic achievement. Specifi- cally, subjects who had used signs and fingerspelling were superior to their matched counterparts by 1.2 to 1.6 years. No differences were found between the groups on speech, speechreading, or psychological adjustment. As expected, the researchers concluded that obtained differences reflected the effect of using manual forms of communication. In summary of these studies which have used ex post facto designs,two points should be made. First, within each study several rival hypotheses exist which could account for obtained differences. A few of the more 21 obvious ones are: (1) deaf parents may be more accepting of their children: (2) deaf parents usually send their children to residential schools whereas hearing parents most frequently send their children to day-school programs in their home community. It has been postulated that the deaf children of hearing parents in a residential school are less capable than deaf children of hearing parents in public school day programs: (3) despite Vernon and Koh's (1970) study, etiology may still be a partial cause for the consistent superiority of children of deaf parents. The interaction of these and other factors may well account for observed differences also. The second point is that despite the arguments used against the studies individually, the combined effect upon the profession has been acceptance of the assertion that deaf children can learn more through use of finger- spelling and signs combined with speech-~combined methods-- than they can through oral communication alone. Consider- ing the constraints under which research must be conducted-- it is usually impossible to randomly assign children to treatment groups because of parental resistence, etc.-- the evidence supporting combined methods is rather com- pelling a Atteppts pp Egperimental Research. The researchers who have carried out the aforementioned studies have 22 recognized, in part, the limitations of ex post facto designs. Three other studies have been identified which approximate eXperimental research: that is, where the researcher had some control over treatment conditions. Johnson (1948) conducted a study wherein she attempted to assess the ability of deaf children at a residential school to assimilate 10 simple sentences presented under different modes of communication: (1) manual (signs and fingerspelling): (2) Oral (speech only: no hearing aid used): (3) Accoustic (speech: hearing aids were used): (4) Speech-hearing (no speech- reading: audition only)3 and (5) Fingerspelling. Johnson did not indicate whether or not speech was used with the "manual" and "fingerspelling" modes. With an N=253, it appears the entire pOpulation of the residential school was studied. However, the students at the school were segregated according to their ability to use their residual hearing and speech. Hard of hearing children were in the Acoustic Department. Deaf children unable to benefit from acoustic training were put in the (Iral Department, and students who could not function in eisther the Oral or.Acoustic Departments were put in the Manual Department. All 88 received all treatments. 23 Results indicated that all three groups of students understood approximately the same amount of information through fingerspelling, or fingerspelling combined with signs. Both the oral and manual groups understood signifi- cantly more through fingerspelling than any other mode of communication. The acoustic group understood more through acoustic communication (audition). The oral group did significantly better in the use of “speech- hearing', lipreading, and acoustic communication than the manual group. The acoustic group did significantly better than the oral group on the same three variables. In summary, Johnson drew three major conclusions: (1) using a hearing aid can significantly increase comprehension: (2) fingerspelling, and signs combined with fingerspelling are more effective than oral communication, and (3) finger- spelling alone was the most effective mode of communica- tion for both oral and manual students. Johnson exercised control in the study by designa- ting the kinds of treatments which were administered to subjects and the conditions under which treatment was received. Basically, she assessed the ability of children to assimilate information under various modes of input: however, she failed to focus upon the contribution of speech in simultaneous communication. Johnson's findings of the superiority of fingerspell- ing are interesting in light of later studies conducted 24 with the Rochester method by Quigley. Quigley (1967) selected the Rochester method as the independent variable of interest in a longitudinal study carried out between 1963 and 1967. (This is the same study cited earlier (p.12) where ”t" scores on speech were combined.) Three "experimental” schools were selected which used the Rochester method with all the children in the school. Each of these experimental schools was matched on the basis of relative geographical proximity, and similarity in the size and composition of student population with a control school which advocated a combined system of instruction. Subjects from the three experimental schools were then matched with subjects from the three control schools: matching was extensive. The mean age of students at the conclusion of the study was 13.18 years for the experimental and 13.25 years for the control group. Student performance on the Stanford Achievement Test, written language, and speech were compared every year for five years. Initially no statistical difference existed between groups: however, at the conclusion of the study the experimental subjects scored significantly higher on all variables except speech. The researcher concluded that obtained differences reflect the superior influence of the Rochester method. 25 In the same official report Quigley eXplains that a second study, called an ”eXperimental study" was begun near the completion of the larger study. The experimental study also focused on the influence of the Rochester method. Sixteen preschool deaf children in a residential school using the Rochester method were selected and matched with 16 preschool children at a comparable resi- dential school which used only the oral method with pre- school and elementary children. The researcher imposed control on the study by securing a commitment from the school administrations, dorm personnel, and parents that only the Rochester and oral methods would be used with each of the respective groups of children. For the oral group the children's dorms and classrooms were physically separated from the rest of the student popula- tion to reduce the possibility of contamination. The results of the study revealed that the Rochester subjects were statistically superior in fingerspelling: on two measures of lipreading: on five of seven measures of reading ability, and on three of five measures of written language. The only difference in favor of the anal group was in “gramatical correctness”, but this, the researcher concluded, reflected the fact that the oral subjects wrote shorter and less complex sentences. This study has been criticized on the grounds that the researcher failed to randomly select subjects and that 26 he matched subjects. Both procedures fail to provide unbiased estimates of the treatment effects. Further, eXperimenter bias may have had an effect upon critical personnel involved in operating both programs. Never- theless, considering the practical constraints under which a researcher must labor, it seems that this later experimental study constitutes a respectable attempt to study the effects of the Rochester method and the oral method experimentally. Sensory Overloading In addition to studying the contribution of speech in total communication, and improving on the ex post facto design, the present study also attempts to provide addition- al information regarding the ability of segments of the hearing impaired population to decode information through various modes of input. Close observation of total communication reveals that it is possible for a person to say one thing and sign or fingerspell another word with synonymous meaning. For example, a person may say: ”I'm going home“, but sign, "I am going home." Another example might be for a person to say: “I felt nauseated', but sign, ”I felt sick“. In the latter example the sign for sick and nauseated is the same. Even though there are now ways of putting the appropriate inflection and tense on words in sign language, it takes a great deal of concentration to do so. Very 27 seldom are all inflections and tense changes perfectly displayed through signs in total communication. The effect of such slight variations in the total communication pro- cess may very well be different for various segments of the hearing impaired population. Most adult deaf persons assert that they understand more through the combined use of speech, fingerspelling and signs than they do through just manual communication alone. However, for some deaf peeple, particularly those who are less verbal- ly competent, the variations between the manual component and the verbal component may slightly inhibit assimilation of what is being said. Such a hypothesis seems even more plausible when one considers the work of Gaeth. Gaeth (1966) studied several aspects of verbal and nonverbal learning in normal children and hearing impaired children. He found that when word lists were presented to subjects their recall was significantly greater when the words were presented either visually or auditorily than when the words were presented through visual and auditory modes simultaneously. He also found that his normal and profoundly deaf subjects quickly adapted to the mode which was most meaningful to them: the normals listened, and the deaf watched. Interestingly enough he found the hard of hearing trying to shift back and fourth between listening and watching: this strategy resulted in depressed scores. 28 It maybe that this same phenomenon exists with regard to total communication. Will attempts at shifting from one mode of communication (manual) to another (speech) result in depressed scores on assimilation of factual material? Respecting the assertions of many adult deaf peOple who say that they receive more through total communication than they do through manual communication, it is hypothesized that average and bright deaf students are sufficiently adroit in shifting attention from the hands to the lips so that under total communication more information will be assimilated: however, it is further hypothesized that lower functioning deaf chil- dren will not be able to process the slight variations between the manual component and speech component as effectively and in fact their attempts to do so will result in depressed scores on the dependent variable of assimilation of factual material. Rgading g§_g Method pf Commun cation. v hEven'though all educators of the deaf use reading a great deal in their efforts to communicate with the deaf, the literature reveals no attempt to compare the ability of deaf students to assimilate information through reading and through oral, or total communication. It is generally accepted that when syntactic patterns are not unduly complex and when the vocabulary is not 29 beyond the ability of the student more information can be assimilated through reading than through oral communication. many peOple would say this assertion holds true for combined methods as well, although some educators feel that more information is transferred through total communication than through reading. Inasmuch as no empirical evidence exists to support these assertions, reading was included as a means of communication and was compared to the oral, total, and manual methods of communication. Hypotheses It is hypothesized that hearing impaired children will assimilate more factual information when it is pre- sented to them through total communication and manual communication than they will through oral communication. It is hypothesized that average or higher function- ing hearingimpaired children will assimilate more factual information when it is presented to them through total communication than they will through manual communication. It is hypothesized that lower functioning hearing impaired children will assimilate more factual information when it is presented to them through manual communication than they will through total communication. It is hypothesized that hearing impaired children will assimilate more factual information when it is 30 presented through reading than they will through oral, total. or manual communication. CHAPTER II DESIGN OF THE STUDY Method of communication of factual material, or mode of input, was the independent variable of primary interest in this study. Four methods of communication were