AN EXPEWMEN?AL STUDY 0?- Ti-‘EE EFFECT 01"" TWC DESCUSSEQN TECHNIQUES Ofi EDUCATiONAL OUTCOMES W A BEGEHNEHG EDUCATiOHAL PSYCHOLOGY CLASS TAUGHT BY CLOSEWCRCUE? TELE‘HSlON Ti'méa for tho Dogm of Ed. D. MECHIGAN STATE UNIVERSITY Marion D. Dunham 1958 JHESIS This is to certify that the thesis entitled An Experimental Study of the Effect of Two Discussion Techniques on Educational Outcomes in a Beginning Educational Psychology Class Taught by Closed-Circuit Television presented by Morton D. Dunham has been accepted towards fulfillment of the requirements for Ed.D Education degree in x’_ ‘ . /l.' . 1., '1'. / ALLLLJ 111;- Ht ——,{ Major professor /\‘./ ' ‘ " kf/ / , Chit/«(1' (AL Date October 20, 1958 0-169 LIBRARY Michigan Stan University l\\\\\\\\\\\\\\\\\\\\\\\\\\ \\\\\\\\\\\\\\ W“ m 312931038382 469 RETURNING MATERIALS: MSU Pla bkd pt LIBRARIES rem eth's ceh cok ut fro ”- y rrrrrrrr d. FINESw ‘i'l‘lm b: Shir :89 ed if be “To stam mpe ed be low \. "1'“! 1 W's—"‘1; if ‘ W 3 \ l i \ ___________r_,_—-—r inf V__ __-_fl _ OUTC CHI-:5 Sum‘i AN EXPERIMENTAL STUDY OF THE EFFECT OF TWO DISCUSSION TECHNIQUES ON EDUCATIONAL OUTCQES IN A BEGINNING EDUCATIONAL PSYCHOLOGY CLASS TAUGHT BY CLOSED-CIRCUIT TELEVISION By Morton D. Dunham A THESIS Sutuitted to the School for Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DCETOR OF EDUCATION Department of Administrative and Educational Services 1958 Grateful 3C nurse of the 5:11; To Dr. Jo'm To Mr. ClayE it course, for hi: Isrmile and eni; ii ACK NOi’dIEDGMENTS Grateful acknowledgment is made to Dr. William W} Farquhar, my'major professor, for his advice and assistance throughout the course of the study. To Dr. John D. Krumboltz for his advice and suggestions. To Mr. Clarence M. Williams, my colleague and cOdworker during the course, for his kind assistance and many helps that made the year worthwhile and enjoyable. To the students who participated in the course. And especially to my wife, Mary, for the many hours spent in typing the manuscript and without whose understanding and patience the study could not have been completed; and my sons, Marc and Martin, who sacrificed much. AN EXPERIMENTAL STUDY OF THE EFFECT OF TWO DISCUSSION TECHNIQUES ON EDUCATIONAL OUTCOMES IN A BEGINNING EDUCATIONAL PSYCHOLOGY CLASS TAUGHT BY CLOSED-CIRCUIT TELEVISION By Morton D. Dunham AN ABSTRACT OF A THESIS Submitted to the School for.Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF EDUCATION Department of Administrative and Educational Services 1958 It ras the . tat the absence 0: ference in eiucati. iaszrgction under . iiscussmn. The expe rim (FE 200) at iiichi: 11g room, t“ View fisian Production. The subject ms of me: 1, Morton D. Dunham 1 It was the purpose of the study to objectively assess the theory that the absence of maximum feedback would result in a significant dif- ference in educational outcomes between two groups of subjects receiving instruction under conditions of television discussion and face-to-face discussion. The experiment was conducted in an educational psychology course (FE 200) at Michigan State University. Facilities included an originat- ing room, two viewing classrooms, and the necessary equipment for tele— vision production. The subjects were one hundred and sixty-seven undergraduate stu- dents, of which h6 were males and 121 were females. The subjects were randomly assigned to two groups designated the television (TV) group and the non—television (NTV) group. For purposes of analysis the sub- jects were further divided into sixteen subgroups based on the two levels of the four independent variables. A four was factorial design was employed with two levels in each of the classification variables. The independent variables were (1) the two discussion techniques, (2) ability level as measured by the ACE Psychological Examination, (3) preference for type of instruction as measured by the Preferred Instructor Characteristics Scale, and (h) sex. The combination of two levels for each variable made possible the formulation of fifteen testable null hypotheses of which four were con- cerned with main effects and eleven with interacting effects. The depen- dent or criterion variables used to measure educational outcomes were (1) achievement as measured by a Pre and Post Achievement Test, Morton D. Dunham 2 - (2) attitude toward teaching as measured by the Minnesota Teacher Attitude Inventory, (3) misconceptions about education as measured by the Misconceptions about Education Scale, and (h) course opinion as measured by the Course Evaluation Scale. Discussion was conducted over the closed-circuit television system for the TV group and in a face-to-face classroom situation for the NTV group. Discussion was conducted in small groups of six students, in panel groups, and in the entire class situation. Discussion periods were fiftyeminutes in length. ‘All discussion periOds were preceeded by fiftyhminute lectures which were televised to both TV and NTV groups. Evaluation of results was accomplished by means of the Chi- square test, “t" test, and analysis of variance and covariance. No sig- nificant differences were found for the achievement and misconceptions about education criteria. A significant interaction effect in terms of the attitude toward teaching criteria was found for the interaction of the discussion techniques and preference for instruction variables. tA significant difference was obtained in terms of the opinion criteria for the discussion techniques effect. The conclusions based upon the study: 1. There is no evidence to indicate the superiority of either the face-to-face discussion methods or the television discussion methods for IPrOducing changes in educational outcomes in terms of achievement and lnisconceptions about education criteria. 2. The combination of face-to-face discussion and students with a "cognitive" preference for instruction is more effective in producing Morton D. Dunham 3 changes in attitude toward teaching then is the combination of face-to- face discussion and "affective" students. 3. Television discussion is about equally effective for "cogni- tive" and "affective" students in attitude change, but less so than face-to-face discussion with "cognitive" students. h. Face-to-face discussion produces more favorable course opinions than television discussion. 5. Face-to-face discussion results in fewer course criticisms than television discussion. ,\ l'_’ \l \ I I] ‘ )\ Alta W A ’13:»;th aw Dr. William w. Farqthmr iii TABLE OF CONTENTS CHAPTER PAGE LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . vi LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . I. DEFINITION OF THE PROBLEM . . . . . . . . . . . . . . . . . Importance of the Study . . Unique Aspects of the Study ix 1 The Problem . . . . . . . . . . . . . . . . . . . . . . . l 9 11 II. REVIEW OF TI-IE IIITWTURE O O O O O O O O O O O O 0 O O O O O 13 Evaluation Criteria . . . . . . . . . . . . . . . . . . . 1h General Background Studies of Educational Television . . . 15 Studies of Achievement and Retention of Course Content . . . . . . . . . . . . . . . . . . .' 16 Studies of Student Opinion of Television Instruction . . . . . . . . . . . . . . . . . . . . 37 Specifically Related Studies of Instructional Television . . . . . . . . . . . . . . . . . . . . . . 38 Studies of Achievement and Retention of Course Content . . . . . . . . . . . . . . . . . . . 39 Studies of Student Opinion of Television Instruction . . . . . . . . . . . . . . . . . . . . S7 Summary.o....................... 6’4 III. TEE DEIGN OF Tl‘lE EXPERIIENT O O O O O O O O O O O O O O O O 68 The mSign O O O O O I C O O O I O The Independent Variables . . . . . . . . . . . . . . . 68 The Dependent variables . . . . . . . . . . . . . . 7O DiagrammaticPlanoftheDesign............ 71 The Control of Extraneous Variables . . . . . . . . . . 72 The Null Hypotheses . . . . . . . . . . . . . . . . . . . 76 Main.Effects . . . . . . . . . . . . . . . . . . . . . 76 First-order Interaction Effects . . . . . . . . . . . . 77 Second-order Interaction Effects . . . . . . . . . . . 77 Third-order Interaction Effects . . . . . . . . . . . . 78 The Statistical Treatment . . . . . . . . . . . . . . . . 78 Summary . . . . . . . . . . . . . . . . . . . . . . . . . 81 Iv. WTALWTHOWANDPMEDURfi 000000000000 82 The Course and the Facilities . . . . . . . . . . . . . . 82 M I" mvn" *- V. L'L‘U .Ab.‘ 1 Measures 0 Amerits. Era: PTEfer: fleasures ( an 31111583: a‘i’isccn. CO’U‘Se T59 C0119 IULEI‘QOI-r SilInary CHAPTER The Population and the Sample The Randomization Procedures The Discussion Techniques . . Television Discussion . . Non-Television Discussion The Instructors . . . . . . . Summary . . . . . . . . . . . O O O O C O 0 O I O O O O O O O O O O O O O O O O O 0 O O O O O O C O O O O O O 0 O O O O O C O O 0 O O O O O O O O O O O O O V. EVALUATION INSTRUIJITNTS e e e e e e e o e e e e 0 Measures of the Independent variables . . . . . American Council on Education Psychological Examination (ACE) . . . . . . . . . . . . . . . Preferred Instructor Characteristics Scale (PICS) Measures of the Dependent variables . . . . . . . . . Pre and Post Achievement Test (PAT) . . . . . . Minnesota Teacher Attitude Inventory (MTAI) Misconceptions about Education Scale (MES) Course Evaluation Scale (CES) . . . . . . The Collection and Recording of the Data . . Intercorrelation of the Instruments . . . . . SWW . C O C O O I O O O O O C O O O O O 0 VI. tANALYSIS OF THE EXPERIMENTAL RESULTS . . . . . . The Computational Procedures . . . . . . . . . Analysis of the Achievement Test (PAT) . . . . Analysis of the Minnesota Teacher Attitude hwmmmy(MMI).. ... ... ... ... Analysis of the Misconceptions about Education Scale (MES) . . . . . . . . . . . . . . . . Analysis of the Course Evaluation Scale (CBS) . Summary.................... VII. SW AND CONCLUSION e e e e e e e o e o e o 0 Summary . . . . . . . . . . . The Problem . . . . . . . The Design . . . . . . . . Experimental Procedures . Evaluation Instruments . . Results . . . . . . . . . Limitations of the Study . . Conclusions . . . . . . . . . . Suggestions for Future Research BIBLIOGMPM O O O O O O O O O O O O O O O O O 0 iv 137 ”2:311 L. Swan L...w:1 B. orizil L???“ , a... K. . kClE :2?“ Law] 3. (fl 0 [J L) ’r1 ’4’ L; APPENDIX A. APPENDIX B. APPENDIX.C. APPENDIX D. Measuring Instruments . . . . . . . . . . . . . . . Personal Data Card . . . . . . . . . . . . . . . Preferred Instructor Characteristics Scale (PICS) Pre and Post Achievement Test (PAT) . . . . . Misconceptions about Education Scale (MES) . Course Evaluation Scale . . . . . . . . . . . OrigiMI Data 0 O O O O O O O O O I O O O O O O Explanation of Computational Procedures . . . . Computations to Test Assumptions of Analysis of variance and Covariance . . . . . . . . . PAGE 1h6 1h? 1&8 151 158 162 166 17h 183 TABLE I. II. III. IV. V. VII. VIII. X. XI. XII. XIII. XIV, LIST OF TABLES Number of Studies Reviewed, Classified by Conclusions, Adequacey of methodology and Purpose Number of Students in the Sample, Classified by Class, Sex and Major Curricula . . . . . . . . . . . . . . Intercorrelations of the Evaluation Instruments . . . Results of Tests of Assumptions of Covariance for the Achievement Test Analysis of Covariance - PAT Post-test Scores Adjusted for Pre-Test Scores . . . . . . . . . . . Results of Tests of Assumptions of Covariance for the Minnesota Teacher Attitude Inventory (MTAI) . . . . Analysis of Covariance - MTAI Post-Test Scores Adjusted for Pre-Test Scores MTAI Scores and Adjusted Mean Scores on the MTAI for the PICS x TECH Int/emotion C O O O O O O O O O O 0 Results of Tests of Assumptions of Covariance for the Misconceptions about Education Scale . . . . . . . Analysis of Covariance - MES Post-Test Scores Adjusted for Pre-Test Scores Result of the Test of Homogeneity of Variance for the Course Evaluation Scale (CES) . . . . . . . . . . . Analysis of Variance - CES Scores . . . . . . . . . . Results of the "t" Test of the Difference Between Mean Scores for the TV and NTV Groups on the CES . . . . Result of the "t" Test of the Difference Between Mean Scores for the TV and NTV Groups on the CES and the Total Sample Comparisons of the TV and NTV Groups on the Individual Items of the CES vi PAGE 65 85 100 105 107 108 109 110 112 113 1114 115 116 117 118 0-10 C-ll C-12 0-13 6-11. vii PAGE Chi-square Analysis of the Free Comments on the Course Evaluation Scale for the TV'and NTV Groups . . . . . . . . 125 Chi-square Analysis of CES for TV and NTV Groups on the Basis of Comments and No Comments . . . . . . . . . . . . 126 Arrangement of Subgroups for Analysis of Ability Level (ACE)Effect.......................l76 Arrangement of Subgroups for Analysis of Preference for Instruction (PICS) O O O O O O O O O O O O O O O O O O 176 Arrangement of Subgroups for Analysis of Sex Effect . . . . 176 Arrangement of Subgroups for Analysis of Discussion TeChniques O O C O O O O O O O O O O O O O O O O O O O O O 17 6 Arrangement of Subgroups for Analysis of ACE x PICS Interaction O O O O O O O O O O O O I O O O O O O O O O O 177 Arrangement of Subgroups for Analysis of ACE x SEX IntemCtiono000.000.00.000.0.0.0.0177 Arrangement of Subgroups for Analysis of ACE x TECH Interaction I O O I O O O O O O O O O O O O O O O O O O O 17 8 Arrangement of Subgroups for Analysis of PICS x SEX InteraCtion O O O O O O O O O O I O O O O O O O O O O O O 17 8 .Arrangement of Subgroups for Analysis of PICS x TECH Interaction . . . . . . . . . . . . . . . . . . . . . . . 179 Arrangement of Subgroups for Analysis of SEX x TECH Inte metion O O O O O O O O O O O O O O O O O O O O O O O 179 Arrangement of Subgroups for Analysis of ACE x PICS x SEX Interaction . . . . . . . . . . . . . . . . . . . . . 180 Arrangement of Subgroups for.Ana1ysis of ACE x PICS x TECH Interaction O O O O O O C O O O O O O O O O I O O 0 O 180 Arrangement of Subgroups for Analysis of ACE x SEX x TECH Interaction O O O O O O O O O O O O O O 0 0‘ O O O O O 181 Arrangement of Subgroups for Analysis of PICS x SEX x TECH Interaction I O O O O O O O O O O O O O O O O O O O O 181 0-15 D-l D-2 13-3 3-5 D-6 ‘Arrangement of Subgroups of Analysis of ACE x PICS SEX x TECH Interaction . . . . . . . . . . . . . Computation Linearity Computation Subgroups Computation Linearity Computation Subgroups Computation Linearity Computation Subgroups Computation Subgroups to of to on to of to Test Hemogeneity of Regression and Regression of Subgroups on PAT . . Test Homogeneity of variance of PA T O C O O O O O O O O O O O O O 0 Test Homogeneity of Regression and Regression of Subgroups on MTAI . . Test Homogeneity of variance of on MTAI O O C O O O O O O O O O O I O to of to on to on Test Homogeneity of Regression and Regression of Subgroups on MES . . Test Homogeneity of variance of LES O O O O O O O O O O O O O O O 0 Test Homogeneity of variance of CES O O O O O O O O O O O O 0 0 O 0 viii PAGE 182 18h 185 186 187 188 189 190 '1‘ 3 1 no .1 8 a .Q .N v. “in! a e S - 2 LIST OF FIGURES FIGURE PAGE 1. Relationship of the Four Independent Variables with Two levels for Each Variable and the Sixteen Subgroups Formed by the Various Combinations . . . . . . . . . . . . . . 73 2. Schematic Diagram of the First and Second Period ArrangementsoftheTVandNTVGroups . . . . . . . . . . . . 9O “P111! in t. head in the me" it Can "I atteapt t °f hamlet: and. CHAPTER I DEFINITION OF THE PROM Tb role of instructional television in education has increased rapidly in the past few years. In many instances it has been intro- duced in the classroom setting with little knowledge of how, when and where it can be used nest effectively. The present investigation is an attempt to evaluate the effectiveness of this relatively new medium of instruction as an aid in teaching large classes by the discussion method. I. m PROBLH The appearance of television has brought forth mam divergent views and claims about the potentialities of this new medit- as an instrument of emacation. Soles view it as a virtual panacea for the probleas of education shile others more modestly see television as an adjunct to sound educational practices. Dunham and Losdenailk (16) in concluding a brief review of in- structional television research state that: Finally, experimental studies have tended, generally, to shes that television can serve as an instrumentaiity thereby every single device, technique and process known to the art of teaching can be brought to bear in group instruction—in short, that its educational potential is limited only by the creative imagination and resourcemhess of those who undertake to use it. President Minis of Iestern Reserve University in an address in 1952 at State College, Pennsylvania stated that television was "one of the mostumagnificent instruments for raising the quality of teaching whether it be college teaching or elementary teaching."1 On the other hand Telford Taylor, former general counsel for the Federal Communication Commission, told the new York commission on educa- tional television: Television is not a substitute for the teacher, but a new tool for the teacher to use...Its proper use will not contract, butwwill greatly'expand, the scope and opportgnity'flbr question and answer between the teacher and the student. .And Milton Eisenhower has observed that television will “supple- ment rather than replace the classroom."3 The present study was undertaken in the hope of initiating as lleast some clarification of such diverse views as those expressed above, on one aspect of instructional television, the use of the discussion method. The discussion method is used at all levels and for a wide variety of subject areas as an instructional method. There was no attempt in the present study to determine the effectiveness of the dis- cussion method as contrasted with the lecture method or other teaching methods not using discussion techniques. It was an underlying assump- tion of the study that benefits are to be derived from the use of the discussion method. .Although this assumption was not tested in the 1A3 quoted by Martin Packman, "Educational Television," Edi- torial Research Reperts, May 18, 195b, Vol. 1, No. 10, p. 376. 2Ib1d, Pe 3724e BIbid, pe 37km -‘* —-A~ -...~__, .- present experiment, a number of research studies indicate that the dis- cussion method is more effective than the lecture method in the learn- ing of certain kinds of materials, and a brief review of this research will be made here. Studies of the acquisition of information where the criterion has been the ability to recall or recognize factual material on ob- jective type tests have shown the lecture and discussion methods to be equal in effectiveness in investigations by Bane (h), Carlson (ll), Eglash (18), Gerberich and Warner (23), Hudelson (211), and Johnson and Smith (26). On the other hand in the same type of investigations Hudelson (2h), Ruja (SO) and Spence (53) have reported the lecture to be mre effective than discussion methods. Studies by Bane (h) and Rickard (h?) of the retention of learned material have shown that knowledge acquired in discussion classes was retained significantly better than that resulting from reading and lectures without discussion. Investiytions by Bloom (6) ,A Brinkley (10), Edmitson and Braddock (17), Ruja (so), and Ward (57) in which the objective was con- cerned with aiding students in the ability to evaluate, draw inferences, synthesise, perceive relationships, and make application of learned material it has been found that discussion is significantly superior to the lecture. It has also been shown that discussion is superior to the lecture method in affecting changes in attitudes and behavior in investigations by Bond (7), Levine and Butler (33), Iewin (3h), Maier (to), and Ruja (so). Although there is no agreement about the relationship between learning and classroom morale, it has been demonstrated that an impor- tant relationship exists between the emotional adjustment of students and between teacher-pupil relationships and classroom morale. Studies by tech (3), Bovard (9), Few (22), LeWin (35), and Ruja (50) indicate that the discussion method is more conducive to promoting high morale and good interpersonal relationships between the teacher and the pupil than the lecture method. The assumption made previously that benefits are to be derived from the use of the discussion technique would seem to be supported on the basis of the above research. An explanation for the findings that in certain kinds of learn- ing the discussion method is superior to the lecture method may be found by reference to current learning theory. In simplified terms, according to the association theorists such as Thorndike, Guthrie and Bull learning occurs by the association of cues (stimuli) and responses. The association or connection of cue and response is enhanced by vari- ous factors such as exercise (practice), effect (reward or non-reward), the strength of the cues, the distinctiveness of the cues, the rele- vancy of the cues, and others. Although relatively little is known about higher reasoning pro- cesses in human learning, it may be theorized that this type of learning is reducible to simple components of a stimulus-response nature. Thus the discussion method may be more effective than the lecture method hum there is ml expense“ Manned tn the W lunar to t m hunted to M Mr lectu “bar of cm , will: emu, . ‘° “In. trial m“ £11m 2: M1“) “urea: trim nth 1m “nut MMtoh'“! because there is a greater opportunity for the association of the cues and responses. In the discussion method a greater umber of ones are presented to the learner for appraisal, more opportunity is afforded the learner te try-out and select specific cues, and more occasions are presented to the learner to practice the material. 0n the other hand under lecture conditions the learner is presented a limited number of ones, he is given less opportunity to try-out and select specific case, and there are fewer opportunities to practice. Reduced tosimple trialanderror terms, itmaybe saidthatthediscussion method allows for a greater number of trials plus the knowledge of results, whereas the lecture method allows for a limited number of trials with little or no knowledge of results. It will he recalled from the research cited above that in the studies of the learning of factual intonation the lecture method was foundtobeatleastequaltoandinsomeexperiments superiortothe discussion method. has learning of factual information as opposed to the learning of probln-solving abilities each as drawing inferences, applying learned information, synthesising, and the like, is more in the nature of rote memorization. lhere the information is not of a highly abstract nature it may be theorised that the association of one sndrespense is easilymede bythe learmrandthat the discussions-y actually interfere with learning the to the presentation of too many ambiguous or irrelevant cuss. the results of a related stuck in communication theory by Ieavitt and Iueller (32) is of interest. The authors reasoned that according to the information theory of cyberneticists and the trial and error theories of psychologists in order for A to hit some target 3 it is necessary for A to be constantly informed of Us own progress. Therefore 1: A attempts to hit a with some information, A will be more successful if B provides 1 with some cues which 1 cannot obtain di- rectly. In other words where comaication between a and B is the goal, feedback in the form of expressive or verbal language should lb for greater effectiveness, especially if the material being com- municated is new or relatively abstract. A In the experiment by leavitt and lueller a series of rectangular figures of equal area but of varying shapes were commnicated to a group of students under four degrees of feedback. The students were required to produce the sise and the shape of the patterns. The four conditions of feedback were: (1) sero feedback in which the experi- menter described the figures to the students, but the subjects could not see the experimenter nor comnicate with him in any way, (2) partial feedba& in which the experimenter described the figure to the students and the students could co-mrnicate to the experimnter by facial expressions, (3) partial feedback in which the experimenter described the figures to the subjects and the experimenter could answer questions of the subjects by a 'yes' or 'no', and (1;) free feedback in which the experimenter and the subjects could freely com-micate with one another. Of course, under all conditions the geometric figure was never visible to the subjects. The subjects were assigned to groups corresponding to the four conditions of feedback and all groups experi- enced each of the feedback conditions in varying orders. It was found that the mean accuracy score for the subjects in reproducing the patterns increased steadily from the conditions of sore to free feedback. It was also found that the subjects displayed more hostility when the condition of sore feedback followed the other patterns. The authors report the following conclusions: (1) Feedback increases the accuracy with which infomtion is tram-itted. (2) Feedback increases sender and receiver confidence in what they have accomplished. (3) Increased feedback results in more time spent in expla- nation. (1:) Feedback experience improves subsequent sero feedback trials considerably. (S) Sender experience contributes more than receiver experience to improved accuracy of c-unication. (6) Zero feedbart engenders hostility in the receiver. (7) Zero feedback engenders doubt in the sender. Insofar as teaching is the communication of information from the instructor to the student, it would seem on the basis of the research andarguments presented above, that learningwillbeatamaximt-where feedback from the student to the instructor is also at a unima. The above principle would as. to be particularly applicable to those situ- ations where the material to be learned is of an abstract nature or where an attempt is being made to teach students the ability to make tpplicatiom of In situations I ticm between a in attitude and all: it but , “mm to in Unchr co ”5". that fee tt"rust. ideal applications of learned naterial, to drel inferences, and the like. In situations where ugh classroom morale and good interpersonal rela- tions between student and teacher are deemed important or where changes in attitude and behavior are objectives it would also appear, theoreti- cally at least, that feedback between student and teacher should be at a nxinun to insure sexism effectiveness. Under conditions of television instruction it is reasonable to seems that feedback from the student to the instructor, under even the nest ideal conditions, would never be at a lexicon. The physical and technical probl-s involved in arranging cansras , monitors and sicrephonss so that the instructor and the students would be in con- plete visual and audio conunication at all times are such, that for practical purposes, ideal conditions could not be obtained. In situ- ations where the communication is principally one-way; that is, where factual information is being conveyed by mans of the lecture, the absence of feedback is probably not a serious handicap. However, in situations where it is inportant that two-way co-unication be Iain- tainsd; that is, where the subject-aatter or the objectives to be not are of a core abstract nature such as the learning of problea-solving techniques, the ability to apply principles, the absence of feedback or the condition of partial feedback would, according to the arguesnt developed here, prove a handicap to both the instructor and the ' student. It is also theorised that the absence of maxi-us feedback would result in the lowering of class norale and a corresponding in- crease in the hostility and insecurity displayed by the students. Can a discussion conducted by closed-circuit television between an instructor and students in two different races produce as effective educational outcues as those produced by conventional face-to-faoe necessionf It .1... the purpose of the study to ascertain the relative effec- tiveness of two discussion techniques, one eaploying conventional face- to-face nethods ad the other using closed-circuit television, on educational outcome as neasured by four criteria: (1) achievennt of course content, (2) attitudes toward teaching, (3) nisccnceptions about education principles and practices, and (h) student opinion about the course. A caplete state-ant of the specific null hypotheses to be test- ed will be found in section two of Chapter III, I"the Design of the h- perineat.‘ I II. WWTHBSTUDI .the recent and predicted increases in the anchor of students at all educational levels in proportion te the present and predicted supply ef available teachers poses serious problem. It is inth that nere efficient teaching techniques be identified and used. Inas- nuchas themberof studentsisincreasingaore rapidlythanthe III-her of teachers, it is also inportant that nethods be devised for providing instruction for classes of larger sise. Theadventoftelevisioninrecentyearshasprovidedaneans of increasing Iarkedly the runber of students that can be reached by a ling]: intructor. culminate an 1 mt. be ac mica mid 1m (1) (he 1 Motion nth a 1 We I solution (2) The r hm,“- Misc 1.0 single instructor. There are a nubor of advantages in being able to co-unicato with large groups by a television ad loud-speaker ar- renmt. The nore enthusiastic advocates of instructional tele- vision would list the following as ieportantr (l) are instructor is provided with a means of two-way co-u- nicatiou with a large mabor of students oieultanoously, thereby pro- viding a solution to the probl- of teacher shortages. (2) The nunber of instructors required can be reduced and better qualified and trained instructors can be exployed as a result of the savings. (3) lbs quality of instruction will be rare uniforu than can be expected fr. several instructors. l (h) a better quality of instruction will result by tho use of specialists in particular area of instruction. (5)1'hoaeountofsuppleeentaryequiputnoededsuchasaudio— visualaids anddononstrationnaterisl canbe reduced. (6) The nuberof nall classroons needed canbo reducedwith a resulting savings in construction costs. Irho above advantages for the use of television instruction are based on the belief that television instruction is at least equal in offectivemss to the more conventional teaching procedures and that all of the present techniques can be perforeod as effectually over television as they can be in the face-to-face situation. lush of the research in recent years in instructional television has clailed that instructionbythenediuoftelovisionis at leastequalandinnany 11 instances superior to regular instruction, and one of the sore perti- nent of these studies will be reviewed in Chapter Two. Inch of this research does not support the clains that are sade for it, and new of tho claiss are based on hasty generalisation fro- inadequate research. It is iaportant for practical purposes tint a clearer understanding of the strengths and weaknesses of this new nedim be investigated, and the present oxporiuent was designed in an attoapt to answer a practi- cal and inportant question concerning the use of instructional tole- vision. III. UHIQUE ASPECTS W THE STUD! The design of the present study was a unique feature, in that it node use of the sore recent statistical techniques. The design was a four-factor classification design with four classification variables and two levels or categories in each of the variables. Such a design allows the oxporinsnter to sinultanocusly neasure the Iain and inter» action effects of several control variables under different treat-oat conditions without resorting to tho traditional procedures of several independent oxperinents of the single variable type. Such a design increases the precision of the expoth because the subjects are randuised within the treat-ant groups and it per-nits a separate stem of the treat-out effects at different levels of the control vari- able. By adapting this design to an analysis of covariance technique it was possible to statistically control for initial differences in subjects on the control variables. A couplets description of the de- 12 sin is given in Chapter Three) Another unique aspect of the sxperiaont was the attempt to in- vestigate the problon in a realistic setting. Inasmuch as one of the problns confronting educators at the present tine is that of providing instruction to large nuabers of students, the present study was con- ducted in a large class setting. Although otter investigators have included the use of large classes in the franowork of their orperinen- tation, so far as is known to the present writer, there have been no investigations in which the television and discussion techniques have been used sinultaneously in the large-class setting. a review of the sore pertinent background studies of instruc- tion]. television and of studies closely related to the present study will be presented in Chapter Two. Chapter Three will contain a descrip- tion of the experdnontal design, a statement of the null motheses, and a description of the statistical procedures used. The experinental nothods and procedures will be reviewed in Chapter Four, and Chapter Five will contain a description of the evaluation instruents. The results of the analysis of the data will be discussed in Chapter Six. 3hr a detailed discussion of nodern uperinental designs see I. 1". Lindquist, Desi and is of gin-onto g mg Education. Boston: u gfihffic" . CHAPTER II m 0? THE LITEMTUIE much has been written on television communication of which relatively little is concerned with research. Prior to 1950 nest of the research was devoud to the techno10gical and scientific aspects of television, but from 1950 to the present an increasing amount of oxporinontation has been related to the sociological and educational probleas of television. These latter studies may be classified into four areas: (1) studies of the general effects of television in the everyday lives of people, (2) studies of the content of television progress, (3) studies of the technical aspects of using television, and (1;) studies of the effectiveness of television used as a means of instruction. The literature appropriate to the scope of the present study concerns experimentation with educational television and will be restricted specifically to instructional television. In the review that follows studies will be classified under two nain headings: (1) general background studies and (2) specific studies closely related to the present investigation. The foraer classifica- tion will contain studies that provide background and infernation which will aid in the understanding of the present study, and the latter will comprise studies which are directly related to the present study. The reviewed studies in each of the above classifications will be further divided on the basis of the purpose of the oxperiaent. (a I i l: k; _ I I. EVALUATICH CRITERIA The diversity of purposes, exporiaental designs, exporinental nethods, analyses of results, and interpretations of findings make it necessary for the reviewer of research to employ an objective standard for evaluating studies. Such a standard is necessary in order that subjective factors on the part of tie reviewer say be eliminated as each as possible, and to canonicate to the reader the evaluation standards employed. Farquhar and Krunbolts (21) have devised a useful check list for this purpose, and it is the one used in the present no- view. It is reproduced below. A Check List for Evaluating Experimental Research in Psychology and Education Wfi- w ‘ ‘ factor; factor; Questions A A ‘e Th ”0th l. Ias tie proble- clearly bfincd? 