EFFECT OF SPEED OF PRESENTATION AND IRRELEVANT CUES ON RECALL OF TELEVISION MESSAGES Thesis for the Dayton of Ph. D. MICHIGAN STATE UNIVERSITY Robert W. Schlafar 1.966 ll min; {lizllfllfllflll Lm M {Li 1mm} i ll This is to certify that the , ‘ thesis entitled Effect of Speed of Presentation and Irrelevant Cues on Recall of Television Messages presented by Robert W . Schlater has been accepted towards fulfillment of the requirements for Ph. D. degree in Communication Date Sthembe 0-169 ABSTRACT EFFECT OF SPEED OF PRESENTATION AND IRRELEVANT CUES ON RECALL OF TELEVISION MESSAGES by Robert W. Schlater The Problem Interference in the video channel of television messages and its relationship to recall of message infor— mation was the problem studied in two experiments. Two video channel variables were isolated for investigation. The thesis was that television production base lines must be established to facilitate scientific investigations of television production techniques. The findings from research of production variables should assist in predicting those techniques which will product strong message effects on tele— vision viewers. Design Experiment I studied speed of visual presentation. Speed was varied in five treatments and ranged from one visual per 30 seconds to nine visuals per 30 seconds. Experi- ment II investigated the ratio of irrelevant to relevant visual stimuli. Irrelevant stimuli were varied in six message treat— ments. The ratio of irrelevant visual cues to relevant visual cues ranged from 0:5 to 5:5. Robert W. Schlater Audio was held constant for all treatments in both experiments. The identical sound track consisted of audio information which was related but was not redundant with the visual information. The criterion variable for both experiments was re- call of visual information. Recall was tested by two types of multiple choice questions. Pictorial video questions utilized pictorial sketches of visual information. Verbal ziggg questions were verbal descriptions of visual inform- ation. Visual and audio stimuli were recorded on videotape and transmitted by closed-circuit television to 910 subjects who viewed the messages on television receivers in 30 differ- ent classrooms. Subjects were randomly assigned to classrooms, groups, treatments, and experiments. High school students attending a 4-H summer conference at Michigan State University. constituted the experimental population. Analysis of variance was used to test the significance of differences among group means of recall scores. Findings 1. Interference, introduced by increasing the speed of video presentation, produced a significant difference (p <.05) when recall of visual information was tested using verbal giggg questions. No significant interference resulted when- the relationship of speed and recall was tested using pictorial xiggg questions. Speed of visual presentation had no signi- ficant effect on comprehension of information presented in the audio channel. Robert W. Schlater 2. Increasing the ratio of irrelevant to relevant visual stimuli produced a significant difference (p < .05) on recall of gagig information. The random pattern of the recall scores, however, precluded a meaningful interpretation. No significant effect was produced on recall of visual inform- ation measured with either pictorial video or verbal video questions when the ratio of irrelevant to relevant visual stimuli was increased. EFFECT OF SPEED OF PRESENTATION AND IRRELEVANT CUES ON RECALL OF TELEVISION MESSAGES By Robert w. Schlater A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1966 ACKNOWLEDGMENTS The author wishes to express his gratitude to the many individuals who c00perated in this research. Because of the large number who assisted, most will have to remain unnamed. Especially helpful were the staffs of WMSB-TV, Closed Circuit Television, and the M-H program of the Co- Operative Extension Service at Michigan State University. The encouragement and assistance of the members of the author's doctoral committee are gratefully acknowledged. Professors Walter B. Emery, Randall P. Harrison, Colby Lewis, Leo A. Martin, James B. McKee, and Verling C. Troldahl each made a Special contribution. Special thanks are due Drs. Harrison and Lewis for the time they Spent viewing the experimental television messages Their criticisms and suggestions contributed significantly to the message quality attained. The skill in research design and methodology exhibited by Dr. Troldahl who directed the study is impressive. To him, a sincere ”thank you" for finding the time in a very busy schedule to make the valuable suggestions which brought this study to completion. Finally, to my wife and son: that stranger in the house is your husband and father. ii TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES LIST OF ILLUSTRATIONS. Chapter I. INTRODUCTION Purpose of Study Research Review II. METHODOLOGY General Research Design Experiment I Message Treatments Experiment II Message Treatments Pre- -Testing Sample of Subjects. Test Instrument Construction Message Presentation Test Administration III. FINDINGS Experiment I Experiment II IV. SUMMARY, CONCLUSIONS, AND IMPLICATIONS FOR FURTHER RESEARCH . . . APPENDICES Appendix I. Appendix II Appendix III LIST OF REFERENCES iii Page ii iv vi 18 18 22 27 30 HO All 118 51 52 57 67 78 79 86 92 97 Table IO. ll. 12. LIST OF TABLES Experiment I Visual Requirements by Treatment, Number, Rate, and