. - ”z ‘_.i.JJ.1.-._: As - .m— __— m .-a.. inky-2994’; A Av‘L—‘J ._ qumw -A 4....._.-_._.—. <-.-'vw.-c$- AN Amvsxsan cemumsmws as A A .. LISTENING TRAINING PR‘QGRA-M ‘ ' ' cauwcma mom: cams: FRESHMEN A: mum»: arm comes 2 :z'mt‘faa‘w - “—‘nn- JflA..‘ _. ._n-\-«-. i . , , _ ’J'llgesfig‘fAr 'fi‘lb 'QAAAAA As: :4. o; A _ Malawian come: 3 . Chum Elwin 1932 Ithgfvf~iwn f sun-«1'4: 1-: ... .0' "I V _ a. Mg, v u, v 7... ....—~- xwffi“‘ff‘“w‘i "“ .fi,‘ .hw‘ahfi; - i Aw}! , .-7'4- h'vvyau‘mgnup .~‘-. .4 a .. w unvrz.< r—_.. fifty-«lair " "C 3.9% “4'1: m- 45:5? I i J“; 'i ijnssnsfi \ 'V ' H ‘ . This is to certify that the thesis entitled An Analysis of Certain Aspects of 5 Listening Training Brogram among College Freshmen at. Michigan State College presented by Charles E. Irvi n A of the requirements for r f \ A A ‘ has been accepted towards fulfillment ‘ - PheD. degree in Edugggion ' i I I l-iilosh Iiuntxan Major professor , Date MIL 0-169 REMOTE STORAGE PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE 2/17 208 Blue FORMS/DateDueForms_2017.indd - 99.5 AA ANALYSIS OF CmeAIN ASPL‘JCTS or A LISTENING TRAINING Pnown CONDUCTED mono comer; Fithlt-mN AT MICHIGAN 31mm; COLLEGE by Charles E. Irvin A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOK OF EDUCATION Department of Education 1952 Jnc‘d" - A .A.v u . . ‘. u. e ,e... '0 _,....~ ‘ ,' -.v1 1.. 0" “‘ ."V ‘ . A" -.u- I A A. '- V A ~ . -..«- ..- t 'h u.‘-- . A , .- .., A. ‘ .a .«. m... a- .» -r‘ .. .4 -A.‘~‘. .--, - . .' ‘1 AI. . .-.. ,. “—x-l vivUVQI A. .,.,,.. ,.- - - —A -..a..vu. 4 .- .. .._.a ‘A.‘ Val.---. .\ ,..-, v- ..-..-- 3a. . -- . 4v..1.. a. ~; ‘ 0-! A... ._ _ . . , 5* ’.-I ,.. . -' s, -y-\ .1..:- ~ ~ '-o.., _. V ... c. —. -V'C‘ Av ,- -. v , . . ‘Y‘ -0- -. C‘. --§ e ,,’ ,. "A' ‘n..... "O . I] I ‘J n. ‘- ~.. #5 ACnhOQLdDGmnNTS The author wishes to eXpress his sincere thanks to Professor Paul D..dagwell for his gracious permission to conduct this study in his department, for his steadfast support of the entire program, and for his vision and inspiration which made it possible to hurdle many obstacles that might have resulted in failure. The results of this study are ‘edicated to him. Grateful acknowledgment is also extended to Dr. Cecil V. Millard for his excellent COOperation in permitting studies of this type to be included in the field of stucy for which he is primarily responsi- ble. To Dr. hilosh.huntyan goes a sincere thanks that defies ex- pression. his expert guidance, sympathetic attitude, and genuine interest in the study and the personnel involved provided both the direction and the atmosphere which insured its successful completion. This author is deeply indebted to Dr. Wilson B. Paul, Dr. Clyde Campbell, and Dr. B. h. Sundwall who, as members of his guidance committee, gave invaluable suggestions and assistance throughout the conduct of the study. He also deeply appreciates the advice and expert statistical aid given by Dr. John h. Schmid without which neither the format of g the experiment nor the evaluation would have been possible. j To Professor Clyde Dow, Professor Charles Hill, and Professor Donald blending goes a sincere thanks for their loyal support as members of the listening committee and as teachers in the experiment. The author is also grateful for the COOperation of the following members of the Department of Written and Spoken English who permitted the integration of listening instruction with their regular staff lectures: Dr. Merton Babcock, Dr. Johaninburne, Dr. Fred Reeve, Professor Judson Perkins, and Dr. Arthur heilman. "m‘ -\:.'- He wishes to express his gratitude to all other members of the Written and Spoken English staff for their helpful suggestions that ‘ brightened many a dark hour along the way, i Finally, to a person who prefers to remain anonymous goes the author's sincerest thanks for the understanding and inspiration with- out which this study could neither have been started nor completed. Hill 35-31.?) " a E -..-h-. — v." I..- d-d—v-t p >.o _'. Hay—-n: _.- ‘4:7 .--;. ‘_ r 5. v- “4. VITA Charles Edward Irvin candidate for the degree of Doctor of Education Final examination: 10:00 A.h., Monday, hay I9, 1952. Dissertation: An Analysis of Certain Aspects of a Listening Training Program.Among College Freshmen at.hichigan State College. Outline of Studies: major subject -- Education minor subjects~— Speech bducation BiOgraphical Items: Born, May 21, 1913, Zanesville, Ohio Undergraduate Studies, Oberlin College, 1931-35 Graduate Studies, Oberlin College, 1935-37; Ohio State University, 1938 Michigan State College, l9hE-S2. Experience: Graduate Assistant, Oberlin College, 1935-36; Instructor, Oberlin College, 1936-37; Instructor, Allegheny College, l937-hl; Instructor, Ohio State University, summer, 1938; Assistant national Director, Military Welfare, American National Red Cross, l9hl-h2; Regional Director American hed Cross, Mediterranean Theatre, l9h2-h5; member United States Air Force, l9h3-h5; Assistant Director Personnel Administration, American National Red Cross, l9h5; Instructor, Michigan State College, 19u5-ts; Assistant Professor, Michigan State College, 19uo-52. All ANALYSIS OF CERTAIN ASPECTS OF A LISTENING TRAINING PROGRAM AMONG COLLEGE FRESW AT HICHICEN STATE COLLEGE By Charles E. Irvin AN ABSTRACT Submitted to the School of Graduate Studies of Michigan State College of Agriculture Ind Applied Science in partial fulfillment of the requirements for the de gru of DIETOR OF EWCATION Department of Education Year 1952 Approved W / _ bydi been able Hit} . .v n iii-BI! k .l. of: of on 1: WEI .’ .U‘f’l 1‘. “ Charles E. Irvin 1313313 ABSTRACT This study m conincted to test the efficecy of teeohing listening by direct instruction. Among the new probleus that evolved, three became most important: 1) the actual selection and Want at teach- shle uterine; 2) the integration of seven units of listening instruction Iith the college course of Written and Spoken English which is required a: an tram at Michigan State college: 3) the evolution or instrec-v tionsl results. Heteriels were selected and errenged by e consittee of four nonbers e! the Britten end Spoken English Deperuent ststf. Seven units or listening instruction of ten ninetes eeeh are then integrated with the one-hour lecture sections of the course in the toll Quarter, 1951. These lecture sections were eight in mnbsr Ind each contained between 200 end - 300 students. . To tuilitste evolution of results, the design of the study no kept silple. Em lecture sections, two nesting in the morning end we in the siternoon, were designated es the experimental group which received the seven units or listening instruction. These seven units were tought, one per seek, ever s period or seven weeks. The other four lecture sections, else evenly distributed between morning end ntternoon periods, were designated es the control peep end received only the regulsr work of the course. Both groups were tested twice: once prior to the initis- tion of the treining program once imedietely after the completion of the prom. The study was tightly controlled in order to elininste or aim-in the new verisbles operative in the listening activity. -1- flwli it the Uni-j, mum-e ii! and two we- ant sabfiect consisted . r "Wading It v: 01' not s l‘eceived in: am listen’u In 13‘3th mien °f coy Grist ”any Inez.a high 12-3 to s be Charles E. Irvin The listening tests employed were those constructed by Rslph Bichols st the University of Hinnesote. Four lacquestion tests, designed to unsure listening comrehension, were used; two were given es the pretest end two were given as the post-test. ‘ Each of the tests covered e differ- ent subject-matter tree. The naterials over which the tests were given consisted of short recorded lecture «meats in the subject sres cor- responding to «on can: It was hoped that the testing would revesl three things: 1) whether or not s difference would exist between groups liter one poup hsd received listening instruction; 2) whether or not 11 difference in listen- ing skill existed between the some; 3) whether or not s difference in listening skill existed according to the time of do: thst students listened. Two nejcr statistical procedures were utilized to evaluate the instructional results. An analysis of variance was used to weigh the differences which might exist becsusc of sex and time of day. An snelysis of covariance was used to weigh and edJust the differences thst light exist between groups due to pro-existing differences in listening sptitude. Hsny of the verisbles operative to produce such pro-existing differences were thus minimised in effect. In oddition, the extras low and extremes high scoring brackets were unlined for on analysis of instructionel results. While the experimentsl and control groups each contained spproxinstel; 1200 students, s rnndon sampling of 500 for each group was used as the test populstion for the evolution of results. As an outgrowth of this stucv, the following conclusions appear to be justified: l) s sufficient mmber of the processes involved in T.‘ ' '9” -“ .Ik listel this (O! I the 61‘5“; g11 Charles E . Irvin listening can be positively influenced by teaching es to result in immanent in listening es measured by behavior on the tests used in this study) 2) with reason, the listening training given could be credited for creeting e highly significant difference between the experimental end the control groups; 3) with reason, the listening training given could be credited for resulting in e very appreciable gain among the below-avenge listeners, es mam-ed by pretest behavior ; h) the listening treining given wes apparently ineffective mong the shove-everege listeners, es defined by pretest behavior; 5) male students eppcered to be significantly superior to female students in perfomsnce on the listening tests used in this study; 6) time of day eppeers to have had no significant influence upon listening sbility es named by the tests used in this study. IIIILN . m; I 1y“ 2".) U,,.>. . aloe II. S lie-Ill‘ ‘.1“|Qllll..‘l 3‘11 vu l r. . luv“ TABLE OF CONTENTS CHAPTER PAGE I. INTRODUCTION ................ . ............................. 1 Statement of Problem ...................... . .......... 2 Definition of Terms .................................. 3 Basic Assumptions .................................... 7 Justification of Problem ............................. 12 II. SURVEY OF LITIRATUP? ...................................... 20 III. BACKGROUND or THIS STUDY .................................. 32 IV. PROCEDURE ................................................. 36 The subjects of the Study ............................ 36 The Integration of Listening Instruction with the Lectures ........................................... 38 The Instructional Materials .......................... Db The Listening Tests .................................. hh The Personnel ................................ . ....... 55 The Variables and the Control of Them ................ 56 The Selection of Test Population for Statistical Analysis ........................................... 6h V. ANALYSIS OF RESULTS ................... . ................... 67 VI. CONCLUSIONS AND RECOMMENDATIONS ................ ...... ..... 101 Conclusions ..... . .................................... 101 Recommendations ...................................... 10h BIBLIO GPJPHY ................................................... i. '. 105 APPENDICES ....................................................... 118 APPENDIX A, Instructional Materials .................. 118 APPENDIX B. Tests .................................... 156 LIST OF TAJLES Tam ma I. Item Analysis of Listening Pre-test ........... . ............ . D7 II. Item Analysis of Listening Post-test. ........... . ......... .. hS III. Means of the Experimental Group by Lecture Section .......... 71 IV. Means of the Control Group by Lecture Section. .............. 72 V. Mean Scores for Both experimental and Control Groups........ 73 VI. Differences Betweeniflean Scores on Post-test. ............ ... 76 VII. Data on son Students Involved in a Study to Determine the Educability for effective Listening. ....... . ........... ... 78 VIII. Analysis of Variance of Final Listening Test Scores ....... .. 79 IX. Analysis of Covariance of Adjusted Means..... ......... . ..... 81 X. Calculation of Adjusted Final Listening Test Scores ......... 82 XI. Analysis of brror Variance in.Listening Study. ............ .. 53 XII. Summaries of Percentages in Various Scoring Brackets ........ 55 XIII. Summary of Statistics of Control and Experimental Groups in the 15 to o Scoring Bracket ...... . ...... . ...... . ......... . 89 XIV. Post-test Distribution of Pre—test Scores 15 and Below ...... 9O XV. Amount of Gain and Loss on Post-test of students in the IS and Below Category.... ................................ .... 92 XVI. Summary Statistics of Control and Experimental Groups in the 16 to 2b Scoring Bracket. ..... . .......................... . 96 XVII. Post-test Distribution of Pre-test Scores It and Above ..... . 98 XVIII. Summaries of Percentages in Various Scoring Brackets ..... ... 99 LIST OF FIGURES FIGURE PAGE 1. Distribution of Pre-test Scores. ................. . .......... ol '2. Distribution of Listening Test Scores Among the Experimental Group 0 O O 0000000000 O ...... O ....... O OOOOOOOOOOOOOOOOOOOOOOOO 7b 3. Distribution of Listening Test Scores Among the Control Group.. .................................................. . 75 h. Experimental XE Control Groups on Pretest in Scoring Bracket 0-15 oooooooooooooooooooooooooooooooooooooooooooooooooooooo 87 5. Experimental vs Control Groups on Post-test in Scoring Bracket 6-15. .................................... . ........ 88 6. Experimental vs Control Groups on Pretest in Scoring Bracket 18-214 ...... T'.’ ......... . .................... . ...... . ...... . 9h 7. Experimental vs Control Groups on Post-test in Scoring Bracket lS-EH ............................................. 9S CHAPTER I CHAPTER I INTRODUCTION hordsworth was perhaps over-Optimistic when he wrote: "The eye, it cannot choose but see; We cannot bid the ear be still." He implies that we are constantly listenino. Through the years, other implications have been made. The first line of a famous poem about Paul Revere begins, "Listen, my children, and you shall hear". The white cross-bars that mark the intersection of railroad and highway bear the words, Stop, Look, Listen. One would assume, therefore, that from poetry to safety warnings, something accrues from listening. This is further emphasized in classrooms. Teachers in the primary grades clap their hands and admonish the children to listen. In later stages of education, instructors request students to take notes in class lectures. The twilight years of graduate study find professors beginning their lectures without benefit of admonition or reguest, merely expecting mature students to listen. No matter where the observer may turn, to classroom, business office, conference room, living room, United Nations assembly, or Korean truce meeting, the implication is clear. Listening is an important phase of our lives; listening results in the acquisition of information and the understanding of spoken symbols; listening is often followed by changes in human thinking and behavior. This study, therefore, concerns itself with a communications skill which is familiar, important, much used, and functional. Statement of the Problem Specifically, the problem with which this study is concerned is: I'AnAnalysis of Certain.ASpects of a Listening Training Program Conducted.Among College Freshmen at the Michigan State College." From this analysis, it is hOped that some light may be shed upon what can be taught in such a program and upon the results of such teach— ing. The study will be probed to discover three things: I. If there is a difference between the listening test results of the group receiving training and the group not receiving it. 2. If there is a difference in test results between the sexes. 