2. Ias tie problon framed in the for: of hypothe- ses which were oxperiuentally verifiable? B. The Design 1. Has the selected statistical design appropri- ate to the particular experimental. nethods, conditions, subjects, and hypotheses under test? 2. Isa the population free which the sample was drawn clearly specified? 3.Isthensthcdofdrswingths samplefrouths population clearly specified? h. Ias the control group chosen in the ease scorer and frm the ease population as the uperiaental groups? 5. Ioro the various treat-outs (including control) assigned at random to the groups? 6. Did the experiaent include a replication? 7. Has the level of significance necessary for re- jsction of the null lvpothoses specified before the data was collected or analysed? 15 c. the Procedure 1. Were the treat-ents and sethods of collecting data described so that an independent investi- gator could replicate the experisent? 2. Iere the sise and characteristics of {the sanple adequately described? 3. Here the treat-cuts achinistered so that ex- traneous sources of error were either held con- staut for all treataent and control groups or randmised along subjects within all groups? D. The Analysis 1. Has the criterion neasure appropriate? 2. Ins any evidence of the reliability of the cri- terion seasure given for the experieental sample? 3. Were the statistical assumptions which are necessary for a valid test of the all hypothe- ses satisfied? I. The Interpretation 1. Iere the conclueiens consistent with the results obtained? 2. Were generalizations confined to the population free which the sanpls was drawn? Ihechecklistwillbsusedinthepresentreviewtoevaluate only those studies that are specifically related to the present study. 11. m sacrum STUDIES cs DISMCTIORAL TELEVISIOI Instructional television research he been sponsored by several agencies and groups, fore-est of which are colleges and universities , secondary schools, and the silitary services. Bus of the research has been supported by the agency itself and sons has received support fra foundations. lost of this research has been concerned with cupariscns of instruction by live television, kinescope projections, and conven- tional teaching nethods in the classroom The najority of these studies have pertained to achiev-ent and retention of course content 16 and to opinions and evaluation of television instruction. For purposes of understanding the background for the present study, a summary of ease of the more pertinent studies will follow. Studies 2; Achievement and Retention 2; Course Content A study to determine the comparative effectiveness of tele- vision, kinescope recordings, and classroom instruction reported by Rock, Duva, and Murray (148) was conducted at the Special Devices Center of the United States Navy beginning in l9h9. Nine naval air stations with approximately 140 Naval Air Reservists at each station were used in the experiment. Three stations received instruction by means of live television, three stations were presented kinescopes of the same lesson, and three stations were instructed by local instructors so that for each treatment there were approximately 100 to 120 men. The course content consisted of two series of eight lessons each; one series for officers and one series for enlisted men. In the live television groups, sixteen-inch receivers were used with ten men to each receiver, and a feedback microphone was provided so that the men could ask ques- tions during the lecture. In the groups viewing recordings, the films were projected through a motion-picture projector on a screen with from 20 to 140 men assigned to a room. In the groups taught by local in- struction, the instructor gave his lesson prepared from a lesson-plan rather than a verbatim script of the lecture. Pro-test and post-tests were given to all subjects. The tests were of the multiple-choice type and consisted of 30 items on the yrs-test and post-test plus an ad- ditional 30 items on the post-test. Results of the study are given in 1? terms of percentages. In comparing the live television and local in- struction groups, the television groups were superior for 501 of the officers and 53% of enlisted men, and equal to local instruction for 385 of the officers and an: of the enlisted men. Television was in- ferior for 131 of the officers and 27% of the enlisted group. The authors conclude that 801 of the cupsrisons show television as good ’ as or better than local instructors. In comparing the television recordings with local instruction, the authors conclude that the re- cordings were superior or as good as local instruction in 75$ of the ccparisons. In comparisons of live television and recorded tele- vision it is reported tlmt live television is superior to recordings in a great majority of the cases. Several criticism can be made of this stow. Comparing live television received on a sixteen-inch re- ceiver with recorded television projected onto a movie screen intro- duced uncontrolled variables. Comparing the smee instructor in the live television and recorded television groups with three different . local instructors also introduced a variable that was not controlled. menisnoevidesceinthexeportthatthemeninthsthresgroups were randcmly assigned to the experimental groups, nor is there evi- dense that indicates the men in the nine naval air stations sure fru the can population. The additional 30 items in the post-test could not be used in a comparison of pro-tests and post-test scores. The statistical model used was inappropriate. The design of the empori- ment was such that a treatment by levels design analysed by covariance technique would have been more appropriate. lo conclusive results can 18 be drawn from the percentage figures reported by the authors. In another experiment reported by Rock , Duva, and Murray (h9) an attempt was made to determine the effect of television on achievement and retention of instruction. A series of eight one-hour telecasts were broadcast to 160 groups of approximately 3,000 U. S. Army reserv- ists ranging from private to colonel. The reservists viewed the pro- grams in 10 different cities in groups of varying sizes. Receivers were of various types ranging from 7" screens to theater-type projec- tion screens. Pro-tests and post-tests were given and a delayed recall test was administered three to six weeks after the original lesson. Each of the eight programs was produced with professional actors and was concerned with the phases of an Army's division Operations. The authors report, 'All grades of officers and enlisted man made higher scores on test questions after the telecasts than they did before the telecasts." Figures for the gains are not shown. 0n retention tests it was reported that officers retained 85% of newly learned material and enlisted men retained 65% of newly learned material for a period of six weeks. Much the same criticisms made of the previous study by these same authors can be made of this experiment. No attempt was made to equate or randomize the subjects. The effect of using different sized groups and different receiving methods was uncontrolled. The .fact that “-1 grades made higher scores after receiving the telecasts would be exPfitted, but whether or not these gains were significantly different 301‘ different ranks, under different modes of reception, and in differ- out lined groups was not answered. The investigation is more nearly a my than it i A study 1 39658 by televi: m Elven to fo: tion, and to si “19 come was true-£113. teat m “If! instrc tion. Student: Possible 32, a: 265‘ Th. anti In “Willa u “at, tl‘e class, 1Inmates t hat scum. It 18 aux-11m“ to 19 survey than it is a scientific study. A study of the feasibility of teaching Quartermaster Corps sub- jects by television was conducted by Allen (1) in 1951;. Instruction was given to forty-seven ROTC students by means of television instruc- tion, and to sixty ROTC students by regular classroom instruction. The course was four hours in length. A 32 item multiple-choice and true-false test was given four days after the conclusion of the course. The same instructor was used for both groups in any one day's instruc- tion. Students in the regular classroaa group scored 27 .7 out of a possible 32 , and the students receiving television instruction scored 26.5. The author gives no data on the significance of the difference. In cuparing the academic standing of the two groups the author found that the classm group was higher than the television group, and indicates that this could account for the differences in final test scores. It is possible, however, that such small differences can be attributed to chance. The findings in this stub are limited because the subjects were not randomly assigned to the two groups nor equated by academic standing prior to the experiment. In a closed-circuit television experiment at the naval Academy at Anmpolis reported by Boone (8), six battalions of about lhO mid- ship-en each were given training in two courses of 25 minutes length. The battalions were split into two groups: one group receiving tele- vision instruction and one group receiving normal lecture instruction. Inthesecondcoursethe gromwerereversedforthenodeofinstruc- tion. All participants were administered a 10 minute prognostic test to fltsrnine 1 courses all we lecturer for 1 television gm mm the re 20 to determine the initial learning level and upon completion of the courses all were given post-tests. Instruction was by the regular lecturer for the lecture groups but by a different instructor for the television groups. An analysis of covariance technique was used to analyse the results. In the first course the television group was significantly superior at beyond the 12 level of significance. In the second course the non-television group was significantly higher than the television group at the SS level. This study has several weak- nesses. An uncontrolled variable was introduced by using different instructors in the different groups. Caper-ing the results of the two methods on the basis of a 25 minute class session does not comprise an adequate test of either television or regular classroaa procedures. Finally the fact that one group did better first under conditions of television instruction and then under conditions of lecture instruction indicates either an initial difference in the two groups or a motive- tion effect that was carried-over from the novelty of eaposure to television. Dowell (15) reports the resultsofanexperimentbytheAir Force to evaluate the effectiveness of closed-circuit television as an instructional media. A group of 266 Air Force trainees were given the last three days of an eight week course in Electronics Fundamentals. The trainees were divided into two groups: one group receiving the in- stmction by television and the second group receiving conventional classrou instruction. lash of the groups were further split beteeen two instructors so that each instructor taught the trainees by both the 21 television and the conventional methods. An experimental group was taught by one of the instructors and the experimental population was approximately doubled two weeks later by repeating this procedure with the second instructor. A control group, composed of students equally matched with those in the experimental group, was selected from subse- quent classes in the course. The students were matched between groups byusimg grades obtainedinphase Iand II oftheirearliertraining. The matching variable was obtained by computing simple and multiple coefficients of correlation between tentative matching variables- grades, aptitude scores, and combinations of these—and the criterion variable for the control group. The criterion variable was a 50-item multiple choice achievement test. The correlation coefficient between the matching and criterion variables was given as .66. The report of findingsinthis stubarepoorlyorganisedand confusing, and will be discussed in some detail. The author first reports on a methods experiment using two methods and two instructors with90students ineachofttm TVgroups andControl mum-9o students for each of the two instructors; a total of 180 students. The author then refers to the precision with which students were matched between instructors by reference to the following table: Table IV lean Scores on htching Variable Instructor TV 0p Control 0p 5. 1.6.73 50.62 B h6.76 50.71 22 Reference is next made to an analysis of variance table as follows: Am F ratio test was used in analyzing the variance of criterion scores. In this analysis, which follows, the apparent variance of the effects of methods is misleading. It must be remmbered that matching was bstnen instructors and not bsmen “methods groups." As can be seen from the mean scores, the control group was the superior group. ‘ Table V (btainsd Required 15 Source d.f. Variance I level Methods 1 1.66.52 8.11: 6.8!. Instructors 1 6e0 e10 6e8h lithin Classes 176 57 .3 Total 179 It is not clear whether this table Presents an analysis of variance of criterion scores—as the author indicates—or an analysis of scores on the matching variable as given in the proceeding table. If it refers to criterion scores, a highly significant difference existed between the TV group and the Control group. If it refers to the scores on the matching variable, a highly significant difference existed between the twogr'oupsonPhaseIandIIgredesandindicates thatthegroupswere not equally matched initially. Since no mean scores are presented for the criterion variable, the author's reference to mean scores is ap- parently to those on the matching variable presented in Table IV. Dowell further reports on a methods experiment involving two methods and three levels. Again the original study is quoted: The umber of students utilised for analysing the variance of criterion scores on a methods experiment involving 2 methods and 3 groups of students classified according to initial ability levels on the matching variable is shown in Table VI. The sub- jects taught by both instructors were pooled since the variance between instructors and the interaction between instructors and 23 methods was found to be insignificant. Table VI Number of Students lean Score on m No. of Subtcts latch}? Variable TV 91 9.2 Control 91 ’49 03 Total .152 The nunber of students in this table is 182 as compared to 180 in the previous reference, and the mean scores on the matching variable are not in agreement with those given in Table IV. Apparently this experiment invole different students than those used in the previous methods experiment. If this is true then the reference to pooling sub- Jects taught by both instructors is not clear. Is the author referring to Table V when he says the variance between instructors and the inter- action between instructors and methods was found to be insignificant, or is he referring to a different analysis for which no tables are given? If the former, he is using data collected in one experiment to perfon operations in a second experiment, and if the latter, the tables should have been printed to make clear his methods. Dowell presents the analysis of criterion scores for a second methods experiment involving two methods and three levels. It is re- produced below. Table VII Obtained Required 11 Source d.f. Variance I" level Iethods l 57 .37 1.31 6.78 Levels 2 Mel; “le3 he73 lethods x levels 2 h2.57 1.17 M73 lithin Classes 176 36.23 Total 181 2h is the above table indicates a highly significant difference was found to exist between students grouped by initial ability level. Since no mean scores are given for the groups by ability levels the table has little meaning. Ioreover, inasmuch as this experiment apparently cu- prised a different group of students than those in the previous experi- ment it reveals nothing about the differences that exist between tele- vision and conventional instruction and there fore adds nothing to the original purpose of the experiment. Referring again to the original study: The criterion measure was administered as a pre-test to a third group of students immediately before they entered the experimental block of instruction. The pre-test scores for this group were compared with post-test criterion scores achieved by matched ex- perinental students and with matched control students. The analysis of variance of criterion scores between matched pee-test groups and poet-test control groups is shown below: Table VIII Value of t for criterion scores between matched pre- test and post-test control groups Pro-test Post-test 0.1392... Control Group lo. of students in each group 98 98 Mean Score on matching variables 50.1 50.3 SeDe 0: matching “mbl. 6e3 6e2 lean Score on criterion variable 15.36 28.39 S.D. of criterion variable 3.97 7.0 Difference in means criterion variable 13.03 Standard error of meme criterion vambl. e3h9 e530 Coefficient of correlation matching variable 8: criterion variable .149 .66 Standard error of difference in means of criterion variable .635 t: 20.05 p: .001} The analysis of variance of criterion scores between matched 25 pie-test group and post-test sxperisental group is shown below: Table II Value of t for criterion scores between matched pro-test and experinental post-test groups Pre-test Post-test Group Experimental Group No. of students in each group '89 89 Kean Score on matching variable 16.3 19.2 8.1). matching variable 6.2 6.2 lean Score on criterion variable 15.26 26.h5 8.1). on criterion variable 3.58 7.14 Difference on mean criterion scores 11.19 Standard error of means criterion variable - .332 .592 Coefficient of correlation latching variable a: criterion variable .h9 .66 Standard error of difference in mean of criterion variable .679 t! l6.h8 p: .0011 The study is even acre confused at this point. The author re- fers to an analysis of variance in reference to the above two tables, but presents figures based on a 't" computation. Were the analysis of variance tebles omitted or did theauthor make a mistake in terminology? Here again the ushers of students does not agree with the number as reported in the previous tables. Evidently these figures are based on studies with different students than the previous experinent. W were comparisons made between the pro-test and post-test experimental group and the pre-test and post-test control group? It would have been nuch more meaningful to have compared the post-test experimental and post- test control groups. The present writer computed a 't“ value for the difference of the mean scores on the criterion variable for the control and experieentel groups froa the data given in the above tables, and 26 found that there was a significant difference beyond the 5% level of confidence in favor of the control group. Considering the ambiguous- ness of this study, however, no conclusions can be drawn from the above computations. The conclusions for the above study follow in the words of the authors The use of live television proved to be equally effective as conventional methods of instruction. Use of the television median of communication was limited to transmission of lecture-demonstra- tion for viewing by students in remote classrooms. This evidence tends to further confirm the premise that unlimited numbers of students can learn satisfactorily by viewing televised programs of good instructors performing as they normally would in the classrom. This proved to be true for all levels of student ability. The amount of learning that took place in the experimental block of instruction was significant for both the control and experimental .tmue In view of the many questionable features and practices in the above experiment, such conclusions are totally unwarranted. The stu- dents were mot randomly assigned to the experimental and control groups, but ' . ..a control group caposed of students equally matched with those in the experimental group was selected from subsequent class- es.‘ This practice and the fact that the control group was shown to be superior to the experimental group (Table VI) on the matching variable leads one to suspect that the students were from different populations. The method of matching is not entirely clear in itself. A coefficient of correlation of .66 was obtained between the matching variable and the criterion variable, and this matching variable was then used to match-out subjects to use in the control group. Such methodology seems certain to bias experimental results. Although the author states, I'In this experiment, equal results were obtained by matched groups of stu- 2'? dents who received training by the same instructor but by different methods ," there is no evidence presented to warrant such a statement. FinaJJy, there is no evidence presented to show that the control groups and experimental groups were actually cupared, but instead a third pro-test group is compared with the post-test experimental group and the post-test control group. These differences are found to be highly significant, as one would expect them to be, but this in no way proves that the experimental group performed less well, equally, or better than the control group, and in fact the differences were shown by the present writer to be significantly in favor of the control group on the basis of the data given. The experiment was poorly designed, the pro- oedures were dubious, and the analyses and conclusions were ambiguous and warranted. in experiment in am training is reported by tanner, Runyon and Desiderato (27). This study was designed to investigate differences between television and regular instruction, differences between kins- scope and regular instruction, differences in retention between tele- vision and regular instruction, differences between high and low apti- tude trainees on learning and retention through television and regular instruction, and the effects of kinescope review on retention compared with effects of no further training for high and low aptitude trainee. Fourteen hours of training were selected from the first eight weeks of basic training. Two experienced instructors were chosen for each of the fourteen hours of training. For 11 of the ll; hours one instructor taught by television while the second instructor taught in the conven- mm “to U I instructors me * instructor taught 01151:. Basic tra into to groups , Test, ind m, us Sim-”1! Other c instruction and r as used in this ltmcud for the . 28 tional manner. lhen a second group was taught the same subjects, the instructors reversed their roles. For the remaining three hours, one instructor taught both the television and the control group simultane- ously. Basic training cupanies used for the experiment were split into two groups, matched for scores on the Am General Classification Test, and were used to compare television and conventional instruction. Sisilarly other companies were matched for comparisons of kinsscopa instruction and regular instruction, however only 7 of the 11; hours were used in this comparison. Seventeen criterion tests were con- structed for the fourteen hours of instruction, including multiple- choice, fill-in, picture identification, and perforunce tests. Post- tasts only were given. The regular instruction group received instruc- tion in a large lecture hall, and the television group was sub-divided into small groups of 12 to 16 men and received instruction in proctored rows. In comparisons of the live television and regular instruction groups it was found that there were no significant differences on mean scores on 12 of the 17 tests. In five of the tests a significant dif- farenca was found in favor of the television groups. No differences were found between the television and regular instruction for the high aptitude group. For the low aptitude group significant differences were found in 10 of the 17 tests favoring the television instruction. Ihsn retention tests were administemd one month after the instruction significant differences were found in four of the 1h tests in favor of the television group. Detention tests were not made for high aptitude 29 men, but for low aptitude men on retention tests, significant differ- ences were found for h of the 11; tests favoring television instruction. In comparing kinescope instruction and conventional instruction significant differences were found in only one test, and this favored the kinescope group. Using kinescope review prior to taking retention tests resulted in the low aptitude men scoring significantly higher than on their immediate post-tests. The authors conclude that tele— vised instruction was at least as effective as regular instruction, that televised instruction was more effective for low aptitude groups, and that televised instruction was remembered at least as well as regular instruction. The authors also conclude that television effec- tiveness may be related to subject-matter content. a criticise of this study is the manner in which comparisons were made between companies. Although the groups were matched within companies the fact that the live television, kinescope television, and conventional instruction groups were not selected from the same companies introduced uncon- trolled variables such as differences in morale between companies. The fact that the regular group received instruction in a large lecture hall whereas the televised group was divided into small groups of 12 to 16 trainees for one received in a proctored room would in itself account for some of the differences found. A study conducted by the Educational Testing Service for the American Red Cross is reported by Shimberg (52). The study was de- signed to measure the differences in learning and opinion of the course under three conditions: (1) instruction by television, (2) instruction 3O bw'tslevision plus a weekly practice session, and (3) instruction in a regular classroom group without television. The two television groups received 13 half-hour programs over a period of seven weeks with the practice group receiving in addition a weekly practice session of one- hour. The regular*classroom group attended two one-hour sessions weekly'for a period of seven weeks. Subjects were members of organized Red Cross groups. The television groups viewed the programs at home. In terms of performance scores the television instruction was found to be as effective as classroom instruction and no differences were found in the two television groups. On the written test the television group did slightly less well than the classroom group and again there were no differences between the two television groups. Nb data on tests of significance are reported. Inasmuch as the populations in this study were volunteers and.no attempt was made to equate or randomize the sub- Jects between groups, no conclusions can be made. Another'weakness in this study is the fact that the classroom group received almost twice as much classroom time as the television groups. Seipman.(51) reports an experiment in teaching elementary school music in the public schools. Three-hundred sixth grade pupils in 7 different schools participated in the experiment with 177 Pupils in the experimental group and 123 students in the control group. Pre-tests and post-tests were given to all groups. The results of this investi- gation are reported in terms of the percent of gain between the pre- test and post-test examinations. The average percent of gain for the experimental group was 20.5 and for the control group was 21.8. It was 3]. concluded that the control group was numerically superior to the ex- perimental group. Although the original publication of the study could not be obtained, there are several criticisms that seem war- ranted. Apparently no randomisation nor equating was made of the sub- Jects within the two groups. No attempt was made to control the vari- able of different schools. The method of analysing the data was inade- quate. Ho statistical analysis was made to determine if the differ- ences obtained were significant, and although the study lent itself to a covariance design of testing methods by schools by classes the results are merely reported in terns of percentage gained. For the purpose of generalisation, the conclusions drawn are not warranted and the facts are inconclusive. Another study conducted in the public schools was reported by Anderson and Vander leer (2). The purpose of the study was to compare television and regular instruction on the use of the slide rule. latched groups frat five classes of high school sophomores were foraed on the basis of scores on the California Test of lental Maturity and the Stanford Achievement Test. Forty-one students comprised the tele- vision group and another forty-one students were in the conventional group. Six one-half hour programs were taught over a period of 6 weeks. The same instructor was used for both groups. a five-item test was given at the end of each class session and a final examination consisting of the items on the daily tests was given at the end of the series. No significant differences were found between the to groups on the basis of final examination scores, nor were there significant 32 differences between final scores when analysis was made to take into account differences in sex and intelligence. A study conducted in the Chicago Public Schools on the teaching of algebra and physics was reported by Willis (60). Nineteen different schools were used in the evaluation of algebra and 2).; schools were used in the teaching of physics. Ten lessons were given in all followed by an examination over the subject-matter covered. Instruction was by one instructor for the television sections but by different instructors for the classroom groups. There were no significant differences between the television md classroom groups in either algebra or physics. When the scores were adjusted for the ability levels of the students it was found that the television students in both courses did slightly better, but the difference was not significant. The report states that in- struction by television is slightly more effective than instruction in the regular classroom. This conclusion is not Justified on the basis of the reported data. The study may be criticized in that comparisons were made without attempting to control the variable of different in- structors and different schools. The results are inconclusive. Comparisons between classes under four different conditions of instruction is reported by Husband (25). The subject was a course in the psychology of adjustment at Iowa State College and the four con- ditions were: (1) television instruction received at home, (2) tele- vision in the classroom received by a monitor, (3) kinescope instruc- tion at a later time , and (it) normal classroom instruction. The in- structor was the same for all groups. The television sessions were 30 minutes in length while the classroom sessions were £0 minutes in 33 length. All groups received the same number of sessions. The group receiving television at home were not regularly enrolled college stu- dents, their median age was 37, and the length of time since graduation from high school was a median 20 years prior to taking the course. The other students were regularly enrolled college students. Grade point averages for the course were used for comparisons. The kinescOpe group did the best, the group receiving television at home next best, and the other groups were next in achievement. No tests of significance are reported. Inasmuch as the comparisons were made between groups receiv- ing differential amounts of instruction, and between groups that were obviously not from the same population no conclusions can be made from this study. An interesting experiment by Williams (59) was designed to com- pare instruction by television, lecture, radio, and by reading. One hundred and eight undergraduate students were divided into four groups so that each group contained an equal number of high, average and low ability students. Each of the four groups was assigned to one of four classes: (1) lecture, (2) television, (3) radio, and (11) reading mineographed copies of the lecture. The subject "Thinking Through Language," was not familiar to the students. The same lecturer pre- sented his material simultaneously to the lecture, television and radio