Duration Experiment II Visual Requirements by Treatment, Number, Ratio, and Duration Experiment II Irrelevant Inclusions by Treatment and Number EXperiment II Irrelevant Inclusions by Type, Treatment, and Relevant Visual Order . Pre- test Messages by Groups, Duration and Ratio Analysis of Variance of Rate Tested on Video Recall by High School Pre- test Subjects . . . . Analysis of Variance of Rate Tested on Audio Recall by High School Pre— test Subjects . . . Comparability of Experimental and Control Groups by Sex, Age, and Education Test of Homogeneity of Subjects by Chi Square Between Treatments by Experiment. Control Group Means by Group Number on Test of Video and Audio Questions Net Differences of Means of Groups With- in Treatments Observing and Not Observing Accident Comparison by ”t" Test of Control Group and Pictorial Video, Verbal Video, and Audio Means By Experiment iv Page 23 29 31 32 33 35 35 37 MO 47 SO 52 Table Page 13. Experiment I Obtained Means and Analysis of Variance of Rate Tested on Video Recall Using Pictorial Video Questions . 5A IA. Experiment I Obtained Means and Analysis of Variance of Rate Tested on Video Recall Using Verbal Video Questions. . 55 l5. Experiment I Obtained Means and Analysis of Variance of Rate Tested on Audio Recall . . . . . . . . . 58 I6. Experiment II Obtained Means and Analysis of Variance of Ratio Tested on Video Recall Using Pictorial Video Questions . 60 1?. Experiment II Obtained Means and Analysis of Variance of Ratio Tested on Video Recall Using Verbal Video Questions. . 61 18. Experiment II Obtained Means and Analysis of - Variance of Ratio Tested on Audio Recall 63 19. Obtained Means of Interest in Television Program by Experiment and Treatment. . 65 20. Experiment I Analysis of Covariance of Rate Tested on Video Recall Using Pictorial Video Questions as Dependent Variable and Interest in Program as 5 Control. . . . . . . . . 65 21. Experiment I Analysis of Covariance of Rate Tested on Video Recall Using Verbal Video Questions as Dependent Variable and Interest in Program as Control. . . . . . . . . 66 22. Experiment II Analysis of Covariance of Ratio Tested on Video Recall Using Pictorial Video Questions as Dependent Variable and Interest in Program as Control. . . . . . . . . 66 23. Experiment II Analysis of Covariance of Ratio Tested on Video Recall Using Verbal Video Questions as Dependent Variable and Interest in Program as Control. . . . . . . . . 66 LIST OF ILLUSTRATIONS Figure I. Effect of Speed of Presentation on Video Comprehension. . . 2. Effect of Irrelevant Stimuli on Audio Comprehension. vi Page \fl Ch CHAPTER I INTRODUCTION A goal of multiple channel communicators is to obtain predictably strong message effects on receivers through con— trolling the audio and visual aspects of the message. The deveIOpment of techniques for achieving this goal is espe- cially crucial to educators who are employing multiple channel media——television and film-—in the teaching—learning process. Of the two media, television has gained wide accept- ance in the field of education in a comparatively short time. The Federal Communications Commission reserved channels for educational television purposes in 1952. By 196A, nearly 37 million enrollments in educational television classes were reported. Media and education experts predict that increasing numbers of students will receive instruction by television in the years ahead.1 Television possesses qualities which makes it a val- uable medium in the educational process. These include the capabilities of l. transmitting both verbal and visual symbols 1Lawrence E. McKune, National Compendium of Tele- vised Education (East Lansing: Michigan State University, 19 J) l 2. attracting and focusing attention 3. rapidly disseminating information to large numbers of receivers at one time. When employed by educators, television is usually referred to as Instructional Television or ITV. The tele— vision message may be transmitted to receivers in class- rooms by closed-circuit cable or by general broadcast. The message is usually subject matter designed for either direct or supplemental teaching. Although instructional television is widely used at all educational levels, research supported techniques for producing strong message effects are notably lacking. Re- search in the past fifteen years has consisted largely of duplications of studies demonstrating that television teaching is generally as effective as classroom teaching. Instructional television differs from many other communication situations in at least two ways: I. The receivers are usually a captive audience and are probably more highly motivated to focus on the message than are typical television viewers.2 The ITV receiver probably attends because of the knowledge that he will be tested over the information in the message. The arrange- ment of the viewing room and the presence of a classroom teacher or proctor also contribute to the focusing or attention "set.” 2J. E. Klapper, "The Comparative Effects of the Various Media,” The Process and Effects of Mass Communi— cation, ed. W. Schramm (Urbana, Illinois: University of Illinois Press, 1954), pp. 92-93. 