CGNSYRUCTIQN AND E‘i’ALUA‘s’iON 05' AN EIGHT MiLLSME‘e’Ea F§LMED LiPBEfiaDéNG TEST W.) far the Bayes 0f M. A. MICHIGAN STATE UN?‘.’ERS!TY ferry L Katt 1.967 HHHHH nmmnuuzmguflmlWW ‘ 3 Michigan St.- cc University C .1 v. f. V. _. A r .. k ._ - , ax p; VLF W“ “U . k .w u m. [C r1 A~V ol 0‘, K mmk ..L V. ‘ o. ‘ dd ll QM . ¢ .1 C‘ . r1 ad \ “\h - 1r“ .7 * #9.. .. A a; C.“ ..i. W. \ U... n Rh n1 .1. 30 ._ u E m V at .7 e 0 e rL. 5 n1. 2 9 fix v. S C V. 3 ‘ o ._...I. .F. SH Qt. u C; ‘T; erH .9», win g . ‘ v rG Ci. v‘. Q; * . nd 7; n V‘ ;. .,.. :1 T. M... rH: .3 L» .7 .a. .. ‘ .. .. :7 PC ..C :A' l. r u .w._ _.. . 5 . 7 5C a u a; Du f1 \ . AESTRAOT CONSTRUCTION AND EVALUATION OF AN EIGHT MILLIMETER FILMED LIPREADING TEST by Terry L. Katt The purpose of this study was to construct a test of lipreading which could be used as a reliable measuring in— strument. A second purpose of this study was to evaluate the test of lipreading to determine if there was a relationship between lipreading performance in a live situation and lip— reading performance using eight millimeter black and white film. Three hundred familiar sentences were administered to fifty junior high school students as the first phase of the research. Two speakers alternately presented these sentences unvoiced to the students who recorded what they had seen. An item analysis was done providing item difficulty data. Thirty-two sentences were selected representing the entire range of difficulty. These thirty—two sentences were divided into sixteen-sentence forms (A and B), each representing the entire range of difficulty. These thirty—two sentences were visually presented to twenty—four college students in the second phase of the research. fonm, A ant ha“: reliak fthe stud other was e jects were ssh speake 01‘ filmEd s Presentatio Criterion S FETISOn of made, The f Bertenn Terry L. Katt These subjects viewed two speakers who alternately presented the lipreading sentences unvoiced from each sixteen-sentence form, A and B. From the scores of these subjects the split- half reliability of the tests was determined. Fifty—seven college students viewed two different presentations of the lipreading test in the third phase of the study. One mode of presentation was live and the\ other was eight millimeter black and white film. The sub— jects were instructed to write down what they determined each speaker had said when they viewed either the live or filmed sentence. After the subjects viewed the two presentations, the answer sheets were collected and the criterion scores determined. From these scores, a com- parison of the effectiveness of the two presentations was made. The following conclusions were reached: A thirty—two sentence test of lipreading was constructed which could be used as a reliable measuring instrument. A modest relation— ship was found to exist between a live and an eight milli- meter black and white filmed presentation of lipreading material. CONSTRUCTION AND EVALUATION OF AN EIGHT NILLIMETER FILMED LIPREADING TEST BY Terry L. Katt A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Speech 1967 TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES LIST OF APPENDICES Chapter I. STATEMENT OF THE PROBLEM Introduction Statement of Problem and Purpose of Study Questions Posed Importance of the Study Definition of Terms Organization of the Thesis II. REVIEW OF THE LITERATURE Eight Millimeter Film Use of Film in Lipreading Lipreading Tests III. TEST CONSTRUCTION Procedure Test Administration Results V. SUMMARY, CONSLUSIONS, AND RECOMMENDATIONS Summary Conclusions Recommendations for Further Research ii Page iv vi 23 143 ‘.' 5C]. (T; l. I \ Comparisor of DEVISCIDLS. Comparisoh of “(’1 LIST e a {”1 TOta 1 Scores and and Mean iv Scores FIT H- H"; ‘LWJ LU s ,— k 4 Lu *1 Cl "I: fig }—1 C3 C1 SO [:1 (D Item Analysis of Sentence Difficulty, Showing the jumper of Sentences Correctly Identified by Various Proportions of the Group . . . Item Analysis of Sentence Difficulty Showing the iumcer of Sentences Correctly Identified my Various Proportions of the Group (Form 5) Item Analysis of Sentence Difficulty Showing the Number of Sentences Correctly Identified by Various Proportions of the Group (Form A) A Drawing of the Speech and Hearing Science .Latoratory on the Day of the Test. . . 31 .:\ 37 LIST OF APPENDICES Oral Instructions . Lipreading Tests 'Form A tn Form Instructions and Answer Sheet Individual Percentage Score Obtained on Forms A and B of the Test (Live Presen— tation) Individual Percentage Score Obtained on Forms A and B of the Filmed and Live Test of Lipreading . . . . . vi KN J: UT \P \ \fi FD \\ p (I) K, II CHAPTER I STATEMENT OF THE PROBLEM Introduction T in the past several years, lipreading has been inter— . . l . . woven With the education of the deaf. As training pro- grams were established for the deaf, lipreading began to be f *— -aught.“ It would appear from the literature reviewed that some persons involved or interested in the rehabilitation of the deaf and hard of hearing are seeking ways and means by which to improve and expand this knowledge. O’Neill and Oyer3 state that through the years thousands of people have profited by training in lipreading, but much remains to t if D discovered about lipreading itself. While reviewing the literature, it can be noted that some attention has been given to the use of motion picture film. Some well—known lipreading tests and lesson materials ale filmed, two prominent ones being the Mason's films, and J. O'Neill and H. Oyer, Visual Comiunication (New ;ersey: Prentice Hall, Inc., 1900), p.9. LU IH CT H 3.2.; \A ’U H KO 2 the Morkovin and Moore5 film series. The majority of these tests and lesson materials were filmed on sixteen mil- limeter motion picture film. One problem