used: (1) oral communication, (2) total communication, (3) manual communication, and (h) reading. Factual material was presented to a stratified random sample of students from a residential school for the deaf under each of these four levels on the fixed independent vari- able. Two other fixed independent variables, age and IQ with three levels on each were incorporated in the design to gain greater precision and to investigate the effects of these variables on the dependent variable which was assimilation of factual information. To control for bias in treatment groups, subjects were exposed to repeated treatments: each treatment was one method of communication. Thus subjects were compared to themselves and not to a random or matched group. It ‘was reasoned that by controlling as many extraneous variables as possible and by allowing the method of commu- nication to be the only systematic difference between groups, test differences observed on the dependent vari~ able should reflect how much information was transferred from the presenter-of—the-material to subjects, or stated 31 32 differently; how much information was assimilated under the various methods of communication. Generalizations made from this study should be rep”.5. stricted to populations which reflect the composition of the random sample and the residential school from which they were drawn. . A description of the sample, scheduling of subjects for testing, selection of test material, and a description of the experimental design are presented in this chapter. Sample A stratified random sample of 45 hearing impaired children was drawn from the maryland School for the Deaf in Frederick, Maryland. The school has a student population of 330; 185 male and 1&5 female. Students range in age from four to 19. All but seven students attend as resident-students. Approximately 195 students are eleven years of age or older. The random sample was obtained by first assigning all students in the school between the ages of 11.0 and 18.7 years to the age and IQ groups shown in Table 2. Of this population, those with a mean hearing loss in the speech range (500,1000,2000 Hz) less than 65 db were eliminated from the study. Students whose IQ was below 60 on the performance scale of the Wechsler intelligence Scale for Children were also.eliminated 33 Table 2 Stratification of the Population of Hearing Impaired Children Intelligence Quotient Age 60-89 IQ 90-110 IQ ill-Above (L) (M) (H) (L) 11.0-13.5 years Group-l Group-2 Group-3 (M) 13.6-16.1 years Group—4 Group-5 Group-6 (H) 16.2-18.7 years Group-7 Group-8 Group-9 LcLow34M=Middlez HzHigh from the sampling population. school records. IQ scores were taken from From the remaining population of 145, five subjects (Ss) were randomly selected from each of the nine groups in Table 2 using a table of random numbers; hence, N=45. The mean age, IQ, hearing loss, and reading level for each of these nine groups are presented in Table 3. The means and standard deviations for each of the three levels of age and IQ are presented in Table 4. The grand mean and standard deviations for the entire sample on hearing loss, reading, and IQ are presented in Table 5. 34 Table 3 Means for Age, IQ, Hearing Loss and Reading Level for the Stratified Sample Age-IQ Age in IQ Hearing Reading Groups Years Loss(db) Level * Group-1 12.50 79.60 90.60 2.36 Group-2 12.50 103.80 93.40 2.58 Group-3 12.10 122.20 96.60 3.26 Group-4 15.30 78.20 82.60 2.12 Group-5 14.90 98.60 81.60 4.08 Group-6 14.50 120.40 90.60 4.08 Group-7 17.40 74.80 85.00 3.10 Group-8 17.50 102.60 81.80 3.68 Group-9 17.50 123.00 92.00 4.30 ;_§ub-test on Stanford Achievement Test Table 4 Means and Standard Deviations for Each Level of Age and IQ Age IQ _. A-L A-M A-H IQ-L IQ-M IQ-H Means 12.42 yr. 14.92 yr. 17.49 yr. 77.33 101.66 121.87 S.D. 14.78 mo. 9.41 mo. 6.15 mo. 8.08 6.73 6.96 35 Table 5 Means and Standard Deviations for the Entire Sample on Hearing Loss, IQ, and Reading Level Hearing Loss IQ Reading Level Means 88.24 db 100.29 3.28 S.D. 11.40 db 20.00 1.13 In an effort to determine if hearing loss was confound- ed with age or IQ, a 3 x 3 fixed effects model analysis of variance was carried out with the two fixed factors being age and IQ. The results of the analysis in Table 6 indicate that hearing loss was not significantly related to age or IQ. Table 6 Analysis of Variance Table on Hearing Loss for the Factors of Age and IQ Sources of Variation df SS MS F P* Age 2 642.72 321.36 2.58 NS IQ 2 524.85 262.42 2.11 NS Age x IQ 4 80.75 20.13 .17 NS Subjects: Age x IQ 36 4,474.00 124.27 * alpha = .05: NS=Not Significant 36 Histogy g§_ngple and School All the Se in the sample were born deaf except for three, and those three were deafened before the age of three years. Of the 42 88 born deaf, 21 are congenitally deaf and 21 are deaf as a result of unknown causes. 0f the entire sample all but three were at the mary- land School for more than three years. Two of the 45 88 were residents of two years and only one was a resident student for one year. Historically the Maryland School has employed a combined system of education: however, three years prior to the study the school embraced total communication and since that time has used total communication at all age levels. From observation it appeared that the entire staff enthusiastically practices total communication at the school. Design and Experimental Methods A 3 x 3 x 4 factorial repeated measures design, fixed effects model was used. The first factor, method of communication, consisted of four levels: (1) oral communication, (2) total communication, (3) manual communication, and (4) reading. The second factor was age. It was primarily included as a blocking variable. Three levels were specified on this factor: (1) 11.0 to 13.5 years of age (Age-Low): 37 (2) 13.6 to 16.1 years (Age-Middle): and (3) 16.2 to 18.7 years (Age—High). The third factor, IQ, was also used as a blocking variable: however, more interest was invested in it. Particular interest in this variable existed because of the hypothesized interaction between method and IQ. Three levels were specified on this factor as well: (1) 60 to 89 IQ (IQ-Low): (2) 90 to 110 IQ (IQ-Middle): and (3) 111 IQ and above (IQ-High). A data matrix is presented in Figure 1. Definition 9: Methods Four methods of communication were presented to all 53 in the study. The four levels on the first factor constituted treatment. All material under each of the methods was administered to 33 by the same person who was a certified interpreter for the deaf: the inter- preter was not a member of the school staff. 38 r . . _ . _Method-1r Method-2' Method-3 Method-4 Age IQ~ Ss s- S- IQ - L S- S... S.— 3. Age - L IQ - M S-1 1 IQ - H 1 1 IQ - L NN Age - M IQ~- M NN IQ - H UK») bob) Age - H IQ ' M #H?’ gIQ-H \naaaPOO-am“...HanoO'mJ‘aaaHOaaommooaHOQQIanwmt—t 4:- Figure 1 Data matrix of the 3 x 3 x 4 Repeated Measures Design 39 Method-1. Under method-1 Ss were presented factual information orally. The oral method was defined in Chapter 1 as: 0.0. a BL x S xEXi By definition speech (S) is only one part of the oral method: however, it is the most important part of the method and thus the component of prime interest. Material was presented to the Ss through speech by the interpreter who was encouraged to use all natural body language (BL). The students were encouraged to wear their hearing aids. Proponents of oral methods have always argued that I'talking" alone does not constitute the oral method: a host of other techniques are drawn upon. However, the techniques and inputs (Exi) are common to all methods of communication: therefore, it was decided not to deal with this component in any form in the various methods of communication. In short, no visual aids or other communicative levers were used in any of the methods of communication. The primary component within the various methods was focused upon. The measured rate at which the material was present- ed is provided in Table 7. 40 Table 7 Mean Length of Time in Which Material was Presented* Oral-1 T.C-Z manual-3 Means 3.43 min. 3.62 min. 3.59 min. S.D. 30.75 8900 “2.23 830. 55.47 sec. 7IMean and S.D. for reading was not computed because of variability between students in reading time. Method-2. Under method-2 Ss were presented passages of factual information by the interpreter using total communication. The interpreter presented the material in signed English using Gallaudet endings, such as:-—ion, ---ment,--ing, etc. The interpreter was encouraged to use all the components of communication specified in the formal definition: T.C. - SS x FS x BL x S x (G) x E11 Inasmuch as the vocabulary in the passages of factual material was not difficult there was little need for contrived gestures: thus non-standardized gestures (G) were deleted along with (3x1). It will be observed that with (G) deleted, total communication and simulta- neous communication are equivalent. The words which were signed and fingerspelled in the passages used under total communication are presented in Appendix A. It will be noted that some words 41 were both signed and fingerspelled. The rate of presentation of material under total communication is also presented in Table 7. Method-3. Under method-3 Ss were presented the material by manual communication. manual communication was defined as: M.C. - SS x PS X BL x (G) x EXi For the same reasons cited for method-1 and method-2, (G) and EXi were deleted from manual communication during test- ing. With these components deleted, the only difference between manual communication and total communication was speech (S): thus a statistical difference observed between total and manual communication will be attributed to the effect of Speech: M.C. = SS x FS x,BL T.C. = SS x FS x BL x S The words which were signed and fingerspelled are presented in Appendix B and the rate of presentation is presented in Table 7. Method-4. Under method-4 the Ss were handed a dittoed 00py of the factual information and given a specified time in which to read it. For passages B, three minutes were allowed, and for passages D, four minutes were allowed. It should be noted that there appeared to be rather broad variability in time taken by the students in reading. 42 Besides setting a maximum time limit which approximated the mean length of time it took to present the other methods no control over reading time was imposed. Scheduling gfwgg for Testing Once the 45 Se were selected, each of the five 88 within the nine age-IQ groups specified in Table 2 was given a numerical code, so that $3 in group-1 were coded 1,2,3,4, and 5: Se in group-2 were coded 6,7,8,9, and 10 and so forth, up to group-9 which was coded 41,42,43,44, and 45. Once coded, the $8 from the nine groups were then systematically assigned to testing groups. A testing group was defined as a group of seven or eight 83 who were tested together. Ss were assigned to testing groups as presented in Table 8. Table 8 Composition of Testing Groups A Testing Age-IQ Groups 2 Groups G-l GéZ G63 G44 G35 G-6 G-7 G-Ef G-9 TG-1 3: S7 S13 s19 S25 S31 S3? 544* TG'Z s2 S8 S14 S20 s26 s32 S38 TG‘3 33 39 S15 S21 S27 S33 S39 545* TG'5 s6 812 S18 S24 S30 s36 S42 Slu=Subject coded 14 from age-IQ group 3: assigned to testing group 2 *83 were randomly assigned to testing groups. 43 To eliminate ”time of day”, ”day”, and ”order of presen- tation” of method as confounding variables testing groups were systematically assigned to receive testing according to the schedules in Tables 9 and 10. Table 9 Schedule of the “Day” and Order of Method of Presentation* Testing Day of Testing Groups ’Day-i Day42 "“"fi§§:§ TG-1 Oral T.C. manual TG-2 Oral Manual T.C. TG-3 T.C. Oral manual TG-4 T.C. Manual Oral TG-5 Manual T.C. Oral TG-6 manual Oral T.C. * Reading was nested within cache? the three days. See Table 11. Table 10 Schedule Indicating ”Day", “Time of Day”, Testing Groups Were Tested and the Method Received Time of Day of Testing Day Day-1 Day-2 Day-3 Period-1 TG-1,2* TG-1,5 TG-1,3 on]. T o C a Manual Period-2 TG-3,4 TG-2,4 , TG-4,5 T.C. manual Oral Period-3 TG-5,6 TG-3,6 TG-2,6 manual Oral T.C. * This entry should be read, ”Testing groups 1 and 2 received the oral method at period-1, on day-1. 44 materials Uggg As mentioned, the dependent variable was ”assimilation of factual information”. To assess this variable 16 factual passages were taken from Gettipg ppg,§gpp§: Specific Skill Series, Bppk_§_and Bppk Q (Barnell Loft, Ltd 1966). According to the publisher the eight passages from Bpplg _B_ were at the second grade level and the eight passages from gppk,2_were at the fourth grade level. Each of the eight passages from the set of passages taken from.§ppk_§_was randomly assigned for use with a single method of communication. Each of the eight passages from set Q,was also assigned randomly for use with one of the four methods of communication. When this procedure was completed each method of communication had four passages randomly assigned to it: two from B_o_gk_ B_ and two from M]; 2. Each passage from.