2. The television director plays a prominent role in the composition of the ITV message. The instructor controls the content but the TV director is highly involved in its treatment, sometimes without the instructor realizing it. The director controls many of the stimuli con- tained in the message through his control of the television cameras and related equipment. He makes the over-all critical decision as to which of the available visual elements will be allowed in the field of view presented the receiver at a given time. He is usually the primary decision maker regarding the setting in which the television instructor will appear. The source of the instructional television message may be said to be, in reality, a source-pair. In many in- stances, there is little correSpondence between the instruc- tor and the director. Each attends to his own Specialty. This unawareness of each other's functions and each other's contributions to the television message may lead to inter- ference in the transmitted message. The fact that television provides two communication channels—~video and audio——can also lead to problems. Mul- tiple channel communication interference is predicted when (l) the information in the various channels is unrelated, (2) the cognitive difficulty or rate of presentation is such that successful alternation of attention among channels is not possible, (3) contradictory cognitive relationships 3 exist among information units in the various channels. ¥ 3Frank R. Hartman, "A Review of Research on Learning from Single and Multiple Channel Communications and a Pro- posed Model with Generalizations and Implications for Tele- Vision Communication,” Research on the Communication Process (University Park, Pennsylvania: Pennsylvania State University, 1960), pp. 6-27. Evidence, primarily from educational film research, suggests that the audio and visual channels of films are often "overloaded" with information. As channels become "overloaded," the findings Show that learning decreases. The concept of channel "overloading" is undoubtedly derived from the generalizations cited previously which suggest that interference occurs when the information in the various channels is presented at a too rapid rate making alterna- tion between channels difficult. The need for further study of the interference con- cept has increased in the past decade with the increased use of television for both commercial and educational purposes. Because of the many similarities between film and television, the interference generalizations applied to educational film probably also apply to television. It is a generally accepted fact that television communicators, both commercial and educational, often fill the visual channel of television messages with as much information as possible in the expectation that the additional cues will result in additional information. Little research is available regarding interference in the Visual channel of television communication. Single “Charles F. Hoban, "The Usable Residue of Educational Film Research," New Teaching Aids for the American Class— room (Stanford, California: Institute for Communication Research, 1960), p. 107. channel investigations have concentrated on the audio chan- nels of sound motion pictures and radio. Audio research is somewhat easier to conduct because of the nature of the code. Verbal code is ordinarily utilized in these studies and, because of its sequential nature, measurements have been easier to obtain. Visual information does not require sequential presentation. A vast amount of information may be presented in one visual display. One of the ultimate goals of research in the tele- vision medium is to discover knowledge about the two channels employedijltelevision communication. Research findings can be used to assist in obtaining predictably strong message effects on receivers through control of the design of the message. Since the component parts of television messages-- the video and audio stimuli—~are not known to be homo— geneous by nature, quantification of the effects of the channels when used in combination is difficult to obtain with present measuring instruments. It was determined, therefore, that the research design of the study reported here would investigate the video and audio channels sepa— rately. Until more is known about the effects contributed by each of the channels, there is a possibility that little additional knowledge can be added to what is now known about the channels when they are used in combination. Purpose of the Study The primary purpose of the research reported in this paper was to investigate interference in the visual channel of television messages. Two main questions were studied: 1. What is the maximum rate at which relevant visual information could be transmitted before recall of visual information was impeded? 2. What is the maximum ratio of irrelevant to relevant visual information which could be transmitted before recall of the relevant visual information was impeded? Because television messages normally employ both the video and audio channels, two secondary propositions were also investigated: 1. What is the maximum rate at which relevant visual information could be transmitted before recall of the audio information in the tele— vision message was impeded? 