involving the use of film is the high cost.6 Included as a part of this cost are the basic cost of the film, the cost of the standard film equipment, the cost of the developing process, and the cost of professional 7 Many lipreading tests, lessons, and programs photography. appear to have been filmed with little thought to economy. However, economy itself can become feasible if eight mil- limeter motion picture film is used, although there is little precedent for testing its accuracy for lipreading purposes. Morkovin and Moore advocated the use of film in lip- reading. Their film series, The Morkovin and Moore Life_ Situation Motion Picture Series,8 consists of a series of life—like situations which are interesting to both children and adults. This series was filmed on sixteen millimeter film. There have been many studies, lessons, and projects that have used sixteen millimeter film, in addition to the two mentioned above. 5Boris Morkovin and Lucelia Moore, ”A Contextual— Synthetic Approach for Speech Reading,” Life Situation Motion Pictures. (Mimeographed Manual, 19A9). 6Sid Norinsky, The Amateur Eight Millimeter Movie Guide (New York: Universal Photo Books, 1950), pp. 1—5. 7Ibid., p. l. 8 Morkovin and Moore, 0 . cit. The general use of sixteen millimeter film brings into focus the problem of this study. It is generally assumed that sixteen millimeter film is the best film for con— ducting research and presenting lessons on film. For instance, 9 Postove not only assumed that sixteen millimeter film was superior, but that color film was superior to black and white because she thought that color photography more nearly simulated the actual task of a lipreader. There was no statistical evidence offered in the study to support this point of view. There is some precedent for the use of eight mil- imeter motion picture film in a testing situation,and these precedents will be discussed in the second chapter. The present status of eight millimeter film is improving be- cause the motion picture industry itself is concerned with the future of this film. Recent photography journals con— tain imformation concerning the development of this film. The motion picture industry, for example, is spending large sums of money to develop and perfect eight millimeter filmlo in hopes that it will perform as well as sixteen millimeter film. The primary reason for promoting eight millimeter film is to make it more appealing to amateur photographers. 9M. Postove, "Selection of Items for a Speechreading Test by Means of a Scalogram,” Journal of Speech and Hearing Disorders, XXVII (1962), pp. 71-75. 10M. Willard, The Complete Photographer, An Encyclo- pedia of Photography (New York: National Educator Alliance, Inc , IQAQ), p 737. Amateur photography is becoming more popular, thus creating an increasing market for photographic supplies.ll The eight millimeter film has grown so much in popularity that it is how gaining the title of the "amateur standard.”12 This research and development within the film in— dustry could make feasible filmed tests, lessons, etc., that were once too costly to produce. This could also make possible amateur lipreading films made within clinics and schools, instead of in specially equipped studios. It is hoped that this study will demonstrate the possibilities and importance that eight millimeter film can have in the area of lipreading Statement 9f_Problem and Purpose of Study The general area of concern of this study was to determine whether a test of lipreading, filmed by an amateur photographer on eight millimeter black and white motion picture film, could produce the same results that a similar test produced when presented "live” to a group of subjects. One purpose of this study was to determine whether there is any relationship between lipreading performance 11Ibid., p. 736. 12Ibid., p. 736. in a face-to-face situation and lipreading performance using similar materials filmed on black and white eight millimeter film. A second purpose of this study was to develop a re- liable test of lipreading which could be used as the measuring instrument for both the ”live" presentation and for the filmed presentation. Questions Posed The following questions were posed at the outset of the study. Can a reliable test of lipreading be constructed that will have two equivalent forms? Is there a relationship between eight millimeter filmed presentation of a lipreading'test and the "live” presentation of the same test when both are alternately presented to the same subjects? Importance of the Study_ Eight millimeter film is very economical when compared to sixteen millimeter and other size film. Eight mil- limeter film could make possible many worthwhile projects that would be too costly if done on sixteen millimeter film. It is hoped that this study will aid in drawing con- Clusions about the effectiveness of filming and the im- portance in determining the future role of this size film in the area of lipreading. Definition of Terms The following terms are used in this study and are defined for clarification: Lipreading--”. . . lipreading is the correct identi- fication of thoughts transmitted via the visual components of "13 oral discourse. Live Presentation--A live presentation is defined as the audience being eXposed to the speakers in face-to-face— communication without sound. Film Presentation-~A film presentation is defined as the audience being exposed to the images of the speakers on film without sound. Eight Millimeter Film-~Eight millimeter film is defined as film being eight millimeters in width. Panchromatic Film--Panchromatic film is a film that is acceptable to color. Brightness—Meter (Photo Grid)¥—A brightness meter changes spectra to reflected light for measuring brightness of