The experiment dealt with the question of the comparative learning effects of television and radio. It was determined that both of these media impart knowledge to the same extent and that the forgetting of the knowledge is not media-specific. However, emotional impressions were found to be different for television viewers and radio listeners. These media-specific emotional impressions do not change with the passing of time after the program presentation. Thus the media-specific emotional impression seems to remain constant over time and seems to be largely independent of the retained or forgotten cognitive content. MEDIA-SPECIFIC LEARNING EFFECTS An Empirical Study of the Effects of Television and Radio REINHARD HELMREICH University of Constance The use of radio and television for educational purposes raises the question of what medium-specific capabilities these conveyors of educational material have-whether they serve as additional unintended educational offerings or as conveyors of learning goals set by official educational institutions. If we want to use the right medium in the right place to obtain optimal learning, it is first necessary to clarify fundamental questions such as: What is each medium capable of, and with what limits? Which medium should be used for which kind of information? AUTHOR S NOTE: The comprehensive final report on this study, containinf many methodological details, has been published in German: Sturm, Hertha; Haebler, Ruth von;2 Helmreich, Reinhard: Medienspezifische L erne ffekte. Eine empirische Studie zu Wirkungen von Fernsehen und Rundfunk. Munchen: r/?-t/ef<a~un/on, 1972 COMMUN ICATION RESEARCH, Vol. 3 No. l, January 1976 [53]
[54] Which medium is preferable for which level of learning and comprehension? Do different media have different short-term and long-term effects, and if so, what are they? In looking for basic research to facilitate answering these questions one comes across many research approaches dealing with possible media variations as well as with investigations of mediated versus traditional instruction (e.g., Chu and Schramm, 1967; Skowronek, 1969; Dallmann and Preibusch, 1970). However, one infrequently finds comparative studies which are oriented toward media-specific issues. This is the starting point for the study dealt with in this article. RESEARCH DESIGN The aim of this study was to determine the effects of identical programs broadcast either by television or radio. Thus we have introduced the first variable of the experiment: The medium, defined as the mode of presentation. Two problem areas were to be examined: What amount of testable knowledge is retained by the subjects after presentation in either mediumhow do they react emotionally to the program in either medium? Consequently, the learning effects of the media, television and radio, were compared in the cognitive as well as in the emotional sphere. In this respect learning and forgetting have to be examined in detail. It was to be expected that the recipients would forget some of the things they had learned if they were not questioned immediately after the program presentation. The emotional impressions which the recipients link with the program could also change as time passed. Forgetting of program content and emotional impressions-as well as learning in these two areasmight differ for the two media. Hence a test strategy was adopted. The recipients had to be questioned as to their knowledge about, and their impressions of, the program at different dates. The period of time between the presentation of the program and the questioning had to be varied in length.
55] We have thus introduced a second variable, time of questioning, which was operationalized as questioning immediately after presentation of the program, and questioning one, two and three weeks later. The variable medium and the variable time of questioning were thus the two independent variables created by the experimenter. The effect of the program was the dependent variable to be measured. The precise questions for the planned study were formulated as follows: (1) Do television viewers differ from radio listeners with regard to the increase in their knowledge? (2) Does the quantity of knowledge change in relation to a growing interval between the presentation of the program and the date of questioning? (3) Does the assumed decrease in knowledge change to the same extent for viewers and listeners in the course of time? (4) Do television viewers differ from radio listeners with regard to their subjective impressions? (5) Do the subjective impressions change presentation of the program and the date of questioning grows? as the interval between the (6) Does the possible change in the subjective impressions viewers and listeners in the same way? occur with For each stage of the two independent variables we had to have one group of recipients. The groups viewing the television program had to consist of recipients other than those listening to the radio program, and the various groups of recipients had to be questioned on the different dates. I n order to make possible a comparison between these groups the subjects in the eight groups had to be comparable. First the subjects had to be divided into eight groups matched with regard to potentially influential variables other than the independent variables. The second step was to present these matched groups with the appropriate experimental program, either an audiovisual or an audio presentation only. The
[56] third step was to question the individual groups about the program at the appropriate times after viewing, with one television group and one radio group being questioned at the same time as to their knowledge and their subjective impressions. In addition to the experimental groups it was necessary to have another matched group serving as a control group to determine the recipients knowledge of the subject of the program without their having seen it or heard it. Consequently, the control group furnishes a starting point which can also be assumed for the matched experimental groups. PRESENTATION MATERIAL One of our considerations on the selection of an appropriate experimental program was the strong argument that each recipient. should obtain the same subject-related information from the program. We decided on a discussion program without supplementary visual information. Such discussion programs are usually found both on television and radio. The subject of the program would have little relationship to any topical problem to minimize the delivery of such information via other uncontrollable communication channels. At the same time, however, we did not want the subject of the program to be too obscure. It was finally decided to present the recipients with a discussion program about &dquo;life and Work in Australia.&dquo; It was planned that the program would last between 40 and 45 minutes and that the five participants in the discussion, two women and three men, should introduce themselves to make it easier for the recipients to link the voice with the person. A further condition was that none of the panel members would be known to a wider public. The program was produced in one of the studios of the Bayerischer Rundfunk (Bavarian Broadcasting Corporation) in Munich.