2. What is the maximum ratio of irrelevant to relevant visual information which could be transmitted before recall of the audio in- formation in the television message was impeded? Although television is a widely employed medium, research directly applicable to the propositions cited above is scarce. A recent (196A) review of research and theory relating to audiovisual information transmission by Travers, et_al., points out serious flaws in studies which have investigated the transmission of redundant information through two sensory channels. The absence of any test of significance ‘1 is the most serious error reported and runs through an entire series of studies beginning with one by Munsterberg and Bigham in l89u.5 Some of the studies were replicated by Travers and his associates. They introduced proper controls in the research designs in treatments which included audio, visual, and audiovisual channels of communication. The replica- tions failed to demonstrate any particular advantage for transmitting redundant information through more than one sensory channel, which the earlier studies purportedly showed. In the visual ”rate" and "ratio" investigations reported in this paper, information in the video and audio channels were related but were, by design, not completely redundant. The experiment was not designed to check the relationship between the two channels. The audio portion of the message was kept the same for all experimental versions of the message. The "rate” and ”ratio" manipu- lations in the video channel were expected to influence recall of information presented in the video channel and recall of information presented in the audio channel. 5Robert M. W. Travers (editor and principal inves- tigator), Research and Theory Related to Audiovisual Information Transmission (Salt Lake City, Utah: Univer- sity of Utah Bureau of Educational Research, 196“), pp. 6.17-6.26. 6Ibid. Research Review Rate of Information Transmission Available related research will be reported sepa- rately for each of the two main propositions reported in this study. PrOposition 1 investigated the rate of transmission of visual information in television messages. Two educa- tional film studies are pertinent and have been analyzed by Miller, et a1. Rate of development is not, of course, a simple, unitary variable; rather it is a term that compre- hends a number of more Specific factors such as rate of speaking, amount of pausing for emphasis or observation, amount of oral or visual redundancy, and repetition within a film. ”Rate of development" treated as an undefined composite has been Shown by JaSpen (1950) to be related to the amount of learning from films demonstrating a mechanical assembly. Similarly, KOpstein, Sulzer, and Lumsdaine (1952) have demonstrated that intra-film repetition by use of added examples increased learning in films on micrometer reading. These investigations showed that the ”point of diminishing returns” for adding of further instructional examples depended on dif— ficulty of the material and intelligence of the learner, but that in any case there was value in giving more examples than often is done in the 7 ”once over lightly” type of training—film treatment.‘ Further examination of the JaSpen study, cited by Miller, revealed the finding that film presentations 7 ‘Neal E. Miller, et al., ”Graphic Communication and the Crisis in Education,ll AV Communication Review, 5 (Winter, 1957), p. 96. containing fewer shots were superior to presentations containing more shots in terms of information gain. GrOpper, in a research proposal to investigate the difficulty level and tempo of instruction, held that an optimum rate of presentation for efficient learning is one that paces students so that they are able with high frequency to reSpond correctly in the presence of appropriate cues. A less than optimum rate is one that results in a high frequenc of student errors in the presence of the same cues. Several television studies have investigated the over- all area of presentation rate but in less specific terms than were present in the film studies. Schramm found in a review of 393 television studies that television mes— sages may be "overloaded" with information. He implies that ”complexity of presentation and a great variety of visual cues may distract a student from the main principles of the presentation."10 Holmes analyzed and correlated a number of television 2’ I studies and found that students who received 'simple” pres— entations showed greater achievement on information tests 8 Nathan JaSpen, Effects of Training of Experimental Film Variables: Study 1, Verbalization, Rate of DevelOp— ment, Nomenclature, Errors, WTHow it Works,Ir Repetition (Technical Report SDC 269-7-17, Special Devices Center, Office of Naval Research, Port Washington, N. Y., 1950). 9George L. Cropper, ”An Experimental Evaluation of Procedures for ‘Individualizing' Televised Instruction,” Television and Human Behavior, ed. Leon Arons and Mark A May (New YOrk: Appleton—Century—Crofts, 1963). loWilbur Schramm, "What We Know About Learning from Instructional Television," Educational Television the Next Ten Years (Stanford, California: The Institute for Communication Research, 1962), p. 66. 10 than did students who viewed "highly visualized" presen— tations. He defined "highly visualized" those presentations which made maximum use of the visual potential of the tele- vision medium. "Simple" presentations made minimum use of the visual potential.ll Carpenter and Greenhill found that lecture and black— board alone made for more learning than lecture plus charts, plus models, plus training films, plus visiting Speakers, plus dramatizations, and so forth in an Air Science class. A replication of the experiment in a psychology class resulted in no significant difference.12 At New York University, Klapper found "simple" tele- vision presentations produced greater learning than did "highly visualized" presentations in a course entitled "Man's Cultural Heritage."13 Rock, et al., found the least effective programs in terms of information gain were those in which considerable time was Spent dramatizing the lesson material while the instructor who described the action was not shown. This llPresley D. Holmes, Jr., Television Research in the Teaching—Learning Process (Detroit, Michigan: Wayne State University, 1959). 