objects. Organization of the Thesis Chapter I contains the statement of the problems of this study. It also includes an introduction, the questions being investigated, the importance of the study, and the definitions of the terms used. 13O'Neil and Oyer, op. cit., p. 2. Chapter II contains three sections. The first part reviews the literature relative to research now being carried out with eight millimeter film and to the im- portance of this size film. Part two contains references which demonstrate the effectiveness of a filmed presentation of lipreading. Part three contains a review of the lit- erature pertaining to research that has been carried out with various lipreading tests. Chapter III includes an introduction, a discussion of the testing procedure, the subjects used, the speakers used, the results, and a discussion of the results in- volved in developing a lipreading test to be employed as the measuring instrument in this study. Chapter IV includes an introduction, a discussion of the testing procedure, the subjects used, the speakers used, the results, and a discussion of the results obtained when eight millimeter film is compared to a live presen- tation of lipreading. Chapter V contains a summary of the study, the con— clusions which can be drawn from this study, and recom- mendations for further study. CHAPTER II REVIEW OF THE LITERATURE This chapter will be devided into three sections. The first section will concern itself with the reported research regarding eight millimeter film, and will discuss the importance of eight millimeter film and the advantages and disadvantages of its use. The second section will dis— cuss the use of filming with lipreading. The third section of this chapter will contain a review of some lipreading tests currently available, and will limit itself to a discussion of the various lipreading tests, their con— struction and validation. Eight Millimeter Film Filming is a way in which research can be improved and expanded, but filming can be expensive and limit itself because of this cost. Amateur filming with eight mil- limeter film is economical and effective; it offers quality comparable with, or exceeding that of sixteen 14 millimeter film. Lately, because of its wide usage and acceptance, eight millimeter film has been gaining a new title and becoming known as the "amateur standard.”15 “Willard, loc. cit. Ibid. In order to discuss the advantages of this size film, it will be necessary to begin the discussion with a com- parison of eight and sixteen millimeter film. The decision as to which film size to use is dictated by various factors. These points, which help decide the size film to use, are as follows. 1. The purpose and use of the film is a consideration. 2. The projection equipment already available is a consideration. 3 The amount to be filmed versus the available funds is a consideration. A. The film processing services available is a con- sideration. Since the introduction of better quality eight mil— limeter emulsions and sound equipment, this size film is used more and more in the educational field because of its economy.17 For example, color footage for a sixteen mil- limeter film with a sixteen minute running time costs $60.60. An eight millimeter film for the same time costs $17.70.18 Therefore, when considering economy, it can be concluded that eight millimeter color or black and white film, can be made at one—third the cost of its nearest rival which is sixteen millimeter film. 16”Eight and Sixteen Millimeter Pictures in Education and Sports Training,” Bolex Technical Information Bulletin, XXXVII (196A), pp. 1-3. l7ibid., p.l. 18Ibid. 10 If high quality equipment is employed, the sharpness of eight millimeter will be ideal for small group studies or small audiences.19 Sound can only be used in con- junction with eight millimeter film by recording the sound on a tape recorder. This sound may be taped either on location or after the film has been edited.20 There are advantages and disadvantages to consider when using eight millimeter film. Flory21 points out many of the advantages of eight millimeter film in the following statements: 1. The equipment used in the operation of an eight millimeter camera is less expensive than is the equipment used to operate other film sizes. 2. The upkeep on the film is less expensive than that of sixteen millimeter film. 3. The eight millimeter camera is smaller and more compact than larger size cameras. U. Extra release prints of eight millimeter film are more economical than those of other speeds. 5. The eight millimeter projector is more economical than other film size projectors. 6. The prints from an eight millimeter film are less bulky and the release prints are easier to ship. 2OWlllard, 09. cit., p. 736. 21J. Flory, ”The Challenge of Eight Millimeter Sound Film,” Society of Motion Picture and Television Engineers Journal, LXX (19607, pp. 581—586. 