3. If there is a difference in test results between the times at which groups convene for listening; morning versus afternoon. The end result of this study is not to justify this one particular prOgram of listening training. Nor is it to justify the materials or methodology employed. Rather, this study represents an adventure into the teaching of listening. The accumulation of data will serve as a foundation for further study; for further deveIOpment of materials and methods leading toward a sound program of instruction. Michigan State College includes a course in\£ritten and Spoken English in its basic curriculum. This course obligates the department to teach the four communications skills: reading, writing, speaking, and listening. The background of information in listening is meager compared to the other three skills. Studies such as the one under dis- cussion can broaden the horizons, extend the information, and build a firmer basis for instruction in this much neglected area of the study of communications skills. Definition of Terms By the term "Analysis" is meant an exploration and evaluation of the results found to be present among the two groups involved in this training and testing program. The term "Certain Aspects" limits this study to the first Quarter of a three-quarter training prOgram. Michigan State College divides its school year into Quarters rather then semesters. The Department of Written and Spoken English planned and executed three quarters of listen- ing trainino. Each quarter became the reSponsibility of each of three members of the listening committee of the department. The writer accepted the responsibility of the fall quarter which involved: a listening pretest; seven units of listening instruction; and a listening post-test. The winter and spring Quarters are being reported as separate studies by two other persons. The term "Listening" needs a larger definition. In the first place, listening is to be differentiated from hearing, although the layman often uses the two synonymously. Hearing is the awareness of sound. You may hear a foreign language; you may hear the motor of your car. Listening is more than hearing. It is important to remember that, for the purpose of this discussion, listening involves the comprehension of what is heard. Listening to a foreign language involves the understand- ing of that language. Listening to the motor of your car involves the listening for something: a knock, or a jumping spark. There have been several definitions of listening. This writer has, in several printed articles, defined it as "reception and comprehension; reception and evaluation of orally presented materials". This definition takes into account the different kinds of listening as described by Ralph Nichols:l l. Discriminative listening to informative materials. (sometimes referred to as discriminatory listening) 2. Critical listening to persuasive materials. This involves evaluation as well as comprehension. 3. Appreciative listening to any aural presentation gratifying to the senses. 1. Ralph Nichols. "Teaching of Listening", Chicago Schools Journal, XXX (June, 19149), 27h. Another definition is that given by Blewett:l "Listening is the process of attaching meaning to the spoken word". This definition is more general and perhaps is more widely accepted. For the purpose of this discussion, listening will imply compre- hension. The term "Listening Instruction" also needs a larger definition. One of the major projects of the Listening Committee of the National Society for the Study of Communication is to define listening instruc- tion. In the annual report of this committee,2 the following is the consensus about what constitutes listening instruction: 1. Determination of the need for listening instruction. 2. Determination of the objectives of listening instruction. 3. The planning of a series of experiences designed to enable the student to achieve these objectives. h. The establishment of procedures to evaluate instructional results. 'With respect to the planning of experiences, a survey reveals three differing modes of instruction now in use: 1. Thomas Blewett, "An Experiment in the Measurement of Listening at the College Level", Journal of Communication, I (May, l9Sl), SO. 2. Charles Irvin, "Report of Committee on Listening Comprehension", Journal of Communication, I, No. 2 (November, 1951) cc. l. The listening-laboratory technique such as that . used at the University of Minnesota. Students listen to :Eigidinge of lecture segments and take progressively more difficult tests over the materials contained therein. 2. The coordination of listening assignments with speaking assignments in speech classes. 3. The direct instrutional method such as that used at Air University and at Michigan State College. This direct method is accomplished by lectures about listening and lectures on how to listen more effectively. Such lectures may be hour-long and complete within themselves, or they may be shorter and integrated with lectures on other subjects.’ In this discussion, listening instruction is to be understood as the direct-instruction method in which lectures on how to listen are integrated with other informational lectures on entirely different subjects. Specifically, the listening instruction involved in this study was composed of seven instructional units of ten minutes each, each of which preceded a twenty-five minute lecture on a different subject. Where time permitted, the instruction was broadened to include a checking period of ten minutes following each twenty-five minute informational lecture. For the purpose of this study, listening instruction is meant to be training in discriminative listening to informative materials. These details are more fully explained in.Chapter IV. The training materials are fully reproduced in APPENDIX A. The term "College Freshmen" refers to the first year students registered in the required course of”Written and Spoken English at Michigan State College. These students were divided into two groups: I. The experimental group, approximately lhOO students, who received the seven units of listening instruc- tion. 2. The control group, approximately lhOO students, who did not receive listening instruction. During the discussion of this study, the terms experimental and control will be used to identify, respectively, those students receiving listening training and those not receiving it. Basic Assumptions The launching of a program of this sc0pe is preceded by research and thought relative to assumptions about listening and listening instruction. Before listing the assumptions favorable to the teaching of listening, the writer wishes to discuss two basic assumptions that have perhaps been primarily responsible for the apparent neglect of this phase of communications skills. Teachers of listening are often asked why there is such a sudden interest in the subject. Discussion with the average layman reveals that he feels listening to be simply a matter of hearing acuity and intelligence and that education can do little about either. Nichols,l in a series of interviews with educators and non-educators, received such answers consistently; listening is a matter of hearing and intelli- gence. Perhaps, then these are the two basic assumptions upon which the long years of neglect are founded. Nichols insists that to explain listening in these terms is the same as saying that reading is merely a matter of eyesight and intelligence; that speaking is merely a matter of articulation and intelligence. Research up to this point indicates that hearing acuity is related to listening skill, but not to an important extent. At least, this has been reported by Nichols.2 Actually, some students with hearing loss have, in some experiments, done better than those with normal hearing. Blewett3 reported that three students with hearing loss sufficient to justify clinical treatment scored h.23 above the mean score on one listen- ing test. While it is obvious that intelligence and listening are closely related, statistical data indicates that the correlation between the two is not high enough to predict success in listening. Nichols reported a l. Nichols, 22. 3313., 273. 2. Ralph Nichols, "Factors in Listening Comprehension", Speech Monographs, xv, No. 2 (19m) 8. - 3. Blewett, 92:.213-1 56. correlation of 0.5h; Johnsonl reported 0.333 and Knower2 reported a still lower correlation of 0.27. Because of such findings, the lack of concern with which the educa- tional world has treated listening has been dissipated to a great extent, and we are faced with problems of discovery of factors and relationships which are more significant than hearing and intelligence have so far proved to be. Despite the small amount of research that has been done in the field of listening, it is possible to assemble some basic assumptions. It is best to divide these assumptions into two categories: assumptions about listening itself; and assumptions about listening instruction: I. Basic Assumptions about Listening: A. Listening is the process of attaching meaning to the spoken word, or to aural symbols. B. Listening is a very significant medium of learning. 1. While listening does not guarantee learning, learning frequently results from listening. C. Because of differences in experience and intelligence, individuals differ in their abilities to comprehend the Spoken word. 1. Kenneth Johnson, "Effect of Training on Listening", Journal of Communication, I (hay, 1951), so. 2. Knower, Phillips, Koeppel, "Studies in Listening to Informative Speaking", Journal of Abnormal and Social Psychology, XL (February, l9b5), do. 10 II. Basic Assumptions about Listening Instruction: A. Research clarifying good and bad listening habits, and the skills, concepts, and attitudes upon which these are based is desperately needed. B. Although all teachers and parents should be concerned with the formation of good listening habits, teachers of English and speech have a major responsibility for the listening skills of students. C. According to research, listening can be improved through training. 1. Research indicates many similarities between listening and reading comprehension. Because good reading habits can be taught, this assump- tion seems reasonable. Training experiments in listening make this assumption more reasonable. D. Instruction in reading does not provide adeQuate train- ing for listening. 1. Despite the similarities between these two skills, there are some differences: a. The receptive mechanisms are different. b. Listening usually involves sight and hear- ing, while reading involves only the sense of sight. 0. Reading is usually a personalized activity, ll done alone; listening is usually a socialized activity, done in groups. d. Listening involves adjustment to the speaker's rate of speaking, while reading is accomplished by adjustment to the pace of the reader himself. E. Listening comprehension, manifested in test behavior, is measurable in quantitative terms, provided a valid and consistent test is used. F. Because, for this study, listening is defined as attaching meaning to the spoken word, meaningful materials, orally presented, must constitute the basic content of the evalu- ation instrument. G. Because listening comprehension inevitably involves some kind of response, either concealed or overt, the quantity and Quality of such comprehension can be measured in terms of observable human behavior. R. If research; supporting increased efficiency in listening after training can be trusted as either indicative or conclusive evidence, then it seems reasonable to assume that some or all of the processes that are involved in listening can be taught. 1. It seems advisable to include in this study only that research which is immediately applicable and pertinent to listening instruction. References will be made in the last part of this introductory chapter and in Chapter Two to these pertinent materials. However, excellent summaries of listening research, not immediately applicable to this study, may be found in the following sources: 12 Justification of the Problem An attempt to incorporate listening into the curriculum is likely to uncover two kinds of negative attitudes: apathy and antagonism. The national listening committee, in reporting its list of difficulties and problems encountered in the teaching of listening, rates these two attitudes as highly obstructive. Apathy stems, perhaps, from the two assumptions previously mentioned: listening is merely a matter of hear- ing acuity and intelligence, and little can be done about either. Antagonism seems to arise out of resentment, as if the protagonists were trying to legally adOpt into the educational family a child who has no right to be there. Therefore, to continue the metaphor, any justifica- tion of such legal adOption must Show that the Child has, in reality, been there all the time but has been neglected. In short, has listening come of age; is it entitled to a place in the educational sun? Harry Goldstein, Reading and Listening Comprehension at Various Con- trolled Rates, Contributions to Education, No. b2l, New York: Bureau of Publications, Teachers College, Columbia University, l9h0. Ralph Nichols, Factors Accounting for Differences in Comprehension of Materials Presented Orally in the Classroom, Unpublished doctor‘s thesis, State University of Iowa, l9ho. Miriam.E. hilt, A Study of Teacher Awareness of Listening As a Factor in Elementary bducation, Unpublished doctor's thesis, Pennsylvania State College, l9u9. Frank R. Elliot, "Memory for Visual, Auditory, and Visual-Auditory Material", Archives of Psychology, No. 199 (May, l93b) S. 13 l . . . . . . . Brown, in his justification, presents a number of reasons why listening should be taught. he cites Rankin's2 study in which it was found that forty-five per cent of our daily lives is Spent in listening. 