120. R. Carpenter and L. P. Greenhill, An Investi- gation of Closed Circuit Television for Teaching University Courses. Instructional Television Project Report Number 2 (University Park, Pennsylvania: Pennsylvania State Univer- sity, 1958). l3Hope L. Klapper, Closed—Circuit Television as a Medium of Instruction at New York University (New York: New York University, 19587. 11 treatment was compared with one in which the instructor presented the material in a straightforward manner.lu Information theory, posited by Shannon and Weaver, acknowledges the identifiable variable of interference in x// the communication process. Weaver submits that ”it seems highly suggestive for the problem at all levels that error and confusion arise and fidelity decreases, when, no mat- ter how good the coding, one tries to crowd too much over a channel."15 One study which could be said to be contrary to the visual rate prOposition was Aylward’s finding that a dy- namic style of editing a television program had a signi- ficant effect on information gain. In dynamic editing, there was a change of picture on an average of every thirty seconds. This was compared with a static style in which the picture was changed on the average of once every three minutes.16 Regardless of the identification term used in the studies cited above, it is apparent that all are concerned 1“R. T. Rock, Jr., J. S. Duva, and J. E. Murray, Training by Television: The Comparative Effectiveness of Instruction by Television, Television Recordings, and Conventional Classroom Procedures (Technical Report H76-O2-2, Special Devices Center, Office of Naval Research, Port Washington, N. Y., 1951). 15Claude E. Shannon and Warren Weaver, The Mathe— matical Theory of Communication (Urbana, Illinois: University of Illinois Press, 1949). 16Thomas J. Aylward, Jr., "A Study of the Effect of Production Techniques on a Televised Lecture," Disserta- tion Abstracts, XXI (1961), pp. 1660—61. 12 in varying degrees with the rate of visual information variable which was one of the two core variables inves- tigated in the research reported in this paper. Only one of the studies defined rate of development of a television message as a unitary variable, manipu- lation of the rate of visual presentation, as was done in the research reported here. None of the studies cited above reported the optimum level of rate presentation. In this research, rate of visual presentation was treated so that the number of visuals presented per unit of time and their effect on recall of video and audio information could be discretely measured. These factors support the thesis of the reported research that a base line must be developed for a variety of variables present in television messages. By estab- lishing base lines, scientific research can be conducted whose findings, it is hOped, will help predict those pro- duction techniques which can be utilized to produce strong message effects on receivers. Ratio of Irrelevant to Relevant Visual Information PrOposition 2 investigated ratio of irrelevant to relevant visual information in television messages. This variable is one about which both film and television pro- duction experts have developed ”rules,” primarily by intuition. Few studies exist which have subjected their rules to scientific testing. l3 McIntyre has pointed out the need for evaluating the apprOpriateness of visual stimuli. Questions about apprOpriateness of cues most often occur with reference to sets and other elements of design which will usually have some general relation— ship to the subject at hand. However, they may con— tribute little as stimuli pertaining to the instruc- tional message with respect to the intended reSponse of students. At best, they are frequently non— functional; at erst, they may conflict with other essential cues. Seibert concurs with McIntyre's position. On cue, or stimulus, in learning, it would appear that television's great strength is also a potential weakness. The stimulus field which television pre- sents to students is immensely rich, yet within the richness, students must somehow be brought to recog- nize which features of any given scene are relevant to the intended learning. . . . Eventually, I suppose, students may learn to discount those parts of an elaborate set which never are central to the instructor's purpose, yet there would be some wisdom in presenting visual scenes which have few dis— tracting or irrelevant cues within them and, thus, that leave the student unencumbered by fgtile searches or frequent visual excursions.l Deutschmann, §t_al,, investigated relevant and ir- relevant information learned in different communication situations. They compared teaching in a classroom or laboratory with teaching by film or television. They assumed that film or TV would provide prOportionately more focusing upon relevant information and less focusing on irrelevant items than would be the situation in a 17Charles J. McIntyre, ”Applying Learning Theory to 'Televised Instruction,” NAEB Journal, Vol. 24, No. 6 (1965).