9. ll The eight millimeter outfit has more portable projection equipment and the equipment is lighter. The newer eight millimeter projectors are self- threading. Eight millimeter film cost is less expensive. Floryda points out the following as disadvantages to the use of eight millimeter film: 1. Eight millimeter film is unsuitable for reverse action. Single perforated eight millimeter film must be turned back to its beginning frequently. A mirror image goes out of focus when the film strip goes through the projector. Eight millimter has a definite lack of sharpness with Kodachrome, especially in distance shots. There is no built—in sound track yet developed for eight millimter film. There are also certain other disadvantages to consider when using eight millimeter film. Eight millimeter has a small image size. Although high quality prints are possible from eight millimeter film, it is possible to 22Ibid. l2 23 produce more prints from greater film sizes at one time. Another factor to consider is that there is more sixteen millimeter equipment available for use in schools and clinics, thus dictating the size of most amateur filming size.24 These factors, plus the disadvantages pointed out by Flory,25 have in all probability slowed the growth of eight millimeter film. It could possibly be assumed that the lack of know— ledge about eight millimeter film and the high cost of filming with other size film has affected research in lip— reading. Most of the filmed studies and lessons about lip— reading were filmed on sixteen millimeter film although there has been some precedent for using eight millimeter film successfully in a lipreading study. Reid26 used eight millimeter Kodachrome film in a study completed in 1946. She found that lipreading ability was not cor—~ related with length of training, mental age, intelligence, or grade status. Franks27 employed eight millimeter Kodachrome color film to study factors that influence the identification of English sounds in lipreading. 23Bolex Technical Information Bulletin, 0 . cit., p. l. 9 2l‘lIbid., p. l. “5Flory, op. cit., p. 584. 26 L. Reid, ”A Preliminary Investigation in the Testing of Lipreading Achievement,” American Annals of the Deaf, XIC (1946), pp. 77—82. 27R. Franks, ”Factors that Influence the Identification of English Sounds in Lipreading" (unpublished Ph.D. thesis, Michigan State University, 1964). 13 Use of Film in Lipreading There is research available which demonstrates the effectiveness of a filmed presentation of lipreading. Simmons28 stated that filmed lipreading tests do measure something related to lipreading ability, but that this some- thing is only a part of the general ability that can be scored in a face-to—face situation of lipreading. Cavender29 justified the use of filming when she concluded that it was not feasible to argue against filmed tests as results proved consistent when compared to face-to—face situations of lipreading. 30 Di Carlo and Kataja offer some theoretical consider- ations which must be considered when using filming. They listed the following as advantages to consider when filming: 1. Filming provides control of the constancy of speech movements, the visibility of a Speech presentation, and the rate of speaking. 2. Filming provides the lipreader more freedom to concentrate attention on the visual patterns. 28A. Simmon, ”Factors Related to Lipreading,” Journal of Speech and Hearing Research, IV (1959), pp. 340-352. 29B. Cavender, "The Construction and Investigation of a Test of Lipreading Ability and a Study of Factors Assumed to Effect the Results” (unpublished Masters thesis, Indiana University, 1949). 30L. Di Carlo, and R. Kataja, ”An Analysis of the Utley Lipreading Tests,” Journal of Speech and Hearing Dis- orders, XVI (1951), pp. 2264240. l4 Di Carlo and Kataja31 listed the following as dis— advantages to consider when filming: l. The degree to which filming approaches a real life situation is not known. 2. At optimal level, the degree of approximation may be very close, though probably never as good as a well conducted natural situation. 32 Di Carlo and Kataja cast doubt on the effectiveness of filming when they said that the degree to which the medium approaches a real life situation is not known. Cavender33 approached and explored the relationship of lipreading skills to other factors related to the lipreading situation. Cavender constructed a sentence test including words which were determined to be within the reading vocabulary of the first three grades. Her tests were given both live and filmed to her subjects. She concluded that it was not feasible to argue against filmed tests as the results proved consistent when compared to the face—to-face situations. Cavender seemed satisfied that a filmed presentation of lipreading and a face-to—face situation of lipreading achieved the same results. Willard3u states that results 311bid., pp. 227-228. 32Ibid. 33 Cavender, op. cit. 3[(Willard, op. cit., p. 737. 15 obtained from studies where eight millimeter film was tested against sixteen millimeter show that the two 35 sizes compare favorably with each other. Campbell, 36 37 Stanwix, 'Arcy, and others have shown how eight mil— limeter film is now being used in many other areas where this size filming previously was not used. Lipreading Tests The initial step in the development of this study was to find a test of lipreading that could be filmed and used in the study as a measuring instrument to test the dif— ference between a live and filmed presentation of lipreading. A test was needed that would be reliable. O'Neill and Oyer38 say that a reliable test of lipreading is a useful research tool. A test was needed that could be used both for the filmed and live presentations and by which compari- sons of the effectiveness of each could be made. -r Snere were several lipreading test (I) {1) <1 {11 F4 'r_: Q) C . l. _l rD S 3‘ H C V‘ i_. H ('1‘ <4 '7 s f_ r. A O F—O (T (1‘) solid be ‘sed for testing lipreading abi q) lipreading tests r viewed consisted of either consonants, l: 3’J. Campbell, "Eight Millimeter Sound and the Distri- bution Bottleneck,” Society of Motion Pictures and Tele- vision Engineers Journal, LXXXI (1961), pp 573-574. Q6 A f“ O O . O V O I J d. StanWlx, ”The Application of Eight Millimeter Film for Commercial Use," Society of Motion Pictures and Television Engineers Journal, LXXXXI (1961), p. 600. 37E. D'Arcy, ”Progress Report on Eight Millimeter 'ic Sound Standard and Methods of Test Film Production,’ SOClaty of Motion Pictures and Television Engineers Journal, (1961), pp 105—109. 