3 Brown's article analyzes a study made by Walter Barnes in which it was discovered that ninety-eight per cent of the out-of—school activities of 7000 school children was Spent in oral activity. A quick survey of one's personal daily schedule will further substantiate claims that the frequency of listening activity is high. But frequency is not the sole justification for including listen- ing in the curriculum. Brown offers, further, that listening is more difficult than reading; that it is as important as, if not more important than, reading; that peOple do not automatically and without effort learn to listen effectively. To further substantiate the worth of listening, one can turn to studies in the field of education, such as the‘vviltlL study. This in- dicated that, even in the elementary grades, pupils are expected to listen 57.5 per cent of the day. One can turn to the field of business and human relations and find quotations like these from W. h. bennett, Coordinator of‘fraining for the Cities Service Petroleum Corporation: l. James Brown, "Why Not Teach Listening", School and Society, LXIX (February 12, 19b9) lhl-S3. 2. Paul Rankin, "Listening Ability.. Its Importance..heasurement..and DevelOpment", Chicago Schools Journal, XII (June, 1930) hl7-20. 3. Brown, 9p. cit., 150. h. wilt, 92. cit. ILL "In our group discussion analysis of mutual problems, we have found that whether any problem is in the field of organization or in human relations, in 98 per cent of the cases, the root of the problem has been a failure of communications. Someone has failed to understand what someone else meant."l "There are spoken communications and there are written communi- cations. But our basic problem is with spoken communications. And I think there is good reason . . . good man-to-man rela- tions and understanding cannot be achieved by written manifestos and memoranda."2 We can, in fact, turn to any phase of modern life and find ample proof that listening is important. But, granted that all of Brown's reasons, and others' reasons, for the teaching of listening are acceptable, there are two questions which remain to be answered before a justification for its inclusion in the curriculum is complete: 1. Are we so ineffective in our present listening skill that we need to teach it? 2. Can it be taught? In answer to the first question, most of the research tends to indicate that listening is superior to reading for comprehension of EB" complicated and relatively eagy materials, but, that for more complicated and difficult materials, reading surpasses listening. It follows, there- fore, that we are either to be content with this situation or devise ways and means to keep orally presented materials simple and uncomplicated. l. W. E. Bennett, "The Need for Effective Speech in a Technological Society", Journal of Communication, I (May, 1951) 17. 2. Ibid., 18. Further, research seems to indicate that an individual's listening skill slips below his reading skill when the teaching of reading is introduced. The Nashville1 study Shows that as pupils progress from the third grade on through High School, their listening ability, superior prior to this stage, falls behind reading skill. It must be remembered that while the pupil's listening skill begins to slow up, his need for listening does not diminish. Much of his further education is given him by the lecture method of instruction. His dependence upon speaker-listener relation- ships in the business and professional world is easily apparent. In a study made at Michigan State College in l9h8,2 the writer discovered the following things: 1. Less than 27 per cent of the students tested could pick out the main points of a well-organized informational talk. 2. Students tested had approximately 60 per cent accuracy when asked to draw inferences and make judgements based on information presented orally to them. Summarizing the various studies that have been made of the efficiency of listening, it is possible to see these two indications: 1. Participants in the studies seem to be from 25 per cent to 60 per cent efficient in their ability to comprehend orally presented materials. 1. Listening Abilities of Pupils in the Nashville Public Schools, Nashville Public Schools, Nashville, Tennessee, September, 1950, h. 2. Charles E. Irvin, Unpublished Study, l9h8. 2. Efficiency will vary among peOple and within the same person among varying materials. Some peOple will listen very efficiently to directions, but they will be less effective in other types of speaker-listener activities. Therefore, in answer to the first question, listening efficiency, as far as present diagnosis can detect, is at such a level as would merit attempts to improve it. Can listening be taught? Chapter II will concern itself with the studies that have been made of training programs. There is evidence to Show that such programs have more than an outside chance of success if properly conducted. The chief concern seems to have been whether the information about listening was teachable and whether the processes in- volved in listening were of a teachable nature. Obviously, we do not teach listening just by making students listen. To argue that an increase in the amount of listening activity will result in an improvement of listening efficiency is not pedagogically sound. The poor listener may simply continue to cultivate his bad habits. Educators are often prone to say, "Practice makes permanent, not perfect". Bess Sondell has written that a good listener will do five things: 1. "See" an idea when he hears it. 2. Distinguish between essential points and details. 3. Distinguish between facts and Opinions. l. Bess Sondel, "Everybody's Listening“, National Parent-Teacher, (January, 1951), lb. 17 h. Distinguish between information and persuasion. 5. Make up his own mind about what has been said. The first four of these five lend themselves to instruction. The fifth might be helped by instruction in the first four. In Nichols' doctoral study, he listed 38 factors which, to some degree, influenced the listening comprehension of his test group. Of these, 19 seem to lend themselves to instruction, either directly or indirectly. These 1? factors are: Reading comprehension Recognition of correct English usage Size of listener's vocabulary Ability to make inferences Ability to structuralize a speech Listening for main ideas as Opposed to specific facts Use of Special techniques to improve concentration Emotional adjustment to the speaker's thesis Ability to see significance in the subject discussed Judging speaker effectiveness Respect for listening as a method of learning . Susceptibility to distraction High school Speech training of the listener EXperience in listening to difficult expository material Social ease of the listener Self-satisfaction of the listener Economic attitudes of the listener Worries of the listener about personal problems Note-taking ability A similar list of factors, 30 in number, were discovered at Michigan State College in a study of "critical incidents"1 in listening. Of this list, 20 factors seem to lend themselves to instruction, either directly or indirectly: 1. "Critical incidents" refers to the Flanagan method currently being used in the five year study by the American.Council on Education. See bibliography. 18 Clarity of points Ability to grasp central idea Ability to see some significance Clearness of the organizational plan Ability to see connections between parts of the talk Clearness of sentence structure Vocabulary Necessity for compensating for Speaker's weaknesses Amount of experience in listening to difficult expository material Feelings about benefits to be derived from efficient listening Ability to make inferences Ability to anticipate the nature and direction of the deveIOpment of talk Previous discussion of the subject Emotional attitude toward the subject Distractions Note-taking ability Necessity for adjusting to varying rates of speech Ability to memorize Amount of previous direct listening training Personal worries about grades, etc. Therefore, upon diagnosing the process of listening and studying the factors which seem to be present, one is lead to believe that much of listening may be influenced by teaching. If a comparison is made between reading and listening, the same conclusion is apparent. The correlation between reading and listening has been.measured from a low of 0.27 to a high of 0.82. It is generally believed to be about 0.78; this was the figure reported by Goldstein:L in l9h0. Since it is well established that training in reading Can increase effectiveness in reading, and Since reading and listening, despite their differences, are Similar skills, it can be assumed that listening can also be taught. 1. Goldstein, 32. git. 19 One final justification of this problem needs to be made. It is one of obligation. Institutions, like Michigan State College, which have courses in communications skills, have announced their intentions of teaching the four skills: reading, writing, speaking, listening. Therefore, any study which attempts to find out why, how, and with what success any one of the skills can be taught is justified in the light of the announced objectives. More effective communications demand that we forsake the comfortable position which assumes these skills to be acquired with a usable competence at an early age; that we continue to probe and analyze; that we continue to teach and test our way toward better materials and better methodology. The implication throughout this introductory chapter has been that an attempt will be made to teach listening and evaluate the residue of that instruction. CkhtPThR II CHAPTER II SURVEY OF LITERATURE A survey of the literature in the field of listening instruction cannot be very extensive. The experiments and prOgrams have been very few in number and very recent. AS was explained in Chapter One, the discussion here will be confined to studies that have been made in the teaching of listening. The great bulk of literature available in the field of listening covers comparisons between reading and listening, analyses of listening habits, and the like. We shall be concerned with nine Studies in instruction that appear to be most significant for the purpose of this study. It will be re- called that such instruction has generally been categorized as follows: 1. The listening laboratory. 2. The coordination of listening exercises with speech ' assignments in speech classes. 3. The direct-instruction method. Most of these nine studies will fall into one or more of these classifi- cations. Chronologically, the first study was begun in 19h7 by Kenneth 0. Johnson1 at Macalester College in St. Paul, Minnesota. The study is 1. Kenneth Johnson, "The Effect of Classroom Training upon Listening Comprehension", Journal of Communication, I (May, 1951), S7. 21 being reported in verbatim form because this author feels that the detail as given by Johnson is important enough to the results achieved to warrant no deletion. The report also serves as an example of the direct- instruction technique in the teaching of listening. According to Johnson: "The purpose of this research was to determine the effect of an experimental course in listening comprehension on the listening comprehension ability of a group of 112 second semester college freshmen . . . . "The experimental course reported here is intended to be no more than perhaps a guide or indication as to the direction which listening instruction should take in the future . . . . The author considered the situation in which the college student was engaged, and determined that a course in listening designed to help the student in the classroom lecture type situation would constitute the most beneficial approach. The materials used in the exercise-lectures which comprised the experimental course were selected with this approach in mind. Nine exercise lectures were deveIOped and systematically recorded on a Brush-tape recorder. They consisted of: l) a short lecture on the listening technique to be practiced that day or a review of the technique deveIOped the preceding day; 2) series of numbers, presented in a manner similar to a digit-memory test; 3) brief lectures and paragraphs selected for practice of the listening techniques. Each exercise- lecture was composed of approximately thirty minutes of recorded material. "Besides standard instruction in posture and general physical attitudes most conducive to good listening, the students were motivated by describing to them the need for better listening techniques. One of the brief lectures discussed listening for tOpic sentences. . . others presented the technique of repeti- tion . . . additional suggestions were made as to how to check new information for accuracy. "The listening tests were recorded abridged forms of the American Council on Education COOperative English Test on Reading Compre- hension (C2). All of the statistical results were arrived at 1. Ibid., SB-ol. 22 through the use of raw scores taken from these tests. The listen- ing tests were recorded under conditions similar to those of the exercise-lectures. "The experimental group was composed of 112 freshmen students selected at random from the second semester communication program at Macalester College. The control group was selected by matching each member of the experimental group with a student from the communication course according to sex, intelligence, and listening ability. "The sequence of administration of this experiment was: 1) a listen- ing test was given to all of the second semester freshmen; 2) the groups were matched on the bases previously stated; 3) the experi- mental group was given the nine exercise-lectures at the rate of three per week; h) . . . the control group occupied itself with a project devoted to pronunciation and Spelling; 5) the second listen- ing test was administered to both experimental and control groups at the conclusion of the experimental course or approximately three weeks after the presentation of the first listening test; 0) the last of the series of listening examinations was given to both groups during the last week of the regular semester or approximately eleven weeks after the second listening test. The eleven weeks constituted a rest period in that no listening instruction of any kind was given. "An analysis of the scores achieved by the experimental group on the second listening examination, after having been subjected to the experimental course, as compared to the first listening test clearly «demonstrates that the experimental group made significant improve- ment at the 1% level on the second examination. The standard error was 0.58 and the critical ratio was h.l. These statistics can be interpreted as meaning that a difference as large as occurred could result from chance less than once in 10,000 times. The control group, during the same period, showed only such improvement as could be attributed to chance 7h times out of 100. The standard error was 0.50 and the critical ratio 0.33. "A comparison between the second and third listening examinations produced equally striking results. It will