groups, while the reading group read the lecture at the same time. The lecture group was in the studio where the lecture originated. Examina- tion consisted of 19 multiple-choice questions and an essay-type ques- tion to be answered in 200-300 words. The results were tested by analysis of variance and "t" tests on the objective part of the examina- 3h tion only. It was found that television instruction was superior to radio instruction beyond the 13 level of significance, that radio was significantly better than reading at the 5% level, and that there was no significant difference between the reading and lecture groups. Testing the data on the basis of academic ability, it was found that the same order obtained for the television, radio and reading groups. However, the lecture group was last in the amount learned by the high and low ability groups, but as high as television for the average groups. This study is well designed and its principal weakness, as the author points out, is the fact that the lecture group was not a 'true" lecture group, but rather a studio group and was undoubtedly influenced by the distractions of lights, cameras and equipment. Paul and Oglivie (1:5) conducted a follow-up of the Williams study reported above, eight months after the original study. Using the same multiple-choice test as was used in the first study, they adminis- tered the test to 7h of the 108 original students. An analysis of variance of the results showed that the television group was still highest, that the studio group (previously last) was second on the re- tention test, followed by the radio and the reading group. Again the authors in this stow state that no conclusions can be made about the studio group, because of the conditions obtaining in the studio. Another study conducted on the university level was reported by Evans, Honey and Icidams (l9). Cuparisons were made of two classes. In an elementary psychology class three groups of students were com- pared: (l) ninety-six students in an on-campus lecture section, (2) a television lecture session plus correspondence work consisting of 17 35 subjects, and (3) thirty students enrolled in a television lecture plus discussion section. In the biology course there were 78 subjects matched for college class, grades in college and sex. The criterion in the psychology course consisted of scores obtained on an 150 item multiple-choice type examination administered at the end of the course. lid-semester grades on a 70 item test were used as the criteria for the biology course. (he instructor taught each course for all groups. No significant differences were found between groups in either course. Inasmuch as no attempt was made to randomise or equate students within groups in the psychology course, it is impossible to draw any conclu- sions frm the experiment in that course. In the biology course the practice of using mid-term grades rather than final examination grades is questionable. It is conceivable that different results could have been obtained had final examination scores been used. Pasewark (uh) conducted a study to determine the effectiveness of television as a medit- in teaching typewriting. Forty-four college students were split into two matched groups on the basis of scores received on the American Council for Education Psychological Examina- tion. One group received instruction by television and the other group received instruction in the conventional classroom. The same in- structor taught both groups. Instruction was given one-half hour per day, five days a week for a total of 1:8 days. Achievement was measured by a timed typing test given at the end of the course. Students in the television section typed significantly faster than did students in the regular class section. There was no significant difference in the num- ber of mean errors between the two groups. The only criticism to be prco ngtO‘ Ute h. *7 Q n \ z .11: a 36 ude of this study is that in the television section there were no proctors or instructor, whereas the instructor was present in the con- ventional classroom. The absence of an instructor or proctor may have provided less disturbance and thereby enabled the television group to perfom better in this type of course. A study reported by Tannenbaum (55) was conducted at the Univer- sity of Illinois Medical School. Students in a basic physiology course were split into two groups equated on the basis of mid-term grades. Three 50-minute lectures were given to the two groups of students. One group received conventional lecture in the presence of television cameras and the second group received instruction by means of monitors. One week after the lectures an examination was given consisting of 19 multiple-choice items. Analysis of variance was used. to analyze the results. The television group was found to be superior to the lecture group at the 7% level of confidence. Although this is not the usual level of acceptance, the author suggests that the novelty of television learning may have accounted for the difference. The author suggests also that "nearness‘ to the instructor was a factor in learning reason- ing that the television students were ”nearer“ to the instructor than those students sitting in the rear of the lecture room, however such a hypothesis was not supported. 1 series of studies conducted at Purdue University (ID) in closed-circuit television instruction was carried on in classes in political science, analytic geometry, general chemistry, and general bacteriolog. In only one of these courses, that in political science, were the students randomized between the television and conventional 37 classroas groups. In the other three courses comparisons were made with other sections enrolled in the same course. In none of the courses were significant differences obtained between students in the two groups. These were considered preliminary studies and the results are not conclusive. Studies of Student Opinion of Television Instruction . In the previously reported study by Rock, Duva, and Murray 048) an attempt was made to evaluate the television instruction by asking the participants to make comments about the course. It is reported that cements by the trainees were favorable to the course. The staff felt that more visual emphasis should be made and the talkback micro- phone was felt to be inadequate because many trivial questions were asked. No objective data was obtained for the basis of making a statistical study of the evaluation. In the second study by the same authors U49) a series of ques- tions were asked the reservists about the course. Four-fifths of the group reported the series as interesting or very interesting, and a majority of the mup said the series was good or excellent. About three-fourths of the group felt that they would rather be taught by television than by the regular method. Again no objective data or analysis is made. In the study by Allen (1) a questionnaire was used to evaluate the course. A majority of the ROTC students reported that the tele- vision instruction was as interesting or more interesting than other types of instruction. Thirty-one out of fifty-three said the material 51' 15 C e ch!“ ‘ l" '1 u Ugl ‘1 38 was about as easy to learn as regular instruction, ll thought television learning was easy or very easy, and 11 stated that learning by tele- vision was more difficult than conventional instruction. No statistical analysis was made of the results of the questionnaire. Shimberg (52) in the American Red Cross study states that those who viewed the television instruction were overwhelmingly in favor of this type of instruction. In the study reported by Willis (60) on teaching by television in the Chicago Public Schools, the students reported that too much material was presented by television, but they favored the summary of key points at the end of the lesson. They felt that television was most effective when followed by discussion afterward in the classroom with the regular class teacher. Evans attempted to evaluate the attitude of students towards television by a questionnaire (19). He found that 70% of the students enrolled in an elementary psychology course would enroll in another course taught by television, 16% were undecided, and 13% said they would not enroll again in such a course. In the Purdue study previously cited (13) , the students were generally favorable to the demonstrations in chemistry and bacteriology over television, but in the courses in political science and analytic geometry the students were not so favorable. No statistical tests of differences were made. III. SPECIFICALLY RELATED STUDIES OF INSTRUCTIONAL TELEVISION Only a few studies have been designed to measure the effective- 39 ness of television instruction in relation to classroom discussion. ‘All of the studies in which this has been one of the purposes have been sponsored by educational institutions, and in the majority of instances these have been universities and colleges. These studies have been mainly concerned with achievement and retention of course content, and ‘with student opinion of the course. .A review of several studies closely related to the present study will be examined in the following section. Studies 23 Achievement and Retention 23.9223E2 Content Kumata (31) reports a study by Parsons designed to compare out- comes in a psychology of child development course under three different learning conditions. Forty university students were used as subjects and were randomly assigned to all experimental groups. Twenty students were assigned to a control group with no instructor. The remaining twenty'students were assigned to three groups. One group was designated as a kinescope-correspondence study group without an instructor; a second group was composed of students studying in a conventional class- room with an instructor; and the third group was an independent corres- pondence study group. All subjects used the same textbook, a workbook, and.manuals. The kinescope group met weekly to view the kinescope lectures and the classroom group met twice a week with the instructor at which time class discussions were held in addition to the lecture. Standard examinations were given to all groups. Pro-tests, post-tests, and a retention test was given approximately four'months after the close of the course. There were no significant differences among the three groups on pre-tests and post-tests and all three groups were signifi- ho cantly higher than the control group on the post-test. m the retention test the correspondence group was significantly higher than the class- rocn group, and the kinescope group scored lower than the correspondence group and higher than the classroom group, but was not significantly different from either. An evaluation of all written work for the course revealed no differences in any of the groups. A measure of group cohesiveness administered to the kinescope and classroom groups only—since the others had little contact—revealed no significant differences. A sociometric test designed to measure the group structure showed that the total mmber of choices increased sig- nificantly from pre-test to post-test administrations for the classroom group, and decreased slightly but not significantly for the kinescope (1‘0“?- Tbe original copy of this stucw could not be obtained so it is impossible to make a critical evaluation of the study. The study ap- pears to be well designed and carefully controlled except for one feature. The classroom group met twice a week with the instructor while the kinescope group met only once a week without the instructor. This feature of the study introduced a variable that could not be controlled. The author offers no explanation concerning the superior retention of the correspondence group. It may be that this group being an indepen- dent group became more highly motivated during the course of study, and reinforced their learning during the period intervening between the postdtest and the retention test. This finding might also be explained by vaothesising that overlearning occurred for this group because of hi the manner in which they were required to study the course, therefore their rate of forgetting was less. During the year 1955-56, experimentation in the use of closed- circuit television was conducted at New York University (h6). The pur- pose of the experimentation was to learn how to make the best use of television in college instruction, to compare the quality of instruction in televised and non-televised courses, and to investigate costs of in— struction using television. Two courses were selected for experimenta- tion, a freshman College Composition course and Literature of Englan -- a sophomore course. Students were those who regularly enrolled in the course. No attempt was made to randomize students within the two groups, and comparisons were made between several sections in each course, some of which received television instruction and some of which received normal.instruction. The composition course met three times per week, receiving a televised lecture for two of the meetings while the third hour was a 'tutorial” hour during which an instructor worked with the students on preparation of papers and to lead class discussions. The normal classroom groups met three times per week with their regular in- structor. The literature course was handled in the same manner. The television lecturers were members of the staff. No attempt was made to have one lecturer for all programs, but instead lecturers were alter- nated so that as many as eight different people participated in the lectures for one term. ‘Achievement was evaluated by comparing final grades at the end of the instruction with final grades of students in non-televised sec- ’42 tions during the same day and term. It was concluded that students per- formed about equally well in the television and non-television sections. It is suggested tentatively that the average and poor students in the composition course under television instruction did not perform as well as the average and poorer students in the conventional sections. In the literature course, however, the average and poorer students did better than their counterparts in the non-television sections. As the report states, however, these conclusions are merely tentative and sug- gestive, and more experimentation is needed with a more care fully con- trolled evaluation of results. It is difficult to make a critical review of the above study in- asmuch as it was not considered to be a final conclusive experimental study, but more nearly a 'pilot" study to precede a more carefully con- trolled study. It was not a carefully controlled study: subjects were not randomized nor equated, the instructor variable was uncontrolled, the criterion measure was not adequate, and no statistical measures were used to analyze the results. Even as a preliminary study it could have been considerably improved. A carefully planned and conducted study was made at Pennsylvania State University in l9Sh-SS (12). This experiment had several purposes but for the present study the discussion will be concerned with compari- sons of the relative effectiveness of conventional instruction with the same instruction presented over closed-cirouit television. Three courses were used in the study; General Psychology, General Chemistry and Psychology of Marriage. Four sections of General Chemistry were h3 used in the experiment. Students in these sections were matched on the basis of interests (major curricula) and grades in the proceeding chem- istry course (Chemistry 1). Two sections of 120 students each were designated as television sections, and two sections of 100 students each received the lecture in the regular manner. Three lecturers were used in such a manner that the students in all sections were exposed to all of the lecturers an eqrsl number of times and for the same subject- matter area of the course. (he of the experimental groups was sub- divided into four viewing classrooms, and one met in the lecture—origi- nating row with the television equipnent present. The students in General Psychology were randomly assigned to experimental or control groups. Two instructors were involved. One instructor lectured to a group of 1:0 students in the television origi- nating room, and to two classes of to students in the television re- ceiving rot-s, while a second instructor lectured to a control group of ho students in another classroom. later in the day the roles were re- versed for the same procedure with a different group of students. The Psychology of Iarriage course was not originally included in the experinent, but enrollnent was so large that it was decided to di- vide the class into four equal groups of about 30 each with one group in m originating room and the others in the receiving rooms. There was no control group in this course. One lecturer was used for the entire course. Proctors were provided for all television viewing sections whose duties were to take roll, adjust monitors, and perform otlmr procedural 14h duties, but who were instructed to do no teaching nor to answer ques- tions. Questions were referred to the regular lecturer. The criterion neasures in the courses consisted of examinations given throughout the course and the final examinations, all of the ob- Jective type. In addition a neasure of student attitudes towards the courses were given, and these will be discussed in the next section of this review. (See page 58 ) Analysis of results was computed by analysis of covariance and analysis of variance techniques. Reliabili- ties of tests were computed by [Mar-Richardson or by analysis of vari- ance of it. scores. In addition to the above neasures, an Opinion or 1" Scale was given to the students in General Psychology; The larriage Happiness Prediction Inventory was given to the students in the Psy- chology of larriage class; and a Student Reaction Schedule was adminis- tered to all students. » Except for one test adlinistered during the tern there were no sipificant differences found in any of the courses on the basis of achievement test scores. The study suggests that differences nay have existed but that the neasured samples of total learning may have lacked the appropriateness and sensitivity to detect the full differences. It is also suggested that sources of infornation such as library books, in- fernal student discussions, and other resources nay have reduced the probabilities of getting assemble results. It was also proposed that the television classes nay have assumed sore responsibility for their awn learning and hence conpensated for an differences in the two nodes Of instruction. The general conclusion made was, 'It is logical to con- 1:5 clude that there is no basis in the evidence found on information learn- ing for rejecting the use of instructional television for teaching courses and students like those used in this experiment.“ It is also concluded that the evidence available indicates that decisions to use or not to use instructional television in such courses as were studied and for defined student populations must be made on the basis of ad- ministrative policies, acceptability of televised instruction to stu- dents and faculty, -.and other practical considerations. The above study has much to calend it. It has the following strong features: ‘ 1. The problea was clearly defined and was framed in the form of objectives which were experimentally verifiable. 2. The statistical design was appropriate to the conditions. 3. Tie population and the sample was clearly specified and the manner of drawing the ssaple was clearly indicated. h. The control groups were chosen fro- the ssme population and inthe sassssnnerastheexperinental groups. 5. Il'he level of significance for test was set at the beginning of the experiment. 6. The treatments and methods of collecting the data were ade- quately described. 7. ktrsneous sources of error were controlled. 8. The criterion measure reliability was given. 9. The conclusions were consistent with the results obtained. 10. Generalisations were confined to the populations and condi- he tions of the experiment. The stucw could have been inproved in that it displayed the following weaknesses: l. The experiment did not include a replication. 2. The characteristics of the sample were not fully described. 3. The sssuaptions underlying the statistical tests used were not tested. In spite of these linitations , the Pennsylvania State University study is one the nest couplets and carefully emscuted studies that has been attempted. Another thorough and well-designed study is reported by Iissi University of Ohio (39). The purpose of the study was priuarily for studying the effectiveness of certain types of large group instruction and to dnonstrate the feasibility of these procedures at the college level. Courses included in the study were courses taught by tele- vision, large courses other than television courses, courses taught by graduate students, and conventional (control) courses consisting of 25-35 students. The first three types were considered as experimental courses. The control sections were taught by the same instructor who taught the corresponding television or large class section. The control sections in u:- graduate student phase of the study were taught by mn- tiae faculty seabers, but were of caper-able sise to the graduate stu- dent assistant sections. The variables employed for equating students included the scores on the Cooperative Test of English Achievenent, Cooperative Iathsastics Placement Test, and the herican Council for M Education Examination for College Freak-an. The equating was completed after registretion and the experimental and control groups represented only saaples of the total enrollumt. neither the instructors nor the students were aware of the students who were not included in the ssaple. It is important to note also, that decisions about collection of data and saspling technique were sade prior to the collection of data. lost of the courses were full year courses, and the sane procedures were enployed during the second semester as were used in the first semester except that students to be included in the sssple the second senester rust have been used during the first semester. Evaluation was made over four areas: achievement, student reaction to the course content and the instructor, student attitudes about instruction, and instructor attitudes . about teaching television and large course sections. is with the pre- vious stub only the achievement area will be considered leaving student attitudes for the next section. Criterion tests for achievement were neasured by subject-utter knowledge tests; tests on synthesis, problsa- solving, and critical thinking; course related tests, such as “Stereo- types in Social Studies' and 'Iisconceptions in Psychology." Seas of these sessures were objective and others were essay tests. Essay tests were graded by sultiple readers after precautions were taken to guard against "halo effects'. Graders were not aware whether the assays were for students in the experimental or control sections. Reliabilities were coaputed for all tests. Analysis of covariance and "t" ratios were used in analyzing the data. X The findings relative to the achievemnt area are as follows: h8 In the acquisition of subject-matter it was found that in general, stu- dents in television sections perform about as well as students in con- trol sections. Exceptions to this were observed in the second semester of full.year courses which was due possibly to the motivational decline as the novelty of television instruction was dissipated. Achievement in large course and control sections was about the same. In the com- parisons of achievement in critical thinking, problem solving, and synthesis it was found that television instruction was significantly inferior to conventional instruction in one course (Economics) but that no differences existed in the other courses. In the large course sec- tions no differences were found as compared to control sections, except again in one economics course. The achievement of course related atti— tudes was not investigated in the television sections. In the large course sections differences favoring the conventional type of instruc- tions were found in only one course (Economics), and in all the others there were no differences. In the comparison of large course sections and conventional sized classes, it was found that the large sections did not do as well in the Introductory Psychology course. This finding was also true for the high ability students in Business and Government, but not for the low ability students for they perform as well in large classes as in conventional size classes. In the Social Studies course there were no differences in.achievement. In.analysis of achievement results in which the inter- action of academic ability and.type of instruction was considered there were no differences reported except in one section. In general it was 119 concluded that no interaction existed. This stuck has many good characteristics. 1. The problem was clearly defined. 2. The problem was stated in the form of questions which were experimentally verifiable. 3. The statistical design was appropriate. 14. The method of drawing the sample was clearly specified. 5. The control groups were chosen in the same manner as the ex- perilental groups. 6. The treatments were randomly assigned. 7. The level of significance was specified prior to experinen- tation. 8. The methods of collecting data were adequately described. 9. htraneous sources of error were held constant or randomized. 10. The criterion measures were appropriate. 11. The reliability of criterion measures was given. 12. The conclusions were consistent with the results. 13. Generalizations were confined to the setting of the experi- ment. The stub has the following weaknesses: 1. The population was not clearly defined. 2. The characteristics of the sample were not given. 3 . The experieent did not include a replication. h. The assumptions underlying the statistical tests were appar- ently not made or tested. So In spite of these weaknesses the Miami University study is one of the most thorough and comprehensive to date because there was an attempt to find the relationship between ability and achievement in large and small classes. The final study to be considered in this section is one conducted at the State University of Iowa during the academic year 1956-57 (5). American Government, a three hour course meeting for three fifty-minute sessions per week, was selected for the experiment. The course is re- quired of majors in the department of political science as well as stu- dents in other areas particularly those majoring in teacher training. It is open to all classes, but the largest enrollees are Freshmen and Sophomores. The study was designed to answer several questions: 1. Can knowledge of and certain attitudes toward American Government be taught better by the discussion method or by the lecture method? 2. Does the method of instruction differentially affect the acquisition of knowledge and attitudes of students at different levels of academic ability? 3. What methods of instruction are preferred by instructors of the American Government course? 1;. What methods of instruction are preferred by students in the American Government course? 5. Can a class with an enrollment of sixty to eighty be taught by the discussion method and achieve the same results in course exami- nations as a class of twenty to twanty-five taught by this method? 301 the 131 81": M 51 6. Can an instructor, using closed-circuit television, promote more discussion and secure better results in course examinations with the sue umber of students as in a conventional classroom situation? 7. Hill the performance of students who can see and hear a television discussion, but who cannot participate orally in it, be differently affected as compared to those of students who can partici- pets? The treatment groups consisted of five types of sections: 1. Television discussion sections composed of a comparatively large number of students—about sixty to ninety-4nd divided into three groups. Group (he consisted of fourteen students in the television originating room whose voices and images were transmitted to the view- ingrooms. Groups TwoandThreewereintwooftheviewingroomsand could com-Inicate with the instructor and students in the originating roa vocally but not visually. lhen they wished to enter the discus- sion a signl was sent to the instructor via remote control and he could then call on tha. Students were rotated among the three groups approximately every two weeks, and all students spent approximately an equal amount of time in each group. 2. Television observation sections composed of twenty to thirty students who could see the instructor and the students in Group are in the originating room, and could hear the discussion between Groups One, Two, and Three. Students in this group, however, could not participate in the discussions. 3. ball group discussion sections varying in sise from eighteen 52 to twenty-nine. These sections were considered the control. ’4. Large group discussion section composed of seventy-five students. 5. lecture sections composed of one Imndred and twelve students one semester, and one hundred and thirty-two students the second semes- ter. No systematic discussion of topics was allowed by the sise of this section although the instructors allowed questions and interruptions during the class hour. Rania assigrments of students to the various sections could be followed only in the case of the television observation and television discussion sections. Assignments of the other students depended on the hour at which the course was chosen. The assmnption was made and tested that students who registered for the course at different times of the day did not differ significantly on any of the variables that would affect the criterion measures. Four teachers were involved in the experiment. One instructor taught all of the small group discussion sections and one of the large lecture sections during the first sweeter. A second instructor taught a lecture section the second semester. The third instructor taught a lecture session each of the two semesters. The large group discussion sections and the television sections were taught by the fourth instruc- tor. Although this arrangement was an obvious wealmess in the experi- ment, scheduling difficulties necessitated such an organisation. In order to promote discussion and to provide a common core of topics covered in the course, a syllabus including assigments, objec- 53 tives, and questions which served as the basis for discussion on any given day in the discussion sections was prepared by the members of the depart-ant. Evaluation was based on four criteria: 1. Achievemnt as eeasured by two mid-tern and one final examin- ation. Each nid-tere exasination consisted of forty multiple-choice its-s plus an essay question. The final examination for the first senester was composed of eighty multiple-choice questions and one essay question. The second semester final examination was lads-up of seventy eultiple-choice itees plus one essay question. 2. Attitudes toward concepts of "liberal democracy,“ as leasured by an instruent under development by one of the faculty nembers of the political science departeent. The eeasure contains forty-eight Likert- type questions. 3. Opinions concerning the various nethods of instruction and the perceived effects of each aethod. The opinions were appraised on a questionnaire. 1;. Attitudes of the discussion instructors toward these methods of instruction. These were measured on the basis of subjective iepres- sions suhaitted by the instructors. Replication of the stw was obtained by comparing the results of the study for two identical semesters. In order to test the assuaption that the various groups did not differ significantly on an of the variables assueed to affect the cri- terion aeasures, the hceogeneity of the groups on the basis of the En- Sh trance Capacite Percentile Bank, which each student had received when he entered the university, was tested by analysis of variance. In none of the groups was a significant difference obtained and the assumption of haogeneity was accepted. The study does not explain what tests are used for computing the Entrance Cmposite Percentile Rank. An analysis of variance of mean scores on the three achievement examinations and the total of the three examinations revealed no sig- nificant interaction between the measures and the sections and no dif- ferences among sections on achievement tests. The only significant difference was that between acasures—tests—and this was expected in- asmuch as a much higher score was possible on the third achievsmnt examination. In order to examine the relationship of ability level with the method of instruction, a 'treataents by levels“ analysis of variance was made. Two separate analyses were made using final mnination scores as the criterion. In one analysis, the Entrance Composite Por- centile Bank was used as the control variable and in the other the Pre- test examination score'was used. Three levels were used for each of the control variables, the top quartile, the two middle quartiles, and the bottu quartile. The analyses revealed no differences and the hypothe- sis of no differences between the effectiveness of the various methods of instruction was accepted. A significant difference was found be- tween the various ability levels as was expected. It was concluded that achievement in the Anerican Government course, as measured by the mid-term and final examinations for the 55 course was improved by all methods of instruction, and there was. no differences due to the method of instruction. Analysis of the results on the attitude criterion showed no sig- nifioant differences between groups during the first semester. The at- titude data for the second semester was analyzed by levels using the pro-attitude score as the control variable and the post-attitude score as the criterion measure. A significant difference between levels was found, as was expected. The differences between groups were also found to be significant. the of the lecture groups obtained the highest lean and the lowest was obtained by the television observation group. None of the other groups differed significantly. Inasmuch as only one of the two lecture groups was significantly different, and a comparison of these results with first suester results showed no similar trends, it was suggested that this difference was probably a chance effect. The authors conclude that there were no significant differences between methods of instruction on the learning of attitudes towards the concepts of I'liberal democracy“ as measured by the criterion instrments used in this study. Theabove studyhasmuchtocmmendit. l. The problem was clearly defined, and was framed in the fona of questions which were easily transferrable to null hypotheses. 