§ppk,§_also had a standard set of eight questions associated with it. Each passage from pppg,2_had ten questions associated with it. These questions were used to assess how much information had been assimilated under each method of communication. Each question was a threeefoil multiple choice question. The total number of questions presented to 8 under each method was 36: two sets of questions from.§ppk.§,(8 + 8 = 16 questions) and two sets of questions from.§ppk_2,(10 + 10 = 20 questions). Under each method of communication the interpreter presented a passage of factual information to two testing 45 groups. Immediately following the presentation, the questions associated with the passage which was given were distributed to the Ss. The Ss were required to read the questions and circle the right answer. On day-1 and day-3 five passages and their respective questions were administered to the Ss. On day-2 six passages and their respective questions were presented. The variation in the number of passages present- ed per day existed because only one reading passage was given on day-1 and day-3: however, on day-2 two reading passages were presented (See Table 11). A schedule indicating which passages were presented under each method of communication is presented in Table 11. Table 11 also illustrates how the reading passages were inserted into the alloted three day testing period. Table 11 Schedule Indicating the Insertion of Reading Treatment Over the Three Day Testing Period and the Passages Presented Under Each Method of Communication Time of Day of Testing Day Day-1 Day-2 Day-3 Oral T.C. manual Passage-1¥-B* .Passage 20-B Passage 18-B Passage 2 -B Passage 2448 Passage 23-B Period 1 Passage 22-B Passage 23-B* Passage 23-D Passage 19-D Passa e -D Passage 25-D Passage 24-D Passage ZI-D Passage 20-D* Passage 22-D T o C 0 Manual 9222: Passage 12eB* Passage 18-B Passage 21-B Passage 20-B Passage 23-B Passage 22-B Period 2 Passage 24-B Passage 28-B* Passage 1 -D Passage 21-D Passage 1 -D* Passage 2 -D Passage 22-D Passage 23-D Pasgage 20-D* Passage 25-D Manual Oral T.C. Passage 18—B* Passage 21-B assage ZO-B Passage -B Passage 22-B Passage 24-B Period 3 Passage 23-B Passage 21-D Passage 23-D Passage 25-D Passage zg-B* Passa e -D* Passage 19-5 Passage 24-D Passage 22-D Passage 20-D* * These passages were presented through reading The material used in the study was reviewed by the re- searcher, two teachers working with the Ss, the school prin- cipal, and an outside consultant: this team agreed that the material appeared to be suitable for the purposes of the 47 study. Several factors were considered in selecting the material: (1) the age range of the sample, (2) the reading ability of the Ss (Examination of school records revealed a mean grade level of reading achievement of 3.28): and (4) the degree to which the material focused upon teaching facts which the 83 were most likely unfamiliar with. Procedures 22; Testing ‘Rpppp .All testing was conducted in a single room made available by the school administration. The roOm was large enough to accommodate comfortably two testing groups. Two groups were tested together in order to reduce the number of presentations made by the interpreter and the number of times teachers' classes were interupted. Care was taken to ensure a dark background so that the interpreter could be easily seen. Lighting in the room was adequate for speechreading. Lpptructions. On day-1, before any passages were pre- sented the interpreter, using total communication, gave the following instructions. We (speaking of herself and the research- er) are trying to understand better how deaf stu- dents learn. We want you to read this story and when you are finished answer some questions about it. Read the story carefully. Don't raise your hand if you don't understand. We can't answer any questions. You have three (four) minutes. We will pass out the papers face down. When we blink the lights turn the paper over and begin reading. When you finish we will pass out some questions for you to answer. 48 The following instructions were also given through total communication on day-1. Oral. Now I'm going to read you a story. I'm not going to use my hands to sign or finger- spell. You will have to lipread what I say. Don't raise your hand if you don't understand: I can't stop to.answer questions. Watch care- fully and try to understand what I say. When I finish I will pass out some questions for you to answer. Total Communication. Now I'm oing to read you' a story using total communicat on. Don't raise your hands if you don't understand: I can't stop to answer questions. Watch carefully and try to understand what I say. When I finish I will pass out some questions for you to answer. manual Communication. Now I'm going to read you a story using just signs and fingerspelling. I'm not going to speak so you will have to watch my hands. Don't raise your hand if you don't understand: I can't st0p to answer questions. Watch carefully and try to understand what I say. When I finish I will pass out some questions for you to answer. While the interpreter presented the passages the re- searcher sat at the rear of the room and timed the rate of presentation for each passage: this information is presented in Table 7 (P. 40). The variable of "rate of presentation“ was analyzed using method and day of presentation as fixed independent variables, by means of analysis of variance. The implied null hypothesis of no significant difference between time 1,2, and 3 across methods, on the dependent variable of rate of presentation, was not rejected. The results of the analysis are presented in Table 12. 49 Table 12 Analysis of Variance Table for Rate of Presentation of Factual Information (F Source of Variation df SS m3 P“ Time 2 109 54.5 .02 NS Method 2 53 26.5 .01 NS Time x methOd LI' 918 22905 067 NS Rate: Time x Method 12 40,848 3,404.0 ? Alpha set at .05 Reading was not included in this analysis because the time alloted for reading was not controlled: each S read the passages at his own rate of speed. As mentioned above, the maximum time given to Ss to answer the questions on;§ passages and Q_passages was three minutes and four minutes respectively. Observation of Ss under reading indicated clearly that only a very few 83 ever took the alloted time: hence, if anything, Ss took less time to assimilate infor- mation under reading, as compared with the other methods of communication. CHAPTER III RESULTS OF THE ANALYSIS It was hypothesized that: (1) hearing impaired children will assimilate more factual information when it is present- ed to them through total and manual communication than they will through exclusively oral communication: (2) average and higher functioning hearing impaired children will assim- ilate more factual information when it is presented to them through total communication than they will through manual c0mmunication: (3) lower functioning hearing impaired chil- dren will assimilate more factual information when it is presented to them through manual communication than they will through total communication, and (4) hearing impaired children will assimilate more factual information when it is presented through reading than they will through oral, total or manual communication. In addition to a description of ancillary findings, a decision of acceptance or rejection was made regarding each of the hypotheses following analysis and these results are presented in this chapter. Prior to the main.analysis of the data,.consideration was given to the assumptions underlying the use of analysis of variance in a repeated measures design. Four such assumptions are essential: (1) normality of the population from which the sample was taken: (2) equality of variance across treatment conditions: (3) independence between 50 51 subjects, and (4) equal correlation between all possible combinations of treatment conditions. Where sample size is equal across treatment groups (methods of instruction) violation of the first two assumptions is of little conse— quence. In order to insure non-violation of the third assumption, subjects were closely observed while they answered the questions to assure that they worked indepen- dently. The fourth assumption, an assumption specific to a repeated measures design, was more difficult to accept. Inasmuch as it is crucial to meet this assumption before proceeding, correlations were computed to determine the appropriateness of making this assumption in the subse- quent analysis. The Ss' scores across the nine levels of age and IQ were first correlated across methods of communication. This resulted in 54 correlation coefficients (nine levels, and six possible method combinations at each level). Each of the 54 correlation coefficients were transformed using an 'r' to ”z” transformation. The mean of the z scores was computed for each of the treatment combinations and that mean was then transformed back to a correlation coefficient. These adjusted correlations were computed because they are slightly more accurate than a simple correlation computed across all Ss independent of levels. Adjusted correlation coefficients are reported in Table 13. 52 Table 13 matrix of Correlation Coefficients Between Methods of Communication Oral T.C. Manual Reading Oral Communication - .49 .41 .32 Total Communication - .80 .64 manual Communication - .80 Reading - Inasmuch as the range in correlation coefficients (.32 to .80) was large, which renders the assumption of equality of correlations across treatments suspect, the. Geisser and Greenhouse Conservative F-test (Kirk, 1968: pp. 142-143) was used throughout the analyses, both in the over-all F-test and in generating specific contrasts between means using the Tukey method. The mean number of correct responses and the standard deviations under each method of communication are presented in Table 14. Table 15 presents the grand means for the factors of age and IQ across all levels of communication. It can be observed that the mean number of correct re- sponses under oral communication is conspicuously lower than the other three methods of communication. Likewise, the means for the younger group (A-L) and lower IQ group (IQ-L) stand apart from the other two levels on this combined measure of communication skill. 53 TabIe 14 Mean Number of Correct Responses and Standard Deviations for Methods of Communication 4r v __.v_ Oral T.C. Manual Reading Means 15.67 23.73 25.51 27.36 S.D. 4.82 6.08 5.31 6.21 Table 15 Grand Mean Number of Correct Responses for Levels of Age and IQ Age ' IQ Levels A-L .A-M A-H IQ-L IQ-M IQ-H Means 20.85 23.97. 24.38 120.13 24.48 24.58 L-Low: M=Middle: H-High _M_gi_p Analysis In order to test for the statistical significance of the differences between means and possible interaction effects the data were subjected to an analysis of variance. Table 16 presents the sources of variation, the degrees of freedom, the adjusted degrees of freedom used in looking- up critical values, the sum of squares, the mean squares and the F-values. 54 Table 16 Summary Table for 3 x 3 x 4 Analysis of Variance Sources of Variation df SS df** MS F Method (M) 3 3,580.40 1 1,193.62 94.32* IQ (I) 2 774.70 2 387.35 6.67% Age (A) 2 447.43 2 223.72 4.022 A x I 4 423.47 4 105.87 1.82 A x M 6 152.92 2 25.49 2.01 I x M 6 255.79 2 42.63 3.37*’ m x A x I 12 73.22 4 6.10 .48 $83 A x I 36 2,092.20 36 58.11 M x Ss: A x I 108 1,366.70 36 12.65 * Significant at the .05 levEI of significanbe **Adjusted degrees of freedom As indicated in Table 16 there was a significant effect due to method of communication, age, intelligence, and a method by intelligence interaction. The interaction effect was plotted to see if the inter- action was ordinal or disordinal before post hoc procedures were carried out. As indicated in Figure 1, interaction was ordinal and thus generalizations across levels on the factor of IQ for methods of communication were possible. That is, mean scores on the four methods of communication have the same rank order at each IQ level. The same holds true for age levels. 55 Observation of Figure 2 reveals that low, middle and high IQ groups do not differ significantly under oral communication, whereas they do differ under the other three modes of communication. Having determined that differences existed between methods of communication, age and IQ it was necessary ' to construct contrasts on the means in order to locate specific differences. Constructing contrasts implies a null hypothesis of no difference between two or more means: that is, the difference between two or more means is zero. If the difference between two means, plus or minus a stretching factor, spans zero, then the null hypothesis of no difference should not be rejected: if the interval generated does not span zero then the null should be rejected and a significant difference is indicated between pOpulations from which the sample means were drawn. Confidence intervals, of the kind specified above, were constructed and are presented in Table 17 for the main factor of method of communication. Alpha was set at the .05 level for all contrasts. Mean Number of Correct Responses 56 7 (29.87) (28.20) I 27 r— (26.93) i ..... 