[57] THE QUESTIONNAIRE In order to obtain a representative catalogue of questions about the program the first step was to divide the broadcast material into content units according to formally defined aspects. Since the program was a group discussion we chose to use Bales (1950, 1968) method for the recording and analysis of social interactions. For our purpose the isolation of comprehensible subjectrelated content units from the broadcast material, we modified Bales scheme. We did not analyse the total group behaviour, we only analysed the verbal statements. Only verbal statements are accessible in the same way to both groups ~of recipients, to television viewers as well as to radio listeners. We isolated only those statements from the broadcast which related to the subject. Such subject-related contributions belong to Bales category, &dquo;gives opinion, evaluation,&dquo; and to his category, &dquo;gives orientation, information.&dquo; And finally we only identified the speakers, not the persons addressed, since the radio recipients were not always able to identify the addressed persons. Questions and multiple-choice alternatives were then formulated for the different units. Altogether 93 questions were formulated, each of them with six alternative answers-one don t know any more.&dquo; correct, four incorrect, and one &dquo;i After a pretest the questionnaire was subjected to an item analysis. The final version contained 34 questions concerning &dquo;orientations&dquo; and 25 questions concerning &dquo;opinions.&dquo; A reliable and valid instrument for measuring emotional impressions was found in an adaptation for German-speaking areas (Ertel, 1964, 1965a, 1965b, 1970) of Osgood s Semantic Differential. In its brief form this procedure records essential aspects of emotional experiences by means of 9 pairs of contrary attributes which are arranged at both ends of eight-step scales. The dimensions involved are those which are described by the criteria &dquo;excitement,&dquo; &dquo;valence,&dquo; and &dquo;power.&dquo; As objects for classification we selected the entire
[58] program as well as each of the five persons participating in the discussion. SUBJECTS We endeavoured to form parallel groups of subjects. For our investigation we recruited first-term students whom we selected at random from the student files of the University of Freiburg, Germany. Five hundred students were contacted with an initial letter, 238 of them came to participate in the pretest. This pretest was to help us to obtain data for the matching of the groups of subjects. We used the following main criteria: sex of the subjects, television viewing and radio listening habits, interest in the subject of the program (investigated by means of an internally developed &dquo;procedure of program selection&dquo; in accordance with the method of paired comparisons), knowledge of the subject areas related to the program (investigated by means of scales selected from the &dquo;d ifferentieller Wissenstest,&dquo; Differential Knowledge Test, by Furntratt and Jager, 1969) and an intelligence test (carried out with the aid of a shortened form of Amthauer s &dquo;intelligenz-struktur-stest,&dquo; Intelligence Structure Test, 1955, abridged version by Lienert and Leuchtmann, 1958). The matching procedure ensured that each criterion was equally represented in each group. Finally we assigned 21 subjects to each of our eight experimental groups and 40 persons to,the control group. Two of the groups of recipients were questioned on their knowledge about the program and on their emotional impressions immediately after the presentation, the other groups participated in the further investigations which took place at weekly intervals. The participants of the control group who had neither heard nor seen the program were presented with a slightly modified version of the questionnaire dealing with questions from the category &dquo;orientations.&dquo;
[59] After the interrogations had been brought to a close we reexamined the matching of the groups of subjects. Since there was some loss of subjects over the course of the study, it was necessary to reestablish the matched groups by an additional elimination of the data of some subjects. For the further evaluation of data there remained 17 test persons in each of the eight groups of recipients, 34 in the control group. The structure of the investigation allowed an evaluation by means of a two-factor analysis of variance followed by the Duncan Test. The two factors were &dquo;medium&dquo; and &dquo;date of questioning.&dquo; Comparisons with the control group were calculated with the help of a one-way analysis of, variance. All the results mentioned below are statistically significant. RESULTS The volume of knowledge measured after the reception of the program &dquo;life and Work in Australia&dquo; was the same for both groups of recipients. This holds true for all items of knowledge summed as well as for certain sub-groups of items such as those which only ask for &dquo;orientations,&dquo; &dquo;opinions,&dquo; those which only ask for numerical data, and those items relating to information given in the first or the last third of the program. The amount of knowledge also decreased to the same extent among television viewers and radio listeners as the interval between the presentation of the program and the questioning grew. Between the first and the second date of questioning a strong decrease in the retained knowledge imparted by the program is noted, between the second and the third date the decrease is less and between the third and the fourth date of questioning no further notable decrease was registered. But even on the last date of questioning, three weeks after the presentation, the television and radio recipients obviously still knew a great deal more correct answers to questions about the program content than did subjects in the control group. The emotional impressions wh ich the television viewers and the radio listeners linked with the discussion on &dquo;life and Work