38 1 f1 O‘Neill and Oyer, 07. Cit., p. 21. l6 syllables, words, sentences, or stories as test items. Many of the tests combined two or more of the above items as test materials. All of the tests reviewed were con- structed and standardized differently, but all were presented either live or on film. o Conklin3“ constructed a standardized test of lip— reading. This test utilized an objective scoring method. It consisted of eight consonants, fifty-two words, and twenty sentences. The subjects were provided with a prepared test form and wrote down what they thought had been said to them. An analysis of the test scores revealed a high correlation (.90) between the test scores and rankings assigned by teachers. The analysis did not reveal any Sigruificant correlation between lipreading ability and chrcniological age. These tests administered by Conklin were presented in a face-to-face manner to the subjects. 11/ . Mason 0 developed a test of lipreading by using “DUKES; for her test material. The purpose of Mason's WOKWC ‘was to develop a standardized test of lipreading and t“) fkilnlthis test on sixteen millimeter for use with larg ‘\ 63 groups of acoustically-handicapped people. It was Mafscnfi's desire to develop a test which might measure the .liIDrwsading ability of handicapped children at various .\ Ede 39E. Conklin, "A Method for the Determination of ID lative Skill in Lipreading,H The Volta Review, XIX (1917), -p' 2916-220. l(OM. Mason, ”A Cinematographic Technique for Testing ch Comprehension,” Journal of Speech Disorders, 271-278. \0 7:5 age levels. Mason started to develop this test by first constructing two forms. She standardized the first two forms on 700 subjects and later added a third form. In standardizing the new third form she administered the two forms filmed to 187 subjects in a large residential school for the deaf. Then she administered the newly developed and filmed third form to the same 187 subjects. The correlation between the two existing forms of the Mason 41 Test and the new third form was .954. One conclusion reached by Mason in this study was that the cinematographic technique of administering a test of visual speech compre- hension provides reasonably objective measurement.“2 Cinematography controls the variability of visual movement by presenting the visual components of speech on a silent screen. It controls distance, sound source, and rate of speed. Mason also concluded that the coefficient of cor- relation for all the forms suggested they all possess a certain degree of r€liabilify as measures of lipreading achievement. UtleyuB developed a test of lipreading consisting of three parts. These three parts consisted of a section of .worsz, of sentences, and of six stories containing assumed u‘ 2 4 Ibid., p. 278. lBJpUtley, ”Factors Involved in the Teaching and I‘ést ing of Lipreading Ability Through the Use of Motion Pictilres,” The Volta Review, XXXXVIII (1946), pp. 657-659. l8 common everyday experiences of children. The test was filmed on sixteen millimeter film. The word and sentence tests were filmed in color. Utley administered her test to 761 school children. She concluded that it is possible to construct a reliable and valid test of lipreading having a wide range of items with which to score. Utley concluded that there was interrelation among the skills of word, sentence, and story recognition. She concluded that the ability to lipread sentences was more reliably predicted from an ability to lipread stories than words, but the ability to lipread words was more reliably predicted from an ability to lipread sentences than stories. The ability to lipread stories was more reliably predicted from an ability to lipread sentences than words. Morkovin“l reported on the development of a series of ten films designed as training materials for the acousti- cally-handicapped. These films were built around everyday experiences. Morkovin stressed the importance of both verbal and non-verbal clues in developing a test of lip- reading. When teaching lipreading, Morkovin stated, "all ideas of learning must be utilized stressing the areas -such as speaking, kinesthetic movement, auditory, visual, ll tactile, and the visual kinesthetic. “5 When training some— 44B. Morkovin, ”Rehabilitation of the Aurally Handi— capped Through the Study of Speech Reading in Life Situ— ations," Journal of Speech Disorders, XII (1947), pp. 363-358. 45Ibid., p. 367. 19 one to lipread, Morkovin feels that these areas must be utilized and that clues, gestures, and a real situation in which the lipreading takes place are extremely important. Morkovin and Moore“6 developed a series of ten films on sixteen millimeter film. These films were designed as training materials for the acoustically handicapped. These tests were designed by using a series of everyday incidents that were realistic and interesting to both children and adults as a basis for the test material. Lowell“7 reported research on a filmed lipreading test constructed at the John Tracy Clinic. This test con— sisted of sixty unrelated sentences which were simple in context and might have occurred in everyday conversation. These sixty sentences were then administered to 408 college students with normal hearing. It was found that half (30) of these sentences were as reliable as all sixty, so two forms were developed. These two forms were administered to 173 college students, and no significant difference was found between the two forms. Kitsonu8 reported an investigation carried out to examine factors assumed to be related to lipreading ability. ‘ Morkovin and Moore, 0 . cit. 47E. L. Lowell, ”A Filmed Test of Lipreading,” John Tracy Research Papers, John Tracy Clinic, Los Angeles, California, November 1957. “BM. D. Kitson, ”Psychological Test for Lipreading Ability,” The Volta Review, XVII (1951), pp. 216-220. 