1‘(25531 24 .— . (24.00) 21 _ i D ........ 0 Oral (21. oooooooo ‘ ‘ T.C. . Manual ‘ l 7 Reading (19.73) 18 _. —I ..................... D ................. ‘....D.‘..........~. (16.00) D .............. (15.60) 15 P- (15.40: V 4/ Low Middle High Intelligence Quotient Figure 2 Mean Number of Correct Responses for Low, Middle and High l0 Groups in the Maryland Study ( o . T .x M .X .X R M .X c of. . X .YMN \W “L 0r th In cc to PC OI he 57 Table 17 Contrasts on the main Factor of Methods Of Communication Contrast Difference Confidence Decision between Means Interval ii;c.-ib 23.73-15.67=8.06 6.02 to 10.10 Sig.Difference Eh - Eb 25.51-15.67=9.84 7.80 to 11.88 Sig.Difference Eh - I6 27.36-15.67=11.69 9.63 to 13.71 Sig.Difference Eh - If 0. 25.51-23.73=1.78 -.26 to 3.82 No Difference ifi - 25.0. 27.36-23.73=3.62 1.58 to 5.66 Sig.Difference ifi - 2M 27.36-25.51=1.85 -.19 to 3.89 No Difference 0:0?EiT—T7573T6tEi-CBmmunication: m=manual: REReading Epggp Hypotheses It was hypothesized that hearing impaired children would assimilate more factual information through total and manual communication than they would through exclusively oral communication. As indicated in Table 17 this hypo- thesis was supported by the results of the analysis. Inasmuch as the mean amount of information under oral communication was significantly lower than the mean for total or manual communication it was concluded for this population of hearing impaired children that they do indeed assimilate more information when it is presented using total or manual communication. It was hypothesized that average and higher functioning hearing impaired children would assimilate more factual 58 information when it was presented through total communi- cation than they would through manual communication. This hypothesis was clearly not supported by the research find- ings. This fact can be observed in Figure 2, and Table 18. The means for both the middle and higher IQ groups under total communication were lower than they were under manual communication. Since these results are Opposite to what was predicted no test of significance was necessary. It was hypothesized that lower functioning hearing impaired children would assimilate more factual information through manual communication than they would through total communication. As indicated in Table 18, the lower func- tioning children did assimilate more through manual communi- cation. In order to determine if that difference was significant, a confidence interval was constructed around the difference between the means of the lower IQ group under total and manual communication: this information is presented in Table 19. Table 18 Mean Scores Across Low, Middle and High IQ Levels for Methods of Communication Oral T.C. manual Reading IQ-M 15.60 25.93 28.20 28.20 IQ-H 16.00 25.53 26.93 29.87 59 Table 19 Contrast Comparing Means of Low IQ Ss on Total and manual Communication Contrast Difference Confidence Decision Between Means Interval im‘zrp. 21.40-19.73=1.67 -2.54 to 4.88 No Difference As indicated in Table 19 the low IQ 83 did not do significantly better under manual communication than they did under total communication. It was generally concluded, therefore, that manual and total communication were not significantly different for lower, average, or higher functioning hearing impaired children. It was also hypothesized that hearing impaired chilf dren would assimilate more factual information when it was presented through reading then they would through oral, total, or manual communication. This hypothesis was par- tially supported. As indicated in Table 17 performance on reading did not differ significantly from performance under manual communication: however, it did differ signi- ficantly from both oral and total communication. It was thus concluded that reading was significantly superior to oral and total communication as a method of presenting factual information to hearing impaired children. Ancillary Findings Contrasts were also constructed using the means for the three levels of age and IQ. The results are presented in 1201 functio in perf the mid groups (2) yox middle from ti differ informu Contra >4: 3 l r n. I >21 >41 >4! :2?"::?" / 60 111 Tables 20 and 21. These results indicated that: (1) lower functioning hearing impaired children differ significantly ix: performance from the higher functioning children, and 1316 middle functioning children. The middle and high IQ groups did not differ significantly from each other: (2) younger age 88 did not differ significantly from the middle age Ss, but the younger 88 did differ significantly from the older age Ss. The middle and older age Ss did not differ significantly in their ability to assimilate factual information. Table 20 Contrasts on the Factor of Age Contrast Difference Confidence Decision Between Means Interval 23.97-20.8523.12 -.31 to 6.57 No Difference . ' ' 24.38-23.98: .40 -3.03 to 3.83 No Difference 24.38-20.85-3.53 .10 to 6.96 Sig.Difference N N N lmlmlgll l tgflzgaltrfl 61 Table 21 Contrasts on the Factor of IQ Contrast Difference Confidence Decision Between Means Interval f), - 2L 24.48-20.13=4.35 .92 to 7.78 Sig.Difference if], - 2M 24. 58-24.48= .10 -3.33 to 3.53 No Difference 56H - i1. 24.58-20.13a4.45 1.02 to 7.88 Sig.Difference Summapy Based upon the results of the analysis and restricted to the population from which the sample was drawn, certain generalizations are possible: 1. Hearing impaired children assimilate more factual information through reading than they do through oral or total communication. 2. All categorical sub-groups of hearing impaired Children assimilate more factual information through total communication; and manual communication than they do through oral communication. 3. High, middle and low IQ groups assimilate approxi- mately the same amount of factual information through oral conlmunication. 4. Bright (IQ-High) and average (IQ-Middle) hearing 1mPaired children assimilate more factual information than lower functioning children (IQ-Low) when that information is 62 presented through total communication, manual communication and reading. 5. Older (Age-High) hearing impaired children assimilate more factual information than younger children (Age-Low), but not significantly more than children in the middle group. 6. The speech component in total communication does not significantly increase the amount of factual information assimilated beyond that assimilated through manual communi- cation. CHAPTER IV REPLICATION OF THE MARYLAND STUDY A replication of the maryland study was conducted in order to determine the degree to which the findings of that study can be generalized to populations of hearing impaired children who have not been exposed to the systematic practice and philosophy of total communication. There are likely to be some differences between popu- lations of children educated in different parts of the country, under different personnel with different education- al philosophies. Recognizing the potential for differences in learning styles between the students at various resi- dentials schools, it is difficult to make sweeping general- izations based upon the results of the Maryland study without further examination of other residential popula- tions. The desire to more broadly generalize from the previous study and at the same time further verify the findings of that study constituted the primary purpose for conducting this replication. The population of the Michigan School for the Deaf was identified as a population which was sufficiently different from the Maryland population to make the repli- cation worthwhile. In contrast to the Maryland administra- tion, the Michigan administration has professed a much greater affinity for the oral method of instruction. 63 64 The primary department is maintained as a purely oral educational system. As part of the school curriculum, both upper and lower grades take a speech class, and considerable emphasis is placed upon the development of speech and speechreading skills. In the upper grades the Michigan School is much like many other residential schools throughout the country in that the administration makes a strong effort to develoP functional speech, but has not forbidden the use of signs and fingerspelling among staff and students. The superintendent stated, in private conver- sation, that ms.a school they adhere.pr$marily to the crab philos0phy of education. Consistent with their emphasis upon oral education, 106 of the 348 student population are listed as day students (American Annals p§.ppg 223;, 116, 1971, p.j173).. The fact that such a large proportion of the population do attend the school as day students suggests that they have more exposure to the hearing society than students in the Mary- land school. If the claims of oral pr0ponents are correct, this exposure should increase the speech and speechreading capabilities of the students. Procedures In conducting the replication, the same procedures were followed as outlined in Chapter 2 with the Maryland 65 study except for the following changes: (1) a different interpreter was used: (2) the Ss in each of the nine age-IQ groups was reduced by one: hence lid-36, not 45, and (3) alle students} were given additionally a set of questions to answer without having been presented any information. The additional set of measures were included as a control measure to determine if the students' scores on oral communi- cation differed significantly from scores obtained when they guessed the answers. The “guessing” questions were given at the end of each period on the first day of testing and the beginning of the third day of testing. With these exceptions the study was conducted in the same manner: the same factors and levels were included in the design: a stratified random sample was taken from the population: the same passages were presented in the same order using the same scheduling procedure specified in Chapter 2. Initially 45 Ss were included in the study, but due to various school activities the S attrition rate was high. In each of the nine age-IQ groups at least one 5 missed some testing. As a consequence, the S who missed the most testing in each of the nine groups was dropped from the study. In those few cases where a missing cell value was needed the cell mean was substituted and one degree of free- dom was sacrificed for each such substitution. 66 Sample Thirty-four of the original 45 Se sampled were deaf at birth: 14 were deaf as a result of genetic factors and 20 as a result of unknown factors. Six of the 83 were deafened later in life: four before the age of two years and two after the age of two. Five of the 88 did not have the cause of deafness specified in their files. Of the entire sample all but two students had been at the school for two years or more. Tables 22, 23, and 24 present statistical data defining the characteristics of the sample drawn from the Michigan population. All the means for these descriptive variables have been calculated on the original sample of 45 so that a more accurate comparison can be made with the Maryland sample. Because of the similarity in mean values of hearing loss for the nine age-IQ groups between the Michigan and maryland sample and inasmuch as it was previously determined that hearing loss was not confounded with age and IQ further analysis was not conducted: for the Michigan sample the assumption was made that hearing loss was not a confounding variable. w Testing of the Michigan Ss was conducted as far as possible in the same way maryland Ss had been tested. The mean length of time for each presentation of material under each method of instruction is presented in Table 25. 67 Table 22 Means for: Age: IQ: Hearing Loss: and Reading Level for the Sample Hearing Age-IQ Age in IQ Reading Groups Years Loss Level Group-1 12.25 78.80 92.00 1.92 Group-2 12.13 106.00 84.00 2.78 Group-3 12.38 116020 90060 2.54 Group-4 14.60 77.40 85.60 1.98 Group-5 15.05 101.80 93.20 3.54 Group-6 14.76 121.20 97.00 3.32 Group—7 218.02 82.20 94.80 2.74 Group-8 16 o 63 103 a 20 94 a 20 4 a 90 Group-9 16.88 122.00 91.00 4.68 Table 23 Means and Standard Deviations for Each Level of Age and IQ Age IQ A-L A-M A-H IQ-L IQ-m IQ-H Means 12.25 yrs. 14.80 yrs. 17.77 yrs. 79.4? 103.67 119.80 S.D. 3.13 mon. 2.76 mon. 8.09 mon. 5.82 4.67? 