be recalled that the third examination was given to both the experimental and control groups eleven weeks after the conclusion of the listening course and the second listening test. An analysis of the scores of the experimental group indicates a statistically significant difference at the 1% level. However, the difference was in the direction of regression . . . the control group, however, had regressed to a position significantly lower than that attained on the pretest. The experimental group at the conclusion of the third test was just 23 about at the same level of listening ability as it was before being subjected to the course. The control group had signifi- cantly less ability than it had at the beginning of the semester. " Johnson concludes that the experimental group made significant improvement after training and held its own far better than the control group after the leng rest period of eleven weeks. In l9h8, halph Nichols began a laboratory method for the improve— ment of listening at the University of’Minnesota. In an article1 pub- lished in l9h9 he writes: I'Do students actually improve in listening ability through direct training? he have no evidence of any kind to the contrary, and considerable evidence is accumulating which points to a strongly affirmative answer. . "In terms of results obtained on the two types of measuring in- struments we are now using, we find that in a typical term the total freshmen pOpulation on the average respond correctly to about 50 per cent of the items on the tests. Special training is given the lowest-scoring 10 per cent. At the end of twelve weeks, when all freshmen are retested with the same measuring instruments, the average score on the retest of the group given direct training in listening has each term been found to approxi- mate closely the average retest score of those who have not had direct training . . . the inference that the gains made by the group given direct training are real ones is strengthened when it is remembered that the directly trained students are drawn from considerably less than one-fourth of the total range in listening ability." The methodology employed by Nichols at Minnesota is clinical in nature. Those students indicating a need for remedial training in listening on the basis of the first test are subjected to laboratory techniques in which they listen to records of lecture segments, take 1. Ralph Nichols, "Teaching of Listening", Chicago Schools Journal, xxx (June, 19u9), 278. 2h progressively more difficult tests over the recorded materials, and receive instructional aid along the way. The fact that the lowest 10 per centgcan attain a score approximating the average for the entire ) freshman class{after such training: and can repeat this performance ' term after term, lends a heavy endorsement to this type of listening instruction. Stephens College uses a type of instruction that could best be called indirect counselling. The students listen to college lectures, gather afterwards and discuss the varying information and attitudes each gets from the same lecture, analyze their own personal listening abili- ties and deficiencies, and confer with members of the speech staff. Such a system perhaps fits into the category of integrating listening exercises with speech class assignments, although the conferences with instructors are unique with Stephens as far as this writer can ascertain. Recently, Stephens has employed a listening test composed of two short talks, excerpts of longer speeches; one is on the subject of Literature and the Other is taken from a Baccalaureate address given in May, 19h9. Each talk is followed by a 20 multiple-choice question test. Miss Betty Bebout, of the department of speech, reported the following range of scores and their means in successive months: October, l9h9: l2 - 3b with a mean at 25. January, 1950: lb - 37 with a mean at 28. February, 1950: 15 - 37 with a mean at 2b. 1. Personal letter received from.Miss Betty Bebout, April 21, 1950. 25 Although very little else is known about the tests and testing procedure, if these results are statistically dependable, one might say that there is indication of improvement reflected here. In l9h9, KennethFrohardt,l a student at the University of Nebraska, conducted a study in the teaching of listening. Fifteen listening projects were given to 89 students in regular college speech classes. One of these projects was given only to an experimental group, consist- ing of h3 of the total of 69 students. It was a 32-minute lecture dealing with the significance of listening and the factors involved in this skill. The experimental group was further motivated by being told that the project was designed to improve listening skill. Two tests were given to all 89 students. Using a formula2 of Garrett's, Frohardt found that on one test no significant difference was revealed between the experimental and control groups, but on the second test, the score difference of h.h6 proved to be significant in 73 chances out of 100, favoring the experimental group. While the study is meager in detail, and the instruction given was short, the residue of improvement seems to fit the pattern of all such attempts, no matter how expert or inexpert the instruction or test administration. Also in l9h9, the writer conducted a small experiment in listening training at Michigan State College. Based on an earlier analytical 1. Kenneth Grohardt, A Study of the Teachipg_of Listening, Unpublished master's thesis, University of Nebraska, 1939. .— 2. H. B. Garrett, Statistics in Psychology and Education, Longmans, Green and Co., New York, 1920, 129-l3h. 20 study of some 2h00 students, a five-Question test was prepared over a h5-minute talk on the subject of "Radio As A Means of Communication". This test was designed to ferret out the weaknesses discovered in the earlier study. These weaknesses appeared t3 be in the following areas: 1. Students could not visualize the structure of informative talks. 2. Students remembered detail but failed to associate that detail with the major points of such talks. 3. Students began to listen too late to get vital initial material. Three of the five questions in the test were aimed at these specific weaknesses. Prior to this talk on "hadio As A Means of Communication" and the administration of the five-question test, a matched group of 75 students was divided into groups: twenty students received one hour of in- struction in listening; fifty-five students had no such help given them. After the talk and the subsequent test, the group which had received the one hour of listening instruction had a mean score 10.2 points above that of the group which had received no instruction. To be sure, neither a group of such small size nor a test of so few items would withstand the withering inspection of acceptable statistical procedure. However, there seemed to be some significance in the fact that the. "trained group'. did an average of 3.2 points better on each of the three test questions which had been aimed at the weaknesses already mentioned. 27 One lecture on listening, no matter how long, can scarcely be called adequate instruction. Yet, in the Frohardt study mentioned earlier, the instruction consisted of but 32 minutes. At Maxwell Field Air Base, the listening instruction given by the staff of the Air University con- sists of a single one-hour lecture. The Air University staff feels that much is accomplished in that one hour. It is regrettable that no measurement of the efficiency of this instruction has been made. At the convention of the National Society for the Study of Communi- cation, held in New York in December,-l950, hr. Ralph Widener, of the University of Missouri, reported as follows: "My own study, completed this year at the University of Oklahoma, was primarily concerned with develOping a manual of instruction on 'how To Listen', that could be used along with the testing of listening comprehension, that might result in some positive effect upon the skill of comprehending orally-presented materials. "From the evidence received, it was discovered that there was some significant gain in comprehension after specific instruc- tion had been administered." There seems to be no valid reason to go into the detail of Mr. hidener's study to isolate statistics. his conclusion that a significant gain was made is his own and was apparently acceptable to his graduate committee. More important at this point might be the outline of his manual of instructions: 1. Paper read before the Listening session of the national convention of the NSSC, New York, December, 1950. 2. Ralph Widener Jr., Instructional.hanual on How to Listen, From an unpublished master's thesis, University of Oklahoma, 1950. 2b Section 1 Introduction Section 2 Knowing the Rules of Good Speaking Section 3 Preparation for Listening Section h Awareness of the Techniques of Persuasion Section 5 Beginning Your Listening Section 6 Reacting to'hhat is heard Section 7 Note-taking and Outlining Section 6 Conclusion At that time, December, 1950, Mr. hidener's manual was the only one of its kind in a form presentable for distribution. his approach to the teaching of listening falls into the category of direct instruction, as did the Frohardt and Irvin studies mentioned earlier. At shittier College, in 1950, David Kruegerl conducted a study in listening instruction. There were 72 students in his experimental group and 72 in his control group. They were matched on the basis of scores achieved on the National Council of Education intelligence test. his listening tests consisted of some items from the tests used at the University of Minnesota and some true-false items based on classroom lectures. Krueger's instructional materials consisted of two lectures, one on the basic eight points of good listening and the other on the psycho- Ilogical aspects of good listening. In addition, students in the experi- Inental group wrote five papers based on listening to college convocation speakers . Using arithmetic means, he reported gains up to at per cent in favor inf the experimental group. Again, this study is of the direct-instruction tripe . 1“. David H. Krueger, AgStudy of the Results of Teaching Factors of Listen- ing'Comprehepsion to College Freshmen in the Basic Communications 11 Course, Unpublished master's the51s, Whittier College, 1950. Up to this point, it is obvious that these studies are lacking in sufficient control of the many variables that are Operative in the listening situation. Some of the variables have been handled through statistical manipulation; some have not. The admitted presence of all forms of motivation in the training sessions might lead us to suspect that part of the reported gains are not genuine; at least they may not be direct results of the training itself. This is perhaps less true of the Johnson and Nichols studies than of the others. A final study to be reported here is the best controlled among the nine. The writer comes to this conclusion not only from the study itself as reported, but also from his own observation of the experiment. Reference is being made to the work of Heilman,l conducted during l9h9 and 1950 at Michigan State College. His students were selected from the second term of the Basic College course in'hritten and Spoken English. The experimental group numbered 23h; the control group numbered 220. These students were matched on the basis of listening pretests and psychological tests. The experimental group received 20 minutes of listening instruction each week for six weeks, totalling two hours. The instruction, as well as the tests, were recorded. The listening tests, administered prior to and after the training period, had reliability indexes of 0.60 and 0.8h respectively. l. arthur Heilman, An Investigation inlfleasuring and Improving the Listening.Ability of College Freshmen, Unpublished doctor's thesis, State University of Iowa, l950. 30 Heilman found that there was a total gain through the entire ex- perimental group, but statistically significant gains were registered only among those students who placed in the lowest quartile on the first test. In short, Heilman, following the experience of Nichols of the University of Minnesota, concludes that the original "ineffective" listener benefitted most from this training. He further concludes that the results justify the time and effort that the training involves. 0f the nine studies reported here, seven were of the direct- instruction type, one was of the laboratory type, and one was a combina- tion of the integration of listening with speech class assignments and the individual conference. All nine studies reflect, in varying degrees, a gain in listening effectiveness after training. While many of these studies leave much to be desired in the way of expert control of vari- able Operative in the listening situation, there is still some hepeful basis for further research in listening instruction. Information is not available as to the present status of such instruction at the institu- tions where these studies have been conducted. The exception to this is the University of Minnesota, where the listening laboratory has been expanded beyond the needs of the remedial group even as far as servicing the faculty who wish to avail themselves of its help. There are other studies under way, but to this writer's knowledge, none have progressed to the point where they might be publicly appraised. As was stated in.Chapter One, a retarding factor in experimentation has been the lack of instructional materials and measuring devices. The 31 reader should judge such studies as those reported here in the light of the comparative newness of listening research. Those involved in such infant research ask the indulgence of those who judge, until such time as listening "comes of age." ChAPl'iz‘Jt III CHAPTER III hISTOhICAL BACKGhOUNU OF THIS STUDY When.Michigan State College began its course in written and Spoken English in l9hh, it included, as one of the objectives, the improvement of listening skill. Listening is incorporated into the four skills of communication. At that time, the planners of the course permitted one hour per week to be devoted to the training of listening. It was looked upon as a listening laboratory. Therefore, the course was set up in the following fashion: A. Four hours per week in class devoted to speaking, writing, and reading. B. One hour per week in a lecture session devoted to listening. In order to get the course underway, the original staff planned and gave a series of informative lectures in this one-hour period. The content of these lectures consisted of materials about language, its develOpment, its use, and its analysis. When this writer joined the staff of Written and Spoken English in 19US, the lecture program was as described above. The syllabus for the course announced, however, that training in listening was an integral part of the curriculum, hithout being unduly critical, it can be said that at this time the only listening training extant was simply practice in listening. In this respect, it could be said that this de- partment had done little more toward training listeners than any other department that offered the lecture experience to its students. As interest in listening began to grow among certain staff members, and as listening began to attract attention at other institutions, certain timid steps were taken toward the develOpment of a more positive program. In l9h7, this writer conducted a diagnostic study, to which reference has already been made, and discovered a fairly accurate listen- ing skill profile for the average freshman at this institution. Similar studies were reported at other colleges and universities. Results of such studies indicated that the listening skill of college students was far below the level on which college teachers supposed it to be; particularly those teachers who used the lecture method of teaching. The inference was drawn that mere practice in listening was not suf- ficient to improve the skill. At about this time, the department of Written and Spoken English created a listening committee composed of four staff members who were deeply interested in this field. This committee was charged with the responsibility of probing whatever literature was available and setting up some positive program of instruction. After months of study and discussion, the committee planned a four- unit training pregram to be tried out in the winter term of 1950. Accordingly, four of the regular staff lecturers were asked to reduce the time of their lectures from fifty minutes to twenty—five so that 3h listening instruction could be given in the remaining time. Each com- mittee member taught one of the listening training sessions. ‘uhile no evaluation was made in terms of statistical measurement, many things of value were learned: 1. The students were interested and the motivation seemed sincere. 