20 Kitson constructed no formal test but used lipreading apti— tude, as judged by teachers, as a measuring criterion. He presented his materials using a tachistocope. The materials presented were a reading test and a completion test. The results of these tests were compared to the results obtained on a specially constructed face—to-face lipreading test. His results were that the subjects who had high scores in visual skills also rated high on lipreading scores. It was assumed a good lipreader had a large visual span. Di Carlo and Kataja49 critically evaluated the Utley lipreading test. They desired in this study to determine if the Utley film is a valid and reliable instrument to test achievement in lipreading. They also wished to learn if the Utley test is able to discriminate between good and poor lipreaders. They administered the Utley test to fifty- seven aurally handicapped and forty-four normal hearing subjects. The conclusions reached from this study were as follows: 1. There was no significant difference between the scores of the aurally handicapped and the normal hearing in lipreading ability. 2. The test discriminates well within the group, but it fails to differentiate the two groups. 3. The Utley test does not measure lipreading achievement. ugDi Carlo and Kataja, op. cit., pp. 237-238. l—--~ " Di Carlo and Kataja “Qfelt that the test failed be- cause Utley did not take into consideration the psychology of the lipreading situation. this being the importance of situational clues of a real life Situation. This weakness was particularly evident in Utley's third section, dealing with six stories. The lack of clues or gestures contri— buted directly to the low scores in this part of the test. Only one story, (.77) the Family Dinner, seemed to corre- late well in relationship to the entire test. From the first section of the Utley test, Di Carlo Cl and Kataja’ report that the test forms containing only words correlated internally poorly because of the lack of clues or gestures. This weakness, in all probability, made the identification of the words more difficult. Morkoving2 reports from his study that the analytical approach to . I lipreading using individual words develops poor lipreaders. C .9 0 P1 , ' l ‘ . 0 , J Di «arlo and hataja report that the section in Utley's test that contained unrelated sentences appeared to correlate the best in relationship to the rest of the test. The lipreaders tested as a group had the best over- aJQl results in the identification of Utley's sentences. \ I: . J P... Ibid., p. 232. ’ *Morkovin, op. cit. \fl L) \O "O Q) K- 4 l b CZ. * . .. "D ‘i ’3‘ 7- ”3147—12—1- , pp. C’3i_‘r;3j. 22 From Di Carlo and Kataja's report, many of the weak- nesses found in the Utley test would, in all probability, appear to be in evidence in some of the other lipreading tests. After reading this report and reviewing other lip- reading tests, it was decided: l. to construct a test of lipreading using sentences as the test materials; and 2. to construct a test that could be used with a wide age range (twelve years old or above). CHAPTER III TEST CONSTUCTION This chapter will concern itself with a discussion of the development of the lipreading test. The initial section of the chapter will be concerned with an explanation of the procedure involved in constructing a test of lip- reading. The test construction, the subjects involved, and the test administration will be discussed. The second section will concern itself with results obtained and con- clusions reached. Test Construction In order to construct the lipreading test, it was de- C; cided to follow the outline suggested by O'Neill and Oyer/4 as a guideline to follow in constructing a lipreading test. This section will review that outline and demonstrate how this lipreading test adhered to the outline. When constructing a test of lipreading, O'Neill and 55 Oyer state that we should consider six basic points con- cerning test construction: 1. The Population to be Tested. This test was to be constructed for use with people twelve years or older). ‘ 24 2. The General Format of the Test. It was decided that this test was to be a write down response test allowing the subjects thirty seconds in which to write their responses to each sentence. 3. Selection of Speakers. In order to give the subjects some varying degree of lipreadability, it was decided to use two speakers, one male and one female, to read the test sentences to the subjects. 4. Selection of Test Items. It was decided to select sentences that were frequently used in either every- day conversation or sentences that would be familiar to the subjects involved. The sentences were to be selected from two source 6 One source was a high school English textbook 2 Two hundred frequently used or familiar sentences were chosen from this source. The second source was a committee of graduate students who selected spoken sentences which they used frequently in everyday conversation. 5. Test Conditions. The variables to consider when constructing and later filming the test were con— sidered and controlled when the test was constructed and later filmed. These variables will be discussed later in this chapter. 6. Scoring Procedure. A percentage scoring system was devised for the scoring of the test. Three hundred sentences were selected as potential test items. These sentences contained from three to six words each. In order to obtain an item analysis to aid in the elimination of many of these sentences, it was decided to test all three hundred sentences on a group of fifty junior high students. The fifty students were enrolled in the eighth grade of DeWitt Junior High School, Clinton County, Michigan. 