£L 59. 18Warren F. Seibert, ”Comments,” NAEB Journal, Vol.. at, No. 6 (1965), p. 66—67. 1A laboratory. The results of the study provided evidence for the hypothesized greater efficiency of mass media over non— mass communication media but did not support the hypothe- sized difference between film and television.19 Deutschmann and his associates tested for the decoding of both the relevancies and the irrelevancies in the verbal learning tests they administered. Ordinarily, the learning of irrelevancies are discovered post hoc when the hypothe— sized learning of relevant material failed to occur. In their research, the key dependent variable was the relation- ship between the measures of relevant and irrelevant information. The data showed considerable variation in the amount of relevant and irrelevant information learned under several testing conditions. The fact that they found learning of irrelevant information did occur supports the proposition investigated in the research reported in this paper. Kumata tested a relevant—irrelevant hypothesis in an investigation of advertising students. His hypothesis stated that a face—to-face group which would see visual aids (advertisements) presented in color and enlarged upon a projection screen would remember a great deal of the detail of the ads compared with a TV group. The TV subjects would see the same ads in black and white on a twenty—five inch ¥ 19Paul J. Deutschmann, Lionel C. Barrow, and Anita MCMillan, "The Efficiency of Different Modes of Communica— tion,” AV Communication Review, Vol. 9 (November- .December, 1961), pp. 263-30. 15 TV monitor. It was hypothesized that the TV group would remember the principles involved better than the face-to- face group on the basis that there were less distracting irrelevant cues for the TV group. The hypothesis was confirmed.20 Educational film studies have included findings of stimulus relevancy. Roshal found in a film that was pro- duced to teach knot tying that the film in which the rope appeared to tie itself proved to be superior to a film in which a pair of hands tied the knot. One interpretation of this finding was that the hands were irrelevant cues and that they obscured the essential or relevant cues.21 Neu investigated the relevance of visuals in a film. He found that irrelevant additions appeared to lower potential learning. His study was based on the use of devices to direct attention to the relevant or critical information as contrasted to the irrelevant information.22 Visual presentations designed for concept learning can be presented with varying amounts of relevant and 2OHideya Kumata, ”Two Studies in Classroom Teaching by Television,” The Impact of Educational Television, ed. W. Schramm (Urbana, Illinois: University of Illinois Press, 1960). 21S. M. Roshal, Effects of Learner Representations _in Film—mediated Perceptual—bebr Learning (TéchnicéI Report SDC 269-7—5, Special Devices Center, Office of Naval .Research, Port Washington, New York, 19A9). 22D. M. Neu, Effect of Attention-Gaining Devices on iiilm Mediated Learning (Technical Report SDC 269-7-9, 53pecial Devices Center, Office of Naval Research, Port iVashington, N. Y. 1950). . 16 ... I irrelevant information. Hunt has summarized the evidence 1 on irrelevant dimensions in concept learning citing a serieS/ of studies conducted by Bourne and his associates. These studies indicated that as the number of irrelevant dimen- sions increased, the number of errors in concept learning also increased.23 The visual presentations in Bourne's investigations consisted of geometric patterns which were varied in size, shape, color, and position in the diSplay. He found that while an increase in the number of irrelevant dimensions, redundant or non-redundant, increased the number of errors, an increase in the number of non-redundant relevant dimen- sions also increased the errors.24’25 Travers analyzed the Bourne studies and held that the effect was apparently one in which the amount of infor- mation represented by the stimulus array was of utmost importance. According to Travers, ”the addition of non- redundant dimensions, whether relevant or irrelevant, 23E. B. Hunt, Concept Learning (New York: Wiley, 1962). 2LIL. E. Bourne and R. C. Haygood, ”Supplementary Report: Effect of Redundant Relevant Information upon the Identification of Concepts," Journal of Experimental Psychology, Vol. 61 (1961), pp. 259-60. 250. M. Walker and L. E. Bourne, "The Identification Of Concepts as a Function of Amount of Relevant and Ir— relevant Information," American Journal of Psychology, Vol. 7A (1961), pp. Alb—17. 