2. The statistical design was appropriate to the methods and hypotheses under test. 3. The population and the method of drawing the sample was clearly specified. 1:. The experimnt was replicated. 56 5. The level of significance for rejection of the null hypotheses was specified before the data was collected. 6. The treatments and the method of collecting data was ade- quately'described. 7. The criterion measures employed were appropriate to the pur- poses of the study. 8. The statistical assumptions underlying the tests of null hypotheses were satisfied. 9. The conclusions were consistent with the results, and gen- eralizations were confined to the population from which the sample was drawn. The following weaknesses in the study should be indicated: 1. The control group was not chosen in the same manner as the experimental groups. 2. The characteristics of the sample were not adequately de- scribed in that the various curricula, classes, age and sex of the sample was not given. 3. The fact that different instructors were used in different groups introduced an extraneous source of error that was not controlled. h. There was no evidence that the reliability of the criterion measures was computed. Although the subjects in this study were not randomly assigned to all groups, the fact that all subjects were tested for homogeneity of variance on the Entrance Composite Percentile Rank and.found to be homogeneous indicates partial control of this variable. The fact that 57 one instructor taught the tehvision sections and the large group dis- cussion sections lessens the seriousness of the criticism of the un- controlled instructor variable, for the comparisons between the above groups a Studies g_f_ Student Qpinion of Television Instruction In the Parsons study (’43) previously cited, student evaluation of the course was rated on a three point scale. A score of 1.0 on the scale was defined as 'more valuable" , 2.0 was the midpoint, and 3.0 was defined as "less valuable.‘ Mean ratings for the classroom group were 1.0, the correspondence group scored a 1.1; and the kinescope group aver- aged l.7. No tests of significance were reported for these figures. A student questionnaire was used to assess student opinion about the course in the How York University study (146). Although no objective statistical su-ary was made of the results, the following general sm- mary is quoted from the stuck: I. There was no change during the year in the students' attitude toward the usefullness of television as an educational device except for an increase in the favorable attitude of the Literature of England group. Both groups were favorable toward the use of tele— vision as an educational device. 2. The College Composition group shifted during the year toward a preference for one instructor. There was no change in the Liter- ature of England course . Both groups were preponderantly favorable to the idea of several instructors . 3. Both groups indicated dissatisfaction with the course. Dur- ing the year the College Composition group shifted to an even more negative attitude. 1;. There was a tendency for students generally negative to com- mercial television to be less satisfied with the course and more negative toward television as an educational medium. In general, they were less favorable toward all aspects of the program. 58 5. No difference occurred between the Composition and Litera- ture groups with respect to their satisfaction with the course, and in general, there were no significant differences between these two groups with respect to the whole program. 6. No differences occurred between any of the groups with re- spect to the felt 'closeness' of the subject matter. The majority did feel that television prevented close contact with subject matter. Several neasures were used to evaluate the students' attitudes toward television instruction in the Pennsylvania State University stub (12). One of these, a Student Reaction Schedule, was administered to the students in the television receiving roan only. The students wen asked to rate their estinated amount of learning and interest in the course compared to what it probably would have been in a conven- tional face-to-face classroom situation. They were also asked to write a paragraph explaining the reason for their ratings, and the advantages and disadvantages of the course. Although a najority of the students estinated they were learning “about the same“ or "a little less" by television no clear trends were evident. This scale was used in all three courses (General Psychology, Psychology of Earrings, and General Chemistry) and the results were sinilar from course to course. In the General Psychology course the students were asked to com- plete a questionnaire purportedly distributed by a “Curriculum Survey Co-ittee' in which they were asked to rank all of the courses they were taking during the senester of the experiment in the order in which (1) the course was contributing to their academic objectives and (2) the courses were liloed by them. The results of this rating showed that the students in the conventional classroom ranked the course significantly 59 higher than did students in the television receiving room both in the contribution of the course and in their liking for the course. No differences were found between groups in the originating classes and those in the standard classes. A Personal Relevance scale was administered to the students in the General Psychology course. This scale was a Guttman-type scale and was designed to measure the perceived immediacy or remoteness of the use of the information by the students. Analysis of variance of the results revealed no significant differences between the three methods. There was a significant interaction between the method of instruction and the instructor. The authors suggest that more investigation needs to be done before definite conclusions can be made about this inter- action. in 1" Scale designed to neasure "authoritarianism“ was also given to the students in the General Psychology course. . No sigrificant dif- ference was found among the experimental and control groups on this neasure. The strengths and weaknesses of the Pennsylvania State University stw that were listed in the section pertaining to the achiement of course content can also be made of this part of the study. In general it is an excellent study. In the study at Miami University (39) which was previously cited, several evaluative criteria were used to measure the student's attitude about the course and the instructor, and about television and large class instruction. l. The C-Scale, designed to neasure the students' rating of the course, was a Thurstone-type attitude scale with a nine-point con- tinuum. Five was the neutral point on the scale; scores below 5.0 were favorable and scores above 5 .0 were unfavorable . 2. The I-Scale was used to rate the instructors. It consisted of twenty-four items of teaching effectiveness about which the students rated the instructor. Scores below 5 .0 were favorable and scores above 5 .0 were unfavorable . 3. A TV-Scale required the students to evaluate the effective- ness of television instruction in comparison to conventional small class instruction. h. in LC-Scale was used to compare the effectiveness of large class instruction with conventional instruction. 5. A TV-lB-Scale required the students to compare television and large class instruction. The TV and LC Scales were administered in all courses at the end of the first semester and readministered in certain courses at the end of the second semester. The results lead to the conclusion that as a group students assigned to television and large course sections prefer assignment to a conventional section. In two of the courses where the scales were administered at the end of both semesters it was found that the attitude toward television at the end of the second semester was less favorable than at the end of the first semester. In one course the students were exposed to television instruction one semester and large course the next semester with the same instructor. In comparisons 61 of these modes of instruction and of small class instruction, it was found that the students favored small classes over large classes and favored large classes over television classes. The study reports the following conclusions regarding student attitudes about television and large class instruction. 1. Students assigned to TV or DC (large class) sections generb ally do not like them as well as conventional (small) classes. 2. It is possible to teach a TV course in such a way that stu- dents actually prefer it to conventional instruction. This, how- ever, requires a uniqre combination of instructor, course material, and the full use of the potentialities of television as an instruc- tional medium. 3. Students in at least one course (wherein the problem was in- vestigated) preferred large class instruction to TV instruction, although they tended to prefer small class instruction to either large or TV classes. 14. Students in TV courses tend to become disenchanted with tele- vision as a means of instruction during the course of the year. lost students reported that they neither learned as much nor were as attentive during the course as they had originally anticipated. Students in LG courses, however, are better able to anticipate their end-of-the-year reactions than those in TV courses. 5. Attitudes about the means of instruction are much more vari- able between TV courses than between LC courses. The prevailing attitude toward LC instruction as compared to control instruction is mildly unfavorable. The prevailing attitude toward TV instruc- tion as compared to control instruction ranges from strong enthu- siasm to extreme displeasure. 6. The instructor is a major determinant of how students will react to TV and IC instruction. There is a pronounced tendenc for students who dislike their instructor to dislike TV (or LC classes and vice versa. 7. The majority of students would enroll in a TV or DC section (even though they prefer small classes) if it meant that they would be assured of being taught by an excellent instructor. 8. In general, attitudes about TV and LC instruction are inde- pendent of level of academic ability. Two exceptions to this gener- alisation were apparent during the spring 1956 semester wherein an 62 inverse relationship between academic ability and attitude about TV instruction was obtained. The analysis of the results obtained after administration of the C-Scale and I-Scale are sunnarized in the study as follows: 1. Blanket generalizations about student attitudes regarding the worth of a course as a function of class size are not Justified. Other factors, including course content and the ability of the in- structor to handle larger groups of students interact with class sise to affect these attitudes. 2. Student motivation and interest in the subject-matter is not significantly dininished when the course is presented on television or in large classes. 3. There is a pronounced tendency for instructors to be rated as none effective when they teach conventional or small sections than when they teach TV or large classes. Again, however, this generali- sation does not hold for all instructors. Sue teachers are able to teach large groups as effectively as smaller ones. 1;. Sons of the specific "intangible" benefits often associated with a low student-instructor ratio need not be sacrificed as a result of large group instruction. These intangibles are, however, achieved sonewhat lore satisfactorily by faculty members than by graduate assistants. The Mia-i University study was critically evaluated as t o the achievement portion of the experiment in the proceeding section of this chapter. The strength and weaknesses indicated in that evaluation apply as well to the above portion of the study dealing with student attitudes. The only criticise that can be made in addition to the above is that the data and the design were well adapted to an analysis of variance test, instead of the 't“ tests that were used. The final study to be reviewed in this section is the State Uni- versity of Iowa experiment (5) . Student Opinions concerning the various nethods and the effects of each method were gathered during the first sensester only. In analyzing the opinions of each group toward the alter- 63 native methods of teaching the course, two results were shown. One was the importance of sue form of discussion for the students and the otter was the inclination of those students who had some familiarity with tele- vision instruction to prefer discussion by this method. A comparison of the groups on the estimated relative preparation required for the course taught by each method showed that the students in the discussion sections felt that the course required more preparation than did those in the lecture section, the television section or the television observation section. The differences were significant between all groups. In reply to a question about which group had learned the most, it was found that the students in the small discussion section felt that they had learned the most and those in the television discussion were only slightly lower than the small discussion group. Students in the ' television observation group felt that they had learned the least. These differences were all significant at the 5% level of confidence. There were also significant differences between the groups on the extent to which they felt motivated to greater thought.1n comparison with compar- able lecture and discussion courses. The major results as summarised in the study are as follows: 1. Students in American Government preferred the course taught by the discussion method or a combination of lecture and discussion rather than lecture alone. 2. Students who had experienced a course taught by the discus- sion method on television were more favorable toward the use of tele- vision for instructional purposes than were students who had not had 6h this experience. 3. The students' ratings or self-evaluations of the amount they learned as well as the degree to which they were stimulated by the courses were positively related, in the television sections, to the students' opportunities to participate. One of the weaknesses of this study, as indicated earlier, was that of not controlling the instructor variable. In making comparison of discussion sections with other types of instruction, this is an im- portant criticism. The ability to use the discussion method is more closely related to personal characteristics of the instructor rather than the ability to enumerate factual data characteristic of the lee- ture method. It would seem, therefore, that the results on the opinion criteria can be only accepted with reservations until further verifica- tion is made. IV. SUMMARY The literature about instructional television reviewed in this chapter was concerned with general background studies in which the effi- cacy of television as a medium of instruction was explored, and'with specifically related studies in which television instruction was used in discussion methods of instruction. The results of the review are summarized in Table I, page 65. Inspection of Table I reveals that of the general background studies in which the achievement of course content was used as a cri- terion only four of the sixteen studies were satisfactory in terms of 65 TABLEI NUMBER OF STUDIES REVIEWED, CLASSIFIED BI CONCLUSIONS, ADNULCEY OF METHODOImY AND PURPOSE. Findings Findings Findings Favorable not Not Favorable to Television Conclusive to Television Instruction Instruction Critical Critical Critical Evaluation Evaluation Evaluation Satis- Unsatis- Satis- Unsatis- Satis- Unsatis- Purpose factory factory factory factory factory factory General Background Studies Achievement of course content 2 S 2 S 0 2 as criterion Opinion of course as criterion” 0 5 o 2 o 0 Specifically Related Studies Achievement of course content 0 O 3 2 O O as criterion (pinion of course as * O O 2 1 1 1 criterion "The studies reported in this row are parts of the studies re- ported in the row above using the achievement criterion. Not all of the studies used both an opinion and achievement criterion hence the difference in numbers . 66 experilental procedures, methods, techniques and analyses. Of these four, two were found favorable to television instruction and in two the findings were inconclusive. In the seven studies in which the opinion of the course was used as a criterion, all were found unsatisfactory in terse of experhental procedures. In the review of studies specifically related to the present study in which the achievement of course content was investigated, three of the five were found satisfactory and the findings were incon- clusive. In the same five studies in which the opinion of course was used as a criterion, two were found to be inconclusive and in one the findings were unfavorable toward television instruction. The other two studies were unsatisfactory in terms of experimental procedures. Because of the many differences in designs of the studies re- viewed, the diversified backgrounds of the subjects, the variety of sub- fleet-utter courses used in the studies, and the varying conditions pre- vailing fro. one study to another, it is difficult to make specific gensralisations about television instruction. Probably the most that can be said at the present tile is that there is no conclusive evidence to show that television instruction is either more or less satisfactory than ordinary classroom instruction. In spite of such findings eany educators and writers have over- looked the weaknesses of the many studies of instructional television and have assuled that if television is not a better medium of instruc- tion than the sore conventional methods, it is at least Just as good and can be used to reach a greater number of people. Such conclusions 67 have resulted in the installation of television equipment in many schools and colleges as a solution to the problems of teacher shortages. Such practices may be justified where television is to be used for the tran-ission of factual information, visual demonstrations, movies and the like, but there is no clear evidence to indicate that television can substitute for conventional classroom methods where discussion eethods of instruction are the principle ways of transmitting informa- tion and teaching critical thinking, problem-solving, and the like. It is clear that more experimentation is needed along these lines in the use of television instruction. Inasmuch as discussion techniques are an important part of teaching, the present study is an attempt to evaluate the use of instructional television where discussion is used in the teaching methods. The design of the present experiment will be described in Chap- ter IIIe CHAPTER III THE DESIGN OF THE EXPERIMENT There are a number of statistical principles and various de- signs available for the modern experimenter in education. The present experiment employed a four-way factorial design. The use of such a design made it possible not only to determine the effects of'the two discussion methods, but also to explore the effects of certain associ- ated variables and their interaction with the discussion methods and with each other. The control of extraneous variables by randomization procedures were also provided for in the design. In addition, by the use of analysis of covariance it was possible to statistically control for initial differences of the subjects on the criterion variables. I. THE DESIGN The students were randomly assigned to two groups. One group was designated the television (TV) group and received discussion by means of closedpcircuit television. The second group was designated the non-television (NTV) group and discussion was in the face-to-face classroom situation. The randomization and discussion methods will be described in detail in Chapter Four. The Independent Variables In addition to the discussion variables three other independent variables with two categories in each variable were chosen for study. 69 The independent variables were: (1) the discussion techniques, (2) the ability level of the students, (3) the preference for instruction of the students, and (h) the sex of the students. The discussion techniques were, of course, the primary vari- ables under study. There were an almost unlimited number of factors that could be considered for other variables. Among these were age, interests, socio-economic status, personality classification, ecllege grade-point average, class in college, and many others. The choice of the number and the particular factors for study in any experiment is dependent upon the purposes and design of the experiment, upon estab- lished and conventional practices, upon the interests and preferences of the experimenter, and upon the practical limitations of partitioning the variables. In the present study it was decided that more than four variables would become unwieldy in terms of statistical procedures and would be difficult to interpret if higher-order interacting effects were obtained. In the choice of the particular factors to use an attempt was made to choose variables that would be of practical use and significance for the present study and for future investigators of instructional television. Thus in addition to the discussion vari- ables, it was decided that differences in ability level of students, the differences in the preference for instruction that students exhibit, and the differences in males and females would be the most practical variables to study in terms of educational outcomes under the conditions of television and non-television discussion. The decision to use two categories or levels for each of the 70 variables was made in order to determine any interacting effects that might exist among the variables. The choice was, of course, limited to two levels in the case of the discussion and sex variables. In the case of the preference for instruction variable, and the ability level variable the number of levels was arbitrarily set at two in order to minimize the expense and labor involved in the statistical analysis. The Dependent variables In the choice of criterion variables the experimenter’must make decisions about what effects he wishes to measure and then select or devise instruments to measure these effects. In.many cases the choice of dependent variables will be influenced or limited by the types of measuring instruments available, the conditions of the experiment, and the existing facilities. In the present study the major concern was in.measuring the effects of the discussion conditions on educational outcomes. Four criterion variables were chosen. These were: (1) student achievement of course content, (2) student attitude toward teaching, (3) student misconceptions about education, and (h) student opinion.about the course. The use of four criteria was considered adequate to give a variety of measures for determining the effectiveness of the two dis- cussion methods and their relationship to the other control variables. The criteria selected were realistic in terms of the purposes and con- ditions of the experiment. The experiment was conducted in a course in beginning educational psychology} it was required of all students majoring in education; and the majority of the students were sophomores. ha: «611'. < V n l :e A 67* BC "7.. tun- C6? A: m .N 71 In addition to the usual criteria of course content it was thought im- portant to learn about the effect of the discussion techniques on the attitudes toward teaching of beginning students in educational psy- chology. Likewise, it was considered of interest and importance to note any relationships that might exist between the discussion methods and the ability level, preference for instruction, and sex of the stu- dents on his attitude toward teaching. The misconception criteria was chosen to include an area not measured by the usual achievement and attitude test. Misconceptions so conceived are in the nature of ir- rational beliefs or attitudes not based on facts and not attitudes in the usual sense. The Opinion criteria was selected to determine the student's Opinions about the course under the conditions Of the two discussion techniques and their relationship to the other control vari- ables. In addition the Opinion variable would provide a measure Of the morale Of the students. The instruments for Obtaining measures of the independent and dependent variables will be fully discussed in Chapter Four. Diagrammatic Plan Q; the Design The combination of four independent variables with two levels of classification for each variable as employed in the present study is known as a four-way factorial design. It is sometimes described in quantitative terms as a 2 x,2 x 2.x 2 (read as two by two by two by two) or simply a 21‘ design. Such a four—dimensional design as con- trasted to a two or three-dimensional design, is difficult if not im- possible to represent or visualize pictorially. It is possible, howb 72 ever, to represent the relationship and inter-relationship of the four variables in a diagrammatic figure, and these relationships in the present study are shown in Figure l on page 73. By splitting each of the two discussion groups (TV and NTV) into two groups on the basis of scores on the ability variable; by splitting them into two groups on the basis of scores on the preference for instruction variable; and by further dividing them on the basis of sex it can he been that sixteen subgroups can be formed for purposes of analysis. Thus by using vari- ance or covariance analysis techniques, it is possible to examine the effects of various combinations of variables at different levels of each variable and to test the significance of each. It was thus possible to simultaneously measure the main and interaction effects of the fOur control variables for each of the four criteria measures with- out resorting to the tedious procedure of several independent experi- ments of the single variable type. Thg_Control‘gf‘Extraneous variables The design of the present experiment provided a means of control of many extraneous variables that would be difficult to control in a simpler type design. Individual differences of subjects introduces may variables into the experimental situation. In many experiments an.attempt is made to control these variables by equating subjects on those factors that it is thought will bias the results such as intelli- gence, age, academic rank, interests, and.many others. It is probably impossible to select and equate subjects on all of the many variables that might affect the experimental results. It is possible, however, 73 The Independent Variables High.Ability Level Low Ability Level High Low Preference Preference for for Instruction Instruction High Low Preference Preference for for Instruction Instruction Male Female Male Female Male Female Male Female TV NTVTV NTVTVNTVTV NTVTVNTVTV NTVTV NTVTVNTV 1 2 3 u 6 8 10 11 12 13 1h 15 16 $1- ii- * it it i- t? it il- * i!- * i- i!- '* *- * Scores on each of the four criterion.measures are entered in the sixteen cells of this diagram for analysis. FIGURE 1. Relationship of the four independent variables with two levels for each variable and the sixteen subgroups formed by the various combinations. 7h to minimize many sources of error or bias by randomization procedures, thus randomizing the uncontrolled error variables among the treatment groups. In the present study the subjects were randomly placed in either the TV or NTV groups at the outset of the experiment. In this manner any variations and sources of error due to individual differ- ences were randomized among the two methods. The randomization pro- cedures will be discussed in Chapter Four. Another source of error in studies of the single variable type in which a series of experiments are performed, is that introduced by the conditions of the experimental situation changing from experiment to experiment and from the risk of not Obtaining samples from the same population for subsequent experiments. Or, if the same sample is used from one experiment to another there is no way of controlling varia- tions due to intervening activity from one experiment to another. In the present investigation such sources of variation were controlled by the design of the experiment itself, because it was possible to Obtain simultaneous measurements of the effects of the desired variables. Differences in instructors is also a possible source of error in experiments of teaching methodOlOgy. In the present experiment the instructor variable was controlled by alternating the instructors between the TV and NTV groups from one class period to another, thus any error introduced by this variable was randomized among all subjects and treatment groups. No control section in the sense of traditional experimentation 75 was provided for in the present investigation. The experimental design was planned to measure the relative effectiveness of the two discussion methods, each group in effect was both a control and an experimental group in relation to the other. Inasmuch as students bring a wide variety of abilities, inter- ests, attitudes, aptitudes and experiences to the experimental situa- tion it is unwise for the experimenter to assume, even for the most naive subjects, that all subjects are at the same level initially. In order to control this source of error, beyond the control afforded by randomization procedures, pre-test and post-test measures were obtained for all of the criterion variables except the opinion variable. By the use of analysis of covariance it was possible to statistically adjust for differences in initial standing on the criterion variables. In the case Of the opinion variable it was thought that the administration of this measure at the beginning of the experiment would influence the post- test measures. Analysis Of variance, chi-square, and the “t" test were used to test the results on this measure. A discussion of the statisti- cal treatment will be found in section three of the present chapter. Replication was not provided for in the present experiment. Rep- lication could have been provided by duplicating the experiment with a second section of the same course, by repeating the experiment in a sub- sequest term, or by altering the present design and using the present subjects and course facilities. Because of administrative decisions in the matter of scheduling courses, instructor assignments, and classroom facilities the first two possibilities were not feasible. Altering the 76 design of the present experiment would have resulted in a reduction of the number of subjects for each replication or a reduction in the num- ber of variables for study. Inasmuch as this was an initial experi- ment of its kind, there was no way of knowing at the outset what prob- lems and conditions would be met during the course of the experiment. A replication would have enhanced any problems encountered and would have increased the cost of the experiment in time and money. The decision was made, therefore, to maintain the design in its present form and accept the limitations of no replication. In the future more refined investigations of a similar nature should provide for replica- tion in the experimental design. II. THE NULL HYPOTHESES Modem experimental procedure usually concerns itself with the purpose of testing a “null" hypothesis or the hypothesis that there is no true differences amulg the experimental treatmnts as far as the criterion is concerned. In the present design with two levels for each of four independent variables or "treatments" , four main effects and eleven interacting effects may be derived. Thus a total of fifteen different null hypotheses may be formulated for test on each of the criterion measures. These effects and interactions and their corres- ponding null hypotheses are: Main Effects 1. Ability There are no differences between mean scores of the criterion instruments attributable to the ability level of the students. 77 2. Preference for instruction There are no differences between mean scores of the criterion instruments attributable to the preference for instruction of the students. 30 Sex There are no differences between mean scores of the criterion instruments attributable to the sex of the students. h. Discussion techniques There are no differences between mean scores of the criterion instruments attributable to the two discussion techniques. First-order Interaction Effects 5. Ability and preference for instruction There are no differences between mean scores of the criterion instruments attributable to the interaction effect of the ability level and preference for instruction. 6. Ability and sex There are no differences between mean score of the criterion instruments attributable to the interaction effect of the ability level and sex. 7. Ability and discussion techniques There are no differences between mean scores of the criterion instruments attributable to the interaction effect of the ability level and the two discussion techniques. 