9.36 68 Table 24 Means and Standard Deviations for the Entire Sample on Hearing Loss, IQ, and Reading Level Hearing Loss IQ Reading Level Means 91.38 100.98 3.17 S.D. 9.33 18.95 1.57 Table 25 Mean Length of Time in Which Passages Were Presented for Methods of Communication Means 3.14 min. 3.58 min. 3.63 min. S.D. 27.53 sec. 42.27 sec. 28.37 sec. 3 Mean and 8.5. for reading were not computed because of variability between students in reading time. The mean number of correct responses and the standard deviations under each method of communication are presented in Table 26. Table 27 presents the grand means for the factors of age and IQ across all levels of communication. It can be observed that the mean number of correct responses under oral communication is conspicuously lower than the other three methods of communication. Likewise, the means for the older group (A-H) and loWer IQ group (IQeL) stand apart from the other two levels on this combined measure of communication skill. 69 Table 26 Mean Number of Correct Responses And Standard Deviations for Method of Communication Oral T.C. Manual Reading Means 17.33 23.00 23.72 24.86 S.D. 6.08 7.56 8.33 7.59 Table 27 Grand Mean Number of Correct Responsestfior Levels ' on Age and IQ Age qIQ. A-L A-M A—H IQ-L IQ-M IQ-H Means 20.56 19.71 26.42 16.94 25.58 24.17 main Analysis In order to test for the statistical significance of the differences between means and possible interaction effects the data were subjected to an analysis of variance. Table 28 presents the sources of variation, the degrees of freedom, the adjusted degrees of freedom, the sum of squares, the mean squares and the F-values. 70 Table 28 Summary Table for 3 x 3 x 4 Analysis of Variance Sources of Variation df SS df** MS F Method (M) 3 1,213.91 1 404.64 34.90* IQ (I) 2 2,064.29 2 1,032.15 10.26* Age (A) 2 1,280.04 2 640.02 6.36* A x I 4 101.79 4 25.45 .25 A x M 6 52.90 2 8.82 .76 I x M 6 200.65 2 33.44 2.88 M x A x I 12 293.10 4 24.43 2.11 SS: A X I 27 2,715.56 27 100.58 M x Ss: A x I 71 939-19 27 11-59 * Significant at the .05 level of significance **Adjusted degrees of freedom As indicated in Table 28 there was a significant effect due to method of communication, age and intelligence: how- ever, the method by intelligence interaction effect did not reach a statistical level of significance. The means for the low and high IQ groups are remarkably similar to those means obtained for the same groups in the Maryland study. The middle IQ group scored much better than either the low or high IQ groups. To determine if the middle group scored significantly better than the other two groups the scores of the low, middle and high IQ groups 71 under oral communication were subjected to a one-way analysis of variance. It can be observed in Table 29 that there was a significant effect attributable to IQ. To determine which means differed significantly, contrasts were constructed on the means for each of the IQ groups, and are reported in Table 30. Alpha was set at .05. Table 29 One-Way Analysis of Variance Table Comparing IQ Groups under Oral Communication —. Source of Variation df SS MS F IO 23 276.50 138.25 4.49* S: IQ 33 1,015.50 30.78 * Significant at the .05 level of significance. Table 30 Contrasts on the Factor of IQ Under Oral Communication Contrasts Difference Confidence Decision Between Means Interval it 4 it 20.92-14.17:6.75 1.17 to 12.33 Sig.Difference Eh - Xfi 20.92-16.92-u.00 -1.58 to 9.58 No Difference i ._ H ‘ xL 16.92-14.17=2.75 -2.83 to 8.33 No Difference Mean Number of Correct Responses 27 N b N _e 18 15 72 128‘”) (27.92) (27.17) A (24.67) Oral T.C. Manual Reading (18.26) (17.83)‘ ........ 117.50 ° MD (16.92) D" (14.17) Low M‘dcle Him intelligence Quotient Figure 3 Mean Number of Correct Responses for Low, Middle and High )0 Groups in the Michigan Study 73 As indicated in Table 30 the only significant difference between the three contrasts comparing IQ groups under oral communication was between the middle and low IQ groups. Although inspection of the means in Table 26 makes it quite apparent where the significant difference due to method is located, contrasts were constructed on the means for method of communication. The contrasts are presented in Table 31. It is apparent that the oral method of communi- cation differed significantly from total, and manual communi- cation and reading: however, reading, total and manual communication did not differ significantly from each other. Thus it was concluded that Se in this pOpulation of hear- ing impaired students assimilate significantly more in- formation through total communication, manual communication and reading that they do through oral communication. It was also concluded that there was no significant difference in the Ss ability to assimilate factual information under total communication, manual communication and reading. Contrasts were also constructed for the grand means on the factors of age and IQ. These contrasts are presented in Tables 32 and 33, respectively. .Alpha was set at the .05 level for all contrasts. 74 Contrasts on the Main Factor of Methods of Communication Contrasts Difference Confidence Decision Between Means Interval if 0 - i0 23.00-17.33=5.67 3.45 to 7.89 Sig.Difference in - Yb 23.72-17.33=6.39 4.17 to 8.61 Sig.Difference 2h - Yb 24.86-17.33=7.53 5.31 to 9.75 Sig.Difference Ifi - If C 24.86-23.00=1.86 -.36 to 4.08 No Difference ifi - it 24.86-23.72=1.14 -1.08 to 3.36 No Difference ifi - if c 23.72-23.00: .72 -1.50 to 2.90 No Difference Table 32 Contrasts on the Factor of Age Contrasts Difference Confidence Decision ii Between Means Interval ‘ifi - KM 26.42-19.71=6.71 2.12 to 11.30 Sig.Difference ifi - it 26.42-20.56z5.86 1.27 to 10.45 Sig.Difference it - ifi 20.56-19.71: .85 -3.74 to 5.44 No Difference 75 Table 33 Contrasts on the Factor of IQ Contrasts Difference Confidence Decision Between Interval ifi - it 24.17-16.94=7.23 2.64 to 11.82 Sig.Difference KM - Eh 25.58—24.17=1.42 -3.17 to 6.01 No Difference in - ii 25.58-16.94=8.65 4.06 to 13.24 Sig.Difference From the data in Table 32 it was concluded that the older Ss (Age-H) were able to assimilate significantly more information than the middle and lower age Ss. The lower age and middle age Ss did not differ significantly from each other. From the data in Table 33 it was concluded that higher IQ Ss and Middle IQ 33 were able to assimilate significant- ly more information than the lower IQ group. However, the rniddle and higher IQ groups did not differ significantly from each other. The over-all mean number of items answered correctly under oral communication in the Maryland study was 15.67. Inasmuch as there were 36 possible, if S answered all questions correctly, and inasmuch as each question was a three-foil multiple choice question it was theoretically ,possible in terms of probability for S to get a score of 12 (one third of 36) simply by guessing. If S was a 76 "good“ guesser the possibility existed for S to obtain a score slightly higher than 12. However, if the material presented to Ss was difficult a mean score somewhat lower that the chance level (12) would be eXpected. In an effort to determine whether Ss were performing significantly better than that expected if they were guess- ing, they were given two additional sets of questions to answer, but without any prior instruction dealing with the content of the questions. One set of questions (passage 1748) was given to all Ss at the completion of testing on day-1; the other set of questions (passage 17-D) were given to all Ss prior to testing on day-3. The means for the passages given orally and for guessing are presented in Table 34. Table 34 Means for Passages Presented Orally and for Questions on Which Subjects Guessed Passages 8 - Questions 10-Questions 17-B* 21-B 22-B 17-D* 19-D 24-D Means 3.33 3.61 4.33 4.97 4.78 4.56 ; Questions on whiEh Ss guessed To determine if the means for the oral condition and the means for guessing differed significantly from each other the three means based on eight questions were subjected to a 77 planned comparison analysis (Kirk, 1968; p. 76); the same procedure was carried out in analyzing the data from Table 34 based on ten questions. The results of these analyses are presented in Tables 35 and 36. Table 35 Analysis of Variance Table Comparing Means On Passages 17-B, 21-B and 22-B by Plannes Comparisons Source of Variation df SS MS F Between groups 2 15-89 - ' Guessing vs. Oral 1 10.08 10.08 4.25* Remaining Differences 1 5.81 5.81 among groups Within groups 105 248.78 2.37 i—§ig'-n-i'ficant at the .05’leve1 Table 36 Analysis of Variance Table Comparing Means 0n Passages 17-D, 19-D and 24-D by Planned Comparisons Source of Variation df SS MS F P* Between groups 2 3-13 7 ‘ Guessing vs. Oral 1 2.16 2.16 .43 NS Remaining Differences 1 1.07 1.07 among groups Within groups 105 * Alpha set at .05 78 From the results of the analysis in Table 35 it can be seen that the Ss did score significantly better under oral communication than they did on guessing for those passages based on eight questions. The results of the analysis on passages having ten questions (Table 36) reveals there was no significant differences between Ss' performance when guessing and when they had prior instruction through oral communication. Considering these data together, it appears there is only meager evidence indicating that Ss scored significant- ly better when they received oral instruction than when they guessed. 'If a Type II error has not been made, then the amount of information assimilated under oral communication is questionable. It may well be that many of the 88 did not assimilate anything through oral communication, or at best only a very small proportion of what was said. Summary 9: Michigan Study Based upon the results of the analysis and restricted to the p0pulation from which the sample was drawn certain generalizations are possible: 1. Hearing impaired children assimilate more factual information through reading, total communication and manual communication than they do through oral communication. 79 2. There was no significant difference between reading, total communication and manual communication in facilitating the assimilation of factual information. 3. The speech component in total communication does not significantly increase the amount of factual information assimilated. 4. Hearing impaired children score only slightly better through oral communication than they do through guessing. 5. All categorical sub-groups of hearing impaired children assimilated more information through total and manual communication than they did through oral communi- cation. ' 6. Se in the middle IQ group assimilated significant- ly more factual information through oral communication than the low IQ group; however, the middle and high IQ groups did differ significantly. CHAPTER V. SUMMARY AND DISCUSSION The polemic between proponents of oral and manual communication for deaf children has been an active' controversy for many years and it is not likely, consider- ing the emotionalism in the field, that it will be re- solved very soon. Obviously, the issue cannot be overlooked because it holds a central place in the philosophy of deaf education. Rather than trying to ignore the controversy or trying to make ubiquitous claims regarding what is best for all deaf children, this study attempted to focus upon two populations of hearing impaired children at two different residential schools, in order to discover the method of communication under which they can assimilate more factual information. A stratified random sample of 45 88 was drawn from the Maryland School for the Deaf. Ss ranged in age from 11.0 to 18.7 years and in IQ from 60 to 140. These Ss were presented factual information under four methods of communication: (1) oral communication, (2) total communi- catixun (3) manual communication, and (4) reading. The independent variables in the study were: (1) method of cummnunication, (2) age, and (3) intelligence. The dependent variable was the amount of information assimilated as measured on questions based on the factual information. 