17 increases the amount of information which the subject must process in order to solve the task."26 Archer found that when relevant information was ob— vious the subject had an easier time learning the concept. (Obvious was defined as a stimulus with a high probability that the subject would respond to it.) When the irrele- vant information was obvious the task was more difficult. A further finding was that when the relevant information was not obvious, it took significantly more time and errors to learn the concept, but When the irrelevant information was not obvious there was less of an inhibitory effect. The concept identification task in the Archer study used patterns which were varied in form and size.27 From the studies cited, it has been demonstrated that the variable of irrelevant stimuli can be identified and its effect measured. In a majority of the studies reported, the effect of irrelevant stimuli on learning was measured. In the research reported later, the criterion variable was recall of relevant visual information in television mes— sages for both prOpositions investigated. Testing for recall immediately after viewing the television messages was used so that the Optimum transmission rate of visual information before recall was impeded could more precisely be determined. _7 26 Travers, op. cit., p. U.ll. 2(E. J. Archer, ”Concept Identification as a Function Of Obviousness of Relevant and Irrelevant Information,” iflflirnal of Experimental Psychology, Vol. 63 (1962),pp.616—2o. CHAPTER II METHODOLOGY General Research Design The research reported in this paper was explora— tory. Formal hypotheses were neither formulated nor tested. Two problems concerning interference in the visual channel 'of television messages constituted the research questions which were explored. The problems were operationalized as Experiment I and Experiment II. Experiment I This eXperiment investigated the maximum rate at which relevant visual information can be transmitted before recall of that information is impeded. Five different television message treatments were developed which varied in the num- ber and duration of units of visual information. The message treatments were produced utilizing the following Operational definitions: 1. Rate is the number of discrete relevant visual stimulus exposures presented per unit of time. 2. A discrete exposure is the output of one TV camera at a given moment in time. A new exposure occurs when the output of a different camera is switched to. 3. Relevant stimuli are those stimuli, video and audio, in experimental television 18 19 messages which subjects must attend to and decode in order to answer correctly the questions posed in the testing instrument. Experiment II Investigated in this experiment was the maximum ratio of irrelevant to relevant visual information which can be transmitted before recall of the relevant visual informa- tion is impeded. Six different message treatments were developed which varied the number of irrelevant inclusions with relevant visual information. The message treatments for Experiment II were based on the following Operational definitions: 1. Ratio is the fraction derived from the number of irrelevant visual stimulus exposures as the numerator and the number of relevant visual stimulus exposures as the denominator. 2. Relevant stimuli are those video stimuli presented in experimental television mes- sages which subjects must attend to and decode in order to answer correctly the questions posed in the testing instrument. 3. Irrelevant stimuli are those visual stimuli presented in experimental television mes— sages which are unnecessary for subjects to attend to and decode in order to answer correctly the questions posed in the testing instrument. Experimental Television Messages Requirements for the experimental television mes- sages were: 1. A tOpic whose content was unfamiliar to the experimental pOpulation. 2O 2. A topic which could be developed into a television message which logically pro- gressed from a beginning to an end. Stimulus materials such as nonsense syllables, foreign languages, or non- familiar symbols were avoided. 3. Relevant pictorial materials which were available and could be presented in dis- crete exposures. A. Pictorial materials which adapted them- selves to Objective testing without a need for audio explanations. The tOpic selected, from several considered, was Greek temple architecture. It was assumed that experi- mental subjects would have had little exposure to a detailed presentation of this subject. A variety of illus- trated books were available in the Michigan State University library. The books, primarily British and German, were well documented with photographs of temple ruins.28 Slides of Greek temples were also available from the Humanities Department at Michigan State University. 28Illustrated books used: Jean Charbonneaux, Aspects de la Grece (Paris: Les Editions Braun & Cie, not dated); Sir Banister Fletcher, A History of Architecture on the Comparative Method (7th ed., London: B. T. Batsford, Ltd., 192A), A. Trevor Hodge, The Woodwork of Greek Roofs (Cambridge: At the University Press, 1960); Hanns Holdt and Hugo von Hofmannsthal, Picturesque Greece (New York: Archi- tectural Book Publishing CO., not dated), H. M. Schwarz, Sicily (London: Thames and Hudson, 1956); Paul Ortwin Rave, Griechische Tempel (University of Marburg, Germany, 192A); Gerhart Rodenwaldt, The Acropolis (Oxford: Basil Blackwell, 1957). 