8. Preference for instruction and sex There are no differences between mean scores of the criterion instruments attributable to the interaction effect of the preference for instruction and sex. 9. Preference for instruction and discussion techniques There are no differences between mean scores of the criterion instruments attributable to the interaction effect of the preference for instruction and the two discussion techniques. 10. Sex and discussion techniques There are no differences between mean scores of the criterion instruments attributable to the interaction effect of sex and the two discussion techniques. Second-order Interaction Effects 11. Ability and preference for instruction and sex 78 There are no differences between mean.scores of the cri- terion instruments attributable to the interaction effect of ability, preference for instruction and sex. 12. Ability and preference for instruction and discussion tech- niques There are no differences between.mean scores of the criterion instruments attributable to the interaction effect of ability, pre- ference for instruction and the two discussion techniques. 13. Ability and sex and.discussion techniques There are no differences between mean scores of the criterion instruments attributable to the interaction effect of ability, sex, and the two discussion techniques. 1h. Preference for instruction and sex and discussion techniques There are no differences between.mean scores of the criterion instruments attributable to the interaction effect of the preference for instruction, sex, and the two discussion techniques. Third-order Interaction Effect 15. Ability and preference for instruction and sex and discussion techniques There are no differences between.mean scores of the criterion instruments attributable to the interaction effect of ability, pre- ference for instruction, sex, and the two discussion techniques. In the present experiment the primary concern was with those hypotheses in which the discussion technique was either a main or inter- acting factor although all hypotheses were tested. The statistical tools used to test the hypotheses and the assumptions underlying each are discussed in the next section. III. THE STATISTICAL TREATMENT The analysis of variance, the analysis of covariance, the chi- square, and the ”t“ test were the principal statistical tools used in the present study. By the use of these tools the experimenter is able to analyze the total variance into independent components for each 79 experimental variable and test its significance statistically. It is important, however, that the assumptions involved in the use of these techniques be satisfied or the experimenter cannot be sure that the obtained differences are "true" differences, but rather result from a failure to satisfy the assumptions underlying the use of the statisti- cal tests. The assumptions underlying the analysis of variance are: 1. The groups are selected at random from the same population. This assumption was accepted in the present study on the basis of the randomization.methods used and described in Chapter Four. 2. The distribution of criterion scores within groups or cells is normally distributed. This assumption of normality of distribution is seldom tested and it was not tested in the present study. It has been shown by the Nerton study (38, pp. 78-90) that the F-distribution is relatively insensitive to the form of the distribution, and walker and Lev (56, p. 230) report that "...empirical studies of samples from nonrnormal populations indicate that a considerable departure from normality does not invalidate the methods described? in a discussion of the use of statistical methods including the analysis of variance. 3. The groups are homogeneous in variability. This assumption was tested in the present investigation by the Bartlett test as de- scribed in.Lindquist (38, pp. 87-88). The assumptions underlying the analysis of covariance are: l. The groups are selected at random from the same population. This assumption was accepted on the basis of the randomization methods “88d 0 80 2. The distribution of criterion scores within groups or cells is normally distributed. The assumption was not tested and was dis- cussed in connection with the analysis of variance above. 3. The groups are homogeneous in variability. This assumption was tested by the Bartlett test previously cited. h. The regression of post-test scores on pre-test scores is the same in all groups or cells. The assumption of homogeneity of regres- sion was tested by the method described in Walker and Lev (56, pp. 390- 393). 5. The regression of post-test scores on pro-test scores is linear for all groups or cells. The assumption of linearity of regres- sion was tested by the method described in Walker and Iev (56, pp. 396- 397). (be other important condition in designs of this general type is that the number of cases from cell to cell in the same row or column be proportional. Although it is not imperative that this condition be observed, the statistical computations become very involved and burden- some if it is not observed. In the present study proportionality with- in cells was obtained by randomly selecting a portion of the sample from each of the television and non-television groups. In analysis in- volving the Chi-square and "t' tests the total sample was used. The level of significance for accepting or rejecting all null hypotheses was set at the 5% level of confidence prior to the beginning of the experiment. For the tests of assumptions underlying the analysis of variance 81 and-covariance techniques, the level of confidence was set at the 1% level prior to the beginning of the experiment. IV. SUMMARY A fourdway factorial design with four independent variables and two levels of classification for each variable was used in the present investigation. The primary variable in the study was the discussion variable with the two levels represented by'a television (TV) group and a nonstelevision (NTV) group. The other variables were the ability level, the preference for instruction, and the sex of the students. Four criterion.variables were selected to measure the effects and inter- actions of the control variables. These were achievement of course cone tent, attitude toward teaching, misconceptions about education, and the opinion of the course. Null hypotheses were derived for the correspond- ing effects and interactions of the independent variables. The basic statistical tools were the analysis of variance, the analysis of covari- ance, the Chi-square test, and the "t‘ test. The assumptions underlying the use of the statistical tools and the methods or tests for satisfying these assumptions were discussed. The level of significance for testing the null hypotheses and the assumptions were set at 5% and 1% levels respectively. CHAPTER IV EXPERIMENTAL METHODS AND PROCEDURES The methods and procedures used in conducting an experiment are determined in part by the experimental design and in part by the physi- cal setting and facilities. In the present chapter the course and the facilities, the pOpulation and the sample, the randomization procedures, the discussion techniques, and the instructors will be fully described. I. THE COURSE AND THE FACILITIES The investigation was conducted in section h of Foundations of Education 200, "The School and the Individual,‘ a six-credit course in beginning educational psychology offered in the College of Education at Michigan State University. The section.met three times weekly for two consecutive periods of fifty minutes each. The catalog description of the course follows (bl); Mental, emotional, physical, and.personality develOpment of the human organism related to problems of formal education. Concepts of learning, motivation, memory, habits, individual similarities and differences as they condition the learning situation. The section was chosen for the investigation because it provided a large enrollment, varying from approximately 150 to 200 students de- pending on the term, and because it had been previously taught by closed-circuit television and many of the administrative problems in- volved were known. In addition it is a required course for all students majoring in education and probably plays an important part in the for- 83 nation of student attitudes toward education and teaching. It was thus an ideal course for studying some of the effects under consideration in the present investigation. The originating room for the television presentation was a large ampitheater-type lecture room with a seating capacity of 331 students. It was provided with two television cameras with three rotating lenses, necessary lighting facilities, microphones and loud-speakers, a black- board, an easel, a speaker's lectern, and a table and chairs. The control room in the basement was equipped.with a 16mm movie projector, a 2x2 slide projector, and.monitoring and switching apparatus for the program director and engineers. Two viewing classrooms each with a capacity for 96 students were provided for the experimental subjects. Each viewing room contained two 2h' television.monitors, two loud-speakers, a.micr0phone, and the usual classroom equipment. The micrOphone and loud-speakers provided two-way audio communication with the originating room. II. THE POPULATION AND THE SAMPLE 1129, pepulatéea All of the subjects in the present investigation were students enrolled in.Michigan State University during Fall Term 1957, and regis- tered in section h of Foundations of Education 200. The students were believed to be representative of all students enrolling in Foundations of Education 200. Typically the course is composed of male and female students from the Basic College and the College of Agriculture, Business 8h and Public Service, Communication Arts, Engineering, Home Economics, Science and Arts, and Education with the College of Education having the largest representation. Two-hundred and eighteen students originally registered for the course. As there was seating capacity for only ninety-six.students in each of the viewing rooms, twenty-six students were assigned to the originating room. These twenty-six students were not included in the experiment. Of the remaining 192 students, three dropped the course during the second week of the term. Twentyhfive additional students were excluded from the sample because they either failed to report dur- ing the first day of class when pre-tests were administered or because their college aptitude test scores were not available at the University Evaluation Services. The remaining 167 students were included in the sample for the present study, and the distribution of these students by college class, sex, and major curricula is shown in Table II on page 85 In the table it is shown that the largest number of students were SOpho- mores and were majoring in education and that females predominantly outnmmbered.males. III. THE RANDOMIZATION PROCEDURES On the first day of class the students completed a Personal Data Card (Appendix.A, p. 1h?) containing their name, age, class, major and other pertinent data. These cards were collected and before the next class meeting they were thoroughly shuffled and dealt into two piles 85 TABLE II NUMBER OF STUDENTS IN THE SAMPLE CLASSIFIED BY CLASS, SEX AND MAJOR CURRICULA —_ .Lfiajor Fresh- SOpho- Junion Senior Special Total Curricula man more M F M. F M F M F M F M F No Preference 1 2 3 9 1 O O O 1 O 6 ll“ 17 Agriculture 0 O h 1 O O O O O O h 1 5 Business 0 O O O O O 1 O O O 1 O 1 Business Education 0 O O 2 1 9 O l O O 1 12 13 Political Science 0 O O O 1 O O O O l 1 l 2 Social'Work O O O l O O O O O O O 1 1 Engineering 0 O O O O O l O O O 1 O l Home Economics 0 O O 7 O h 0 O O O O 11 11 Fine Arts 0 O 2 1 1 1 O 2 O O 3 h 7 Language & Literature 0 1 O 1 O h 0 O O O O 6 6 Biological Science 0 O l O 1 O 2 O O O h 0 h Physical Science 0 o 1 o 1 2 o o o o 2 2 h Social Science 0 h h 7 6 2 1 l O O 11 1h 25 PreéMedical o o o o 1 o o o o o 1 o 1 Speech 0 O O 1 l O O O O O 1 l 2 Education 1 3 1 uz 2 9 o 1 o o h 55 59 Physical Education 0 1 6 l O O 0 O O O 6 2 8 TOTAL 2 ll 22 73 16 31 5 5 l 1 h6 121 167 86 corresponding to the television and non-television groups. No bias was introduced by this method as each card had an equal chance of being placed in one of the two groups. The television and non-television groups were designated by the toss of a coin. In order not to exceed the seating capacity of each viewing room the last twenty-six students on the class list were assigned to the originating room and were in- cluded in the sample. Although this procedure would seemingly intro- duce error into the randomness of the sample, inasmuch as these twenty- sipreOple were from the latter part of the alphabet, an unpublished study by Krumboltz (29) based on a sample of 12,000 Air Force recruits indicates that no bias is introduced by alphabetical sampling. At the end of the second class meeting, during which pre-tests and an introductory lecture were given, the students were assigned to their respective groups and rooms and for the remainder of the class meetings reported to their assigned rooms. III. THE DISCUSSION TECHNIQUES The discussion method as used in the present study may be defined as student participation by the interchange of opinion, information, illustration, question and answer, and pro and con arguments for the purpose of understanding and clarification, with the instructor playing the role of group leader and moderator. The instructor usually defined the area for discussion, posed questions for consideration, supplied information and illustration where necessary or when asked to do so, clarified and related meanings where necessary, and helped maintain an 87 orderly sequence of discussion aimed around a central theme or problem. The two types of discussion used were six-by-six and panel dis- cussions. In the six-by-six discussion the class was organized into groups of not more than six students per group. The groups discussed the tapic, problem, or question under consideration for about six to ten.minutes after which the recorder from each group reported the views of the group on the t0pic, and these were further discussed by the class with the instructor acting as the moderator and group leader. During the sizebybsix discussion phase the instructor moved from group to group to help in clarifying problems and to act as a resource person. In the panel discussions the students were grouped into panels of not more than six persons. The panels were assigned tapics for pres- entation to the class, and gave a fifteen to twenty minute presentation. After the panel presentation the class further discussed the tepic or problem posed by the panel, or engaged in question and answer and dis- cussion with the panel. The instructor acted as a resource person for the panel and helped the panel to organize their tOpics. During the postepanel discussion the instructor allowed the panel members to con- duct the discussion except to help clarify and to maintain the discus- sion on the tOpic. In the use of each of these types of discussion the last few minutes of the class period were utilized by the instructor for a sum- mary of the day's discussion. The summary consisted of a brief state- ment of the tOpic, an enumeration of the main points discussed, a final clarification of difficult and misunderstood points, and an emphasizing 88 of important points, concepts or principles. Related problems were also mentioned. The use of these discussion methods will be discussed under the description of the television and non-television techniques. The class meeting was divided into two fifty-minute periods. The first fifty-minutes of the class consisted of a lecture over the closed-circuit television system to both the television and non-tele- vision groups. The lectures were supplemented by demonstrations, kine- scopes, movies, and guest speakers. A ten.minute intermission followed the first period. During the second fiftyeminute period the class participated in discussion. Television Discussion During the discussion period the instructor remained in the originating room while an assistant was present in the viewing room. The discussion was conducted between the classroom and the originating room with feedback provided by the two-way audio and one-way visual communication system. In the sixsby-six discussion at the completion of the six to ten minute group discussions the group recorders reported to the originating room where they presented reports to the class over television. The instructor remained in the originating room and acted as a moderator for the ensuing discussion, while the assistant in the classroom proctored the class. In the panel discussion the panels presented their discussion to the class over the television system after which the discussion was con- ducted between the class and the panel members over the television with the instructor again acting as moderator. 89 Non-television Discussion All discussions in the non-television group were conducted in the conventional face-to-face manner in the classroom with the instruc- tor in the classroom. The topics for six-by-six and panel discussion were coordinated with the television section so that for any day the two groups were discussing the same tepics. The primary difference between these two groups was in the feedback arrangement. In the non- television group feedback was obtained in the face-to-face situation while in the television group feedback was obtained by the television and loud-speaker arrangement. The first and second period arrangements are shown in the schematic diagram on page 90. IV. THE INSTRUCTORS The instructors in the course were two full-time Assistant Professors in the College of Education. The lecture presentations were shared equally by the two instructors. Two assistants aided in the administrative duties such as roll-taking, adjustment of television monitors, and proctoring in the two viewing classrooms. One was a part-time Instructor and one was a Graduate Assistant in the College of Education. The Instructors and Assistants were alternated between the two classrooms to minimize any differential bias due to instructor dif— ferences. V. SUMMARY The investigation was conducted in Foundations of Education 200, 90 First Period Lecture Cameras "'— "fi Instructor Television Television Monitors Monitors Coud-speaker ' oud- sp eaker 7/11: LA, -<-— Audio —9 f v H\ Loud-speakers AM a # Microphones 5 #- Second Period NTV Discussion Cameras Television Instructor Monitors o0 0 Panel members ’0 “or ,9, Sb Group reporters ‘v / L—I 1:: l—l \J / L—’ 12:: 13—1 \ L V «- : Loud-speaker :‘ M - Mi crophone a TV NTV Figure 2. Schematic diagram of the first and second period arrangements of the TV and NT! groups. 91 Section h, a course in beginning educational psychOIOgy, required of all students majoring in education. The physical facilities included a television originating room, a control room and two viewing classrooms equipped with the necessary cameras, monitors, lighting and technical equipment necessary for television production and viewing. The sample consisted of 167 students and was believed to be representative of the students enrolling in Foundations of Education 200. The students were randomly placed in the television (TV) or non- television (NTV) groups. The first fifty minutes of the class period consisted of a tele- vision lecture to both groups. The second fifty-minute period of each class consisted of six-by-six and panel discussions followed by general class discussion with the instructor acting as a moderator. In the television group the discussions were conducted over the television system between the class and the panel members or group reporters. In the non-television group all discussions were held in the classroom in a face-to-face situation. The instructors for the course were two full-time assistant professors aided by two assistants. All instructors were alternated between the two groups to minimize differences. CHAPTER V EVALUATION INSTRUMENTS The experimenter can frequently use existing instruments for the purposes of the investigation, but often it is necessary to construct or adapt available instruments appropriate to the objectives being sought in the study. In the present investigation some of the measuring in- struments were in existence and some of the instruments were constructed by the experimenter. The present chapter will contain a description of the evaluation instruments used in the study, a discussion of the admin- istration of the instruments and collection of the data, and an explana- tion of the procedures used in recording the data. I. MEASURES OF THE INDEPENDENT VARIABLES It will be recalled that the independent variables were (1) the discussion techniques, (2) the ability level of the students, (3) the students' preference for instruction, and (h) the students' sex. Since the discussion techniques were fully described in Chapter Four and the students' sex was obtained from the Personal Data Card only the measures of ability and preference for instruction will be discussed here. The American Council on Education Psychological Examination for College Freshmen was used as the measure of the students' ability and the Pre- ferred Instructor Characteristic Scale was used to measure the students' preference for instruction. A description of the above instruments follows. 93 American Council on Education Psychological Examination (ACE) The American Council on Education Psychological Examination for College Freshmen is designed for and standardized on entering college freshmen and is used by about 300 colleges and universities. As this test is so well known and is so widely used no description of its con- tents will be given. The test yields three scores: a quantitative (Q) score, a linguistic (L) score and a total (T) score. Only the T-score was used in the present study. A study of the reliability of the test by the authors, L. L. and T. G. Thurstone, as reported in Super (Sh), indicated an odd-even reliability of .95 for the total score. The ACE is administered to all entering students at Michigan State University by the University Office of Evaluation Services, and the scores for the students in the present study were obtained from the files of that Office. Preferred Instructor Characteristic Scale (PICS) The Preferred Instructor Characteristic Scale is a measure de- signed by Farquhar (20) and Krumboltz (30). The authors postulated a l'cognitive—affective" continuum of instructor characteristics. In the words of Krumboltz (30) "a cognitive instructor was defined as one con- cerned with the intellectual, abstract, subject-matter goals of teaching; the affective instructor was defined as being more concerned with emo- tional adjustments and student interaction in the classroom.' A number of statements believed characteristic of the two types of instructors were prOposed. After validation and refinement the final form.of the scale consisted of six cognitive and six affective statements each 9h paired with the other to make a total of 36 items in the scale. The sub- ject is required to choose one of each pair of statements in responding to the test items. The PICS is scored so that a high score indicates the student prefers a cognitive type of instruction and a low score in- dicates a preference for an affective type of instruction. The authors report a test-retest reliability coefficient of .88 and a reliability computed by Hoyt's Analysis of Variance technique of .90. The PICS was administered to the students in the present study on the first class meeting of the term. The scale and the scoring key appears in.Appendix A. II. MEASURES OF THE DEPENDENT VARIABLES The dependent variables in the present investigation were (1) stu- dent achievement of course content, (2) student attitude toward teaching, (3) student misconceptions about education, and (h) student Opinion about the course. In the sane order as listed above these variables were meas- ured by a Pre and Post Achievement Test (PAT), the Minnesota Teacher Attitude Inventory (MTAI), the Misconceptions about Education Scale (MES), and the Course Evaluation Scale (CES). The description of each of these instruments follows. Pre and Post Achievement Test (PAT) The Pre and Post Achievement Test was constructed from test items in the Instructor's Handbook accompanying the textbook for the course, "Educational Psychology" by Cronbach (1h). Fifty multiple—choice items were selected as representative of the content of the course. Cronbach 95 has ranked the items according to difficulty into three classifications and the items were selected so that approximately one-half were in the easiest and most difficult range and about one—half were of medium.dif- ficulty. The corrected split-half reliability coefficient based on the odd- even items was found to be .53. A second reliability measure computed by‘a Kuder—Richardson formula using the difficulty level for each item and the standard deviation of the test was .52, agreeing almost per- fectly'with the first value. Such a reliability is considerably lower than is considered useful for individual testing purposes and would reduce the standard error of measurement by only'approximately'70% of the value if the reliability were zero. However, for purposes of group measurement as were employed in the present experiment a reliability of about .50 is considered adequate. Kelley states, “If group measurement is all that is undertaken, scores which are individually reliable are not demanded, for the reliability of an average score is such greater than that of the single score....a test so unreliable that it will not be serviceable in making individual diagnosis may be very serviceable for group diagnosis.“ (28). Assuming that the errors in measurement are distributed in both directions from a true value, in a large number of cases they would tend to cancel each other, the group mean would not be adversely affected by'a low reliability, and the test would.measure adequately enough for the purposes of the experiment. The PreeAchievement Test was administered to the students during the first class-meeting of the term, and the Post-test during the final 96 examination period. The test and the scoring key will be found in Appendix A. Minnesota Teacher Attitude Inventogy'gggél) The Minnesota Teacher Attitude Inventory is published by the Psychological Corporation and was developed by Cook, leeds, and Callie. The Inventory is based on the assumption that the attitudes of the teacher are the key to the type of classroom he will be able to main- tain and that these attitudes are the result of the interaction of a multitude of factors including social intelligence, personality traits, values, academic intelligence, general knowledge, and teaching tech- niques. The Inventory is designed to measure those attitudes of a teacher which predict how well he will get along with pupils in inter- personal relationships and indirectly how well satisfied he will be with teaching as a vocation. The Inventory consists of 150 attitude statements about which the subject is asked to reapond on a five-choice scale of “strongly agree", "agree", “undecided or uncertain”, ”disagree", and 'strongly disagree". Scoring is on the basis of "right" minus "wrong" answers. The instrument was validated on teachers in the field and norms appear in the test booklet. The split-half reliability of the inventory in three separate studies was found to be .93, .88, and .93. The MTAI was administered to the students as a pre-test during the second class meeting and as a post-test during the final examination period of the tem. Misconceptions about Education Scale (MES) The Misconceptions about Education Scale was designed by the 97 1 in an attempt to measure an area that is not writer and a colleague ordinarily measured by the usual attitude or achievement test. Miscon- ceptions about Education refer to irrational beliefs or attitudes not based on factual knowledge that are believed to be held by the layman, beginning students, and neophyte teachers. Such beliefs or attitudes are akin to superstitious beliefs or "old wives' tales" but not super- stitious in the usual.meaning of the word. A number of statements were constructed by examining psychology and educational psychology textbooks for discussions of such fallacies. The Wickman study (58) was also a source for the construction of many of the statements. The statements were submitted to seven full-time faculty members of the College of Education and all items about which there was disagreement were dis- carded. .Ambiguous statements were reworded. The final scale consisted of 61 statements such as "A high forehead usually indicates intellectual superiorityz' The subjects were required to answer each statement as a true-false item. A corrected split-half reliability was computed and found to be .63. The scale was administered as a pre-test to the students during the second class—meeting and as a post-test during the final examination period. A copy of the scale and the scoring key will be found in Appendix A. 1Mr. Clarence Williams, formerly Part-time Instructor in the Department of Foundations of Education, Michigan State University; 98 Course Evaluation Scale (CES) The Course Evaluation Scale was an adaptation of a similar in- strument used in instructional television research at Pennsylvania State University. The scale consisted of 22 multiple-choice items about the course, the instruction, the evaluation procedures, about the ways in which television was used, and about the mechanics of the course. The multiple-choice items were designed after the Iikert technique (36). Each response was given a weight on a five-point scale so that a five indicated the most favorable response, a three was a neutral response, and a one indicated a least favorable response. In addition to the 22 items, space was provided for comments, criticisms, and suggestions. The scale was so scored that a high score indicated a relatively favorable opinion of the course and a low score indicated a relatively unfavorable opinion of the course. No reliability was computed for this instrument. The scale and the scoring key appear in.Appendian. The scale was administered to the students on the last class-meeting of the course. III. THE COLLECTION.AND RECORDING OF THE DATA The pre-tests were administered during the first two class periods with all subjects present in one room. No time limit was set for any of the instruments and all students completed the tests before the end of the two-hour period. Instructions were given by one of the instructors and the assistants aided in distributing4and collecting 99 materials and in proctoring the tests. The post-tests were administered in the respective rooms of the TV and NTV groups during the two final examination periods of the course. The procedures for administering the tests were established beforehand and both groups began the tests at the same time. Again no time limit was set for completion of the tests and all subjects completed the tests before the end of the two-hour period. All answer sheets were machine-scored by the University Office of Evaluation Services. The scores were recorded on the Personal Data Card (see Appendix A) and were later keyepunched into IBM cards for machine computati on. IV. INTERCORRELATION OF INSTRUMENTS The intercorrelations of the measuring instruments for the TV and NTV groups are presented in Table III on page 101 The correlations are based on the scores of 86 students in the TV group and 81 students in the NTV group, the 167 students for whom a complete set of scores were obtainable. In comparing the correlation between pre and post scores on the MTAI, the MES and the PAT it will be noted that the MTAI is the highest, the MES is next and the PAT has the lowest correlation of the three. The reliability coefficients for these three instruments (reported in the previous section) ranked in the same order indicating a consistency in this respect. The significant correlations on the post-test measures of these three instruments seems to indicate that a common factor is 100 madam moepmodpm muons maofipmmosoomflzvpmom I mm: Hmom madam Goapmdfim>m mmhdoo I mmo kuopcman mpepfippd hmsomme meomchaz pmom I H492 emom mawom moapmosem peeps meoapmmonoeofi: mam I mmz_mmm mmm. I RH Hwa. 3 mm mm I z I >82 auoeeo>eH mespappa genomes spammeeaz mam I Hag: mam pews pamem>mfino4 pmom I Bdm emom emoe pnmsmeownodronm I 94m mmm .oamom moflpmfinovomnmmo noposnpmaH vonnmmonm I mon moanmonm omoaaoo pom soapmdflemxm Hmofimoaonohwm compmodpm so afiondoo mmoonE< I mo< .Hopoa RH one pm open seam econommwe hapcmoHMHmmHm *# mew. n ma pea. . am em . z I >9 .am>ma mm map as one“ zone nemumemae sapeaoamaewam * no. *ma. oa.- mo. eo. ea. mo. mo. mo.a ea. Ho. NH.I *mm. No. eo.- HH.I wee mo. *ma. *eom.**am. aaee.**am. *aee.**em. *amm.**mm. mo.**mm. team. 50. OH. ANN. was emom oa.- no. **om.**em. wade. Adm. arme.*aem. *ma.**eN. «*Nm.**mm. Ho. Ha. *aH. amm. mm: mam eo. ea. aaes.#*am. same. *HN. game.aaee. *Imm. ma. ma. *NN. *NN. mo.- ma. ma. Haas emom no. mo. Iaes.**mm. **m3.**em. *aae.**ee. *Iam. Ho. Ho. NH. do. 00.- OH. so. Haas mam mo.- ea. AIQm.**mm. *aH.**-. *Imm. ma. *IHm. Ho. aawm.**Nm. No. ma. ma. ma. Eam emom Ho. Na.- mo.**Nm. *amm. *mN. ma. *NN. Ho. NH. ammm.**mm. weem.- NH.I**Em. ea. Ham mam *mN. No. *Aem. Eo. Ho. Ha. INN. mo.- Ho. so. «0. ma. aaom.- NH.- Ho. *ma. muHm 40.- HH.