80 81 It was hypothesized that hearing impaired children would assimilate more factual information when it was presented to them through total communication, manual communication and reading than they would through oral communication. It was hypothesized that there would be an interaction between method of communication and intel- ligence: the lower IQ Ss assimilating more through manual communication (speech deleted) and the average and bright Ss assimilating more through total communication. It was also hypothesized that hearing impaired children would assimilate more information through reading than they would through oral, total, or manual communication. An experimental design was used to eliminate criticism directed against the use of ex post facto designs which employ matching techniques. A 3 x 3 x 4 factorial repeated measures design, fixed effects model was used. Ss were jpresented four passages of factual information under each of the four methods of communication: each S was compared 'to himself across the four methods of communication. The ‘use of a repeated measures design eliminated the need to match or randomly assign heterogeneous $3 to treatment groups. Four different passages were used with each of the four methods of communication: two of the four passages were at the 2nd grade level of difficulty, and the other two were at the 4th grade level. Equivalent passages 82 were randomly assigned for use with the four methods of communication. The same certified interpreter for the deaf was employed to present the material to all Ss under all methods of communication. Information was presented over a three day period, at three different periods each day. ”Time of day”, "day” and “order of presentation” were eliminated-as confounding variables through systematic scheduling. The results of the analysis of the data suggest the following conclusions: (1) Hearing impaired children assimilate more factual information through reading than they do through oral or total communication: (2) Hearing impaired children assimilate more factual information through total and manual communication than they do through oral communication: (3) All categorical sub-groups of hearing impaired children assimilate significantly inore information through total communication and manual communication than they do through oral communication: (4) The speech component in total communication does not significantly increase the amount of information assimi- lated over what is assimilated through pure-manual communi- catirmn (5) Bright, average and low functioning Ss do not differ in their ability to assimilate information under oral. communication: however, the average and bright Ss do 83 significantly better than the low functioning children under total communication, manual communication and reading. A replication of the maryland study was conducted at the Michigan School for the Deaf. Three changes were made: (1) a different interpreter was used: (2) the number of 58 was reduced from N=45 to 36; and (3) Ss were given two sets of questions on which to guess to see if guessing scores differed significantly from the mean score under oral communication. The results of this replication suggest the following conclusions: (1) Hearing impaired children assimilate more factual information through total communi- cation, manual communication, and reading than they do through oral communication: (2) There is no difference between total communication, manual communication and reading in facilitating assimilation of factual information: (3) Hearing impaired children score only slightly better through oral communication than they do through guessing; (4) No significant method by IQ interaction was detected; however, the trend in the data clearly supported the results of the Maryland study. The probability of observ- ing the obtained F-value for interaction when there was no effect due to interaction was .10. In general the results of the replication supported the findings of the Maryland study. 84 Discussion Maryland Study Proponents of total communication have argued strongly that systematic and continual use of total communication would result in greater academic gains for hearing impaired children than would the use of oral-only methods of instruc- tion. Proponents of oral-only instruction strongly disagree with this assertion. TheoretiCally, proponents of total communicatiOn claim that the sum of the information assimilated through various daily educational experiences across many years of school under total communication would be greater for a specified group of children than the sum of the information assimilated through the same daily educational eXperiences under oral-only instruction. In other words, if one could eXpose identical subjects to ”N” number of learning eXpe- riences over time, under total communication and oral communication, the sum of the information learned under total communication would be greater. Of course it is impossible to test this proposition directly. It is possible, however, to sample a single educational eXperience and evaluate what is learned and then send the same children through a nearly identical experience, but using a different method of communication and again evaluating what is learned. If the material used under both eXperiences was equivalent, obtained 85 statistical differences would,of course, reflect differ- ences due to method of instruction. Finally, if the experience sampled was representative of the majority of educational experiences which children go through through- out their education, then it would be possible to generalize the outcome of many such eXperiences, and a hypothesis could be constructed predicting which method would facilitate the acquisition of more information over time. Succinctly stated, it would be very desirable if the cummulative effects of utilization of a method could be predicted, relative to other methods, before children are subjected to that method for the duration of their education. In this study, an attempt was made to make a reality of the aforementioned theoretical experiment: a typical educational experience was contrived: typical educational material was used: the same children were repeatedly ‘ exposed to eXperiences which were equivalent, insofar as could be determined, except for the method of communica- tion used. The results of this study reveal that the quantity of information assimilated through total communication was significantly greater than that assimilated through oral communication. Based upon this study then, it is postulated that the sum of many such eXperiences under total communication would result in the assimilation of 86 more factual information over time than if oral communica- tion had been used. Assuming there is high correlation between academic achievement and the»cumulation of factual information, it is further postulated that at the termi- nation of their school eXperience, students from the Mary- land school will manifest significantly greater academic achievement if instruction is provided through total communication in a systematic and continual fashion, than if instruction were provided through an oral-only approach. Time and space have been taken to deve10p this argument because the conclusions drawn, regarding long term educaa tional achievement, are the same as those reached by other researchers using ex post facto designs. However, in one way the conclusion drawn regarding the efficacy of total communication relative to oral communication is somewhat more defensible because an experimental design was used. An experimental design allows for a oausational conclusion to be drawn: in this instance it can be said that the addition of signs and fingerspelling increased the assimilation of factual information beyond that assimilated through oral communication. Thus, there is evidence both from experimental and ex post facto research that hearing impaired children assimilate more information through methods which employ some kind of manual component. 87 One of the primary purposes in conducting this study was to determine if teachers who fail to use speech in communicating with their students while fingerspelling and using signs deprive those students of information which they might otherwise have received. In other words, does the inclusion of speech in addition to signs and fingerspelling make possible the learning of more infor- mation than if the manual components had been used alone? To educators concerned with academic achievement of hear- ing impaired children this is a crucial issue. Outstanding proponents of total communication have continually asserted that the use of speech in total communication does indeed communicate to the child more of what is said. The mean number of correct responses under total communication was actually less than under manual communi- cation, although not significantly lower. The finding- of no statistical differenceebetWeen.total and manuall communication, and particularly the fact that total communication was lower than manual communication was contrary to what was predicted by the researcher and administrators and teachers at the maryland School. How can such findings be interpreted? Two kinds of interpretation can be rendered: (1) a conservative inter- pretation which would ignore the direction of the results and focus completely upon a statistical interpretation, 88 or (2) a liberal interpretation which would consider the direction of the results and then speculate about having made a Type II error: that is, having failed to reject the null hypothesis of no difference between manual and total communication. The only defensible interpretation is, of course, the conservative one. The power of the F-test was .99 which means that there was only one chance out of 100 of failing to detect differences in the population based on differences comparable in size'tofthoserobserved in the sample. . Taking the conservative approach, it has been con- cluded that there is no difference between total and manual communication and that the observed difference was a product of chance. Future research might produce results slightly favoring total instead of manual communication. The conservative approach precipitates a single con- clusion: the speech component in total communication does not facilitate the assimilation of factual infor- mation. If the speech component does not give students additional information beyond that provided through the manual component then it is possible that (1) students do not attend to speech when information is presented through total communication, or (2) the students do attend to speech, but what they receive from this com- ponent is purely redundant information. 89 Another purpose of this study was to determine which method of instruction, --oral, total or manual communi- cation,-- is best suited as a method of communication for different categorical sub-groups of hearing impaired children. As indicated in Chapter 2, nine groups of hear- ing impaired children were sampled (three age levels crossed with three IQ levels). All nine groups of chil- dren did significantly better under total communication and manual communication when compared to the oral method. This ubiquitous observation might well be viewed as a positive outcome of this study, particularly from an administrative point of view. Certainly, knowledge that children in the population learn best under a single mode ‘ of communication makes the administrative task of develop- ing an instructional orientation easier. One of the major findings in the study was the inter- action effect between method of instruction and intelli- gence. As described in Chapter 3 the interaction was ordinal and thus generalizations made in the preceeding paragraph still hold. The nature of the interaction was such that no difference existed between the lower function- ing students (IQ-L), the average students (IQ-M) and the bright students (IQ-H) in their ability to assimilate information under the oral method of instruction: however, under total communication, manual communication and reading the average and bright students did significantly better 90 than their lower functioning peers. This finding indicates that instructing average and higher functioning students using a pure oral method results in a partial negation of their innate intellectual endowment. Stated differently, the average and bright students in this population can function no better than their lower IQ peers when the oral method of instruction is employed. Such a finding should be given major consideration by administrators and teachers serving hearing impaired chil- dren: particularly is this true for those serving children in residential schools because these are the populations to which generalizations may most legitimately be made. Nevertheless, these findings also make it incumbent upon oral educators in the various types of day and residential programs to begin immediately exploring and examining the assimilative abilities of their students when exposed to oral and combined methods of communication. Obviously, oral.educators true to their philosophy will resist pro- viding their students with manual skills. Respecting the concern of oral educators not to contaminate their students by manual methods it is recommended that they systemati- callyq and empirically evaluate the ability of low, middle and high IQ students to assimilate information through spe echreading . No significant difference was found between the stu- dents' ability to assimilate factual information through 91 reading and through manual communication. Likewise, there was no significant difference between the students' ability to assimilate