21 Several scripts were developed around the available photographs before the final selection was made (Appendix I). Nine major sub—topics which met the criteria listed above and supported the over-all topic of Greek temple architecture were: 1. Location Of temples on hills 2. Floor plans of temples Roof construction of temples Capitals of Doric order columns Capitals of Ionic order columns mmtw Capitals Of Corinthian order columns Bases of Doric order columns \1 O 8. Bases Of Ionic order columns 9. Bases of Corinthian order columns Verbal information presented through the audio chan— nel in the experimental television messages was held constant for all message treatments in both experiments. The verbal cues were non-redundant with the visual stimuli. They did consist, however, of statements which were related to the message sub—tOpics. Three sentences, each approximately ten seconds in length, were read by a narrator for each of the nine sub- tOpics of the basic message. The narrator was a profes— sional television performer who delivered the audio infor— mation at approximately 125 words per minute. Television 22 techniques for presenting verbal information were employed to include incomplete sentences, pauses, and verbal punctuation. Experiment I Message Treatments It was intuitively determined that the base line for Treatment 1 of Experiment I should be 1 visual per 30 seconds. From that base, the other A treatments were de- signed with progressively higher rates of visual presen— tation (Table l). The messages were 4.5 minutes long. TO place the visuals in the context of a ”typical" television presenta- tion, it was decided that a narrator would introduce and close the experimental messages. A 30 second opening and a 30 second closing were developed during which the nar- rator appeared on the television screen. Total length Of the experimental television messages was 5.5 minutes for all treatments. The narrator was seated at a desk with a non- distracting drape in the background. A chest shot which did not show either his hands or the tOp of the desk was used in both the Opening and closing. The camera was positioned and the pan and tilt heads locked. The camera did not move during either the opening or closing. The experiment was designed so that the number of visuals shown per 30 seconds progressed by 2 for each of the 5 treatments. The Opening and closing shot of 23 mocooom m\anm Hm m m m mocooom :\Hn: m0 0 S : mocooom w m: m m m mpcooom OH mm m m m mpsooom om m m H H 4 UONHHmsmH> Hesma> comm mo coepmssm pcoEpoosH cm newsma> Heooe mocoomm om hog gonadz on Op mOHQOPIQSm mamzmfl> mo stEBZ pcmEpmmsB owwwmoz mo monasz .COfipmMSU new «town seeped: «pcoepmosp hp mpcosmsflsoon Hmsmfl> H pcmeflsmmeuu.H mqmoaom nonesz compensa songsz songsz Op psm>oaonnH mCOHmBHOCH pcm>maonnH mHmSmH> unm>oaom pcoEpmosH .COHpms3p new «Oapms «senses «pcoEpmosp mp mpsoEopfiswos anomfl> HH pcoeflnomxmnu.m mqm memos mo mnoono conga so>o whopmow m a m m H smngCHhoo m a m m H OHEOH m a m m H canon ”mHmsmH> mHmpHQmo mo memoso posse no>o mechanc no whoppmq m a m m H mHmzmH> COHposnpmcoo moon so>o 68mph mpmcso m z m m H mHmSmH> CMHQ MOOHm so>o zoomzm Eoom m a m m H deSmH> mHHHL co moHQ8ou so>o pHUmno mo osSmOQEHsOQsm m a m m H pcmapwmse psoEpmosB pcmEpmosB pcoapmose pcoEpmosH COHmSHOCH pcm>OHoshH .honEBC one psoEHQOSp an mCOHmSHOCH pcw>OHoan HH pcoEHSOQXMuI.m mqmmH6psH pcm>OHos new . .socho HmsmH> pcoEomosy «camp HQ mCOHmSHOCH pcm>OHOHSH HH pcoEHHOQXMIn.: mqm housepcom OHUS< masonw mHmsmH> .Oflpwh DEG «coHpEHSU «mmsosw hp mownmmoe unmatchmnn.m mqm mo wfimhamc< as.m mw.m mo.m mH.m mm.m mQOHpmmsv omofl> HmHLOpoflm.onms mmmoom came Hamoms omofl> Ammuzv Ammuzv Ammnzv Aamuzv Aomuzv m pcmspwmse : pomepmose m pcmEpmmmB m pcmEpwmsB H pcmEpmmnB :oflmeEOMCH omofl> pcw>mamm mo mpmm .mCOHpmmSU omofl> HMHMOpoflm wCHmz Hawoms oo©fl> co ompmop moms do mocmflpw> mo mfimzamom cow mammE encampoo H ocmeflsmgxm1u.ma mqm<9 55 -- mm.omm Hm: Hmpoe -- -- mm.o ss.mmm sq: coupmpcmmmsm do mommam shrug: Ho.v Hm.: m©.m ms.qa : COHowpcmmosQ mo mommmm cowzpmm m m .m.E .m.m wo mossom mocmflsm> mo mflmzawc< oo.m mo.m mm.H zw.a :m.a mCOHpmmsv ooofl> Hmosm> wCHms mosoom cmos Hamoms ooofi> Ammnzv Ammuzv Ammuzv Afimnzv Aomuzv m pcospmope : pcmapmmge m pcmEpmmmB m pcmEpmmsB a pcmEpmmsB coamesomcH omofl> pcm>mfimm mo opmm .mooflpmmsv ooofi> Hangm> mcflms Hamomg omofl> :o ompmop moms mo mocwflsm> mo mflmxamcm ocm momma noCHMpoo H pcosfismmeuu.:H mqm to wflmzamc< ©m.m ©m.m 0:.m om.m mmgoom :mmE Hamoog oaosd Ammuzv Ammnzv Ammnzv Afimuzv Aomuzv m pcmEpmmsH : pcoEpmosB m pcmEpmmsB m pcmEpwth H accepmose coamegomCH omofi> pcw>mamm mo momm .Hamomm oaosw co powwow mums mo mocmfism> mo mflmhfiwom pom mcmmE omcflmpno H pcmeflmmmxm:n.ma mqmmamssfiupcm>mfims :flzpflz .m.c mm.o H©.H so.m m mosses pcm>maospflupcw>mams comzpom Q g .m.E .m.m mo meadow mocmflbm> to mammams< ow.m mm.m mm.m m©.m sw.m mo.m chHpmmsa omofl> amflsOp uoflm wCflm: mthUm Emma Hamoms ooofl> Ammnzv Aawnzv Aosuzv Amsnzv Ammuzv Amwuzv m s m m H o pcoEpmose pcospmobH pcmEpmmbE ocmEpwmgB pcoepwobe pcospwose COflmegomCH omofl> pcm>mamm Op pcm>mamssH mo oapmm .mCOHpmosw om©fl> HMflmOpon wCflms HHMoms omofl> co powwow oapms wo mocmflsm> mo mfimhamcm ocm mcmmE omcflmpoo HH pcmsflsomxm|u.wa mqmmammsflupcm>oams afizpflz .m.c os.n mm.a ma.w m manpms pcm>mamssflupcm>mamm cmmzpmm Q m .m.E .m.m mo moszom mocmflsm> mo mflmmawc< sm.a mm.a mm.fi os.a oo.m mm.H mconpmmaa omefl> Hmosm> wCHms mmgoom cme Hamoms omofl> Amwnzv Asmuzv Aosnzv Amsuzv Amauzv Aawuzv m z m m H o pCmEpmosB pcoEpmmsB pcospmose pcmEpwmnB acmepmmpe pcospmose cospmss0@CH ooofl> pcm>mamm ow pom>mamssH mo oapwm .mCOHpmozv omofl> ampsm> mCHms Hawoms ooofl> so powwow owpwg mo mocmwsm> mo mflmxamcm ocm momma omCproo HH pcosflsmmenu.sH mqmmammsflupcw>mfimh oflspflz Ho.v oo.m Hm.s om.mm m mosses pcm>mamssflnpcm>mams cmmzpmm Q m .m.E .m.m mo oopsom moQMHsm> wo mammfiwc< ms.m HH.m om.m sw.m ms.m ss.m mascom cams Hamoms oaos< Ammuzv Aswuzv Aosnzv Amsnzv Ammuzv Ammuzv m s m m H o pcoEpwmsB pcoepmmse pomepwose pcmEpwomB pomEpmmsB pcoapmose :oflumEsOMCH omofl> pcm>mamm ow pcm>mamssH to oapmm .HHMUmH oaosm co ompmmp OHpms do mundagm> mo mflmzamcm 0cm memos nocflmooo HH pcmefistXMun.mfi mqm