2. There were many "bugs" in the program that needed treat- ment and revision before a more ambitious effort could be made. 3. Lecturers who formerly had argued that it would be im- possible to cut their lectures to twenty-five minutes had found that they could do a very satisfactory job in the shortened time. h. Enough of the total staff had become interested that COOperation seemed assured if a larger effort were to be made. The experience gained from this four-unit training program pre- cipitated immediate plans for a full year's training program to begin in the fall term of 1950. The committee began the work of revising some of the previously used materials and sought new materials which would need to be added. Eventually, the committee had on paper the training materials for eighteen units of instruction. Seven were to be given during the fall term, six in the winter term, and five in the spring term. 35 It was decided by the department and the listening committee that such a program should be measured for effectiveness so that evidence could be secured on which to base its continuation or its abolition. Each of the three terms of training were given to each of three members of the listening committee as his complete responsibility, from materials and planning down through the final evaluation procedure. Every effort was made to conduct the program in such a way that the results of the evaluation would be reliable. Advice and c00peration of other college departments were sought in order to insure the success of the program“ Acknowledgment has already been made of the valuable assistance received. All of the detailed problems and adjustments are fully described in.Chapter Four. With no intention of influencing either the reader or the results of this study, it might be added here that the program proved successful enough that it became an integral part of thelflritten and Spoken English course in the fall term of 1951. The materials used originally by the committee of four were rewritten so that all of the staff lecturers could administer the training along with giving their regular informational lecture. CHAPTER IV CHAPTER IV PROCEDURE As indicated in Chapter Three, the organization of the Written and Spoken English Course at Michigan State College provides a natural listening laboratory. The course meets five hours weekly: four hours of recitation during which speeches are given, papers are written, dis- cussions take place, and instruction is given; there is, in addition, one hour of lecture. This lecture is the listening laboratory. Prior to the inauguration of the listening training prOgram, these lectures consisted of fifty-minute talks about communication and language. The listening training program was integrated with these lectures. The entire process of integration as well as other aspects of this study will be described in this chapter. THE SUBJECTS OF THE STUDY All incoming freshmen must take Written and Spoken English. In the fall term of 1950, the enrollment extended slightly beyond 2bOO. All of these students attended one of the one-hour lecture sections provided in the schedule of the course. Each lecture section averaged from 250 to 300 students. Because one of the members of the listening committee was endeavoring to construct and validate a new type of test of listening, it was agreed 37 that he should be given two of the ten lecture sections with which to work. Therefore, eight lecture sections, or approximately 2hOO students, remained to be involved in the listening study under discussion here. These eight sections were scheduled thus: Lecture A ....... Monday ...... . 6-9 Lecture B ..... 3. Monday ....... l-2 Lecture D ....... Tuesday ...... 2-3 Lecture E ....... Wednesday .... 9-10 Lecture F . ...... Wednesday .... 3-h Lecture G ....... Thursday ..... lO-ll Lecture J ....... Friday . ..... . ll-l2 Lecture K ....... Friday ....... 2-3 These eight lecture sections were divided into two groups of four. One group of four was designated as the Experimental Group and was to receive the listening training. The other group of four was designated as the Control Group and was to go on receiving the regular fifty-minute informational lecture as all previous sections had done. Each group, experimental and control, was so picked that each would have both morning and afternoon sections. The resulting division is reported as follows: The Experimental Group Lecture A ....... Monday ....... 8-9 Lecture B ....... Monday ....... l-2 Lecture E . ...... Wednesday .... 9-10 Lecture F .. ..... ‘Wednesday .... 3-h Number: 1200 students. 38 The Control Group Lecture D ..... .. Tuesday ....... 2-3 Lecture G ....... Thursday . ..... lO-ll Lecture J ....... Friday . ...... . ll-12 Lecture K ....... Friday ........ 2-3 Number: 1200 students. As can be seen, each group contained two morning and two afternoon sections. Each group also contained the same number of students. The experimental sections were chosen on.Monday andeednesday so that lecturers who had to give both a twenty-five minute and fifty-minute version of their lecture would not have to give both versions on the same day. THE INTEGPZTION 0F LISTENING INSTdUCTION WITH TdE LECTUEES Since all of the regular informational lectures in the course had been fifty minutes in length, the integration of the listening training program with these lectures involved: 1. The determination of the specific informational lectures with which such training units would be integrated. The reduction in time and content of these chosen lectures [\D o from the previous fifty-minutes to twenty-five minutes in order to provide time for listening instruction. 3. The determination of which two weeks during the term would be devoted to measuring the students' listening ability. h. The determination of which members of the listening committee 39 would handle each of the instruction units. Because each member of the committee was also a regular staff lecturer, conflicts had to be avoided. ‘he final outcome of the integration was as follows: 1. The fall term's listening instruction program would consist of seven units. 2. The first two regular course lectures should remain in full length and not be shortened by the addition of listening instruction. 3. The second week of the term should be utilized for the listening pre-test; the tenth week of the term should be utilized for the listening post-test. remaining six informational lectures should be to twenty-five minutes in length to provide time the listening instruction. seven units of instruction should be divided h. The cut for S . The among the Unit 1 Unit 3 Unit 5 Unit 0 Included in each of A. Ten and listening committee as follows: and Unit 2 ........ Mr. Irvin and Unit h ........ Mr. Dow ...................Puu Blending and Unit 7 ........ Mr. Hill the training units were two types of instruction: minutes of direct instruction about listening how to listen; this to precede the twenty-five minute information lecture. to B. Ten minutes devoted to a "check-out"l period following the informational lecture. The original plans, therefore, included seven direct-instruction talks about listening and seven check-out periods. However, only six informational lectures2 remained with which to integrate seven training units. Consequently, in order to maintain the training program as planned, the check-out periods for units 1 and 2 were abandoned and the direct instruction periOds for these two units were combined into one. Thus, unit 1 was given in the ten minutes preceding the informational lecture, and unit two was given in the ten minutes following the same lecture. This adjustment assured that all seven units of instruction would remain intact. The character of these first two units was such that the deletion of the check-out periods represented no serious omission. Later in the term, during the ninth week, another alteration had to be made. A change in College plans relative to the Thanksgiving holiday left an insufficient number of school days in the ninth week to cover all of the experimental sections with instruction. Therefore, the instruction originally planned for this ninth week was added to the eighth week. Once again, the check-out period had to be abandoned to Inake room for another unit of direct instruction. 1. The "check-out" period refers to ten minutes of oral questioning, after the twenty-five minute informational lecture, by the listening instructor. The purpose of this period was to check on the degree of application, made by the students, of the listening instruction of the day to the informational lecture of the day. 2- See items 2, 3, and h on page 39. Thus, after integration and alteration, the listening training program actually consisted of seven units of direct instruction, each ten minutes in length, and three ten-minute check-out periods. The program was preceded by a listening pre-test and was followed by a listening post-test. Instead of being spread over six weeks, exclusive of the testing periods, it became concentrated in five weeks. In total, the prOgram amounted to seventy minutes of direct in- struction and thirty minutes of check-out time, or one hour and forty minutes. Adding to this the two fifty-minute testing periods, the training and testing amounted to two hundred minutes, or three hours and twenty minutes. Transposing these figures into per cent of the total time, ten hours, devoted to'Written and Spoken English lectures, we find: A. 16.6 per cent spent in direct listening instruction. B. 16.6 per cent spent in the testing of listening. The full ten-week schedule, as it actually took place, will be found on the next page. The Listening Training Program Integrated with the Lecture PrOgram in Fall Term, 1950 Oct. Oct. Nov. 23-27 30 - Nov. 3 6-10 week 1 Description Informational lecture, "The Importance of Communication"; fifty minutes. Listening Pre-test; fifty minutes. Informational lecture, "Emotional Blocks to Communication"; fifty minutes. Listening training unit 1, "Listening As a Fourth Skill"; ten minutes. Informational lecture, “How and Why Language Changes"; twenty-five minutes. Listening training unit 2, "The Kinds of Listening"; ten minutes. Listening training unit 3, "Preparation for ’ ' II. ‘ Listening , ten minutes. .ii‘\‘\\ ._‘ Informational lecture, "Levels of Usage"; twenty-five minutes. Listening checkout period; ten minutes. Listening training unit h, "Exercising Emo- tional Control in.Listening"; ten minutes. Informational lecture, "The DevelOpment of American English"; twenty-five minutes. Listening checkout period; ten minutes. Listening training unit 5, "Structuralizing during Listenin0"° ten minutes. 0 3 Informational lecture, "The Dictionary"; twenty-five minutes. Listening checkout period; ten minutes. Date Nov. 13-17 Nov. 20-2h Nov. 27 - Dec. .34 I (d (D 10 :3.“ A3 Description Listening training unit 6, "Listening for Main Points"; ten minutes. Informational lecture, "The DeveIOpment of Language in the Individual"; twenty-five minutes. Listening training unit 7, "Listening for Comprehension"; ten minutes. Informational lecture, "The Personal Aspects of Communication"; fifty minutes. Listening Post-test; fifty minutes. at THE INSTRUCTIONAL EATERIALS The instructional materials utilized in this study were compiled by the various members of the departmental listening committee. Their sources were few. Nevertheless, each member was able to gather data pertinent to his assigned segments of the seven unit training program. After each member had collected his data, he submitted it to the entire committee where it was re-worked and styled for presentation as instruction. The materials had to be further adjusted to meet the ten- minute time limit for presentation which was imposed upon the committee by the character of the prOgram. In addition to the materials for direct instruction, suitable ques- tions for discussion had to be written. These questions constituted the bulk of the check-out period. They were designed to integrate with the direct-instruction materials presented in each unit and to tie in to the materials presented in the preceding unit. Further, they were de- signed to elicit answers which would reveal the effectiveness or non- effectiveness of the listening that had been done in that particular unit. The questions were to be asked orally by the listening instructor. The training materials and questions for each check-out period may be found in.APPENDIX A. THE LISTENING TESTS As has been stated, two tests were given during this study to all of the students registered in the Written and Spoken English lecture LS sections. The first test, or pretest, was administered prior to the beginning of the listening training program. The second test, or post- test, was administered immediately following the seven—unit listening training series. Adeguate measuring instruments in the field of listening were scarce. There was available to us only the Nichols tests which were constructed at the University of Minnesota and used extensively there. In all, the Nichols tests number six: One in Literature