information through manual and total communi- cation: however, reading did differ significantly from total communication in facilitating the assimilation of information. It was concluded that students can assimilate signif- icantly more information through reading than total communication. Regarding the nonsignificant differences between reading and manual communication and manual communication and total communication no conclusion was drawn: rather, it was decided that the best strategy would be to wait and see if additional information from the Michigan study would elucidate these findings. The research findings in the Michigan study supported the findings of the Maryland study. Ss assimilated signif- icantly more information through reading, manual communi- cation and total communication than they did through oral communication. Although no significant method by IQ interaction was detected the results were clearly in the direction of the results observed in the maryland study (interaction was significant at the .10 level). The difference between the means of the low IQ groups in the Maryland and Michi— gan study was .70 in favor of the Maryland Ss. The differ- ence between the means of the high IQ groups in both 92 studies was .92 in favor of the Michigan group. The major difference was in the middle IQ groups: the Michi- gan group scored 5.32 points better than the Maryland group. When IQ groups at the Michigan school were com- pared the middle IQ group assimilated significantly more information than the low IQ group. The SS in the Michigan study did not differ signifi- cantly in their ability to assimilate information under total communication, manual communication and reading. These findings suggest that there is no difference be- tween the amount of information assimilated under these three modes of communication. The latter generalization may not hold, however, when the kind of information being assimilated is significantly altered. For example, if the style of the language is altered, or the complexity of the factual material is increased then differences may be detected between reading, manual communication and total communication. Another restraint upon the above mentioned generalization is that of "rate of presentation". Differences might also be detected if the rate of presentation were significantly increased. It was also found that the mean number of correct responses for two of the four passages presented through oral communication did not differ significantly from the means obtained through simply guessing. This does not necessarily mean that the Ss received no information from oral communication: however, if they did receive a 93 significant amount of information the materials used in assessing information assimilated lacked sufficient pre- cision to detect it. mgjg§,Conclusions In considering the Maryland study and the Michigan study jointly several points can be made. First, it was found that for reading, manual communication, total communication and oral communication the trends were the same: that is,in both studies the means for reading were greater than the means for manual communication and the means for manual communication were greater than the means for total communication. In neither study did reading differ significantly from manual communication. Likewise, in neither study did manual communication differ significantly from total communication. In only the Mary- land study was the mean for reading significantly greater than the mean for total communication. In both studies the means for reading, manual and total communication were significantly greater than the means for oral communication. Based upon these finding the following conclusions were drawn: 1. Hearing impaired children can assimilate more factual information through reading, and methods which employ signs and fingerspelling than they can through oral communication alone. 94 2. The speech component in total communication does not significantly contribute to the assimilation of factual information: the amount of information assimilated in total communication is primarily a function of the effect of signs and fingerspelling. One should resist the temptation to conclude that there is no difference between students' ability to assimi- late information through reading, manual communication and total communication because no statistical differences were found, except for the one between reading and total communication in the Maryland study. Significant differ- ences, among other things, are a function of sample size and precision of the instrument used in assessing the dependent variable. Because the rank order of the means 'for reading, manual communication, and total communication was the same for both studies it is postulated that by increasing the sample size or the precision of the measure- ment tool, in future replications, that the relatively small differences observed in these studies would be expanded and the amount of information assimilated through reading, manual communication, and total communication would be found to be significantly different. Based upon this postulate three additional tentative conclusions have been drawn: 3. Hearing impaired children can assimilate more infor- mation through reading than they can through manual or total c0mmunication. 95 4. Hearing impaired children can assimilate more information through manual communication than they can through total communication. 5. The interaction of speech with signs and finger- spelling reduces the amount of information which can be assimilated through total communication as compared with manual communication. Regarding the interaction of IQ and method of communi- cation it was found in the Maryland study that the three IQ groups (low, middle and high) did not differ signifi- cantly from each other in their ability to assimilate factual information under oral communication, yet they did differ significantly in their ability to assimilate factual information under total communication, manual communication and reading. The trend in the Michigan study was the same, except for the middle IQ group which did significantly better than the low IQ group under oral communication. Such a finding would not be difficult to interpret if the high IQ group had also scored significantly better than the low IQ group, but no such difference was detected: in fact, the high IQ groups' performance was closer to the low IQ group than it was the middle IQ group. It is possible that in the long run, after many replications with different groups of randomly selected Ss, that the middle IQ group would fall to the level of the other two IQ groups: however, it is also 96 possible that the high IQ group would score better and more closely approximate the middle IQ group. If the latter situation proved to be the case then it would suggest that both middle and high IQ groups are benefit- ing more from the oral environment than their lower IQ peers. Although it seems unlikely, it may be that there is something unique to the middle IQ group which allowed them to speechread better than their higher IQ peers. Such speculation is interesting: however, there is really no way to completely explain this finding without further research with the same or similar residential populations. One broad conclusion can be drawn based upon the results of both studies. 6. Bright hearing impaired children are unable to take full advantage of their innate intellectual endow- ment in assimilating factual information when that infor- mation is presented to them through oral communication. Implications ggd_Limitations A major implication of this study relates to linguis- tic development. In order for children to learn language they must have sufficient eXposure to the target language so- they can acquire the linguistic rules necessary for generating meaningful and culturally acceptable communi- cation. It is obvious that for a deaf child, simple EKposure to oral communication is grossly inadequate. For'the deaf child, exposure must mean a qualitthuantity 97 input. Insufficient exposure to the target language (a lack of quantity) and insufficient exposure to syntactic patterns (a lack of quality) leave a deaf child with a very limited means of acquiring the rules by which lan- guage is generated. In stark contrast to speechreading, it appears that the employment of manual methods--in this study, signed English-~does provide deaf students with a quality-quantity input. This is a reasonable conclusion inasmuch as lan- guage was the medium through which the factual information was transferred from interpreter to students. Further- more, because language learning is the major social- educational handicap of the deaf child, the findings of this study that hearing impaired children assimilate more information--implicitly, more language--through the use of signs and fingerspelling should be given considerable attention. In conclusion, it must be remembered that the major goal or objective of most educators of deaf‘children is twofold: (1) the development of adequate language skills, so that (2) as adults their deaf students will be able to function effectively in society. The development of adequate language skills implies the development of speech skills and the rules of language. Of the latter two skills, knowing and using correctly the rules of our lan- guage is by far the most important. many adult deaf 98 function very effectively in society without speech, but for those deaf adults who have not acquired the rules of language lifecan be miserable. Language, not speech, is required to read and write. Without the ability to read and write a very low ceiling is immediately imposed upon deaf adults vocationally. The kind of factual information presented to Ss in the study might be viewed by some people as a limitation in the study inasmuch as that information is not typical of all types of information presented to students in the course of their education. However, it was felt that the material used was as representative as any material that could have been selected. Furthermore, the communi- cative process was being investigated not the material itself. The use of different kinds of material should not significantly alter any of the conclusions which have been drawn. The precision of the instrument used in assessing the amount of information assimilated is a matter of more legitimate concern. Although the questions used were sensitive enough to detect some differences, a larger set 0f questions with more than three-foil items per question would probably have detected other differences such as those mentioned above regarding reading, manual communi- cation and total communication. 99 Suggestions for Future Research Based upon the results of this study, and the design employed,several possible extensions can be suggested: It is preposed that a study be conducted including "rate of presentation" as a major independent variable. It is important for educators to understand something of the interaction effect between rate of presentation and method of instruction. It is proposed that a study be conducted which focuses upon the difficulty of material and type of infor- mation as major independent variables. It is important for educators to understand the constraints upon reading, total communication and oral communication imposed by the nature of the material being presented. It is prOposed that a study be conducted comparing the effects of the Rochester method with total communi- cation. The time has come to begin focusing upon various combined methods so that the parameters surrounding them can be delineated. (It should be noted that manual communication is not being suggested as a tentative educational method because it ignores Speech and speech remains an important goal of all educators.) It is proposed that a study be conducted which com- pares the ability of students to assimilate information from teachers with varying degrees of experience and 100 training with manual skills. In addition, the ability of deaf and hearing teachers to communicate with deaf children should be studied. Answers to the following questions are presently unavailable: (1) What level of manual skill is necessary to achieve an acceptable level of information transfer? (2) Are deaf teachers any more adroit in teaching deaf children than hearing teachers?~ Finally, it is proposed that oral educators begin to critically and objectively assess the amount of information their students are able to receive through oral communi- cation. Suoh evaluation is possible without violation of their educational philoSOphy. Obviously, if the results of such research reveal that the amount being assimilated is relatively small compared to what was presented, then it is time for a reassessment of the philosophy of educa- tion which advocates using a system of communication which fails to communicate. LIST OF REFERENCES LIST OF REFERENCES American Annals of the Deaf. 116, 1971, 172-173. Boning, R. A. Gettin the Facts: Specific Skill Series, Book A. Rockville Centre, New York: Barnell Loft, Ltd., 1966. Bruce, W. T. Book Notes and Reviews. Volta Review, 1969. 