56Pollack, Sheridan, Williams, Our English Language (New York: MacMillan Company, 1961). 25 The students were presented the sentences in their class- room. The three hundred sentences were administered to the students on two consecutive days; one hundred fifty sentences were presented each day. These tests were administered to the students for an hour and one—half each testing period. Two male speakers took turns presenting the sentences, without voice, to the students. The students wrote down what they lipread from each speaker. After the sentences were given, they were scored to determine the raw score or lipreadability of each sentence. The raw score was converted to a percentage score and this score was determined in the following manner. If the sen— tence was identified, it was given a score of 100%. If any part of the sentence was identified, it was given a percent score based on the ratio of the number of words in the sentence. If the entire sentence was missed, it was given a score of zero. Selection of Stimuli In selecting the sentences to be used in the final test, it was necessary to devise a means of distributing the sentences according to sentence difficulty. This distri- bution was obtained by plotting the sentences on a frequency histogram with sentence difficulty in class intervals of ten along the abscissa. These sentences were plotted on the basis of their percentage scores until all 26 the items had been distributed on the graph according to sentence difficulty. All three hundred sentences were tested and all item scores were plotted on the graph so that they ranged from easily identifiable to unidentifiable as lipreading items. Figure 1 demonstrates this distribution. Starting with the three hundred sentences, it was then necessary to select sentences for the lipreading test in such a way that item difficulty was normally distributed. In theory, this should result in a normal distribution of lipreading scores for the population tested. The sentences were placed according to a Gaussian Distribution using .50 as the midpoint. In order to select the lipreading test items, the sentences were grouped according to difficulty in class intervals of ten along the abscissa. A number of sentences were then randomly selected from each class interval along the abscissa so that each sentence represented a certain degree of lipreading difficulty. Using this means of selection, thirty-two sentences were chosen to be the lip- reading test items. In this way the normal distribution of lipreading scores for the population tested remained and the sentences selected ranged from items easy to identify to items difficult to identify. The sentences were repre— sentative of the entire range of difficulty. Number of Sentences 7o -- 6O 50- 40 30 2O 10 27 IF l fit I_ q' 1— 10 20 3O 4O 50 60 70 80 90 100 Percentage of Group Making Correct Sentence Identification: Fig. l.-—Item Analysis of Sentence Difficulty, Showing the Number of Sentences Correctly Identified' by Various Proportions of the Group. 28 The thirty-two sentences were divided in the same manner as previously stated into two sixteen-sentence tests, so that each test maintained the normal distribution of item difficulty. The sentences making up the two forms of the test are contained in Appendix B. Following this procedure, the next step was to test the sentences on college students. The purpose was to determine the reliability of the constructed test using correlation techniques. Subjects The subjects selected to take the lipreading test were twenty-four college students at Michigan State University. The majority of these students were freshmen and sophomores, but there were some juniors and seniors included. All of the students stated that as nearly as they could determine, they had normal hearing and had not had previous lipreading training. Lighting The lipreading test was presented visually to the subjects without a special lighting arrangement for this room. The standard overhead lighting was used for this room and the intensity of this light was not measured in candle power. Test 0n the day of the test, the subjects came to the Speech and Hearing Science Laboratory at Michigan State University. 29 The subjects were seated in three rows consisting of eight chairs each. The chairs were grouped toward the center of the room. As soon as the twenty-four subjects were seated, instructions were given verbally for the test and questions were answered. A copy of the instructions can be found in Appendix A. When the subjects were ready, a speaker stepped to a spot in the middle of the room in front of the subjects and spoke the first sentence, without voice, from Form A. The subjects wrote their responses on an appropriate response sheet. A copy of the response sheet can be found in Appendix C. The students were given both forms of the tests alternating items from Form A and B. The sentences were presented live by a female and a male speaker who alternated giving all the sentences from each form until they had spoken the thirty-two sentences. In this process, each speaker presented eight sentences from each form to the subjects. The subjects were given thirty seconds after each sentence to write down their responses and then the next sentence was given. When the test was completed, the answer sheets were collected and the subjects were dismissed. Results and Conclusions Results The subjects' scores for each form were obtained using the same scoring procedure as was previously discussed. The 30 raw scores were converted into percents, as previously out- lined. A subject's percentage score for each of the sixteen sentences of each form was summed to obtain the total score for each form of the test. From the subjects' scores, two frequency graphs were drawn demonstrating the