- 0H. *Nm. *aa. amm. ma. ma. 0H. so. ma. ma. seem. an. ac. *ma. woe HE E E2 E Ez E E2 E Ez E E2 E E2 E Ez E Ez E was mm: Ha: Ha: Sm Em ---I mmo emom mam emom mam emom mam mOHa mos mezmzbmemzH onH¢DA<>m mme mo mZOHedqmmmOQMHBZH HHH mammy 101 being measured by these instruments. It is interesting to note that the correlation between the PICS and the CES was significant in the NTV group but not in the TV group, indicating that the students with a preference for a cognitive type of instruction held a more favorable Opinion toward the course in the NTV group than in the TV group. The ACE, the PICS and.the CES, except in a few instances, were not significantly correlated with the other measures nor with each other, and thus were measuring independently of each other. V. SUMMARY The T-score on the ACE was used as a measure of the ability of the students in the present investigation. It is a widely used test and has a reliability that is very high. The PICS was used as a measure of the students' preference for a cognitive or an affective type of instruction. The test-retest relia- bility of this instrument is high. The PAT was designed as a measure of the achievement of the course content. It is a 50 item multiple-choice type examination selected from the Instructor's Handbook accompanying the course text- book. The split-half reliability of this instrument was low, but was considered adequate for the measurement of group means used in the in- vestigation. The Minnesota Teacher Attitude Inventory, a commercially pub- lished instrument, was employed to measure the students' attitudes to- ward teaching. The inventory contains 150 attitude statements about 102 which the subject responds on a five-point scale of agreement or dis- agreement. The reliability of this instrument has been found to be con- sistently high. The MES was designed to measure the students' misconceptions about education. Misconceptions in the context used here relates to irrational beliefs or attitudes believed to be held by the laymen, beginning students and the neophyte teacher. The scale consisted of 61 true-false items and it was found to be of medium reliability. The CES was designed to measure the students' Opinions about the course. It consisted of 22 Likert-type items covering the instruction, evaluation, use of television, and the mechanics of the course. No re- liability was computed for this instrument. The pre-tests were administered during the first two class meet- ings of the course, and the post-tests were administered during the final examination periods. ACE scores were obtained.from the University Office of Evaluation Services. The scores for all measures were machine- scored and entered on the Personal Data Card. The intercorrelations of the measuring instruments were computed and the results presented in tabular form. The correlations between the post-test scores for the MTAI, the MES and the PAT were found to be consistent with the reliabilities reported for these instruments. The ACE, the PICS and the CES were found to be measuring independently of the other instruments, while the MTAI, the MES and the PAT were found to have low but significant relationships. CHAPTER VI ANALYSIS OF THE EXPERIMENTAL RESULTS A description of the computational procedures, the results of the tests of assumptions underlying the statistical techniques, and the results of the statistical tests for each of the dependent vari- ables will be discussed in the present Chapter. I. THE COMPUTATIONAL PROCEDURES The basic data for each individual including his name, student number, sex, curriculum, class, the experimental group (TV or NTV), and the scores on all measuring instruments was key-punched into an IBM card. The cards for each experimental group were then sorted on ACE scores and the cutting-score for high and low groups determined. The cards were next combined and the cutting-scores for the high and low PICS were obtained in the same manner. The cards for each experi- mental group were then sorted into the sixteen subgroups corresponding to the sixteen cells in the experimental design (See Chapter III, Figure 1). In order to maintain prOportionality between subgroups for the variance and covariance analysis, using a table of random numbers, cards were randomly selected from the subgroups so that each male sub- group contained 3 subjects and each female subgroup contained 11 sub- jects for a total of 56 subjects in each of the TV and NTV groups. This reduced the TV group by 30 students and the NTV group by 25 stu- 10h dents. The subgroup numbers were then punched in the cards for identi- fication. The basic data with the exception of the name and student number is tabulated in Appendix B, “Original Data.” The calculation and summation of scores, squares, and cross- products was accomplished by passing the cards through a 60h IBM Elec- tronic Calculating Punch. As a check on the programming of the machine the scores for one subgroup were calculated by hand on a mechanical computer, and as a further check all cards were fed through the machine for two passes. The data from the above calculations were automati- cally punched on a summary card for each subgroup. Appendix C, "Explanation of Computational Procedures,” contains a series of tables showing the manner in which the data for the six- teen subgroups were combined to test each of the null hypotheses. A separate table is shown for each of the main and interaction effects. Each table shows the identification number of the subgroup and the number of subjects in each subgroup and the total number of subjects. By referring to these tables in Appendix C and the original data in Appendix B, the reader can compute the sums of squares for the pre and post measures, the cross-products, the adjusted variances, and the F ratios. For the test of the null hypotheses the sums of scores, sums of squares, and sums of cross-products for each subgroup were combined and entered in its proper place in the table corresponding to the null hypothesis under test. The adjusted scores, variances and F ratios were then computed and entered in the analysis of covariance tables for each of the criterion tests. All of the above calculations were 105 accomplished on a mechanical computer. The above procedures apply to the covariance analysis of the measures for which pre and post tests were obtained. Essentially the same procedures were applied to the variance analysis of the CES meas- ure except, of course, the computation of cross-products and adjusted scores was eliminated as this was a post-test measure only. For Bartlett's Test of Homogeneity of Variance it was necessary to compute an adjusted post-test score for each of the sixteen sub- groups for the three criterion measures involving this assumption. The adjusted scores were computed on a hand calculator and are tabulated in Appendix D, "Tests of Assumptions of Variance and Covariance.” The data for all other tests of assumptions also will be found in Appendix D. The assumptions underlying the analysis of variance and covariance were discussed in Chapter III, Section 3. II. ANALYSIS or THE ACHIEVE-HIT TEST (PAT) The results of the tests of assumption of covariance for the Achievement Test (PAT) are summarized below in Table IV. TABLE IV RESULTS OF TESTS OF ASSUMPTIONS OF COVARIANCE FOR mm ACHIEVIT'ZINT TEST (PAT) Assumption Obtained Value at Null Value 1% level Hypothesis Homogeneity of Regression .56 2.28 Accept Linearity of Regression 1.03 2.32 Accept Homogeneity of Variance 35.66 30.58 Reject 106 The assumption that the variance from subgroup to subgroup is the same, except for chance, must be rejected. Lindquist (38, pp. 96- 7) has shown that heterogeneity of variance will result in a consider- ably larger portion of high F value for tests of significance than the proportions given in the table for F. In effect this would result in the lowering of the level of confidence. Quoting from Lindquist, In general, when the heterogeneity in form or variance is ‘marked' but not 'extreme', allowance may be made for this fact by setting a higher 'apparent' level of significance for the tests of treatment effects than would otherwise be employed. In cases of very marked heterogeneity, for example, if one wishes the risk of a Type I error not to exceed 5%, he might require the effect to be 'significant' at the 2.5% level, or if he wants the risk of a Type I error not to exceed 1%, he might set the 'apparent' level of significance of the test at 0.1%. (38, p. 86) Unfortunately it is impossible to state what the limits are for a "marked“ or "very marked“ heterogeneity, or to what level one must change the level of significance to compensate for the heterogeneity. In the analysis of the Achievement Test that follows it is impossible to ascertain the size of the F ratio required for rejection of the null hypotheses at the 5% level, and therefore none of the null hypotheses can be rejected. However, if the covariance analysis reveals no signif- icant F ratios, the null hypotheses of no differences can be accepted. The analysis of covariance of the PAT will be found in Table V on pageIKW. Inspection of Table V reveals that there were no signifi- cant differences on the Achievement Test due to the ability level of the students, the students' preference for instruction, the students' sex, or the two discussion techniques, nor were there significant dif- ferences due to the interaction effects of these variables. 107 TABLE V ANALYSIS OF COVARIANCE PAT POST-TEST SCORES ADJUSTED FOR PRE-TEST SCOPF Nature of Sum of Null Variation df Squares Variance F* Hypothesis ACE (A) Ability l 3.59 3.59 0.19 Accept PICS (P) Preference for Instruction 1 61.32 61.32 3.29 Accept SEX (S) l 66.83 66.83 3.58 Accept TECH (T) Discussion Techniques 1 0.83 0.83 0.0h Accept A x P 1 9.91 9.91 0.53 Accept A x S l 0.00 0.00 0.00 Accept A x T 1 11.59 11.59 0.62 Accept P x s 1 10.08 10.08 0.58 Accept P x T l 11.25 11.25 0.60 Accept S x T 1 20.73 20.73 1.11 Accept A x P x S l hh.hh hh.hh 2.38 Accept A x P x T 1 18.57 18.57 1.00 Accept A x s x T l 8.02 8.02 0.22 Accept P x S x T l 18.39 18.39 0.99 Accept A x.P x s x T 1 1h.21 lh.21 0.76 Accept Between Cells (15) (306.51) (20.83) ‘Within Cells 95 1772.A6 18.66 Total 110 2078.97 18.90 *F55 (1,95) 3-9h 108 III. ANALYSIS OF THE MINNESOTA TEACHER ATTITUDE INVENTORY (MTAI) The results of the tests of assumptions of covariance for the MTAI are tabulated in Appendix.D and are summarized in Table VI below. TABLE VI RESULTS OF TESTS OF ASSUMPTIONS OF COVARIANCE FOR THE MINNESOTA TEACHER ATTITUDE INVENTORY (KTAI) Obtained Value at Null Assumption Value 1% level Hypothesis Homogeneity of Regression 0.81 2.28 Accept Linearity of Regression 1.57 2.32 Accept Homogeneity of Variance 10.57 30.58 Accept As indicated in Table VI the assumptions underlying the use of the covariance analysis were met for the data obtained on the MTAI, and the data may be validly tested by the covariance technique. The results of the analysis of covariance of the MTAI data are presented in Table VII, page 109. Inspection of the table shows that a significant interaction effect exists between the Preference for Instruc- tion (PICS) and the Discussion Techniques (TECH) variables. For all other effects the null hypotheses of no differences were accepted. The pre-test mean scores, the post-test means scores and the ad- justed post-test mean scores for the PICS x TECH interaction are pre- sented in Table VIII, page 110. Examination of the adjusted post-test scores shows that the "cognitive”-NTV group Obtained the highest scores, TABLE VII ANALYSIS OF COVARIAHCE MTAI POST-TEST SCORES ADJUSTED FOR PhE-TEST SCORES 109 Nature of Sum of M Null Variation df Squares Variance F" Hypothesis ACE (A) Ability l 6h8.53 6h8.53 1.72 Accept PICS (P) Preference for Instruction 1 1298.38 1298.38 3.hh Accept SEX (s) l 1301.0h 1301.0h 3.hu Accept TECH (T) Discussion Techniques 1 118.62 118.62 0.31 Accept A x P 1 0.88 0.88 0.00 Accept A x S 1 23.35 23.35 0.06 Accept -A x T l 15.73 15.73 0.08 Accept P x S 1 26h.8h 26h.8h 0.70 Accept P x T 1 2281.0u 2281.0b 6.0u Reject S x-T 1 30.81 30.81 0.08 Accept A x.P x S 1 3h5.l9 3h5.19 0.91 Accept A x P x T 1 7.25 7.25 0.02 Accept A x S x T l 91.9h 91.9h 0.2b Accept P x S x T 1 767.12 767.12 2.03 Accept A x P x S x T l 1020.8h 1020.8h 2.70 Accept Between Cells (15) (33h7.16 (552-57) Within Cells 95 35898.05 377.87 Total 110 88285.21 *Fsz (1,95) 3-9h 110 the I'affective"--TV group scored next highest, and the "cognitive"-TV and 'affective'-NTV groups scored the lowest. Inspection of the dif- ferences in adjusted mean scores reveals that the “cognitive“ students obtained much higher scores in the NTV group than the "cognitive“ stu- dents in the TV group, whereas the differences for the “affective" stu- dents in the TV and NTV groups is much smaller. It will also be noted that the differences between the "affective" and "cognitive" students in the NTV group is large, while the differences between the "affective“ and ”cognitive“ students in the TV groups is considerably smaller. TABLE VIII MEAN SCORES AND ADJUSTED MEAN SCORES ON THE MTAI FOR THE PICS x TECH INTERACTION High PICS Low'PICS nCOgnitiven "Affective" Adjusted Adjusted Pre Post Post Pre Post Post Technique TV 27.1 50.3 u5.8 ll.h A3.8 h8.3 Technique NTV 22.0 58.7 58.0 2h.5 hh.l h3.h Apparently under the conditions of the present experiment the combination of "cognitive" type students and face-to-face discussion promoted the best scores in attitudes toward teaching as measured by the MTAI. The television discussion technique was not nearly as effec- tive for either the "cognitive” or the ”affective“ students although it favored the I'affective" students slightly. The interpretation of these findings is somewhat hazardous. The 111 rejection of a null hypothesis of no differences indicates only that it is unlikely the observed differences are due to chance effects and does not indicate the cause of the differences. The differences may be due to uncontrolled variables in the design of the experiment itself. In interpreting the above findings it is not possible to say with certainty that the differences were due to the discussion techniques, because the instructor variable could have biased the results. Although the in- structors were rotated between sections to equalize any differences due to instructors, it is possible that all instructors favored the "cogni- tive" type of instruction rather than the "affective" type, and if this were the case, it would favor the students who designated a preference for a "cognitive" type of instruction. Even this bias would not be serious if it were constant for both groups, as both “affective“ and "cognitive" students would be affected equally. It is possible, how- ever, that the instructors in the present investigation displayed dif- ferent degrees of "00gnitiveneSS" or "affectiveness" in the face-to- face and television situations which would introduce a third uncon- trolled interacting factor. Therefore, in interpreting the above findings, it can only be said that under the conditions of the present experiment, the evidence indicates that the students' attitudes toward teaching as measured by the Minnesota Teacher Attitude Inventory are most affected by the com- bination of a preference for a "cognitive" type of instruction and the face-to-face discussion technique. 112 IV. ANALYSIS OF THE MISCONCEPTIONS ABOUT EDUCATION SCALE (MES) The data for the MES was tested for the assumptions and the re- sults are shown in Appendix D and are summarized in Table IX below. As indicated in Table IX all tests of assumptions were satisfied for this data, and the analysis of covariance is a valid test for the data. TABLE IX RESULTS OF TESTS OF ASSUMPTIONS OF COVARIANCE FOR THE MISCONCEPTIONS ABOUT EDUCATION SCALE (MES) Obtained Value at Null Assumption Value 1% level Hypothesis Homogeneity of Regression 0.85 2.28 Accept Linearity of Regression 0.80 2.32 Accept Homogeneity of Variance 17.hl 30.58 Accept The analysis of covariance for the MES will be found in Table X on page 113. The analysis shows that all null hypotheses can be accepted and that there are no differences, other than Chance differences, on the Misconceptions about Education Scale due to the main or interacting effects of the independent variables. v. ANALYSIS OF THE COURSE EVALUATION SCALE (CES) Inasmuch as the CES was employed as a post-test only, the analy- sis of variance was used to test the data and the only assumption tested 113 TABLE X ANALYSIS OF COVARIANCE MES POST-TEST SCORES ADJUSTED FOR FEE-TEST SCORES Nature of Sum of Null Variation df Squares Variance F* Hypothesis ACE (A) Ability l 2.77 2.77 0.18 Accept PICS (P) Preference for Instruction 1 15.5b 15.5h 0.98 Accept SEX (s) 1 0.35 0.35 0.02 Accept TECH (T) Discussion Techniques 1 1.76 1.76 0.11 Accept A x P 1 h.76 u.76 0.30 Accept A X S 1 15.39 15.39 0.97 Accept A x T 1 3.50 3.50 0.22 Accept P x S l 39.03 39.03 2.87 Accept P x T 1 27.h3 27.h3 1.73 Accept S x T 1 0.58 0.58 0.08 Accept A x.P x S 1 12.51 12.51 0.79 Accept A x P x T l 37.36 37.36 2.36 Accept A x s x T 1 26.69 26.69' 1.69 Accept P x S x T 1 0.80 0.80 0.05 Accept A x P x S x T l 1.6h 1.68 0.10 Accept Between Cells (15) (187.52) (12.50) ‘Within Cells 95 1501.79 15.81 Total 110 1689.31 *Fsz (1,95) 3-9h 11b was that of homogeneity of variance. The results are tabulated in Appendix D and summarized in Table XI below. The assumption of homogeneity of variance was met and the analy- sis of variance is a valid test for this data. TABLE XI RESULT OF THE TEST OF HOEOCIIIEITY 0F VARIANCE FOR THE COURSE EVALUATION SCALE (0E5) Assumption Obtained Value at Null Value 1% level Hypothesis Homogeneity of Variance 20.07 30.58 Accept The results of the analysis of variance for the CES are presented in Table XII, page 115. For this analysis the results on items 7, 10, 11, 12, 15, and 16 were omitted in computing the total score as these items refer to particular course procedures and cannot be evaluated in terms of the opinion for cross-comparison purposes. It will be noted that the F value for the effect of the discussion techniques was suf- ficently large for rejection of the null hypothesis at the 5% level of confidence, and that there is a significant difference between the TV and NTV groups on CES scores. Since the effect involves only the dif- ference between two means, it was tested by the "t" test for verifica- tion and the results are shown in Table XIII which reveals that the dif- ference between the obtained means for the two discussion groups is significant at the 5% level. 115 TABLE XII ANALYSIS OF VARIANCE CES SCORES Nature of‘.- SMm.of Null Variation df Squares Variance F* Hypothesis ACE (A) Ability 1 h6.29 86.29 0.2h Accept PICS (P) Preference for Instruction 1 6.08 6.0h 0.03 Accept SEX (s) l 37.85 37.85 0.19 Accept TECH (T) Discussion Techniques 1 880.32 880.32 8.50 Reject A x P 1 66.08 66.08 0.38 Accept A x S 1 lb.31 1h.3l 0.07 Accept A X T l 33.39 38.39 0.20 Accept P x S 1 0.02 0.02 0.00 Accept P x T 1 5.18 5.18 0.03 Accept S x T 1 5.00 5.00 0.03 Accept A x P x S l 6.91 6.91 0.0h Accept A x P x T 1 137.28 137.28 0.70 Accept A x s x T 1 151.88 151.88 0.26 Accept P x S x T 1 132.83 132.83 0.68 Accept A x P x S x T 1 18.71 18.71 0.10 Accept Between Cells (15) (lhb7.07 (96.hh) Within Cells 95 18788.36 195.71 Total 110 20235.83 *F5z (1,95) 3'98 116 TABLE XIII RESULTS OF THE "t" TEST OF THE DIFFEHENCE BETWEEN MEAN SCORES FOR THE TV AND NTV GROUPS ON THE CES Group N Mean SD SEM Diff. t t5% Television 56 5h.27 9.71 1.03 5.71 2.30 2.00 Non-Television 56 59.98 15.h3 2.06 The non-television (face-to-face) discussion students had a more favorable opinion of the course than the students in the television group. This being the case it would be of further interest to learn on what particular items of the Course Evaluation Scale these differences were significant, and therefore each of the items on the CES was tested for differences between the TV and NTV groups by the "t" test. For the computations in these tests the total sample of 167 stu- dents was used. It will be recalled that in order to maintain prOpor- tionality between cells in the analysis of variance and covariance tests the TV group was reduced by 30 students and the NTV group by 25 students. Since the "t" test places no restrictions on the proportionality of sub- jects these 55 students were included. The two groups were first tested for the difference between total means scores and the result is shown in Table XIV on page 117. The result of this test agrees with the findings in Table XII, and further substantiate the findings that the NTV group displayed a more favorable opinion toward the course than the TV group. The "t" test of the individual items on the CES will be found in 117 TABLE XIV RESULT OF THE "t" TEST OF THE DIFIERIJCC BLFIEJI I‘TI‘ZAN SCORES FOR THE TV AND NTV GROUPS ON THE CE" FOR THE TOTAL SAMPLE Group N Mean SD 5551 Diff t 1:5,?» Television 86 53.3h 8.71 0.9b 6.83 3.3h 1.96 Non-Television 81 60.17 15.95 1.77 Table XV on page 118. The items are so scored that a mean value of 3.00 would indicate a completely neutral response, a score of above 3.00 a favorable response, and a score below 3.00 an unfavorable response. The Course Evaluation Scale will be found in Appendix A. For ten of the twenty-two items on the CES a significant differ- ence was found in favor of the NTV group, and no significant differences were obtained favoring the TV group. In examining the results for each item it will be noted that for Item 1 the NTV group considered the in- formation in the course more useful than did the TV group and they also felt it to be more interesting, as revealed by the responses to Item 3. Inspection of Item N indicates that the NTV group were happier about having taken the course than the TV group, and according to Item 5 were more certain about taking an advanced course in Educational Psychology. The TV group felt that they had learned somewhat less than they would have learned in a similar course taught without television, while the NTV group felt they had learned slightly more as indicated by responses to Item 8. Analysis of Item 17 reveals that the NTV group felt that television instruction was somewhat less disturbing than ordinary in- 118 .08 on mcwpmonopaans hum» .o .03 op wawpmohopafiqd panamEOo .v .02 op moHpmohochoo no wqumonopoa nonpfloa .o .08 on moHpmouopoH ganzmsoo .9 .ms Op wnapmouopnw knob .d HHaz pomnmm om.H wH.m HH.o mm.o mm.g mo.: .mH omnaoo man :H aoHpaanooaH one .m .0» oo pom haobfipwmom HHHS H .o .om oo no: hanwnoum HHHS H .U .oo HHHS H van: roux 9.:oo H .o .8 on .5303 HHHs H .n .ow oo hHo>H9Hmom Hsz H .o «omnsoo uHsp :H Hopmetpomhnnm on» pdono whoa ado oaaw op 3.84 34 mm; 2.0 35 ohm 2A 32.38%“. a. 3 82$ 358:: mépfi m5. 5 .N .mnwnowop manpsm he GH mm: 0: Ho on HHH3 pH Mcwnp H .o .mcfinooop whopsm as :H mm: 33.2 .8 3 HHHn pH in» H .c .moflnowop ohspdw he :H Hammad on HHH3 pH go: no ponpmnz was» ac: Ea H .o .mqanomoa enough as 5. Ham? ”.3323 B HHH: pH €20. H .n .mcfisommp mndpsm as 5 Ham»: in.» 3 a": pH xfifi H .a woozpoonoo , poonom om.H mm.m HH.o H©.o mw.a :m.: mH amazon man aw oOHposuomoH map no new «4 .H 383805- mm... p cam .33 fimz 5% .Bfi $3. >92 >9 [1" \ll mmo mme mo mXEBH AanH>HQZH mme zo mmbomu >92 924 >9 mus mo zomHmdmzoo >N_uqmHpmepoHaH mmmH nose mp .o .:0HmmSomHn map swap opHpmspoon mmoH pmnszom mH .o .opHmeHOH:H MHquvo mam QOHmmsOmHo pom .o .GOHmmsome map swap opppwsnomow whoa pmnszom mp .p .cOHmmoomHo map swap m>preuomow opos_no:e mp .m .QHSpomH map pmnp HmoH H «mandoo mflnp Ho mGOprom 04.: :OHmmoomHo pom mpdpomH onp wappdeoo :H .HmoHpomndeH hpm> .o .Hwowpomnmea pmnsmsom .o .HmoHpompmfiH you Hmoppowhm hospHo: .o .HmOHpompm posBmEOm .n .Hmoppomhm huo> .o wH.: .mH ompsoo mHSp CH pmppmelpoonnzm one .om op poo thpruoo HHHS H .m .om oo poo thmQOHQ HHHz H .v .oo HHH3 H pm:3 30:3 p.qoo H .o .0» 0e aHnmpouo HHH: H .p .om 0o thHmphoo HHHk H .w ammo mfinp op anHEHm mmhsoo mm.m omocwpnm cm mxmp op Hpficdppomgo mnp ope: H HH .ompsoo map xoop H hypom huo> so H .o .omhsoo mnp xoop H pwnp Hypom em H .o .mmnsoo map xoop H appom you omHm hmnppoc so H .o .ompuoo opp xoop H owHw ea H .9 .omhsoo map xoop H ome hpo> Em H .m ammMSoo ma.m mHQp :H oocmHuomHo 080m no: o>m£ H pmnp Roz .N .m .: mHmmzporwmm HHdz am» .HHHQ cam: >92 coo: aopH pmme >9 A.e.psoov >H mamae 120 .mmoH none .0 .mme oprHH a .o .poomopm po soap whoa on .o .6905 oprHH d .n .opoa node .a .mpamHOHfiam 8. 8x: 8:3 382 H pmmood om.H mo.o HH.o Ho.o wH.: ~H.4 umchooaoo up coHpMQHopphmm mmmHo no new ad .OH .oHnmSHdb mmoH pop: .0 .oHnddHo> mmoH pdnzmsom .o .oewm map pnond .o .mHszHm> whoa pmn3msom .n .oprst> mnoE_sofiz .w «coma coon poo can nowmwbmHmp HH coon upon deoz pp soap mHnade> pmoood om.H Hm.H 0H.o mm.o mm.m oo.m mmoH no whoa mp omnsoo man xufinp 90% on .m .mmoH nos: .0 .mmoH oprHH 4 .o .oswm opp pnond .o .onos opr Had .n .93.: £052 ..a «nonH>oHop Mo on: onp pooan3_pnmsup pan «ounzoo uoafiewm a up uonhdoH upon oHdos.soh no nopmw>oHop hobo 33$ 93 cog 3.0 2.0 8A HH.N nose 3 3933 32 5% as: 3» so» on .m 33:33 amp ... new. £an 502 5% 89H puma H92 Ez E ”I L TFESV EH mama. 121 .QHon 0: Ho mm: pH anzp H .o .mHma mepHH me was pH xcan H .u .HSHmHon mm: pH HH whom p0: so H .o .HnmmHms passagem mH pH xcan H .n .HngHms Hum» mH pH step H .a «ooqnooooo mH thomz oooo mmHothm Hdnpdon pamooa om.H mm.o mH.o No.0 Hm.m mm.m Ho maHpuoamu and waHemmu we» we emu m4 .4H .mmmHmms hhm> who: .m .mmmHomo pmnzoEom who: .o .mmmHom: Ho: Homom: pmanmo who: .0 .Hsmoms pmnonom mum: .Q .Hommms zuo> onoz .m mmeoE map pmnp Hoop H «SOHmH>oHop no>o moHpos wchcomth ow om>Ho> pdoood pm.H mw.H NH.o mm.o Hm.4 M4.4 IcH moHpHsOHHHHo HwOHozomp map scam oon< .MH .conmdomHo :H oEHp mmoH nods .m .oonmdomHo 2H oEHp mmmH oprHH o .o .GOHmmoomHo oH osHp mmmH on no: once 0: .o .oonmsomHo :H mSHp once oprHH o .p .:0HmmsomHo co mEHp mace moss .m poogm m>mn op omxHH m>mn oHdoz H .mthOO map paooo< om.H om.o 0H.o no.0 mm.m H4.m QH QOHmmSQmHU op oopo>oo oEHp map mo yam w< .NH .mmdopm QOHmmdomHo Hocmm map soap HammHmn mmmH soda mymz .0 .mgsopw :OHmmdomHo Hocmm mop swap HdHQHmn mmoH prSmEom one: .6 .mmoopm :OHmmsomHo Hmcmg esp swap HSHQHon mmoH hos once noanm: who: .0 .mQSOHM cOHmmSomHo Hocwm opp supp HSHQHoz whoa puzzosom who: .9 .mdooum :OmesomHo Hound map comp HommHms whoa nose one? .o madopm :onmSomHo o x 0 map pmzp HomH H «mmdogm COHmmSQwHo o N o pmooo< om.H No.0 4H.o mo.o HH.4 no.4 and masonw :OHmmSomHo Honda map mchomsoo oH .HH mHmmnpomam amp 9 gem .HHHQ camz new: smpH pmmH HHdz >92 >9 A.p.p:00v >N mqmhquoho CH comp honoom pansmeom .o .mmmmmHo >pmcHopo :H soap mmmH no: whoa on .o .mommmHo hhonono :H :mnp noppmp panamEOm .p .mmmmmHo hymnfloho QH comp hmppon 2058 .m mpmnpomocoo coo H pmzp xCan H 44.0 04.m om.m «omnpmoooo mH GOHposppmcH GOHmeoHop mm umH m4 .mmmhsoo humapro omnp mcthdpmHo whoa nods .o .mommeo hpmcHopo comp mcthdpmHo whoe.pm£3mEom .o .momusoo >umnHoho swap mapphspmHo once on .o .mmmHSoo >hmoHoho swap manndpme mmoH pwzsosom .p .mmmpsoo hhonoyo swap mopnudpmpo mmoH nose .m mH pH xcan H .emcumo 4m.o 4m.m o~.m Isoo mH QOHpUShpmGH come>mHop ozp mm 9mm m4 .mxHHup nose cop who; monp .m .oxHHdd pm£3m80m who: >05p .o .panh pdonm who: monp .o .oxHHm pmnsmeom ohms heap .n .mxHHo nose cop who: >on .m «Umapmocoo was HmHumpms xoonpxmp op comemQEoo :H mH.o mm.m 0H.m endpomH onp Ho pmppmsnpoonpnm map mm you m4 .conmSUmHo Hmcwm :H mmoH nods omeHOHpnmm op .o .cOHmmSomHo Hocmd :H mmoH pmnSmEom omeHoHpan op .o .mQOHmmSQmHo Honmm :H mmmH Ho: ohos pmanmc mpmaHoprmm 0p .0 .chHmmSomHo Hoqmm :H whoa pmnaoeom mpmmHoHpumm op .p .w20Hmm30mHo Honda :H whoa nods omeHowpuwQ op .w omxHH o>mn oases H «noonmoooo was mHmamm GH QOHpmmHoHpumm mnp mo mom m4 Nm.o mm.m oo.m owH .HH 00H oMH mHmonpomxm HHsz 9mm coma Baez >92 >9 .HHHQ Epr pmma A.p.pcooV >N mqmdy 123 .pmmnoom mnp wCoEw .m .pmoE Camp hmpoom .0 .mmmom>m p5090 .0 .pmoE Cmnp hoppmn .p .pme esp mCoEw .0 one >me xCHCp H «omChooCoo pomhmm mm.H m4.m mH.O H4.o mm.m mm.m mum omHCoo mHCp CH mpopospmeH map on n09 m< .Nm .me >Hgoom >Ao> .m .pme hHhoom .o .me >HHnopomHmemm .o .pmE HHm3..p .me HHos >Hm> .0 Comp pamooa 00.H 0m.H mH.0 0H.0 H9.m mm.m 0>m4 amusoo map Ho mo>Hpomhno map pmep HmmH H .Hm .Cpmmm COHpomm COHmH>onp mmhwa w omoono poC UHCos >Hm>HpHmom .o .Cmem CoHpoom COHmH>mHmp omCmH m mwoono poC UHCos >Hnmnopm .0 .moCmpmHohm oC m>mn .o .CHmmm COHpomm COHmH>mHmp mmpmH m mmooCo thmnonm .n .CHmmm COHpoom COHmH>onp mmpwa m omooCo >Hm>HpHmoQ .m . oHCos H .3oCM pomnmm 00.H 00.m 00.0 40.0 00.m 0m.m 30s H pan: mcHaoax .qu0m 00 op pH 0mg H HH .00 .owmhm>o BoHon HHos .o .omwno>m 30Hon .U .owmpm>o pCopm .o .owmpm>m o>oom .n .ommpm>w o>ooa HHoz .o mH pp mep HmoH pomnmm om.H m9.m NH.o m4.o Hm.m ©4.m H «ooCuooCoo 0H COHposhmeH ozp mm 909 m4 .mH mHmmeomhm mmp p amm .HHHQ C002 C002 spr pmm9 HHCz >9z >9 9 $053 EH @4599 12h struction, whereas the TV group considered it somewhat more disturbing. Much the same feelings are expressed in Item 18 in which the TV group felt that television instruction interfered with their ability to con- centrate, but the NTV group felt that they could concentrate somewhat better than in classes without television. Although both groups indi- cated that the instruction was somewhat above average, Item 19, a sig- nificant difference was found again in favor of the non-television group. Item 20 shows that the non-television students had no prefer- ence about choosing or not choosing a large television section again, but the television students tended to feel they would not choose a large television section again, the differences being significant. In comparisons of the instructors in the course with other instructors both groups indicated that they were above average, but again the NTV group was significantly higher than the TV group. Although both sections were taught by the same instructors, the students in the television section were not as favorable in their opinion of the instructors as were the students in the non-television section. Neither did the TV students feel that they had learned as much as the NTV students although the results on the Achievement Test showed no differences in this respect. It is apparent that the students in the television group were not as happy with the course as were the students in the face-to-face situation. The free comments on the Course Evaluation Scale were tabulated into three groups as follows: (1) Complimentary Comments such as "I think the instructors in the course were among the best", (2) Helpful 125 Comments such as "Discuss more practical problems", and (3) Complaining Comments such as "I found television boring". The tabulation of these results is shown in Table XVI below. Some of the students made com- ments that were classifiable into more than one category and these were so classified. A Chi-square analysis of these results shows no differ- ences between the TV and NTV groups. TAJLL XVI CHI-SQUARE ANALYSIS OF THE FREE COKEENTS ON THE COURSE EVALUATION SCALE FOR THE TV AND NTV GROUPS Complimentary Helpful Complaining N Comments Comments Comments Total X2 ngg TV group 86 9 26 25 60 0.33 6.00 NTV group 81 9 19 19 h7 Total 107 18 as m. 107 Comparing the number of students in each section who responded by making a comment with the total number of students in each section who could have responded, revealed a Chi-square value significant at the I; level of confidence, as shown in Table XVII on page 126. Evi- dently the students in the television section felt more of a need to make some kind of a comment about the course than did the students in the non-television section, indicating a dissatisfaction with the course. From the results of the findings on the Course Evaluation Scale it is clearly indicated that the students undergoing discussion by means of the television technique held a significantly less favorable 126 opinion about the course than did the students in the face-to-face situation. It may be inferred that the inter-personal relationships TASLE XVII CdI-SQUARE ANALYSIS OF CES FOR TV AND NTV GROUPS ON TEE BASIS OF COME-PETS AND NO COrlffl‘JTS TV NTV Total X2 x219; Comment h8 37 85 7.08 6.6h No Comment 38 an 82 Total 86 81 167 between the students and the instructors in the television group were not as good as those displayed in the non-television group, and that the students in the fonner group displayed more hostility as evidenced by the significantly greater number of comments made. Such findings bear out the argument presented in Chapter One that the absence of maximum feedback would result in the lowering of class morale and a corresponding increase in the amount of hostility and insecurity dis- played by these students. VI. SUNEARY The basic data for each individual was key-punched on IBM cards and the basic calculations were accomplished by the use of electronic calculating machine. All other computations were performed on a hand mechanical calculator. The original data for each student and summari- zation of subgroup data is shown in the Appendixes. 