71. 523- Denton, D. N. Educational Crises. The maryland Bulletin, 17, 1971, 3. Fellendorf, G. W. Face Up or Cop-Out? Volta Review, 1970, 71, 406-407. Gaeth, J. G. Verbal and Nonverbal Learning in Children Including Those with Hearing Losses. Part II. Final Report Cooperative Research Project No. 2207, Wayne State University, 1966. Johnson, E. H. The Ability of Pupils in a School for the Deaf To Understand Various Methods of Communi- cagiona II. American Annals of the Deaf, 93, 1948, 25 -31 a f Kerlinger, F. N. Foundatigns of_§ehavioral Research. Holt, Rinehart and Winston, Inc., 1967. Kirk, R. E. E erimental Desi : Procedures for the Behavioral Sciences. Belmont, California: Wadsworth Publishing Co., 1968. Meadows, K. P. Early manual Communication in Relation to the Deaf Child's Intellectual, Social, and Communication Functioning. American Annals Q; Miller, J. Oralism. Volta Review, 1970, 72, 211. 101 102 Quigley, S. P. The Influence of Fingerspelling on the Development of Language, Communication, and Educational Achievement in Deaf Children. Washing- ton, D.C. Department of Health, Education and Welfare 1967, 99. Quigley, S. P. & Frisina, R. D. Institutionalization and Psycho-Educational Development of Deaf Children. CEC Research Monograph, Series A, No. 3, 1961. Stevenson, E. A study of the Educational Achievement of Deaf Children of Deaf Parents. California Newg, 1964, 143. Stokoe, W. C., Jr. CAL Conference on Si Language. The Linguistic Reporter, 12, 1970, -7. Stuckless, E. R. & Birch, J. W. The Influence of Early manual Communication on the Linguistic Development of Deaf Children. American Anngls of the Deaf. 111, 1966, 452-462. Vernon, M. & Koh, S. D. Effects of Oral Preschool Com- pared to Early manual Communication on Education and Communication in Deaf Children. American Annals of the Deaf, 116, 1971, 569.574. Vernon, M. & Koh, S. D. Early manual Communication and Deaf Children's Achievement. American Annal§ of the Deaf, 115, 1970, 527-536. White, A. H. A Comparative Study of the Articulation of Consonant Phonemes Between Day School and Residential School Deaf Students. Master's Thesis. Salt Lake City, University of Utah, 1969. Winer, B. J. Statistical Princi les in E erimental Desigg. New York: McGraw-Hill, 1962. APPENDICES APPENDIX A Total Communication Passage 20-B.......Mr. Four Eyes Passage 24-B.......The smalleSt Pets Passage 21-D.......Voices from the Ocean Floor Passage 22-D.......Mother Nature's Actors Code a a signed word a a fingerspelled word ( ) a a word that was both signed and fingerspelled a word inserted by interpreter UNIT NO. 20 - MR. FOUR (EYES) Most animals can (get along) with twg eyes. With two eyes they can find their food. With two eyes they can keep from becoming food for some other animal. Fgr the Anableps two eyes are not enough. This (fish) has four eyes. Each 2;,its_eyes lg really two eyes. The (upper) two eyes can see above the water. These eyes are high i its head. With them the Anableps can see bits 2; food that might (float) along the (top) 93 the water. The (bottom) eyes also he 2. With them the Anab- le s can see what lg (going on) below the water. I; Iv g_1arge fish comes near, the Anableps can see him and swim (away). Most often the Anableps gg about six inches (long). Sometimes i3_grows somewhat longer. I; lives ig the fresh water of g_far-away land. Much g: the time the W E on .1313 (lookout) 22;; (bugs). With §_9_ many eyes gg see with, it's the bugs who have :2 look out! Would you like $2 catch such g_fish? You would Laxgtgggguonswaflfmma Egalhaemm mmw- Apeolemmamgmme mmgiemaex- "ggaammmbefore REARbgigggygycomingl” 103 UNIT NO. 24 - THE SMALLEST (PETS) Some people have (pets) that are very small. They re so small that you have 19 take _a_ good look _t_g see them. (Hundreds) p_f_‘ them live ip just one (glass) house. gamsyassmtgsfa? gmmgm goldfish? 11:9, ”cg 533 pg}; goldfish. g m giyg pp? Theygpglittle §._n_t_s_. flflygp‘ymm. greasgmiaaesmmreb gfigm 8- little honey every other day. _A_ few drops p_f_' water run st 333 put .12 their glass cage from time _t_p time. ‘— m must pg kept out p__f the sun. Why don't you get some ant pets pf your own? “\uraalaueiaseisaasiau- 12222215245.“ ht ”\‘b _t_o_ your garden and find g an} hill. = & (queen) ant. She pg bigger than the others. You may had; :03 gig down deep 1:__o_ find her. Put the Queen ant in a glass cage with some (dirt). Then put in lots of the other ants. The other ants are M workers. 59.6.2 192E m .°_n.. Alp—“E 3.1.9.! Let—3' _lThe _Will soon g _t_p 2531. Each worker ant has _a_ job 2 (_12, g E gun :22 ass. flail. rage Leia: 21a ages-ri- _Thex work 1%); hard. They 93 not stop until they have made 3; 104 UNIT NO. 21 - (VOICES) from the (0CEAN)(FLOOR) d’ 23' (D For g_long time people thought that the (floor) (ocean) was 3 very (quiet) place. Some thought that I)” l° c+ H) Hi I) almost without any (sound). (Scientists) were sugprised when they put g_microphone under water and heard all kinds p£_noises. Scientists heard grunts, growls, snarls, and whistles. The bottom of the sea sounded like a zoo. Scientists were puzzled when they found out that these noises were coming from fish. Fish make these noises py (grinding) their teeth. Others scrape their bones together. Still others make the meaning of some sounds. Tests show that he is one of the smartest animals ip the sea. Peeple say that they have been able pp_teach the dolphin many human words. A: even been taught pg sing. They (rise) 32 the (surface) and sing for visitors. 53 least i} sounds somewhat like singing! 105 106 Who knows? Someday scientists may pg able pg talk with the dolphin and learn many secrets of the sea. ~— learn the location of sunken treasure ships. We could learn where pg locate great schools pf fish. W might even learn the location of enemy submarines. It could bea athrilling moment when pg first begin _*“_*”*__ UNIT NO. 22 - MOTHER (NATURE'S)(ACTORS) Have you ever seen Mother Nature's Actors (perform)? _I_f_ you have, you know that animals often put pp a_ show that i_s_ better than anything pp television. These animal actors sometimes protect themselves _by pretending _p ‘93 something they are not. Take the opossum. Hg _i_s_ slow and 5 poor fighter. Bpp 3119:; gt: (enemy) pp p125; L: pretends 3:223 g_ggg. 31:3 gpgpg pg phi ground, without moving 3 muscle. _Wh_e_r; .t_h=e my leaves, £_—h_3_ Opossum g_e_t_s pp gpg gpgg M p_.’_1__s_ business. Other animals learn _t_p act for their dinner. A pair pf (foxes) will sometimes put p_p g strange show 32 catch _woodchuck. One fox walks in front of the woodchuck' 8 —-———~—. g — 5 den while the other hides nearby. The first fox acts 5; if he has gone (crazy). He leaps into the air and chases Nature pgp g_i_v_ep w animals 2 disguise pp h_e_l_p 131% £13; 2313; animals. 11213 g_e_a anemone _lp_g_1g_s_ likp 5 beautiful flower. Bu_t LE lovely ”petals" 2.3g actually (stingers). weanaafaieimeeaf. g'L—etals'eeregafe and (stun) the small fish. Then these same petals shove 108 Perhaps the strangest disguise ig that worn py the vine snake. fig ig long and slim. fig (twists) around 9 H go 0' gg that pg looks just like a vine. Lizards who climb on this “vine” find to their sorrow that they have been fooled py another one 2; Mother Nature's Actors. APPENDIX B manuathommuniCatiOnr Passage 18-B Names for Indians Passage 23-B Little Man in a Pie Passage 23-D Here Comes the Army Passage 25-D The Secret of a Snake Charmer ngg = a signed word = a fingerspelled word ( ) = a word that was both signed and fingerspelled ' ' 2 word inserted by interpreter UNIT No. 18 - (NAMES) FOR (INDIANS) Egg would you like 32 have (just) g,(first)(name)? (Indians) had only first names. They never had last names. Some of their first names were made pp 2; two pp *_——*—_ three words, like Red Cloud, pp (Little)(Bird) Wing. Indians didn't give family names like gg‘gp tod y. IE was g_happy day when pp Indian baby was named. many Indians came to see the new baby. Some 2; them gave long (talks). Everyone ate lots pf food. Some- times they named the baby for something that the father did. Sometimes the name came from something the father saw i3 3 (dream). Indian children were often (given) nicknames. Some pg these nicknames were vepy (funny), like (Flat Head), Big Teeth, and fig Nose. Indian children had pp keep their nicknames until they did g_great deed. (Then they they would get new names that told about what they did. Indians changed their names often. When they got (became) sick, they liked to take another name. When things werenot going well, they would chapge. They thought the new name might chapge their luck. When 5p Indian (died), people ip his family would not say his name (anymore). When they talked about the dead erson, the Indians would say ”my father,"gg 'your friend.“ The name died with the Indian. UNIT NO. 23 - LITTLE N N A (PIE) The (pie) was placed pp the table before the (Queen). I3 was time for the cook jg out the pie 1p pieces. Out jumped g little man. (Up and down) the table pg ran. Ila :3‘ (0 Queen liked the gok . She asked 32 meet the (tiny) man. She wanted to know his name and more about him. Jeff Hudson was the name of the tiny man. Hg was (a foot and a half)(high)'talr. Jeff was the smallest man the Queen had ever seen. The Queen liked Jeff. She called him Sir Jeffrey. She et Sir Jeffrey stay ;p the palace. Life was not easy for Jeff. One time Jeff (fell) gp d' he (wash)(bowl). Jeff could not (swim). ‘flg yelled for el . I; was lucky for Jeff that a man came py and I: pulled him out. started pg (peck) g3 him. Around the r he n. Little If: Jeff took many fast (steps), but the turkey was not far behind. Again Jeff was lucky. ‘é man (came along) and (drove)'chased‘the turkey (off). lik to feed the kitten. The little (D Jeffrey liked joke . One time pg dressed Ilé g (kitten). A lady started kitten looked at her and said, “I can help myself when I,gp hunggy.' The lady almost (fellover). The ”kitten“ (ran out) of the room. 110 UNIT NO. 23 - HERE COMES THE (ARMY) 3‘ ‘22 you know which (animal) ith e most (feared) g2, all pf (Africa)? The (elephant)? 0-3 he (lion)? Egg (gorilla)? If you chose any pf these, you would pg_wrong. The most feared animal pp the continent ;g_much (less) than gp_inch (long). Yet, every animal ig the jungle runs when he is on the (march). He is the army ant. Millions pf these ants (form) an army that is (hundreds) pf ygrds long and many ygrds wide. Army ants eat nothing but (meat). flp_animal lg too big pp too small for them. Even the elgphant begins pg_run when that pg must (move) fast. gp minutes these little insects can (strip his flesh), leaving only bones behind. the ants have (scouts). These are the ants who run out ip_front pf the army and leave g trail for the others 39 follow. This trail is a little liguid which the (scouts) carry ip_their bodies. The army follows this trail. Army ants are almost (blind) and need g (path) pg fgllow. Both men and animals alike can tell when the army ants are coming. The army (drags) dead meat along. The (smell) (carries)(a long distance). When the natives catch this smell ip_the gig, they know ip_ig time pp_gg_into (action). llI 112 Thg,patives (release) all their animals. Then they hide 1p,the (brush) until the army ants (pass by). Even the (bravest) man won't stay around when hg, learng that the army ants are coming! UNIT NO. 25 - THE (SECRET) QE A (SNAKE)(CHARMER) 921; mu. (charm) gg_§é_§_9__lx (snake) 51; 31121.1; will n91 2122mm? musics. ummgngmggng (Indian) M (charmer). Am chamgpggkgflgg large. (cobra) ing. (basket). H; 5.113 scan in firm: nigh; (basket) gm plays % (flute). AS pg plays, pg (sways) _ip (time) 32.92 g 1mm. 3.1.2111 _mg snake (rises) from g (basket), & (swaying)(from side to side). Scmsnsmgsaxmmmunflmmmancing) 'ngm. wmgmmggmear) (sounds). _Thg gay that _Shé m (moves)(back and forth) gmgmmgmmmnmum side to side). Marmammghgmummmhmm _phg gngkg Mg; 1135 (taken out) pg (fangs). In 59mg gaggmrishl- Inmenxmhsrggg. hm. gig (cobra) mmgmmmmgm. gmmmszmmsflmasmakaswamflmmg mggwmmw. gnaw Lbs; house (searching for) snaksa- linen h_g finds one. g_a ifi'naidasmallggg- Thammihcsauhnmmgg mmmmgmmmmm Diana. w has some strapge (power) over snakes. Many (visitors) 1'13 114 agree. All will (admit) that pg matter how the snake charmer (performs) his feat, he must be (brave). fig one can be sure how snakes are charmed. This is the secret of the snake charmer. It's a lot safer to talk about this secret than it is tot ry to charm snake. Don' t Hm ygg,agree? MICHIGAN STATE UNIV. LIBRRRIES IIIJIIIHIIH!m1HIHmlmullIIWI‘IUINWIIIIHIIIIIHI 31293102085143