One in Economics One in Biology One in Sociology One in Psychology One in Chemistry These tests contained questions on materials presented in recorded ten-minute lecture excerpts. They have been analyzed as follows: "To test the listener's comprehension of material, 20 multiple— choice Questions were constructed to cover each of the lO-minute lecture excerpts. In building the original test of 120 items, an attempt was made to construct questions which would test the listener's grasp of important points or ideas. Despite this attempt to emphasize understanding, a substantial number of factual questions appear. "The six subtests were returned to the original contributors for checking . "After the test battery, six tests, were administered to the test population of 200, a complete item analysis was made. To avoid weighting any of the six parts of the battery, 72 of the 120 questions were retained. (this brought the total to 12 questions per test) The discriminating power of the 72 was determined through the use of the Flanagan table of values of the product- moment coefficient of correlation in a normal bivariate be population corresponding to given prOportions of successes. Correlations of .20 or higher, as read by the Flanagan table, after computing the prOportion of successes in the 27% scoring highest and lowest on the continuous variable, are said to indicate test items of "good" discriminating power; correla- tions of .30 "extremely good"; correlations of .50 are so high as to be regarded as fortunate and due to chance. "The Median coefficient of correlation of the 72 retained items, indicative of discriminating power, was found to be .36. The median coefficient of correlation of the 12 retained items on each test was as follows: Literature .30 Economics .U3 Biology .hO Sociology .31 Psychology .30 Chemistry .36 "The degree of difficulty for each of these 72 items ranged from 20 to 9d per cent. The items tended to cluster about the 60 per cent level; slightly easier than would be the ideal examina- tion." Because these tests contained only twelve questions each, it was decided that two tests should be given in the pre-test period and two in the post-test period. Thus, twenty-four Questions were given in each of the testing periods. The tests employed in each period were as follows: Pre-test: Biology Sociology Post-test: Literature Economics An item analysis of each of these four tests will be found on the next pages. 1. Ralph Nichols, Factors Accounting for Differences in.Comprehension of Materials Presented Orally in the Classroom, unpublished doctor's thesis, State University of Iowa, 19am, 10-22. TABLE I I'l‘riv". ANALYSIS OF LI STENII‘J G Pita-TEST b7 Difficulty Item Low 27% high 27% Average in % Power Biolo gy 1 7o 93 8h.5 .305 2 93 100 90.5 .325 3 55 61 be .295 h 61 93 77 .hSS 5 7t 9t 8h .36 5 LA 67 65.5 .LB 7 be 85 65.5 .h35 B 11 33 22 .31 9 26 6S D5.5 .LO. 10 85 100 92.5 .t75 11 At 91 67.5 .sus 12 7C 96 67 .38 median . . ...... . . .hO Soc iolo g; 1 67 87 77 .275 2 55 67 71 .365 3 st 7:2 as .27 u 22 55 38.5 .35 5 89 98 93.5 .30 6 57 96 91.5 .2h‘ 7 91 9d 9L.5 .275 8 9b 100 97 .30 9 7b 96 5o .Le 10 5h 51 67.5 .305 11 07 b9 76 .31 12 he 91 67.5 .53 hedian ............ .31 TABLE 11 L; (J IThh,AhALISIS OF LIS”EhIhG POS”-TEST Difficulty ten Low 27% high 27; Average in % Power Literature 1 70 91 q3.5 .255 2 to 69 57.5 .2' 3 57 65 71 .335 u 65 t9 77 .33 5 65 63 7b .235 6 55 7h 6h.5 .21 7 96 100 96 .23 6 55 7h 6t.5 .21 9 37 61 5' .DC5 10 52 63 07.5 .355 11 55 9c 91.5 .305 12 b7 98 93.5 .355 median ........ .30 Economics 1 52 87 69.5 .L15 2 33 76 55.5 .42 3 39 50 59.5 .h3 u 57 69 73 '50 5 67 100 93.5 .hLS 6 63 E7 75 .315 7 DH 83 63.3 .h25 8 39 63 Cl .b? 9 15 57 36 .hé 10 he 81 eh.5 .305 11 50 93 71.5 .535 12 65 96 80.5 ’.50 hedian ......... .L3 D9 The reliability coefficients1 of the tests used in this study are as follows: Pre-test: r BiolOgy, Sociology ........ 0.7h Post-test: Literature, Economics ..... 0.93 These coefficients compare favorably with the recently completed Brown~Carlson listening tests.2 Their reliabilities are 0.79 and 0.75. Since the Nichols tests and the Brown—Carlson tests are the only extant measuring devices in the field, these reliabilities must serve as adequate. The validity of the Nichols tests has not been established. This very obvious deficiency is primarily due to the lack of suitable evalua- tion devices against which such tests could be contrasted or with which they could be compared. Such a deficiency exists in all other measuring devices in this field. Validity, then, must be assumed. Since the materials over which the tests are given are presented orally; and since 1. Ibid., p. 20. Computed by the Kuder-Richardson formula: r = n . s.d. 2 — fpg‘s n—l s.d.2 n the number of test items d. the standard deviation of the examination f the frequency p the percentage of successful answers 4 the percentage of unsuccessful answers 2. Brown~Carlson Listeninngests, In process of publication, world book Company, Ionkers-on-the-Hudson, New York, 1951. 50 1;}:163 tests themselves are designed to measure the degree of comprehension c);fT these materials present in the listener: and since, for the purpose . ()LfT this study, listening has been defined as "attaching meaning to the 5333<3ken word" it must be assumed, therefore, that we are measuring listen- :i.Iugg ability as expressed in test behavior. The correlations of these “tns:sts with other factors that might be considered variables in the Ilgi.stening situation are not sufficiently high to warrant argument that 1;11ese tests might not be valid at all. Some of these correlations are:1 Intelligence (ACE) ... ..... 0.53hb Reading (Iowa Silent) .... o.uoo2 Scholastic rank .......... 0.2791 Vocabulary (COOperative English) .. o.u9uo As has been stated, the materials over which the tests were given, ‘weezee ten-minute lecture excerpts, each having been prepared by the de- jpéazétment represented by the content; i.e. the Biology department prepared tries materials for the Biology lecture, etc. To eliminate the variable 01? eenmade to maintain a normal classroom lecture situation. This con- iTOI‘ms to the real purpose of this and similar studies; to provide a filrmer basis for further progress in listening traininO. In this study, three speaker personalities were involved: 1) the FHersonality of each week's regular staff lecturer; 2) the personality of 5b the listening instructor; 3) the personality of the "voice" on the recorded materials over which the tests were given. As has been stated, the regular staff lecturers, within any one week, lectured to both the experimental and control groups. Such a system controlled this aspect of personality variation. however, the experimental group had the vary- ing motivation of four different listening instructors. A more adequate system might have been to permit one listening instructor to give the entire seven-unit series. Yet, even though this aspect of personality difference was not controlled, the materials in the actual testing procedure, for both the experimental and the control groups, were de- livered by the same voice. The recording eliminated to an extent the full force of the speaker-present personality; but it cannot be said that a recording does not possess some personality. Studies have shown that it does. Thus, as far as Speaker personality is concerned, the materials actually tested were delivered by one speaker, through records; the informational lectures were given by the same lecturer; and the listening instruction was given by four different staff members. The controls existed in the teeting and the practice, but not in the instruction. Inconsistent attendance at training sessions was controlled by throwing out test results of those students who indicated that they did not attend all such sessions. The reader will recall that the test administrator gave instructions prior to the post-test that would enable the committee to delete those answer sheets which indicated incomplete attendance. This, of course, affected only the experimental group. 59 After such deletions were made, the original number of approximately 1200 students drOpped to 1003. Variation in verbal intelligence among the students was first analyzed by running a correlation between the listening pre-test re- sults and the students' achievement on the linguistic section of the American.Council on Education test series. Later this factor was minimized by application of the analysis of covariance. The possibility that the listening aptitude of the students would vary to the point where matching would be necessary was a variable that had to be controlled. It was originally planned to adjust such differ— ences through application of the covariance technique. However, the scores on the listening pre-test, which would have served as the diagnostic guide to such differences, revealed that the experimental and control groups were so evenly matched that further control of this variable seemed unnecessary. These pre-test scores were as follows: Pre—test Scores (perfect score 2h) Group Number of students Mean Experimental 500 17.078 Control 500 l7.b57 The difference, 0.379, favoring the control group appears to be of such negligible significance that a safe assumption would be that the two groups were as well matched as could be heped for as far as listen- ing aptitude was concerned. The distribution curves of both groups were equally similar. Figure 1 on the next page graphically shows the similar trends of the distribution of scores on the pre—test. The fifth variable is the interest of the student listeners, at the time of listening in the materials being presented. Obviously, 0) such interest can neither be measured nor controlled; at least, it can- not be effectively controlled. Students have, in some studies, been given questionnaires and grouped according to their own subjective judgments as to their interest in subjects. Such a method cannot be trusted. In other studies, materials of a "nonsense” nature have been used in order to eliminate the interest factor. Such investigations neither conform to the so-called normal listening situation nor do they throw any significant light upon the measurement of comprehension. The interest of the listener in the materials being presented is an important factor. Nichols rates it high; the British Broadcasting Company,1 in its studies of educational broadcasts, also rates interest as perhaps the most important factor in comprehension. however, until some acceptable method of measuring and controlling interest is found, studies such as this one must recognize its existence, but admit frankly to an inability to control it. Previous knowledge of the materials presented is similar to interest as a variable. It is difficult to ascertain the extent of such knowledge and the methods of ascertaining it are not dependable. In this study, 1. Joseph Trenaman, Understanding Broadcasts on Science, unpublished study, British Broadcasting Company, September 5, 1950. (DQJCd'UJ HJO HmU‘BCEZ' (Dc-+5 9O 80 70 SO hO 30 61 \. -- Cohtrol ,1 Group \ .\ ‘\ Experimental "\ group --- / I 6 7 8 9 10 ll 12 13 1h 15 lo 17 18 19 20 21 22 23 2h Score Figure 1. Distribution of Pre-test Scores. Five Hundred Students in Each Group. C2 the choice of subjects for testirg, as was explained under testing procedure, was made in the attempt to control this variable. It seemed highly unlikely that incoming freshmen would vary much in their previous knowledge in the areas of diology and Sociology. Thus, these two areas were used in the pre-test. t also seemed likely that such students would be fairly matched, after one term, in the areas of Literature and Economics since few, if any, would have had courses in these fields during their first term. These two areas were used in the post-test. Consequently, control of this variable was made through the choice of materials. The assumption about these materials on the pre—test seemed to be valid inasmuch as the scores and distributions of the scores among the two groups compare favorably. Variable number seven, the receptive mood of the listener at the time of listening, is again similar to the two which have just been discussed. The presence of fatigue, emotional imbalance, challenge of the subject or lack of it, worry, and all of the other human behavior tendencies that tend to make concentration difficult are factors in the listening process. This study recognizes them as important. However, the only attempts to control this “receptiveness” were these: 1. An earnest attempt not to offer rewards in the form of grades, etc. to the listener which might be con- sidered an extra motivating factor favorable to receptiveness. 2. An explanation within the listening training itself of the power of worry, emotional upset, etc. as a deterrent to effective listening. .At this point in listening research, it must be conceded that receptive- ness at the time of listening is a variable that must be considered ti ’ uncontrollable. Some studies have sought to overcome it by extra motiva- tion devices. However, such devices have tended to destroy the normalcy of the listening activity. The same might be said of the last variable to be discussed, the intentional or unintentional motivation provided by the test administra- tor. It would be easy for such an administrator to slip into his instructions some words of encouragement or offerings of grade rewards and the like. Such motivation would tend to destroy the validity of any results achieved. This was controlled in this study by constructing a very brief, step by step, set of instructions and having the same administrator give these same instructions each time. The script for these instructions was given earlier in this chapter. In summary of this section on control of variables, three variables in the listening process seem to defy control: 1. Personality of the lecturer. 2. The interest of the listener in the materials. 3. The receptive mood of the listener at the time of listening. 