127 Analysis of the data obtained on the Pre-test and Post-test measures of the Achievement Test indicates that there were no signifi- cant differences due to the ability level of the students, the prefer- ence for instruction of the students, the sex of the students, or the two discussion techniques. Nor were there significant differences due to the interaction effects of these variables. Analysis of the Pre-test and Post-test data obtained on the Mis- conceptions about Education Scale indicates no significant differences due to any of the main or interaction effects, and all null hypotheses were accepted. Analysis of the Pre-test and Post-test data obtained on the Minnesota Teacher Attitude Inventory showed a significant interaction effect between the preference for instruction and the discussion tech- niques variables. Interpretation of this finding seems to indicate that under the conditions of the present investigation the combination of face-to-face discussion techniques and a preference for a cognitive type of instruction produces the most favorable attitudes toward teach- ing as measured by the Minnesota Teacher Attitude Inventory. All other null hypotheses for this criterion were accepted. Analysis of the Course Evaluation Scale indicates a significant difference between the students due to the discussion techniques. A significant difference between the students in the television and non- television groups were found for the total scale and for ten of the twenty-two items on the scale all of which favored the non-television students. Further analysis of the free comments by the students indi- 128 cated that the students in the television section made a significantly greater number of comments than the students in the face-to-face sec- tion. Interpretation of these findings seems to indicate that the stu- dents in the face-to-face situation held a more favorable opinion of the course than the students in the television discussion situation. The morale of these students was higher and they displayed less in- security and hostility than the television students as inferred from the fewer number of comments. All other null hypotheses concerning the Course Evaluation Scale were accepted. CHAPTER VII SUMMARY AND CONCLUSION I. SUMMARY ThngrOblem It was the primary purpose of the present study to determine the relative effectiveness of face-to-face discussion and discussion by means of television on educational outcomes in a large class. The use of in- structional television has increased rapidly in the past few years, but little is known about its efficacy in the discussion situation. Research evidence indicates that the discussion method is a superior instructional technique for teaching problem-solving techniques and where the objective is the learning of abstract materials. Learning theory and the evidence from.at least one study in communication theory suggests that feedback between the instructor and the students is an important factor in the learning of abstract material. It was theorized that under conditions of television discussion the absence of maximum feedback would prove a handicap to both the instructor and the student, and would result in less satisfactory educational outcomes as well as a lowering of student morals and mu increase in insecurity and hostility on the part of the student. .1 review of the instructional television literature revealed that the type and quality of studies conducted so far has contributed little to an understanding of the problem.and that few studies have been 130 concerned with the discussion techniques. Thg’Desigg The study employed a fouraway factorial design with two levels in each of the classification variables. The four independent variables were: (1) the discussion techniques, (2) the ability level of the stu- dents, (3) the preference for instruction of the students, and (h) the sex of the students. The combination of two levels for each of the in- dependent variables allowed the formulation of fifteen testable null hypotheses of which four were single or:nain.effects and eleven were interacting effects. Educational outcomes were measured on the basis of four dependent variables: (1) achievement of course content, (2) stti- tudes toward teaching, (3) misconceptions about education, and (h) opinion of the course. The design.mads it possible to test each of the dependent variables for the fifteen null hypotheses thus allowing for a total of sixty testable null hypotheses. The control of variable errors was afforded by randomization of subjects, rotation of instructors be- tween the discussion groups, and the use of covariance statistical a- nalysis. Egperimental Procedures The experiment was conducted in an Educational Psychology course. Facilities included an originating room, two viewing classrooms, and the necessary equipment for television production. The pepulation consisted of all lichigan State University students enrolling in Section h, of Foundations of Education 200 in the Fall Term of 1957. The final sample 131 consisted of one-hundred and sixty-seven students. The students were randomly assigned to a television and a non-television group. In the television group discussion was conducted over the closed-circuit tele- vision system and in the non-television group discussion was conducted in the classroom in a face-to-face situation. Evaluation Instruments The ability level of the students was measured by the American Council on Education Psychological Examination. The preference for in- struction was measured by the Preferred Instructor Characteristics Scale, an instrument designed to measure the students' preference on an affec- tive-cognitive continuum wherein "affective” denotes a preference for the inter-personal, student-centered type of instruction and "cognitive' denotes a preference for a subject-centered, intellectual type of in- struction. The PICS has a reliability of about .90. Achievement was measured by a fifty-item multiple-choice achieve- ment test over the course content. Reliability of this instrument was found to be about .52. Attitudes toward teaching was measured by the [insects Teacher Attitude Inventory, an instrument designed to measure the students' score on attitude statements about teaching in comparison with successful teachers. The MTAI has a reliability of about .90. Misconceptions about education was measured by the Misconceptions about Education Scale, designed by the writer and a colleague to measure the students conceptions about irrational beliefs in education and psy- chology. The ES has a reliability of about .63. Student opinion about the course was measured by the Course Evalu- 132 ation Scale, adapted from a similar instrument used in research at Pennsylvania State University. It consisted of’twenty-two items about the instruction.and the course designed after the Likert technique. Additional space was provided for free comments. The basic data was summarized by use of an electronic computer and completed by use of a mechanical calculator. The data was tested by the analysis of variance and covariance, the ‘t‘ test, and the Chi-square test. The level of confidence for all test of assumptions was set at the 11 level of confidence, and the level of’confidence for the criterion tests was set at the 5% level. Results 1. .Achievement criterion - There are no significant differences in achievement due to the ability level of the students, the preference for instruction of the students, the sex of the students, or the dis- cussion techniques, nor are there any significant interaction effects due to the above variables and all null hypotheses are accepted. These findings agree substantially with other research in the use of instructional television. Many'experimenters have interpreted such findings to mean that television instruction is equally effective as con- ventional instruction. It may be, however, that the measuring instru- ments are not sensitive enough to detect differences, and this was possibly the case in the present investigation since the.Achievement Test had a low reliability. 2. Misconceptions about Education criterion - There are no sig— nificant differences due to the ability level, preference for instruc- 133 tion, or sex of the students, or the discussion techniques, nor are there any significant interaction effects due to the above variables and all null hypotheses are accepted. In only one other study has a similar instrument been used. A Hisconceptions about Psychology test was used in the Miami University study (39), and in comparisons of large class and control sections of small classes a significant difference was found favoring the control classes. In the present study the lack of significant differences may have been due in part to the fact that the instrument was week from the stand- point of reliability. 3. Attitude toward Teaching criterion - There is a significant interaction effect due to the preference for instruction and the dis- cussion techniques variables, and the null hypothesis concerning the PICS x TECH interaction is rejected. All other null hypotheses were accepted. Interpretation of this interaction indicates that the com- bination of students with a preference for a “cognitive" type of in- struction and the face-to-face discussion technique is most conducive to the promotion of attitudes toward teaching as measured by the Minnesota Teacher Attitude Inventory under the conditions of the present experi- ment. No other studies to the writer's knowledge have used these in- struments in instructional televisionresearch, and therefore no com- parisons can be made. The effect obtained in the present study could have been obtained on a chance basis alone, however, the phenomenon 13h does merit more study. 1;. Opinion about the course criterion - There is a significant effect due to the discussion techniques variable and the null twpothesis concerning the discussion variable is rejected. All other null hypothe- ses are accepted. Interpretation of this finding indicates that the stu- dents under the conditions of face-to-face discussion held a more favor- able opinion of the course than the students under the conditions of television discussion. Comparisons of the two groups on individual items of the scale verified these findings. Comparisons of free comments made by the students indicated a significantly greater number of criticisms were made by the students under the conditions of television discussion. It is suggested that this is an indication of the lower morale of the students in the television discussion group and evidence of greater hos- tility and insecurity. Limitations 93 _t_h_e_ Study The present study was not without its limitations. Replication was not provided for in the present study. Verification of results and less hazardous generalizations are afforded by replication. It was pre- viously discussed that replication in the present investigation was im- practical, nevertheless the study was limited in this respect and its generalisability less certain. A second limitation of the present study was the weaknesses of two of the criterion instruments, the PAT and the MES, as indicated by their low reliability. It is possible that more significant findings would have been obtained had these instruments been more carefully 135 planned and tested at the outset of experimentation. The findings of any study even negative findings, are more satisfying to the experimenter and more important to his profession, when he is certain that the instruments he is using are yielding maximum results. The most obvious method for rectifying this limitation would be to lengthen both tests. No matter how carefully the experimenter attempts to determine and control sources of error prior to experimentation, frequently during the course of the experiment or in the analysis and interpretation process, additional sources of bias are discovered. As was suggested in the dis- cussion relative to the findings of a significant interaction between the PICS x TECHNIQUES variables, it was possible for instructor bias to have contaminated these results. The instructors may have displayed differ- ent degrees of 'cognitiveness' and "effectiveness" under the conditions of face-to-face and television discussion. However, even if this bias did exist it bears out the original argument that under conditions of partial feedback both the instructor and the student are handicapped and, in effect, merely intensified the conditions of partial feedback already existing. The effect of the lecture period in the present study cannot be separated from the other effects although it was equalized for both groups of students. A decisive study of instructional television dis- cussion must limit the experimental conditions to discussion only. The presence of the lecture may have been the factor that results in no dif- ferences being found on the achievement and misconceptions criterion. Inasmuch as the experiment was an attempt to study the problem in a real- 136 istic and practical situation, this is not considered a serious limita- tion. II. CONCLUSIONS Conclusions based on the findings of the present study cannot logically be carried beyond the present experimental situation without further replication and research. With these limitations in mind the following conclusions seem warranted: l. The results of the achievement criterion and the misconcep- tions about education criterion are inconclusive. There is no evidence to indicate the superiority of either the face-to-face discussion or the television discussion methods for producing educational outcomes on these criteria. Neither is it possible to generalize from this evidence that the two discussion methods are equally effective in producing edu- cational outcomes on these criteria. 2. The combination of face-to-face discussion method and "cogni— tive" students, as measured by the Preferred Instructor Characteristic Scale, is much more effective in increasing student attitudes toward education, as measured by the Minnesota Teacher Attitude Inventory, than is the combination of face-to-face discussion and."affective“ students. 3. Television discussion is about equally effective for "cogni- tive" and 'affective' students in increasing attitudes toward teaching, but is much less so than face-to-face discussion with "cognitive“ stu- dents. h. Face-to—face discussion produces more favorable student 137 opinions about the course than does television discussion. 5. Face-to-face discussion results in fewer student criticisms of the course than does television discussion. III. SUGGESTIONS FOR FUTURE RESEARCH The majority of instructional television experimentation to date, in comparing television and conventional classroom instruction, has found no differences to exist in the achievement of course content. This finding has been interpreted as indicating that the two methods are equally effective. Where achievement is defined as the comprehension of factual and infermational data about the course, and it is so defined in the vast majority of studies, such an interpretation is probably relevant. In another context, however, achievement may be defined as the ability to apply principles and concepts, to draw inferences, to synthesize, to perceive relationships, and to evaluate. Is television instruction as effective as conventional classroom methods where the above objectives are to be met? No definitive studies have as yet beenrmade to answer this question, mainly because the criterion instruments have not been constructed to measure such objectives. Future research should be con- cerned with developing more sensitive instruments to measure the above Objectives and in applying these instruments to the problem.of the effectiveness of the two types of instruction. The closely related problem of attitude changes would seem a needed area for future research. EWidence from.the present study indi- cates that face-to-face instruction is more effective than television 138 instruction for certain kinds of students in producing attitude changes. This interrelationship needs to be investigated much more thoroughly. What subject-matter attitudes are most affected by different types of instruction? What types of students are most affected in their attitude changes? What is the interrelationship between attitudes, student pre- ferences, and types of instruction? These are but a few of the questions that need to be answered. What is the relationship between student morale and learning? Does the fact that a student has an unfavorable opinion of a course mean that he will not learn as well as students whose opinion is more favor- able? If such a relationship exists, is it as true for the learning of factual information as for the learning of more emotionally laden atti- tudes? Is good classroom morale more important for some students than for others? How important are teacher-student relationships? Do poor relationships actually interfere with learning for all types 'of objec- tives? How do they affect teacre rs? Are some teachers affected differ- ently than others? The answers to these mestions would seem to be im- portant for television instruction. The problem of feedback in itself is an intriguing one. Does the instructor acquire cues that alter his instruction in a face-to-face situation that he would not acquire in a less personal situation? What cues are important? Are different cues important for the learning of different kinds of materials? Are some students better able to commi- cate cues to the instructor than other students? Are some instructors lure capable of recognising important cues than other instructors? 139 Again the answers to these questions are directly related to the prob- less of television instruction and discussion.methods. The problems suggested by the Preference for Instruction Charac- teristic Scale are likewise intriguing. This is as yet a new and un- tried instrument that seems to have promise for research. Is there a relation between the students' emotional needs and his preference for certain kinds of instruction? In uhat particular kinds of instruction will students with different preferences profit most? What is the rela- tionship between intellectual ability and.preference for instruction? tire sose instructors better able toumeet the needs of certain students than other instructors? Do different instructors prefer different kinds of students? Is there a relationship between the students' preferences on this scale and his vocational choice? There are many more problems that need to be answered in the use of television as an instrument for instruction than have been answered thus far by research. The use of television in the schools will no doubt continue to grow, and it is vitally important that its strengths and weaknesses be much.more thoroughly understood than they are under- stood at the present time. 1. 2. 3. h. S. 6. 7. 8. 9. 10. lbO BIBLIOGRAPHY Allen, M. R. Quartermaster Training Connnand Educational-Television Study. Fort lee, Van Wartermaster School, Quartermaster Training Command, June )4, 1951;. Anderson, C. R. and Vander User, A. W. "A Comparative Study of the Effectiveness of lessons on the Slide Rule Presented via Tele- vision and in Person,‘ Lhe; Mathematics Teacher. 1951;, 1:7: 323-27. Asch, Horton Jay. "Nondirective Teaching in Psychology: An EIperi- mental Study,“ P_sycholOgical Monographs, 1951, Vol. 65, No. 321. Bane, Charles L. The lecture _in Coll egg Teachi hing. Boston: Richard Ge Badger, 19310 Becker, 8. L. , Murray, James N. Jr., and Bechtoldt, H. P. Teaching a); the Discussion H__e_____thod. A Stuch supported in part by a grant rom the Fund for the Advancement of Education. Iowa City, Iowa: State University of Iowa, 1958. Bloom, B. S. "Thought-Processes in Lecture and Discussion,“ Journal of General Education, 1953, 7: 160-69. "‘— Bond, Betty Wells. “The Group Discussion—Decision Approach—An Appraisal of its Use in Health Education," Dissertation Abstracts, 1956,16: 9034‘s Boone, W. F. Ev__a____1uation of th__e_ U. S. Naval Academy Educational Television as a Teachin— Aid. Annapolis: United States NaVal I—T'icad om Tmmr 29, 1951.. Bovard, Everett 71., Sr. "The Psychology of Classroom Interaction," Journal o_i: Educational Research, 1951, 1:5: 215-2h. Brinkley, S. G. "Mental Activity in College Classes: Student Esti- mate of Relative Value of Ten learning Situations ," J_9_____urnal o__f_ Experimental Education, 1953, 20: 373-78. Carlson, Carl Raymond. "A Study of the Relative Effectiveness of lecture and Directed Discussion Methods of Teaching Tests and Measurements to Prospective Air Force Instructors ,“ Disserta- tion Abstracts, 1953, 13: 1112-13. 13. 15. l6. l7 . 18. 19. 20. 21. 22. 23. 11:1 Carpenter, C. R. and Greenhill, L. P. An Investi ation of Closed- Circuit Television for Teachin UnfieT‘Efi—rs ty ourses.—In_struc- {1051 Television Research ProjectTImnber One. Project spon— sored by the Fund for the Advancement of Education. University Park, Pennsylvania: The Instructional Film Research Program, Pennsylvania State University, July 1955. Closed-Circuit Television Instruction. Four duplicated preliminary reports. Lafayette: Purdue University, 1956. Cronbach, lee J. Educational Psycholog. New York: Harcourt, Brace and Company, 1951:- Dowell, E. C. An Experiment in Trainin by Television. Keesler Air Force Base, Mississippi:_3385th Technical Training Group, Pro- ject 53-32, February 1951;. Dunham, Franklin and Lowdermilk, Ronald R. Television in our Schools. U. S. Department of Health, Education and mlfarefBuII-e'tm, No. 16. Washington: U. 3. Government Printing Office, 1956. Edmitson, R. W. and Braddock, R. W. ‘A Study of the Effect of Various Teaching Procedures upon Observed Group Attention in Secondary Schools,” Journal o_f Educational Psycholog, 191:1, 32: 665-72. Eglash, Albert. 'A Group Discussion Method of Teaching Psychology,'| Journal _o_f_ Educational Psycholog, 1951:, 115: 257-67. Evans, R. E., Roney, H. B. and McAdams, W. J. "An Evaluation of the Effectiveness of Instruction and Audience Reaction to Program- ing on an Educational Television Station ," Journal 2;: Applied Psychology, 1955, 39: 277-79. Farquhar, William W. An Investigation _o_f the Relationship of Three Teachin Methods 39 Student Behavior in a How ETtudyTourse. Unpubfished Ph. . Dissertation, University—3? Minnesota, 1955. Farquhar, William W. and Krumboltz, John D. "A Check List for Evalu- ating Experimental Research in Psychology and Education," Journal 9_f_ Educational Research, in press. Faw, Volney. "A Psychotherapeutic Method of Teaching PsycholOgy,“ American Psychologist, 191:9, h: lob-9. Gerberich, J. R. and Warner, K. 0. ”Relative Instructional Effi- ciencies of the lecture and Discussion Methods in a University Course in American National Government," Journal _o_f_ Educa- tional Research, 1936, 29: 57h-79. 21;. 25. 26. 27. 28. 29. 30. 31. 32- 33. 3h. 35. 36. m2 Hudelson, Earl. "Class Size at the University level,a Problems of College Teaching. Minneapolis: University of Minnesota Press—,- 9 9 PP. 3‘ ' Husband, R. W. nTelevision versus Classroom for learning General Psychology,“ American Psychologist, 19511, 9: 181-83. Johnson, Donald M. and Smith, Henry C. "Democratic Leadership in the College Classroom," Psychological Monographs, 1953, 68: NO. 36le Kanner, J. H., Runyon, R. P., and Desiderato, O. Telegifsion _i_n Am Training: Evaluation of Television in Aggy Basic Training. shington: Human Resources Research Office, The George Wash- ington University, Technical Report No. 11;, November 1951:. Kelley, Truman lee. Interpretation 23 Educational Measurements. New York: World Book Co., 1927-. Krumbolts, John D. Alphabetical Bias 1.11 Certain Aptitudes pl Geo ramhical Areas o_f 3339 United Statgs. Unpublished study. 195 Krumbolts, John D. in Investigation o__f_ the. Effect 3f Three Eggsh- i_n_g Methods 93 Motivationgl Outcomeg in _a_ How to Study Course. Unpublished Ph.D. Dissertation, University-3'? Minnesota, 1955. Kumata, Hideya. g Inventory _o_f_ Instructional Television Research. Ann Arbor, Mich.: Educational Television and Radio Center. Ieavitt, Harold J. and Mueller, Ronald A. H. "Some Effects of Feed- back on Communication,“ Small Groups: Studies in Social Inter- action, ed. by A. Paul Hare, RSbert Bales, and’Edgar F‘. Bor- gotta. New York: Alfred A. Knopf, 1955. levine, Jacob and Butler, John. I'lecture vs Group Decision in Changing Behavior," Journal 2;: Applied Psycholggy, 1952, 36: 29-33- lewin, Kurt. "Group Decision and Social Change ," Readings -i_n_ Social Psychology, ed. by Theodore M. Newcomb and Eugene Hartley. New York: Henry Holt and Company, 19147, pp. 330-hh. Iewin, Kurt, Lippitt, Ronald and White, Ralph K. “Patterns of Ex- pressive Behavior in Experimentally Created 'Social Climates',n Journal _o_f_' Social Ps cholo , 1939, 10: 271-99. Likert, R. I'A Technique for the Measurement of Attitudes ," Archives g_f_ Psycholog, No. 1:0, 1932. 37. 38. 39- 1:0. hl. 1:2. 113. hh. 145. 1:6. h7- h8. 1143 Lindquist, E. F. Statistical Anal is in Educational Research. New York: Houghton Mifflin Co., 1915. Lindquist, E. F. Desi and Analysis 53f Experiments in Psycholoa 399. Education. ston: Houghton Mifflin Co., 195'. Macomber, F. G. , and others. Ergierimental Study in Instructional Procedures. Project under grant from the Fund for the Ad- vancement of Education. Oxford, Ohio: Miami University, October 1, 1957. Maier, Norman R. F. 'An Experimental Test of the Effect of Train- ing on Discussion Leadership," Human Relations, 1953, 6: 161-73. Michiggn State University Catalog. East Lansing, Michigan: Michigan State University Publication, April 1958, Vol. 52, No. 10. Packman, Martin. “Educational Television,“ Editorial Research Reports, May 18, 19514, Vol. 1, No. 19, p. 376. Parsons, T. S. A Comparative Analysis of Some Outcomes of Instruc- tion by Kinescope, Corregpondence Study and Classroom Proced- ures. Duplicated report of a paper read at the APA conven- tion, San Francisco, Sept. 1955. Ann Arbor, Michigan: School of Education, University of Michigan, 1955. Pasewark, W. R. The Effectiveness of Television 313 _a_ Medium _o_f learning Typewritin . Unpublished Ph.D. dissertation, New fir]: University, 19 6. Paul, J. and Ogilvie, J. C. “Mass Media and Retention,“ Explora- tions, 1955, h: 120-23. Pollock, T. C. Closed-Circuit Television as a Medium of Instruc- tion, 1955-36:' A Re-“oort-on New Fork Universit 's Firs't'r' e' _ar g; Wrinentation‘vith TeIEVi's'iBn _i_n_ Collegg Classrooms. _PrEject sponsored by the Fund for the Advancement of Education. New York: New York University, October 1956. Rickard, Paul B. 'An Experimental Study of the Effectiveness of Group Discussion in the Teaching of Factual Content,” Summa- ries of Doctoral Dissertations, Nortlnrestern University. $953,111! 72"? 0 Rock, R. T. Jr., Duva, J. S. and Murray, J. E. Training Tele- vision: _'I_'_h_e_ Comparative Effectiveness _o_f Instruction _y‘Te‘le- vision Recording and Conventional Classroom Procedures. Wt Washington, L. 1., N. T: Special Devices Center, SDC Report 1476-02-2 (NAVEXOS P-850-2) No date. R9- 50. 51. 52. 53. 5h. 55. 56. 57. 58. S9. 111; Rock, R. T. Jr., Duva, J. S. and Murray, J. E. Training py'Tele- vision: A Stud in Iearnin and Retention. Port Washington, 17—1". ., NT YT: ngEia DeviceE‘be'n—ter“, 's‘ED Report 1476-02—3 (NAVEXOS P—850-3) No date. Ruja, Harry. “Outcomes of Lecture and Discussion Procedures in Three College Courses ," Journal 9_f_ Emerimental Education, 1951‘, 22: 368-9140 Seipman, Charles A. Television and Education if} the United States. Paris: United Nations Educational, Scientific and Cultural Organization. 1952. Shimberg, B. Effectiveness of Television _i_n Home Nursing. Prince- ton, N. J.: Educational-Testing Service, Research Bulletin (RB-51:49), August 1951;. Spence, Ralph B. "Lecture and Class Discussion in Teaching Educa- tional Psychology,“ Journal pf Educational Psycholog, 1928, 19 3 11514.62 e Super, Donald M. Appraising Vocational Fitness pry Means pf Psycho- logical Tests. New York: Harper and Brothers, 191:9. Tannenbaum, P. H. Instruction Throu h Television: A Comparative Study. Urbana: Institute of Communication Researcthniversity 0 Illinois, June 1956. Walker, Helen and Iev, Joseph. Statistical Inference. New York: Henry Holt and Co. 1953. Ward, John N. "Group-Study versus Class Discussion in Teaching Educational Psychology ," Journal _o_f Educational Psycholgg, 1956, 11:: 197-210. Hickman, E. K. Children's Behavior and Teachers' Attitudes. New 1011:: Commonwealth Fund, 1928. Williams, D. C. 'Mass Media and Learning-An Experiment," Ezplora- tions, l95h, 3: 75-82. Willis, B. 0. Evaluation Report _o_i_' the ho Week Emriment pf Direct Teaching _o_p Television. Chicago: Chicago Pu ic Schools, 1956. SUPPLEMENTARY REFERENCES Barr, A. 3., Davis, R. A. and Johnson, P. 0. Educational Research gig Appraisal. New York: J. B. Lippincott Co., 1955. 1115 g Manual g_f_ St 1e. 11th ed. Chicago: The University of Chicago 5388. 191: . Campbell, William G. Form and Style in Thesis Writing. Boston: Houghton Mifflin Co., 1951;. Dixon, W. J. and Massey, F. J. Jr. Introduction to Statistical Analysi . New York: McGraw-Hill Book Co. 1951-.- Garrett, Henry E. Statistics 3 P cholo 9.119. Education. New York: Longmans, Green and Co. 19117. Good, Carter V. and Scates, Douglas E. Methods o_f Research. New York: Appleton, Century-Crofts, Inc. 1951;. Hillway, Tyrus. Introduction 3.3 Research. Boston: Houghton Mifflin COe 1956e Johnson, P. 0. Statistical Methods _i_n_ Research. New York: Prentice- Hall, Inc. 1%90 Perrin, P. G. Writer's Guide 199. Index pg Englis . New York: Scott, Foresman and Co. 1955. APPENDIX A EVALUATION INSTRUMENTS 11:7 PERSONAL DATA CARD 53333.82 u >92 can: .. a n35» .. a. aoaafioaoe u E 188m u an Eoéanom .. m censors .. a 320m .. m cesarean .. a 0380 Anamofiosfl negro no: .oonm II. II. II. 1.... mm: henwdm H49: Joe I III III 1| no- .em .8 38¢ .88 11 1, .84 Sousa soon one room ..a .33m :33 3.884 peasant .02 393309 «encased omednoo 62 38m 62 snow 3 ~38 finer not? mash .25 ...E 88o 62 333m 352 PE 2. a x mm a. m m 1148 Krumboltz ' 79 Preferred Instructor Characteristics Scale Directions: What kind of an instructor do you prefer? In the following items you will find two instructor characteristics paired. From each pair choose the one characteristic you most prefer. Then mark your choice in the proper column on the special answer sheet. Do not omit am items. This is to find out your preferences. There gag pp right _o_r_ wrong answers. I prefer an instructor who: 1. a. is an expert. 10. a. is friendly. b. treats us as mature people. b. is well known in his field. 2. a. makes the classroom pleasant. 11. a. covers all the material. b. thinks logically. b. understands our point of view. 3. a. understands our point of view. 12. a. is interested in us. b. is well known in his field. b. is dedicated to his subject. it. a. is dedicated to his students. 13. a. is an expert. b. is dedicated to his subject. b. is dedicated to his students. 5. a. thinks logically. 1b. a. is well known in his field. b. is friendly. b. treats-us as mature people. 6. a. is well known in his field. 15. a. covers all the material. b. makes the classroom pleasant. b. makes the classroom pleasant. 7. a. is interested in us. 16.a. understands our point of view. b. covers all the material. b. is dedicated to his subject. 8. a. is dedicated to his students. 17. a. is interested in us. ‘e. knows the theoretical background b. knows the theoretical background. of his subject. of his subject. 9. a. think! logically. _ 18. a. is friendly. b. treats 118 88 mature people. b. covers all the material. Krumboltz I prefer an instructor who: l9.a..makes the classroom pleasant. 20. 21. 22. 23. 2h. 25. 26. 28. 29. 30. b. b. a. b. a. b. 8. b. a. b. a. b. a. b. a. b. a. b. a. b. a. b. 31. is dedicated to his subject. knows the theoretical background 32. of his subject. understands our point of view. is interested in us. 33. is an expert. is dedicated to his students. 3h. thinks logically. treats us as mature people. 35. covers all the material. is dedicated to his subject. 36. is friendly. makes the classroom pleasant. knows the theoretical background of his subject. is an expert. understands our point of view. is dedicated to his students. is well known in his field. is dedicated to his subject. treats us as mature people. is friendly. knows the theoretical background of his subject. is an expert. makes the classroom pleasant. as b. a. b. b. a. b. a. b. a. b. 1&9 8o thinks logically. is interested in us. treats us as mature people. knows the theoretical background of his subject. is an expert. is friendly. thinks logically. understands our point of view. is interested in us. is well known in his field. is dedicated to his students. covers all the material. Check to egg if 192 left any blanks. 150 Scoring Key Preferred Instructor Characteristics Scale (PICS) Item Cognitive Item Cognitive Response Response 1 a 19 b 2 b 20 a 3 b 21 b h b 22 b S a 23 b 6 a 2b a 7 b 25 b 8 b 26 a 9 a 27 b 10 b 28 a 11 a 29 b 12 b 30 a 13 a 31 a 1b a 32 b 15 a 33 a 16 b 3h a 17 b 35 b 18 b 36 b 1. 2. 3. h. S. 6. 7. .8. 9. 151 P.L r - F.E. 200 People are especially likely to remember (a) startling facts. (b) facts which they expected. (c) facts about which they had no previous background or expectation. Interference between two learning situations can best be avoided (a) by making sure that a correct response is made in the first situation. (b) by warning the learner against transferring responses. (c) by teaching why a certain response is successful in the first situation. (d) by presenting the situations only after readiness has developed. A psychologist trying to understand the significance of a delinquent's acts would be especially concerned about (a) his religious knowledge and beliefs. (b) his history of school achievement. (c) his attitudes toward authority. ((1) his understanding of principles of good conduct. Tension during an activity is eliminated if the person (a) has readiness for the activity. (b) has emotional support from classmates and teachers. (c) is certain he will reach his goals. (d) is genuinely interested in reaching the goals set. When group decision is used to set goals, (a) the group members act in accord with the decision. (b) goals are set to fit the slower members of the group. (c) performance is close to capacity. The chief fault of ability grouping is that (a) it makes no provision for social-emotional differences. (b) it works an emotional hardship on the less able group. (c) it cannot markedly reduce the range of individual differences. Compared to older drill methods modern activity methods (a) place less emphasis on learning through active practice. (b) place more emphasis on learning through active practice. (c) make more provision for evaluation and reinforcement of responses. (d) require more active responses to meaningful complex situations. in action does not show good adjustment if (a) it is unlikely to attain its purpose. (b) the person acts without considering the facts. (c) the person chooses a course of action which increases his tension. (d) the person leaves some of his needs unsatisfied. Overlsarning is primarily the result of (a) extra practice. (b) strong motivation. (c) pressure to pass examinations. (d) r..: of failure. 10. 12. 13. 1h. 15. 16. 17. 