individual lipreading score of each subject on both forms of the lipreading test. Observations on these graphs would seem to indicate that most of the subjects scored in the middle ranges of dif- ficulty (40% through 60%) and that some subjects scored well and some poorly on these two forms. Figures 2 and 3 demonstrate the individual lipreading scores that each sub- ject had on the two forms. The two forms of the lipreading test were correlated to show the relationship between the two forms. The fol- lowing formula for the correlation coefficient was used for this purpose:57 anyg - (XX) (XX) J an2 - (Xx)2 Jfinty2 - (2y)2 Using this formula, the product moment correlation obtained between the two sixteen—sentence forms (A and B) of the lipreading tests was r = 0.85. Utilizing the SpearmansBrown Formula, the extrapolated reliability for the total test was r = 0.92. 57H. Blalock, Social Statistics (New York: McGraw- Hill Book Company Inc., 1960), p. 289. Number of Subjects 31 1 I 0 10 2o 30 no 50 60 7o 80 90 ‘100' Percentage of Correct Identification Fig. 2.-—Item Analysis of Sentence Difficulty Showing the Number of Sentences Correctly Identified by Various Proportions of the Group (Form B). Number of Subjects O 10 2o 30 no 50 60 70 80 90 100 Percentage of Correct Identification Fig. 3.—-Item Analysis of Sentence Difficulty Showing the Number of Sentences Correctly Identified by Various Proportions of the Group (Form A). 33 Table 1 shows the mean scores and standard deviations for the two forms of the lipreading test. TABLE I COMPARISON OF MEAN SCORES AND STANDARD DEVIATIONS Form A Form B Mean Score 0.53 Mean Score 0.50 Standard Standard Deviation 0.149 Deviation 0'130 The mean scores and standard deviations as seen in Table 1 appear similar when they are compared to each other. Conclusions The two sixteen-sentence forms of the lipreading test correlated highly with each other. The means and standard deviations as seen in Table 1 appear similar when they are compared to each other. It was concluded that: l. the lipreading test has two equivalent forms; and 2. the lipreading test can be used with normal hear- ing people above the age of twelve years. 34 CHAPTER IV PROCEDURE, TEST, RESULTS This chapter will consist of a discussion of the pro~ cedure used and the results obtained when an eight millimeter film presentation was compared to a live presentation of the developed test of lipreading. Included in this chapter will be sections dealing with film production, the speakers, the equipment, subjects, and the test administration. Procedure Film Prouuction The film used in the study was produced in the Speech and Hearing Science Laboratory at Michigan State University. Two speakers, a male and a female, were filmed with eight millimeter film while each speaker read a sentence from Form A of the lipreading test. The two speakers were again filmed with eight millimeter film while each read a sentence from Form B. The speakers continued to be filmed until they had alternately read all thirty—two sentences from both forms (A and B) of the test. The speakers were filmed with a camera (Yashica, Eight Millimeter) that was mounted on a tripod in the center of the room. The film used was Dupont Panchromatic film. Two 1000 watt flood lights were 35 illuminate the area where the speakers stood. After the speakers were photographed, the camera continued to run for an additional five seconds. In this way, the film did not have to be cut and extra film spliced between each sentence. When the filming was completed, it was developed and spliced by a commercial film developing company. Speakers There were two speakers used in a attempt to present the lipreading task with more validity. O'Neill and Oyer58 state that the more speakers used in a lipreading situation, the more opportunity there is for validity in the results. This is based on the assumption that some people are easier to lipread than others. Two speakers offered a varying de— gree of lipreadability in the test. Equipment The following equipment was used in order to present the film to the experimental population: 1. Eight millimeter motion picture projector (Bell and Howell, Model 265—A). Light meter (Spectra-Combi, Model No. 500). f0 3. Brightness meter (Photo Grid, Bell and Howell Model 76). 4. 44" x 44" motion picture screen (Radiant Picture Master). BBQ'Nelll and Oyer, op. cit., pp. 30-33- ML‘U "$1.11.“. I' l I El» ”J:- 36 Subjects There were fifty—seven subjects used in the study comparing a live presentation to an eight millimeter pre- sentation of lipreading. The same subjects were used in both the control and experimental groups. The subjects were college students who were students in a freshman level speech course at Michigan State University. The majority Bi of the subjects were freshman and sophomores, but some were at higher academic levels. The subjects stated that they had, to the best of their knowledge, normal hearing and no one had had previous lipreading training. The subjects were given the lipreading tests in the Speech and Hearing Science Laboratory at Michigan State University. The subjects came into the laboratory in three groups. It was decided to divide the subjects into the three groups in hopes of facilitating reliable testing with a smaller group. The laboratory was arranged in a specific manner for the test to take place. The subjects were seated in three rows of chairs facing the front of the room. In the front of the room was placed a motion picture screen and a spot was marked on the floor where the speakers were to stand. To the rear of the room was the motion picture pro— jector. Figure 4 shows how this room appeared on that day. 37 .meB one mo mam one co m90pmaopmq cocoflom wcflpmom can cooomm one mo wcflzmsa