'While some control was exercised over the personality of the lecturer, it is frankly admitted that the other two must be considered as operative in this study in whatever extent they exist. CA THE SELECTION OF THE TEST POPULATION FOR STaTlSTICAL ANALYSIS Since one of the variables in this study was inconsistent attendance at listening training sessions, the first step in the selection of a test population for statistical analysis was the deletion of all test answer sheets on the post-test which indicated that the student had missed one or more of the seven units of listening instruction. Such a step applied only to the experimental group. After this deletion, the actual number of students remaining in the experimental group was 1003. Rather than use the entire group, the writer decided to use a group of 500. This number seemed large enough to make the sample more than adeQuate. In order to select this 500, the writer alphabetized the answer sheets for the post-test, and beginning with the first one, counted three, which would bring him to the first answer sheet of 1000, and then by random choice selected every other answer sheet. This se- lection resulted in the desired number of 500. These 500 answer sheets of the experimental group's post-test scores were then matched with the pre-test sheets for the same students; these two answer sheets were clipped together. A typed list, by lecture sections A, B, E, F, was then made with the scores of each of the 500 students Opposite his name. Approximately 1230 student answer sheets were available for the control group. No deletion for missed units needed to be made, since the group had received no listening instruction. The same process of selection was followed. Beginning with the first one, the writer counted i H‘ three, which brought him to the first of 1227. he then selected by random choice every other one until 227 answer sheets were isolated. These were deleted, leaving 1000, the same number as existed in the experimental group when the random sampling of that group was made. 0 At this point, further selection was made, as was done in the experi- mental group, by taking every other sheet until 500 was reached. This 500 became the sample of the control group. Again reference can be made to the similarity of pre—test scores and the similarity of score distribution as a defense of this method of selection. Lists were made by lecture section of the control group and names and scores were typed into these lists. The lists for both the experi- mental and control groups will be found in.APPENDIX C. These 1000 students, 500 experimental and 500 control, make up the test sample, analysis of which will follow in the remainder of this study . ChiLPi'l-Jh V CHAPTER V ANALYSIS OF RESULTS Earlier in the discussion of this experiment, it was pointed out that the results of the listening training and testing would be probed to discover three things: 1. If there is a difference between the listening test results of the group receiving training and the group not receiving it. 2. If there is a difference in test results between the sexes. 3. If there is a difference in test results between the times that groups convene for listening. Before getting into such an analysis, however, it is necessary to remind the reader of several qualifying circumstances. Because so many variables Operate in the listening activity, every attempt was made to eliminate all possible elements that might influence results. One of those is the test itself. The same test, given more than once, can influence results because students become acquainted with it. To eliminate this re-test factor as a variable, a different listen- ing test was administered at the end of the term, after training, than was given at the beginning of the term, prior to training. This eliminates any possibility of comparing either the experimental or the control group I‘ ' O? with itself. However, in the extreme high and extreme low scoring brackets some kinds of comparisons were possible and were made. It was also pointed out that certain variables Operative in the listening activity could not be controlled at all. These were the pl ' interest and mood of the listener, and the personality of the speaker. Two other variables, listening aptitude and verbal intelligence, have been minimized to great extent by other means. however, all of these variables may be minimized by the application of the analysis of co- variance. This is easier to see by the following small chart: Variables . Controlled by Controlled by ' Analysis of Variance Analysis of Covariance Sex ~ Listening aptitude Time Interaction of these Any or all Of the variables not specifically controlled might be Operative to make differences in listening aptitude. Therefore, the adjustment made in the post-test results to account for differences that appeared in pretest results, by the analysis of covariance, would minimize such vari- ables. In addition to this, with respect to verbal intelligence, a co- efficient of correlation was computed between the test results on the pre- test and the scores made by the same students on the American Council of Education Linguistic test. Ob Using the Pearson Product-Momentl formula, the following results were obtained: Correlation of Pretest with ACE Linguistic test Experimental group 0.1102 Control group 0.1103 The magnitude of both coefficients is not sufficiently high to be compelling evidence that a non-chance relationship exists between the listening scores and the ACE (L) scores. However, these coefficients do have a significance level Of between S and 10 per cent which suggests a possible positive relationship. Our best estimate of this possible relationship is the low correlation of .11. It did not seem that the linguistic factor, other than as it is reflected in the pretest score, should be included as an additional control variable in the study. However, recognizing the possibility that the ACE (L) test might not be a good measure of verbal intelligence as it applies to listening, this variable was not ruled out with the revelation of the insignificant correlation. It was still under consideration as a possible factor in- fluencing test results when the covariance technique was applied. Variation of listening aptitude was also discussed as a variable Operative in the listening situation which might very well be respons- ible for differences in end results. Despite the similarity of the 1. r=N XY-(X)(‘1’) N x2-( x)2 . N i3 -( 102 N number of students X sum Of Pre-test scores Y sum of AOL scores I" 09 results between the experimental and control groups on the pretest, as fully described in Chapter IV, this variable was likewise under con- sideration at the time of analysis of post-test results. The covariance adjustment of the post-test scores was made with this variable, as well as others, in mind. Therefore, with this brief recapitulation of certain qualifying cir- cumstances, the following analysis may be easier to comprehend. In Tables III, IV, V and VI, the arithmetic mean scores are given for all sections of both the experimental and control groups and the combined arithmetic mean scores of the groupings about which we are most concerned: Experimental vs Control hale vs Female horning vs Afternoon In.Figures 2 and 3, the distribution of scores for both the experimental and control groups is given for both the pretest and the post-test. However, before beginning any interpretation of these score results, it is apparent that some explanation should be made of the results shown in Table V. In every instance except the male-experimental, and morning experimental group, the post-test scores are lower than the pretest scores. This might appear unusual to the casual observer, particularly in the case Of the experimental group since they had received seven units of instruction in the interval between the two tests. Table V is not as confusing as it might appear. It must be remembered that the same test was not administered as a post-test as was given as a H" 7O pretest. Therefore, the scores in any group can not be compared, one test with the other. Our only concern, therefore, is with the post-test scores and with the fact that, as the post-test was being administered, one group taking it, the experimental, had had listening instruction, and the other group, the control, had had no instruction. The post-test scores, as they are shown, do not represent a "loss" in listening ability since they can not be compared with the scores on the pretest. Had the tests been the same, the results might prove embarrassing. Turning our attention to Table VI, we find here the scores that may be meaningful in this study. 0n the post-test, the experimental group had a mean score on the 2h-guestion test of 16.9h8 which was 1.190 better than the mean score of the control group. This figure of 1.190 has very little meaning unless it can be transposed into something meaningful. It represents 7.1 per cent of the combined means Of both groups. In the light Of this per cent interpretation it might be said that the experimental group, because of the apparent positive gain represented by the difference in means of 1.190, did 7.1 per cent better on the post-test than did the control group. However, this would only be true if some factors other than listening training were not present. Therefore, further analysis of these apparent gains must be made. In the final analysis, we are only concerned anyway with whether or not any gain, no matter what per cent it might turn out to be, is significant or not. It is interesting to note that in the post—test results the males had higher means scores than the females in both groups. According to H" TABLE III MEANS OF THE EXPmnIhENTAL GhOUP BY LECTURE SECTION 71 Morning Lecture A Lecture E Male Female Male Female I?rwe-test 17.027 17.611 16.969 17.333 Post-test 17.608 17.361 17.565 16.637 Afternoon Lecture B Lecture F Pre -test 17.1t3 1o.3'zo 17 .1185 1o.ou1 PRDSt-test 17.383 16.000 16.865 15.923 '0‘ TABLE IV MLANS OF THE CONTROL CHOUP iBY LECTURE SECTION 72 Morning ._._._..____§L_ Lecture G Lecture J Male Female Male Female IDIVe-test 17.669 17.583 17.972 17.630 Post-test 16.515 15.000 16.903 16.087 Afternoon Lecture D Lecture K Pre-test 17 .731; 16.711; 17 .362 16.600 Post-test 16.1493 lh.661 15 .316 15.0o7 '5 73 TABLE V MLAN SCORES FOR BOTH EXPERIMbNTAL AND CONTROL GROUPS Experimental Pretest Post-test bdale 17.166 17.h15 Iremale 16.969 16.h61 biorning 17.235 17.3h6 lifternoon 16.920 16.5h8 'Total Experimental 17.078 16.9h8 Control iMale 17.73h 16.3lh Female 17.161 15.203 IMOrning 17.763 16.126 .Afternoon 17.152 15.391 Total Control l7.h57 15.758 ‘1 Iw11nnber cxf Students 7b 80 70 50 8 9 10 11 12 13 1h 15 1o 17 18 19 20 21 22 23 2E Scores Figure 2. Distribution of Listening Test Scores Among the Egmfimmmd_mbwn Number of Students 500 Pretest Mean Score 17.078 Pretest Post-test Mean Score 16.9h8 -?---- Post-test i v0 ' Number of Students 75 9O 80 ..4 70 to 50 to 30 20 10 o 7 o 9 10 11 12 13 13 15 16 17 18 19 20 21 22 23 2E Scores Figure 3. Distribution of Listening Test Scores Among the Control Group. Number of Students 500 Pretest Mean Score l7.h57 Pretest Post-test Mean Score 15.758 ------ Post-test vi ‘ TABLE VI DIFFERSNCES BETWEEN MLAN SCORES ON POST-TEST 76 Experimental Experimental Control Gain or Loss Hale 17.u15 16.3lh / 1.101 Female 16.ht1 15.203 / 1.278 Morning 17.3h8 16.126 / 1.222 Afternoon 16.5h8 15.391 / 1.157 Totals 16.988 15.758 / 1.190 Table VI, the 0.93h and the control males had a plus difference of 1.111., Combining both groups, the males had higher mean scores of 1.022. figures have no meaning other than possible significance, which will experimental males had a mean score difference of plus have to be analyzed. Again, these The morning groups received higher mean scores than the afternoon gTOUps; plus 0.800 in the experimental, and plus 0.735 in the control. Significance will also have to be tested here. At this point, then, we are concerned about two things: 1. Are these differences in meansdue to listening 2. Are these differences significant? training? ya Following a statistically acceptable method of applying an analysis of covariance, according to Snedecor,l the following results were obtained: The scores were arranged according to the following pattern: Group Sex Time Pre-test Post-test Experimental Control In order to eliminate supererOgative and troublesome design, the following procedure was used: the smallest number of students in any one Of the above categories was determined and used as the group experi- mental unit. This number occurred in the female afternoon control group and turned out to be 63. It then became necessary, by random choice, to reduce the number of students in all other groups to 63. The distribu- tion is as follows: Experimental Control Morning 63 63 Male Afternoon 63 63 horning 63 63 Female Afternoon 63 63 Total number of students ~- 50h 1. George w. Snedecor, "Two Variates in Two or More Groups", Chapter 12, pp. 318-327. In Statistical Methods, Iowa State College Press, Ames, 19u6. I H' The summary values for each DATA ON 50h STUDENTS INVOLVED IN’A STUDY TO DETEMEINE THE values Opposite (S) are the sums listening tests; (82) represents TABLE VII and (SS)represents the sum of the products. group are shown in Table VII. EDUCABILITY FOH EFFECTIVE LISTENING The 76 of raw scores on the initial and final the sum of sguares of those raw score5° 1 Initial Final Grogp Sex Time Test Test 8 1019 1175 A.n. sz 17075 22203 . 53 19270 “ale 3 1oo7 llh6 P.M. s2 ltSAS 2125b Experi- as 19721 mental S 1073 109h A.M. 33 18735 195o6 ss 156st Female 8 1022 1017 P.M. s2 1&968 17019 83 1o727 s 113H‘ 1036" AJfl. 3 20938 19672 ,7 0 ss 1&9h9 ”T“ s :mfi; 9m P.M. 83 19536 16577 ss 171h3 CODLI‘Ol S 1673‘“ 609 A.M. 52 18793 129u1 . as 15399 Female S 10h9 912 P.M. 52 17913 lAOAZ SS 15565 i It ' The first Question about these figures that is easiest to answer is, "Are differences in the average performance between the students, classi- fied by instructional units, sex, or time of day, sufficiently large to suggest a £331 and not a chance difference?‘ By a straightforward analy- . . l . , Sis of variance we find that some are and some are not. TABLE VIII ANALYSIS OF VAHIANCE OF FINAL LISTENING TEST SCORES * Variation df 88* NS* Significance Groups 1 797.5317 797.5317 Sign. at It level Sex 1 327.0555 327.0555 Sign. at 15 level Time 1 h9.5317 h9.5315 hot significant G x S 1 0.3869 Not significant G x T 1 5.7856 Not significant ‘8 x T 1 b.960h Not significant G x S x T 1 h2.293o h2.2936 Not significant within A96 7655.7770 15.h35o Total 503 8803.3258 * df = degrees of freedom; SS = sum of sguares; NS = mean square. It appears that differences in means between the experimental group as a whole and the control group as a whole are significant at the one per cent level, favoring the experimental group. This means that a difference as large as that found will occur by random fluctuation less than one time in a hundred. Because of the rarity of chance Operating to produce a difference as large as this, it may be assumed that a 1. Snedecor, pp. cit., p. 318—32h. yl 60 non-chance factor is Operating. The same is true about the difference between the male and the female students. This difference is also significant at the one per cent level. The difference in mean performance that existed between the after- noon and morning groups, however, turned out to be not significant. A non-significant interaction, such as shown for GxS, GXT, SKT, GxSxT, indicates little evidence of any relationship between listening ability and the joint effects of the factors of these particular interactions. This evidence can be found in Table VIII Opposite the indicators GxS, GXT, SxT and GxSxT. Having thus established significance in the differences between the means, experimental XE control and male XE female, it then becomes necessary to determine whether or not this significance is due to the variable being measured, listening training, or to some other variable that might have been Operating to influence these results. As has been indicated, verbal intelligence, even though a low correlation was found (0.11), might possibly be at work in these results. Therefore, a test of significance among adjusted group means, or covariance, was run. Table IX indicates the results of this statistical analysis. Table IX will be .found on the next page. According to the results found in Table IX, the regression adjust- nuent made in this analysis shows that group and sex differences in pezrformance on the post-test remain highly significant. Evidently, the dihfferences in listening ability existing between sexes and between the exqverimental and control groups prior to the initiation of listening H' 81 ‘ .0mm .d ..pflo 2mm .hooopmcm .H Hm>mH aH pm .eMHm amen.