18. .2- 152 The accepting teacher (a) does not judge whether the pupil's word is good. (b) considers the pupil's feelings and judgments. (c) tries to provide the pupil with continual success. Constant criticism of the pupil's original ideas is quite likely to teach his (a) to insist on high standards. (b) to rebel against adults. (c) to follow others' suggestions. (d) to take responsibility for himself. The basic vocabulary of fourteen-year—olds is in the neighborhood of words. Good adjustment is a matter of (a) contentment. (b) confidence. (c) co-Opertiveness. (d) conscientiousness. .L traditional testing procedure may have the following bad effect: (a) it provides little incentive to study. (b) it provides the student no basis for judging which outcomes are considered important. (c) it rewards behavior which is contrary to the objectives. The level of aspiration of poor students (a) will be set cautiously. (b) will be set unreasonably high. (c) cannot be predicted. (d) will vary widely from trial to trial. If a student is to gain the right sort of satisfaction from.performing a good action, the situation should be set up so that before the action the pupil is (a) challenged. (b) relaxed. (c) threatened. (d) mildly irritated. Transfer of a behavior pattern learned in one situation to a new situation.will occur if (a) the situations are the same in all perceptible respects. (b) the learner thinks of the situations as similar even if they actually are not. (c) the situations are closely similar and the learner sees this SimilaritYs In attitudinal learning, the model alters (a) the learner's readiness. (b) the situation to which response is made. (c) the provisional try. (d) the consequence of the response. 19. 20. 21. 22. 23. 2h. 25. 26. -3- 153 The most significant function of subject matter learning is (a) teaching specific responses to problems. (b) clarifying situations to be encountered in future living. (c) developing readiness for later subject matter. (d) teaching technical vocabulary. l.democratic home is best described as one (a) where the children have as much influence on decisions as the parents do. (b) where the children.are allowed to make decisions about their own activities. (c) where the children make decisions about family affairs. (d) where the children's desires are freely expressed and considered in making decisions. .Lnecdotal records placed in the school file should be (a) a random sample of the pupil's behavior. (b) incidents representing unusual behavior by this pupil. (c) incidents which raise special questions or give new evidence about the pupil. (d) incidents in which the pupil's difficulties are exemplified. Ps cholOgy contributes to education by (a defining what goals the teacher should strive for. (b) testing commonly held beliefs about learning and educational procedures. (c) showing what methods or techniques the teacher should use. (d) stressing “new methods" of teaching. Frank (age 20) is sour’and rude. He has a chip on his shoulder..A trained counselor would view him in this way: (a) If Frank would change his attitude, he could be helped. (b) Frank evidently doesn't know how to satisgy'his needs. (c) Frank's unfortunate character was probably fixed in early childhood. (d) Frank's behavior would change if people were pleasant to him. .Allowing the student to express his disagreements with authority is es ecially important in teaching subjects (a that students enter with definite ideas. (b) which arouse emotional reactions. (c) where answers are supported by reasoning. The particular advantage of a standardized reading test is that (a) it presents the same task to all pupils. (b) it permits a comparison of the pupil's score'with the score expected for his grade. (c) it presents selections unlike those from everyday books. (d) all reading tests give essentially the same information. The teacheerho obtains a sociogram by asking pupils whom they wish to work with should form groups (a) to put together pupils who choose each other. (b) to break up mutual-choice pairs and other cliques. (c) on the basis of educational needs without regard to the sociometric result. (d) to put together pupils of equal popularity. 27. 28. 29. 30- 31. 32. 330 3h. 35. -h- Practice on a skill is particularly valuable for (a) improving understanding. (b) reducing time required for each movement. (c) developing discrimination. 15h Of these self-concepts, the most desirable one is: (a) whatever I do is good. (b) if I fail at something, it isn't very important. (c) I am capable of reaching my goals. (d) I am constantly alert to my weaknesses. Which of these habits, considered by itself, would be strong evidence of maladjustment to an adult? (a) Driving hard to attain political prominence. (b) Finding fault with the work of others. (c) Enjoying wildly imaginative fiction. (d) Taking poor care of one's property. General scholastic ability is a term used to refer to (a) the pupil's record on standard achievement tests. (b) innate mental ability. (c) performance on a variety of problem-solving tasks. (d) a combination of special abilities and talents. Which of these would be most likely to suggest faulty emotional develop- ment in a ninth-grade girl? (a) Lack of interest in boys. (b) Striving for perfection in all her work. (c) Getting along smoothly with teachers. (d) Strong interest in music, with only passive interest in other schoolwork. The basic question in determining objectives is: (a) What learning will be needed in later courses? (b) What does the community want a young person to know? (c) What behaviors are needed to perform successfully in this society? (d) What can each of the school subjects contribute to development? "DevelOpmental tasks" are those learnings which . (a) the culture expects all members to master'at a particular age. (b) depend primarily on physical maturation. (c) the child must complete before he is ready for school. (d) are nearly complete by the time the child reaches puberty. An.adolescent boy is asked to climb a ten-foot ladder, and shows panic at the danger of falling. This is best interpreted as showing (a) a specific fear conditioned in infancy. (b) a general attitude of insecurity. (c) negativism stemming from.difficulty with adult authority. (d) displaced emotion resulting from some incident earlier in the day. By the definition in the text, attitudes differ from.ather'meanings in this'way: (a) They are less clearly expressed. (b) They are fanned more indirectly. (c) They involve acceptance or rejection. (d) They apply to general classes of objects or events. 36. 37. 38. 39. h0. h3- 65- 155 Practice material for typing courses will prove best, in terms of ability to type later on whatever job the student gets, if (a) the material resembles closely the sort of copy used on most jobs. (b) the material includes a complete coverage of varied words and sentences (for example, pharmaceutical terms like "polymyin") to include all letter combinations even though some are uncommon on jobs. (c) the material is relatively easy, so that there is a high degree of over-learning of speed for the copy practiced. To avoid the faults of tests as an evaluation procedure, tests should be (a) used chiefly for the less important objectives. (b) used regularly throughout the course. (c) reserved for final evaluation at the end of the course. (d) given little weight in marking. When a teacher invents a new teaching procedure, psychology's chief contribution is to provide (a) a scientific method for determining its value. (b) principles that the proposal should conform to. (c) superior procedures which.made the invention unnecessary. (d) ways of studying individuals when the procedure fails. A. rson adopts the standards of the group that (a he is with at the time. (b) accepts him in a friendly fashion. (c) he believes he is like. .A teacher's case study of a pupil'begins‘with (a) administration of a general series of readiness tests. (b) observation in varied situations. (c) a detailed interview. Setting up an individualized program.for each pupil (a) is especially useful in developing definite, easily checked skills. (b) leads to less thorough coverage than a program of uniforn.assignments. (c) is practicable only where pupils have quite similar readiness. (d) provides for social development as well as intellectual growth. Inattention ordinarily means (a) lack of readiness. (b) lack of interest. (c; poor study skills. (d low mental ability. Motion pictures are especially useful in teaching about (a) events involving motion of change. (b) experiences familiar to the pupil, but in which he has been emo- tionally involved. (c) events which the pupil has not experienced. When pupils are graded on the basis of their ability, (a) good grades become an attainable goals. (b) students no longer see grades as rewarding. (c) the better students are doomed to trustraticn. h5. h6. h7. h8. h9- 50. .6. 156 Socialization is concerned with deve10ping (a) social-emotional adjustment. (b) attitudes, likes and dislikes, and beliefs. (c) skills needed.for'keeping alive and well. (d) patterns of action expected by one's group. Governments are often classified into such groups as democracy, dictatorship, monarchy, etc. In advanced subjects, thinking is im roved by (a) stating generalizations in terms of'these concepts. (b) reclassifying on some other'basis before generalizing. (c) subdividing the group before stating generalizations. The advisable relation between early practice and explanation is this: (a) Explanation should be introduced only after the learner has passed the random trial-and-error stage. (b) Practice should be allowed before explanation only if it is needed to make the explanation intelligible. (c) Practice and explanation should be alternated, in roughly equal proportion. (d) Explanation should be offered only when the learner realizes that his form is poor. Among five superior fourth-graders who can do arithmetic at the sixth- grade level, we would expect that (a) four of the five can read at the sixth-grade level. (b) the average reading ability will be at the fifth-grade level. (c) four of the five will have fourth-grade reading ability. Evaluation of school learning refers to (a) preparation and administration of objective tests. (b) assignment of marks on the basis of performance. (c) judging the adequacy of responses. (d) ualdng provision for individual differences. Which of these traits of teachers is most liked by pupils? (a) maintains an objective, impersonal attitude. (d) Has high intellectual ability. (c) Is sympathetic. (d) Sets standards which are easy to attain. 15? Scoring Key Pre And Post Achievement Test (PAT) Item Correct Item Correct Response Response 1 b 26 a 2 c 27 c 3 c 28 c h c 29 b S a 30 c 6 c 31 b 7 d 32 c 8 b 33 a 9 a 3h b 10 b 35 c 11 c 36 a 12 d 37 b 13 b 38 b 1h c 39 c 15 c ho b 16 a hl a l? b h2 a 18 c h3 c 19 b hh c 20 d as d 21 c h6 b 22 b h? b 23 b h8 b 2h b h9 c 25 b 50 c ' 158 M E S - FeE. 200 "f-K’V,‘ (j, Please read each statement carefully and then indicate whether you believe it is TRUE or FALSE by marking the appropriate space on the answer sheet. 1. The number of man's senses is five. 2. Girls are not more moralistic and better behaved by nature than boys. 3. Long slender hands indicate an artistic nature. 1;. Most stealing offenses by Junior High students are not serious. 5. The limits for the development of any student are set more by his desire and persistence than by his heredity. 6. There are characteristic racial differences in intelligence. 7 . Outstandingly beautiful girls are usually not as intelligent as girls who are not so beautiful. 8. The basic needs of adolescents are not different from the needs of people of other ages. 9. Smoking among young people is a sign of maladjustment. 10. Students who are mpertinent and defiant are usually exhibiting signs of personality problems. 11. The speed of forgetting is slowest immediately after material (such as a poem) has been learned. 12. Social acceptance and the need for recognition are not necessarily motivational forces. 13. Fearfulness is not an indication of underlying behavior problems in children but is comon among young children. lb. Subject-matter must have personal meaning to become interesting to the learner. 15. Thoughtlessness in students should be looked for by the teacher as a sign of serious emotional problems. 16. The student who constantly interrupts class proceedings is showing signs of maladjustue nt. 17. One cannot detemine what his life vocation should be by the use of tests. 18. Pupils should be given heavy study loads so as to hasten maturation. 19. Many eminent men were feeble-minded when they were young. 20. You can estimate a person's intelligence pretty closely by his facial characteristics. __ _._ _fi 21. 22. 23. 2h. 25. 26. 27 . 28. 29. 30. M E s 2 159 Pupils who are consistently restless are usually displaying symptoms of an emotional problem. Women (as a group) are inferior to men (as a group) in intelligence. Grades are important for forecasting a student's vocational success. Carelessness in work is usually an indication of an underlying emotional problem. Unsocialness is not a serious problem but is usually characteristic of the "braim" or intellectual student. Students who are usually disorderly are in need of help for some emotional problem. Intelligence test scores cannot be increased by training. Practice nukes perfect. Cruelty and bullying are not serious signs of maladjustment. It is usually helpful in teaching to classify people into types or categories. 31. A person who has a square jaw will usually have a great deal of will 32. 33- pmr. Children who are large for their age are usually not as intelligent as children who more nearly approach the size norms for their age. Dishonesty is indicated when a person does not look you in the eye. 3h. A high forehead usually indicates intellectual superiority. 35. 36. 37. 38. 39. to. ’41. If you have enough will power you can compensate for most defects of w or Mind. The teacher need not concern himself with mental health as there are specialists to do this. Students who are easily discouraged need extra help and confidence but their problems are not usually serious enough to be referred to a counselor or psychologist. The use of profanity by young people does not constitute a behavior pmbleme Since the major concern of teachers is intellectual growth, the growth of the personality is of only minor importance to teachers. One need not understand the culture in order to understand the behavior of young people. Over-study and over-work can cause feeblemindedness. 2.12. ’43- 1:5 . h6- lfl . 148. L9- 50. 51. 52. 53. Sh. 55. 61. M E S 3 160 Adolescence is a biological phenomenon and has no relation to the culture. In most cases, disobedience is not a sign of serious maladjustment. Continued unhappiness is usually not a sign of a serious emotional problem but indicates a lack of interest or the need for friends. Children who are weak and retarded physically are usually especially intelligent. In general, gifted children require less direction and guidance than average or dull children. Social values of teachers are usually the same as the social values of the upper classes. Grades in school subjects are a reliable and valid means for measuring a student's total growth. Working class people are eager to learn and accept "middle-class" values. The study of mathematics is valuable because it helps develop logical thinking. Usually, adolescents need help with problems related to sex maturation. Adolescents are basically negativistic. There are innate differences in intelligence between different races. All men are not created equal in the capacity for accomplishment. Failing a student is a very good way to motivate him to improve his work. Teachers usually have risen into the middle classes from the upper- lower classes. Tardiness is usually an indication of an emotional problem. I.Q. and intelligence are the same thing. Man is an intellectual being and his emotions play an insignificant part in his behavior. Being overly-sensitive is typical behavior of adolescents. An expectant mother can influence the character of her unborn child. 161 Scoring Key Misconceptions About Education Scale (MES) Item Correct Item Correct Response Response UNNNNN NNNN H H owmflmmgwNHOSmgmGFSKESwmdmmrwNH wwwwwwmawweanewmammwwmamwmamem r o wwwmwawawweenmmwmwewmwwwmwwmwmw uwnom nvALUL'i'lUL‘i 5mm 1.13. 200 Sect. 1: . 162 Roan NO. Seat NO. This scale is designed to help your instructors in evaluating the effectiveness of instruction in this course. Your responses will have no bearing on your grade. In order to be of madman value, you are asked to be sincere, honest and objective in your answrs. Please choose the alternative which best completes each statement for £92 and mark the appropriate space on the answer sheet. 1. As far as the information in this course is concerned, a. I think it will be very useful in my future teaching. b. I think it will be somewhat useful in my future teaching. c. I am not sure whether or not it will be useful in my future teaching. d. I think it will be of little use in by future teaching. e. I think it will be of no use in W future teaching. 2. In the mture whenever there is an opportunity to find out more about the sub- ject matter in this course, a. I will positively do so. 7 b. I will probably do so. c. I don't know what I will do. (1. I will probably not do so. e. I will positively not do so. 3. The information in this course is a. very interesting to me. b. somewhat interesting to me. o. neither interesting nor uninteresting to me. d. somewhat uninteresting to me. e. very uninteresting to me. 1;. Now that I have had some experience in this course, a. I am very glad I took the course. b. I am glad I took the course. 0. I am neither glad nor sorry I took the course. . d. I am sorry that I took the course. e. I am very scrry I took the course. 5. If I have the opportunity to take an advanced course similar to this one, a. I will certainly do so. b. I will probably do so. c. I don't know what I will do. d. I will probably not do so. e. I will certainly not do so. 6. The subject-matter in this course is a. very practical. b. somewhat practical. c. neither practicalmr impractical. d. somewhat impractical. e. very impractical. 7. In comparing the lecture and discussion portions of this course, I feel that 190 a. is much more informative than the discussion. 1:- is somewhat mere informative than the discussion. c. and discussion are equally informative. d. is "Mat lass informative than the discussion. e. is much less informative than the discussion. Do you think that you have learned as much.over television as you would have learned in.a similar course, but taught without the use of television? a. Much more. b. A little more. c. About the same. d. A little less. e. Much less. 9. Do you think this course is more or less valuable than it would have been if to? 10. 12. 13. 11*. vision had not been used? 8. Much more valuable. b. Somewhat more valuable. c. About the same. d. Somewhat less valuable. e. Much less valuable. As far as class participation is concerned, I would have liked to participate a. mmeh more. ‘ b. a little more. e. no more than at present. d. a little less. e. much less. . In comparing the panel discussion groups and 6 x.6 discussion groups, I feel that the 6 x 6 discussion groups a. were much more helpful than the panel discussion groups. b. were somewhat more helpful.than.the panel discussion groups. c. were neither more nor less helpful than the panel discussion groups. d. were somewhat less helpful than the panel discussion groups. e. were much less helpful than the panel discussion groups. As far as the time devoted to discussion in the course, I would have liked to have spent a. much.more time on discussion. b. a little more time on discussion. 0. no more nor no less time in discussion. d. a little less time in discussion. e. much less time in discussion. Aside from the technical difficulties involved in presenting movies ever tele~ vision, I feel that the movies a. were very useful. b. were somewhat useful. c. were neither useful nor useless. d. were somewhat useless. s. were very useless. As far as the reading and reporting of journal articles once weekly is concerned a. I think it was very helpful. b. I think it was somewhat heldful. c. I am not sure if it was helpful. d. I think it was of little help. e. I think it was of no help. As far as the participation in panels was concerned,I would have liked a. to participate much more in panel discussion. b. to participate somewhat more in panel discussion 0. to participate neither more nor less in panel discussion. d. to participate sanewhat less in panel discussion. e. to participate much less in panel discussion. .-n . . 4 15h ~38 16. As far as the subject-matter of the lecture in comparison to textbook material was concerned, a. they were too much.alike. b. they were somewhat alike. 0. they were about right. d. they were somewhat unlike. 0. they were too much unlike. 17. As far as television instruction is concerned, I think it is a. much.less disturbing than ordinary courses. b. somewhat less disturbing than ordinary courses. e. no more disturbing than ordinary courses. d. sonewhat more disturbing than ordinary courses. e. much more disturbing than ordinary courses. 18. As far as television instruction is concerned, I think that I can concentrate a. much better than in ordinary classes. b. somewhat better than in ordinany~classes. o. no more nor less than in ordinary classes. d. somewhat poorer than in ordinary classes. e. much poorer than in ordinary classes. 19. As far as the instruction is concerned, I feel that it is a. well above average. b. above average. c. about average. d. below average. e. well below average. 20. If I had it to do again, knowing what I now know, I would a. positively choose a large television section again. b. probably choose a large television section again. e. have no preference. d. probably would not choose a large television section again. e. positively would not choose a large television section again. 21. I feel that the objectives of the course have been a. very well met. b. well met. 0. satisfactorily met. d. poorly met. e. very poorly met.. 22. As far as the instructors in this course are concerned, I think they are a. among the best. b. better than.most. c. about average. d. poorer than most. e. among the poorest. 23. Use the space on the back of this sheet to make any comments or suggestions that you feel would be helpful to the instructors in improving the course. 165 Item Weights Course Evaluation Scale (CES) l. 2. o 3 S. 6. 8. 9. 10. 12. 13. 1h. 15. 16. 17. 18 . 19. 20. 21. 22. APPENDIX B ORIGINAL DATA 167 «18 8333 .880 ... as 3.9 unautflozoom I at 98m 58m manages 33. 833.883....28 .. a $8 «..a 383.3275 I ado a: 33.... 833:3 33.. 8338:8515 I was new 38» 833.30g”. ..Sofieufi echoes .. 8E boaslfi 8333 25...." 38858.28 I HE. smog «8.5.3 .358 new 3383 bounces“ 3333 uses: sponges—Iona .. a: mum ginseng 838:3 no $05.8 sects—4 .. 34 «3:05.33 masons: to: 4308 I 8 nausea .. a 8:38 H3338 .. m~ .3. .. : conga .08 e .83 .35 I NN 33m 38% .. cm 2.5.8 .. 8 scansseo ~83?” I 3 838.3 I 2. 3390.22 I OH non!- oz I I swam no 88:. no 10.3035 I mm ”agony—6 nuance gouging oesoaom Heaven I pm on... no canoe 3 .5 00:30.» Handgun I cm 3 new: 83!: 2E. 853m Haiehwoam .. mm 3.26 .. m agape asvsaoaS o 51H I am .338 I a 3.: 05E I om you?» I m nogoou 86m I o: chosen—now I N eodatqneaIeoz I E weaken-.3 I on nonhuman I H sedated!" I >9 2: £33 engage a 313 ”96.5 .333 0.000 493 gag mug 168 3288 Ha... 3 NB? 33 ..a on no £338 4.. on 3 N4 H3 30 mm on 3 SH R N a m B. an 3 3 S Nmo Nmo 4N RN 8 «NH R m a R E on .3 R NR d: «2 Nn NN go «Na R a a R B «N 8 3 S «a N8 NN oN NN oNN NR N N a E NN Rm 2 R 8a N8 RN NN fi RN R N ..a. a E. NN mm .3 Q :2 NS RN 5 3 8a 2. N .N a E. «N am mm R 80 moo NN ma 3 cm." R N m a 2.: RN om G N: mod 43 RN NH RN ad" NR N h a PE aN 3 on on 03 NS NN 3 NH 03 3 N m a E ...N S 9 an an 8o Rn NN R .3 NN N N a 92 NN 8 .3 R 09 So oN 3 NN ed 2 N N a 22 NN No NR 3 NNN o8 an NN S 8." R m m a E 8 mo 2 Na 2: Bo NN NN 8 RS 2 n m a E «N mo 2 3 EN RRN ..N RN NH 8a 2. n N a E NH 3 3 mm d: a an ad 3 a NN a m m E. 3 mo 3 on N3 a? an N .3 RN NR m m m B. 3 R 3 mm So Ro mm N 3 Na NN n m n 2 RN R om am Re 8o «N 3 NN 8." d. N N n B. .3 mm 3 on an :2 «N 2 NH «2 d. N ..— n E S no N: a. N3 N8 on 3 : RN 2. N .... n 3 NH .3 3 3 .5 m8 NN NN MN NNN R N N n 2. fl R 3 an «8 «no Hm fl NN an R N u n 2.. 3 R S R «Na 33 an 8 «N mun NR N N n E N R 2 2 an RN 3 8 NH a S N .N n B N N0 N4 R R8 9o NN S 0N oNN I. N N m B. N RR 3 2 08 20 NN 8 2 RS om N a N E o R 3 3 mm" 30 NN 2 N m2 2. n x N E m cm 3 an R8 «3 RN fl 8 03 mm a a N E a S R: 3 Re to NN N 8 8” R N a N E. m S S 3 NNN R8 NN 0N RN .1: R m u N E N S cm 3 So «8 «N N 3 SN R m a N B. N flu m3. as La: #3.: Na NE SE 84 :38 «9.8 Em N88 .598 2....an .68 at .88 NE .58 NE . .38 3&5 763.33 4849 ASH—”3H0 I m ”E 169 688. a: 3 N83 83 .2 om no £3.80 at 888238 338338 88. N; R NR «8 m8 NH ..H NH 48 NN N 3 NH E He no on N: 3H NNH NR NH Nm Re R n N a 3 2.3 8 3 mm mm mNo NNo mm NH NH 8H m N N a N E Nm R N; 3 ed mNo NN NH NN NNo NN N H. N Na Nm HN .3 N: Nmo 28 NN mH NH NNo mm N a N NN Nm Nm Hm S NNH Nfi NN NH 8 SH I. N .H N NE on we 3 3 NNH HmH R NN .3 N3 NN m N N Na: mm om N: 9 N8 N8 RN mN 8 8H Nm N ...H N E. am R N: Na NNH NoH NN NN mo 09 Nm N ..H N E R ON 2 R NHH .16 NN NH 8 NHH om m .N N NE Nm mo R R Nmo Hmo NN NH No NNH HN N N N E Hm NN B N; NNH an RN NH 8 3H HN N N N EH om me 5 m4 :8 28 NN NH No HNH HN N N N NE N4 8 N: 3 8H 4.HH HN NN 8 NH om N .N N E. N: N: NR 3 8H «‘8 NN NN Ho NHH 2 N .H N E. S RN 3 HR mNo mmo NN NH no NHH HN H ..H N E 3 R 3 3 m8 Nmo mN HN No mHH 3 n N N Na B 2 S 9 mfi 80 NN NH 8 NNH o: N __H N Na .3 3 NR an NNo mmo NN NH 8 8H NN N N N B. 9 mo 3 mm NNo Nmo mN 8 No oHH HN N N N B 3 RN 3 R N8 N8 NN 8 8 «NH HN N ..H N Na 3 mm Hm mm 98 50 RN NH no NNH HN N N N B. 2 N4 mm NR 50 m8 HN NH Ho oHH HN N .N N 2 NR mm 3 NR NoH mNo NN HN no NHH NN N N N NN NR 3 R NN NNH So mN mN mo NHH I N N N B R 2 3 NR NHH Nmo mN NH 8 NNH .... N .H N E R mo 2 NR :8 Re NN NH 8 NHH .... N .H N E mm NN N: 5 SH SH NN NN No NNH HN N a N NE. .3 Nm 3 NR 80 H8 NH NH 8 NNH I N a N E an no 2 NR HNo oNH NN NH mo NNH 3 N a N 3.. NR an a gin—”delta: Hdm man” no: Mod so amino a ~85 58 gm anon gm awe a snow an IED Imam Léénoov 3.3 5:38 .. N fig 170 £88. HHNNH 3 323 n8: «3 om No 23.38 at. «NNN no 258 8. HHNN H38 88.. 44 44 H4 NNo NNNo NN NH 8 N8 1. H N mH E HN 4m N4 N4 4mo 03 NN NH 8 48 .... N s 4H EN 8 N4 04 NN NNo 4N0 NN NH 8 8o NN N .— 4H E NN HN 44 NN 440 N8 NH NN No 28 NN N s 4H PE NN N4 N4 NN Re Re NN NH oo 48 Nm N 2 NH NN NN 8 N4 44 NH HNo NN NH 8 SH m N N 2 NH NN NN Hm N4 N4 H8 H8 NN NN N0 Re .I N : NH E. N NN 44 N4 NHH NNo NN NH mN NNo HN N N NH NS 4N NN Nm 3 40H Nmo HN 4H 4H mNo cm 4 N NH NNN NN NN N4 N4 NHH 80 NN NH 4H 08 HN 4 N NH NNN NN Nm NN HN N8 N8 8 NH NN oNo HN N N NH NNN HN 4N NN H4 NNH N40 HN NH 8 HNo HN N N NH E 8 8 N4 NN NHH NoH HN NH NH NNo HN N N NH EN NN NN 04 4N NHH 80 NN NH NH NNo HN N N NH NNN NN Nm N4 mN N8 N8 HN NH NH 02 HN N N NH EN NN 4N N4 N4 HOH Re R NH 3 HNH HN N N NH 2.: NN HN N4 N4 NNH NHH NN NH 2 80 NN N N NH EN NN Nm N4 mN 48 80 NN NH N0 N8 HN N N NH NNN 4N NN m4 NN NHH 4N0 HN NN NH NNo .... m N HH E N N mN N4 NN mNo 4N0 NN NH 3 4NH HN N N HH 3 NN N4 N4 R NNH 80 NN NN NH 30 HN N N HH 2. HN NN N4 44 NHH N8 8 NH NH NNo HN N N HH E 2 HN 04 NN N8 80 ON NN NH 80 HN N N HH NN NN Nm o4 4N mNo N40 NN NN oN 40H HN N N HH E NN Nm N4 NN NNH NNo NN NN NN HoH HN N N HH 3 NN HN N4 NN 48 08 NN NH NH 40H NN N N HH NN. NN mN HN ON 80 30 NN 4H NH NNo om 4 N HH NN mN 8 N4 NN NOH 50 NN HN NH NNo HN N N HH 3 4N N4 N4 04 8H 08 NN HN NH 98 Nm N N HH Na NN NN H4 NN NOH NNo NN HN NN NNo NN N : 3 E NN as «.3 N3 g...HHNN 1.3.: NHN N .238 no in no Hon-mm 838» no 25! 94m no gnaw .3 2055 8 Madman—H" B: 39335 ho Hagan BE On. zen—”adage Hun as TABIED-Z COMPUTATION TO TEST HOMOGENEITY OF VARIANCE 0F SUBGROUPS ON PAT 185 Subgroup N 11 n82 82 log s2 n log 82 l/n 1 3 2 0.50 0.25 9.397911-10 18.79588-20 .50 2 3 2 8.50 2.25 0.35218 0.70136 .50 3 11 10 158.99 15.90 1.20180 12.01100 .10 h 11 10 203.91 20.39 1.30982 13.091420 .10 5 3 2 12.78 6.37 0.801.114 1.60828 .50 6 3 2 38.00 19.00 1.27875 2.55750 .50 7 11 10 81.56 8.16 0.91169 9.11690 .10 8 11 10 130.02 13.00 1.1139h 11.139140 .10 9 3 2 0.50 0.25 9.3979h-10 18.79588-20 .50 10 3 2 21.0.63 120.32 2.0803h 11.16068 .50 11 11 10 57 .33 5.73 0.7 5815 7 .58150 .10 12 11 10 282.99 28.30 1.15179 1h.51790 .10 13 3 2 0.68 0.32 9.50515-10 19.01130-20 .50 1h 3 2 29.83 111.92 1.17377 2.3h75h .50 15 11 10 233.82 23.38 1.36881. 13.68880 .10 16 11 10 128.18 12.82 1.10789 11.07890 .10 160.92262-60 112 96 1608.18 100.92262 8.80 N —no. of subjects n =N-1 (df) log 11s2 =log 16011.18 =3.20526 =1.98227 1.52599 log n =1og 96 n82 2 adjusted sums of squares s2 = variance 3' =2.3026 96(1.22299)-100.92262 ‘ =2.3026(16.h8hhz) 37.9570 0 1.06143 __ 1 -1 =y=afim_ C —l 511:1)an 961:1006h3 B 35066 x2 = 30.58 (35.66 Reject 186 abound «noun Ammaflnvmah . "5.th u 5H 8.9% W. m cadauouwam mo hafluofifl you anon. .3304 m~.~ u 3m. mdufin 0 .II. 1:.“ H... WI" 8 o 8.0% a m aoannonmum no hpaquowoaom man .309 mm. «3 3.324. o2 flange: d3 98» 33 «I: am now and” now-name." 9923 3.33.. 3.0mm 3.53 mm in one? :35 8.3” H03 am +mm Haw define-rum: £59. 3303 magnum booth meommmmm mm In 83.: nus—30 5a.! 73.: ~m+amflkm 8.: 3:3an .323 3.30m .98»?- qflpfln £3.33.“- Ioo :oaunoum’n non-non and .3on .35m H 3308 uo woman anowowmuooo :m 3.33%?» 5:08. 0235339 305% mmommmm 5“ 330: :0 v33 .58 aoanuam «Ix mm Ion 33¢ undo. 331309 @3586 «@QO cméam 8 9838 as .393 .5 xulz um and.“ coding?" £59.- .333 3.on mm.~m5 ma 9:33.380 533 £64. 7x .5 naqoaoauooo gunman-a 95.533 sgdb 0% No 25% MU Hana—hm flowed—Huh no 8332 a: zo uncomcmbm .3 Eon—”madam .3 Egg 924 ZOHmmumam ho Haggam Ban. 2.. zen—”81950 mung TABLE D-h COMPUTATION TO TEST HWOGENEITY 0F VARIANCE OF SUBGROUPS ON MTAI Subgroup N n n32 82 log 32 n log 82 1/n 1 3 2 108.81 51.21 1.73808 3.86816 .50 2 3 2 1675.17 837.59 2.92303 5.811606 .50 3 11 10 3587.01 358.70 2.551473 25.5h730 .10 h 11 10 11.93.81 1119.31; 2.17108 21.7h180 .10 5 3 2 1129.17 5614.59 2.75173 5.503146 .50 6 3 2 530.78 265.39 2.1.2388 14.81076 .50 7 11 10 3632.21. 363.22 2.56017 25.60170 .10 8 11 10 2313.00 231.30 2.36MB 23.6h180 .10 9 3 2 608.79 3011.110 2.1mm: 11.96688 .50 10 3 2 13811.11? 692.21; 2.811025 5.68050 .50 11 11 10 3289.15 328.92 2.51709 25.17090 .10 12 11 10 2092.28 209.23 2.32062 23.20620 .10 13 3 2 912.15 1456.08 2.65901: 5.31808 .50 1h 3 2 125.10 62.55 1.79623 3.592116 .50 15 11 10 5157.75 515.78 24121.6 27.121460 .10 16 11 10 3106.22 310.62 2.19223 21.92230 .10 fi 112 96 311115.10 236.18096 h.80 1:73; 23.3“W 23:22:22: 8“” °‘ ...... 10g n32 -1og 3ms.1o=u.19339 B' = 2.3026 96(2.51112)-236.18096 log n =10; 96 =1.98227 = 2.3026(h.88656) 2751-117 11.2518 c =1+—3-d—6:1&80 - 935:1.06113 B=€i=§§i§§fi3§ =10.57' 1299 =30.58 >10.57 Accept 187 I .. am- . i -I- c: 188 6.683 «Gnu 38:3?“ 6.683 8.?” “afimduam o "Hmomfluhmnu o H mflommuw." 8 o $.fi am .2. mm o 8.3 an m noanmfiwom «0 3203.3” you puns 0338me no hpaaowoaom you anon. 3.3” $.83 2H A3685: 3.: 98» 36» «In am no.“ on: 5.33.50.“ .33: 3.390. 35 2.5” E 3 83» £0. 25 aux 4m+mmflnm scannoumon $93..» 388. 96.593 45.3” Seam." mm in one: non-30 5.: ALT: «n+5 "km 05H ”Samoan." 053d 3902.... anew?» 55.? 2325. too noauuohwoh noaaoo van 2.6a 2.6a H 3303 no 6003. 3333000 am coauaopmoa non-$3 3:23.39 meow." E6: 4" . 3.38 no voown 05H dogmas «Ix mm log 38.: 3.603 pucplauoa magma—m 2.3 Riga om nae—mes as has .8.“ 08:2 mm 05H 0338mm." anon: 3.33 3.3 cm. SN ma 9:303:00 qua-60 anon: aux Hm 3:33.380 doing!" gang 88:5; 6833 no 056 .8 H35 835.8» go 8362 a no mmBmomDm B ~8ngan no mass 924 ZOHMEMUMM ho HHH—”Egam Emma 09 20349830 mung TABLED-6 COMPUTATION TO TEST HCMCDENEITY OF VARIANCE OF SUBGROUPS 0N MES Subgroup N 11 n82 32 log 32 n log 32 l/n 1 3 2 12.65 6.33 0.801h0 1.60280 .50 2 3 2 2.67 1.38 0.12710 0.25h20 .50 3 11 10 9h.09 9.h1 0.97359 9.73590 .10 h 11 10 1h1.hh 1h.1h 1.150h5 11.50850 .10 5 3 2 9.53 km 0.67852 1.35705 .50 6 3 2 8.85 8.23 0.6263h 1.25268 .50 7 11 10 211.29 21.13 1.32h90 13.2h900 .10 8 11 10 159.80 15.9h 1.20289 12.02h90 .10 9 3 2 0.01 0.01 0.00000 0.00000 .50 10 3 2 5h.00 27.00 1.h3136 2.86262 .50 11 11 10 90.77 9.08 0.95809 9.58090 .10 12 11 10 1h3.09 18.31 1.1556h 11.55680 .10 13 3 2 8.17 2.0h 0.30963 0.61926 .50 1h 3 2 3.28 1.6h 0.21h88 0.82968 .50 15 11 10 109.63 10.96 1.03981 10.39810 .10 16 11 10 2h9.86 2h-99 1.39777 13.97770 .10 112 96 129h.33 100.80568 h.80 111:3:- czgfgubjects 11:: :zfifizzzg sums of squares log m2 =log 12911.33 = 3.11197 8' = 2.3026 96(1.1297)-100.h0568 log n =10g 96 = 1.98227 = 2.3026 (8.01552) 1212976 18.5256 1 1 C —-l-3-(-1—6_—15 1:680 - a=lo06h3 X399=30 058> 17 .111 Accept 3 = §L=M=17,u C 1.06h3 189 TABLE D-7 C(MPUTATION TO TEST HOMOGFNEITY 0F VARIANCE OF SUBGROUPS 0N CES 190 Subgroup N :1 n32 92 log 32 n log 52 l/n 1 3 2 266.67 133.33 2.12h93 h.2h986 .50 2 3 2 2.00 1.00 0.00000 0.00000 .50 3 11 10 690.99 69.10 1.839h8 18.39880 .10 h 11 10 651.68 65.16 1.81398 18.13980 .10 5 3 2 272.67 136.33 2.13h59 h.26918 .50 6 3 2 50.67 25.33 1.h036h 2.81728 .50 7 11 10 82h.00 82.h0 1.91593 19.15930 .10 8 11 10 670.18 67.02 1.82620 18.26200 .10 9 3 2 158.00 79.00 1.89763 3.79526 .50 10 3 2 186.00 93.00 1.968h8 3.93696 .50 11 11 10 1996.55 199.66 2.302h6 23.02860 .10 12 11 10 372.00 37.20 1-SYOSh 15.705h0 .10 13 3 2 78.00 39.00 1.59106 3.18212 .50 1).; 3 2 266.00 133.00 2.12385 b.2h770 .50 15 11 10 715.00 71.50 1.85h31 18.5h310 .10 16 11 10 230.55 23.06 1.36286 13.62860 .10 112 96 7h30.92 171.30650 h.80 = no. of subjects n52 =adjusted sums of squares n =N-1 (df) 32 =var1ance log n32 = log 71430.92 =3.8710h B' =2.3026 96(1.88877)-171.30650 log n =10g 96 =1.98227 =2.3026 (10.015112) 1.33877 23.0615 _. _ 1 _ __ . -E_M .. C --1 3m [$080 9&4.th B—C - 1.06113 —— 21066 1699—: 30.58 > 21.66 Accept g . L 3 ‘0. \\ . \ Sun“ ‘0 IUVL be? - - =3 mm 10 AQUA ' \ 1 v A ' '.‘“ \ Uh .‘ “...". W. 015.. .4123? '.'} «A! ~E‘H‘J n' 1 . ‘-.-~ it? p 14.3.99; 2:? ~ b g; 9'" f3" '.' ' _ '3 rm-rv-v't-JVQ