@eoH H amOQ.QOOH .o .aea ems ommm.moma ammo.Hm04 asa0.ammm maom.mmsw ems posse masons Hm>mH aH pm .eaHm mews.emm H mews.omm H xmm .aea was momw.aamo aoeH.mwmm a:m3.awmm mmmm.mwma am: hoses xmm pamOHeHemHm poz Hemm.0m H Hemm.om H meHa .aea om: smm®.~oHo 00H:.msmm mmmm.Haem mmOm.mo~a was gonna meHa sceOHchmHm p02 was osoo.HH -H emu .nea ems amms.mmHo sta.oamm maam.mwem :Hao.na0a ems scene emu weoe.mH mas mnem.HqHs om0m.saam smwm.om4m @aaa.mmoa was :HepHe Hoem.om mmHm.mm ommm.ms H a x m x o oops.mH Hemm.s- geom.e H a x m amsm.a mmmm.@- wmsa.m H a x a oosN.om meaH.m- msgm.o H m x e woom.® osoo.Hm aHmm.as H meHa ameo.nH mme.wa mmmo.amm H xmm awoe.©m mHam.mHm- aHmm.ama H masons mmwm.NNHs eowH.mmmm smmm.msss mom Hmpoa moeeoHaHQMHm me an mm New mum mam mmmeHpmfl mo whoppm mposbopm bum mmhmdrm mo mafim mp EOHpmHLw> H NH mqmfia mzmflz omemshmd m0 moz¢Hm<>Oo mo mHmMAHN named. 91 experimental group in score points is nearly twice that of the control group. The loss, on the post-test, which was suffered by those in the experimental group who failed to do better after training, was one full point less than the loss suffered by the control group. In Table XIV, the distribution of post-test scores again highly favors the experimental group. One more set of figures may make this clearer to the reader. Table XV analyzes the percentages of loss and gain of both groups in this category of 15 and below. Table XV shows once again a decided advantage on the post-test of the "trained" group, as far as this below—average category is concerned. The amount of gain in the test results was greater and the amount of loss was far less. It appears, therefore, that, as far as these test results Show, this below-average group enjoyed a substantial residue of gain which may, with reason, be attributed to training in listenino. The gain not only permitted a great number of students originally in this category to move out of this bracket and up the scoring scale, but the gains that were made were substantially higher than those registered by the control group. Logically the above-average group should now be analyzed. Accord- ing to Table XII, the percentages of both groups in this bracket were as follows: Score Experimental Control Bracket Pre-test Post-test Pre-test Post-test 16-2h h5% 30.2% 55.2% 28.2% '6' 92 TABLE XV AMOUNT OF GAIN AND LOSS ON POST-TEST OF STUDENTS IN THE 15 AND SELON CATEGORY Number Pre-test of Percentage on Post-test Score Scores Gaining Losing Experimental 15 ua 65% 31% 1h 39 D9 36 13 29 b5 h2 12 16 SO hh ll 10 20 80 10 3 o7 O 9 2 O 100 8 l O 100 1E5 99219.; 15 h2 30% h9% 1h 33 2 77 13 ll O 100 12 1h O 100 ll 6 O 100 10 2 O 100 9 3 O 100 8 l O 100 H [.4 [\J 93 It is obvious that neither group maintained itself in this above- average category on the post-test. Of the 22S experimental students who were in this bracket on the pre-test, only 151 remained in this bracket on the post-test. This is a loss of 7h students or 32.8 per cent. Of the 276 control students who were in this category on the pre-test, only lhl remained on the post-test; a loss of 135 students or h9.2 per cent. 'While there is a spread of difference between the experimental and the control which favors the experimental, the fact that neither group held its own in the post-test results nullifies any possible advan- tage which might be attributed to listening training. Figures 6 and 7 will show the relative distribution of both groups on both tests. From Figures 6 and 7 it can be seen that while the experimental group had 51 fewer students in this high bracket on the pre-test, it had 10 more students than the control group in this bracket on the post-test. Again, despite the fact that both groups suffered losses in this high bracket, the experimental group's loss was not as great as that of the control group. A summary table for this bracket, similar to Table XIII for the below-average bracket, can be found on page 96. As in the over-all results, there can be found in this bracket a consistent pattern. The experimental group seems to have higher scores than the control on the post-test and the male group seems to have higher scores than the female. Again it needs to be emphasized, however, that i It ' fiosmcawdfij Control \ ------ Group Experimental 20 Group 18 19 20 21 22 23 2h Scores Figure 6. Experimental 3§_Control Groups on Pretest in Scoring Bracket 18-2h. Number of students, experimental 225 Number of students, control 276 Per cent of total 500 experimental h5.0% Per cent of total 500 control 55.2% 9b 1! ' ‘4 0 a m g Jim w u: SO ho \‘5 30 Experimental Group 20 Control Group ----- - .. .. .\ 10 O 18 19 2O 21 22 23 2h Scores 95 Figure 7. Experimental vs Control Groups on Post-test in Scoring Bracket 18-2h. Number of students, experimental Number of students, control Per cent of total 500 experimental Per cent of total 500 control 151 1141 28.2% It ’ TABLE XVI 96 SUMMARY STATISTICS OF CONTROL Ann EXPLnInonTAL onOOPs IN Tan. 15 TO 2h SCORING BnACKnT ,O '0' Experimental Control Pre-test Post-test Pre-test Post-test Mean score 19.566 17.760 19.502 17.23h Mean difference Pre-test / 0.06h Mean difference Post-test / 0.526 beans Male morning 19.667 18.5h7 19.799 17.892 Male afternoon 19.537 18.091 19.u79 17.u52 ‘ Total male 19.602 16.319 19.639 17.072 Female morning 19.673 17.657 19.550 16.6h3 Female afternoon 19.167 16.h68 19.180 16.750 Total female 19.530 17.062 19.365 16.796 Above 18 on post-test h8.5% 38.b% (this group) Holding at 18 1o.5n 12.6% below 16 on post-test 35.0% h8.0% Average loss 2.6 points 3 points Per cent of total of 500 students h5.0% 30.2% 55.2% 28.2% 97 despite this positive difference in mean scores both groups lost ground in this bracket. It might be interesting to see just how the students who were in this high bracket on the pre—test distributed themselves throughout all brackets on the post-test. Table XVII will show this. Little else need be said about this above-average scoring bracket. The distribution as shown in Table XVII further emphasizes the observable fact that the experimental group scored higher on the post-test than did the control. It also further emphasizes the observable fact that both groups suffered losses in this category. The third scoring breakdown that was made was in the 16 through 17 C) O bracket: Experimental Control Score Pre-test Post-test Pre-test Post-test 1o-17 2S.L% 55.2% 22.u% 52.0% Because this scoring bracket encompasses most of the mean scores, it does not merit the detailed analysis given the two extreme brackets. 0n the post-test both groups in this bracket increased their preportions and in approximate amounts. The only real item of interest in this shift is in the make-up of this increase. In order to show this clearly, it becomes necessary to include a variation of Table XII. In order not to confuse the reader, this inclusion will be labeled Table XVIII (A Variation of Table XII). The item of interest previously referred to in the discussion of the 16-17 score bracket is the figure included parenthetically. The H' a' OmN.NH smog Omopupmoa mNom.mH smog pmoplond nu Hoppcoo omb.NH cams pmoplmeQ mbbm.ma same pmmpuood In HOOCOEHpoaNm ...... H O H m O N O OH OH HH OH HH m m H O O Hoppooo OH ................ N H O O O, O a O mH m N N O O anooeHoOakO - ............... N m m m N a a a O N O m N H HonOOoO OH ---- ................ N m H N NH O OH O OH O O O O HnOnosHoooxO £3.51... uuuuuuuuuu .. uuuuuuuuu O O O O O m N O H H H HOBOOO ON ............................ H N O {O Ow, .m N O O H H HnOOoeHnoOxu --1--- H O O O O N H H m .N N N HH N N O O HonOOoO HN ------ .................. H O O O. N m [IO N OH O m O HnOOoeHnoOkO nuununlqunnun: uuuuuuuuu N o N N N m H N m o H o Hohpcoo NN ....................... ------u-- H O O m N N N m H O HOOOoeHnoOxO --------- ....... ------- ............. H O N H m m N O O HonOOOO mN ........ I------1---------------------------- H O H N H m O HnocoEHooaxO . ..... u-----------u--- ........ ------- ........... H O O O O O HooOOOO ON n------u-------a----------------'-:-------- H O H O O O O ancoeHnoaxm wopoom poop Imam O .N O O .OH HH NH mH OH OH OH NH OH OH ON HN NN ON ON nooooo Onop -Onoa a>on< 22¢ ma mfldoom Hmueamxm mo ZOHBDmeamHQ 9mmBIHmOm HH>N mamFJ The speaker has implied that the unfit: 1. are now largely in institutions 2. are now, in part, in institutions. 3. seldom.reproduce their kind h. need not be confined if watched by their families One difficulty with sterilization laws is the problem of: l. constitutionality , 2. surgical difficulty 3. the health of the subject h. the numbers involved That the more able classes are maintaining their numbers in comparimn: with the less able classes seems: l. improbable 2. probable 3. rather certain h. hypothetical Efforts to increase the proportion of able classes in the population have been: universal ltmited to the English speaking nations made by several nations limited to the United States 4-"me End of Sociology Test 'NICHOIS TEST OF LISTENING COMPREHENSION (POST-TEST) LITERATURE You have heard the recorded lecture on Literature. Answer the following questions according to the information given in the recorded lecture. ‘Mark your answers on the answer sheet. Do not make any marks on these questions. 1. The most logical title to give this recorded lecture excerpt would be: 1. A literary Adventure 2. The Importance of Reading 3. A Liberal Education h. Human Nature 2. The 0. Henry story is in itself primarily: l. pathetic 2. moralizing 3. humorous h. tragic 3. Which conclusion can best be drawn from the 0. Henry story? . That man is gregarious in nature. . That women are unpredictable. That reading affects one's personality. That facts are more valuable than poetry. FUJIUH h. The different magazines being published may be numbered in: thousands millions hundredsx scores 43’me 5. Americans ought to: try to read .0031“ of the books in the world tackle the problem.of what to read individually subscribe to a daily newspaper write at least one book each 4?me 6. The speaker implied that the question of what to read and why we do read are: both affected by the materials available closely related too difficult for the layman to solve quite unrelated 43’me O. (OVER) Test B Part 1 '0' 10. ll. 12. LITERATURE In his essay Thomas Huxl y makes plain his conviction that: 1. life is somewhat like a game of chess 2. we ought to learn to play chess 3. giving and getting out of check is most vital in chess k. chess is fascinating Huxley makes it plain that in life our opposing player is: l. Satan 2. God 3. hasty and remorseless, h. always fair and just That enjoyment may motivate reading was illustrated by which one of the following? 1. the work of Stevenson 2. the work of Defoe 3. the story of the Welsh blacksmith 1+. the quotation from Carlyle The author of Robinson Crusoe was: 1. Robert Louis Stevenson 2. Daniel Defoe 3. Charles Lamb h. Sanderson Pratt Of the books mentioned as being of an inspirational type, the meet significant has probably been: . The Rubaiyat Egbinson Crusoe The Bible Pitcairn Island FWIDH 0.. 2 e author of Half Mile Down is: . Nerdoff Hall Daniel.Defoe Robert Louis Stevenson William.Beebe #‘WNH End of Literature Test 159 521 NICHOLS TEST OF LISTENING COMPREHENSION (POST-TEST) ECONOMICS You.have heard the recorded lecture on.Economics. Answer the following questions according to the information given in the recorded lecture. Mark your answers on the answer sheet. Do not make any marks on these questions. Note that the first question is number 13. Begin your answers on the answer sheet with number 13. l3. Which one of the following words, each appearing in the definition of accounting, gives the best clue to that science? rights enterprise statistics method 43"me 1%. Which of the following holds most true of the business partnership? A uniformity law guides its organization. Internal disputes are frequent. It is seldom composed of more than two persons. . One co-owner is usually the senior partner. FOUND-J 15. Which of the following conditions is most true of the corporation? It is synonymous with "big business". It is best represented by Northern Pump Co. It is treated legally as an artificial being. It is made up of many members. 4?me o o e 16. Which of the following is the best example of'a private corporation? The City of St. Paul Bob Feller Enterprises U.S. Steel Twin-City Transit Co. 3'me see. 17. Which of the following is the best example of a public corporation? The Mercy Hospital The General Hospital The telephone company Northern States Power 41"me (OVER) Test B Part 2 fl pt 18. 19. 20. 21. 22. 23. 2h. ECONOMICS Which one of the following makes most frequent use of the accountants products? 1. the potential stockholder 2. many parties 3. bankers approached for loans h. the immediate business executive Which of the following statements best describes the relationship of property to property rights? 1. The two are exactly synonymous. 2. The two are always co-existant. 3. Both are concerned with money values. h. Both involve claims to things. h Whic of the following names describes our common accounting system? 1. the double-entry system 2. the single-entry system 3. the triple-entry system h. the net-worth system Which one of the following statements is most true with respect to the equation that "properties equal property rights"? . It classifies property and property rights. . It simplifies accounting. ’ . It underlies all accounting. . It illustrates the methods of accounting. FUJIUH The speaker stated or implied which one of the following? ' . Property rights equal liabilities. . Property rights equal net worth. . Property rights equal creditor's rights plus owner's rights. . Property rights equal owner's rights plus net worth. 47me Creditor's rights are which one of the following? 1. assets 2. liabilities 3. proprietorship h. value of your credit rating Which one of the following is false? .1. Assets equal liabilities plus net worth. 2. Assets minus liabilities equal net worth. 3. Assets minus net worth equal liabilities. h. Assets plus net worth equal liabilities. End of Economics Test ”1 ‘ SAugSQ ’ MERLZZRARY LOAN '" ' no 24 '53 4W?)- :- r5193! m g JE3 '54 “19 5‘ M7: sob 16'? Ear "To 159 S (