[60] in Australia&dquo; were clearly different depending on whether the program had been received via television or via radio. Radio listeners feel the program to be more exciting, more agreeable and more powerful than television viewers do. These different emotional impressions which viewers and listeners have do not change as time passes. The emotional impressions which the television viewers as well as the radio listeners once linked with the program remain constant; neither intensity nor direction change within the period of time observed. This also holds true for the emotional impressions linked with the individual participants in the discussion. Admittedly some media-specific changes were noted in some cases, but in the great majority of the cases there were no changes in the emotional impressions of television viewers and radio listeners on the different dates of questioning. I n summarizing there are two contrary statements to be made. First, knowledge is imparted to the same extent by both television and radio. Thus the amount of knowledge acquired is independent of the medium. For television viewers as well as for radio listeners the imparted knowledge decreases at the same rate, independent of the medium. On the other hand the emotional impressions linked with the program as a whole, and linked with the individual participants of the discussion, proved to be dependent on the medium. The emotional impressions of television viewers and radio listeners differed. These media-specific emotional impressions do not change; consequently the media-specific emotional impression seems to Bremain constant and largely independent of the retained or forgotten cognitive content. DISCUSSION Before considering the meaning of our results, we should note the restrictive conditions to which the results of this experimental investigation apply. The results are valid for recipients selected according to certain variables; they are valid for the students accomplishments in the context of learning
[61] which has been assessed by a multiple-choice questionnaire; and they are valid for programs which verbally impart knowledge. Going back to the question we asked at the beginning: &dquo;which learning offering is to be conveyed by which medium?&dquo; we could state that cognitive contents were conveyed in an equally effective way by both media. With regard to long-term effects the results of our investigation show the same rate of forgetting information from both media, a rate which parallels that found in other fields of memory research using other methods and other materials. For us it seems important to make clear which kind of learning offering could be appropriate on radio, even if certain visual illustrations belong to the subject matter. The evidence that the media-specific emotional impressions of the program remain unchanged-at least within a period of three weeks-gives rise to a whole range of further questions. First we would have to study the dependence of certain qualities of emotional impressions upon certain variables of the program; we would have to ask whether and how the emotional impressions of television viewers and radio listeners change, if the same actors are seen or heard in several television and radio programs in a different context, and we would have to inquire into the development of emotional impressions of actors. What if, for example, they suddenly do not appear when a program series has been dropped? This last question may be particularly important for children at preschool age who lose identification objects because of such a program policy. Perhaps our experiment has raised more questions than it has answered. The attempt to register not only knowledge but also media-dependent emotional impressions has, in our opinion, produced some unexpected and interesting results. NOTES 1. Hertha Sturm is Professor of Communications at the University of Munich and Head of the research department of the Internationales Zentralinstitut fur das Jugendund Bildungsfernsehen at the Bayerischer Rundfunk (Bavarian Broadcasting
--- (1950) --- (1965a) --- (1965b) --- (1964) [62] Corporation). She studied psychology and received her doctor s degree from the University of Freiburg, Germany in 1948. Subsequently she became head of the educational department of various radio and television stations. In 1967 she habilitated at the University of Freiburg on the basis of her study on "Masse-Bildung- Kommunikation," "Mass-Education-Communication" (published) by the Ernst Klett Verlag, Stuttgart, 1968). At first, Professor of Psychology at the University of Freiburg, and then, from 1974 at the University of Munich. 2. Ruth von Haebler a psychologist has, from 1960-1970, been a collaborator at the Institute of Psychology, Freiburg, Germany. In 1968 she received her doctorate with emphasis on Psychological Diagnostics. REFERENCES AMTHAUER, R. (1955) Intelligenz-Struktur-Test IST. Gottingen: Hogrefe. BALES, R. F. (1968) Die Interaktionsanalyse: Ein Beobachtungsverfahren zur Untersuchung kleiner Gruppen. In: R. Konig (Hrsg.) Beobachtung und Experiment in der Sozialforschung. Koln: Kiepenheuer & Witsch. Interaction Process Analysis. Cambridge, Mass.: Addison-Wesley. CHU, G. C. and W. SCHRAMM (1967) Learning from Television: What the Research Says. Washington: U.S.O.E. DALLMANN, G. und W. PREIBUSCH (1970) Unterrichtsmedien. In: K. Ingenkamp und E. Parey (Hrsg.) Handbuch der Unterrichtsforschung. Teil II. Weinheim: Beltz. ERTEL, S. (1970) "Selbstbeurteilung, Semantik und Personlichkeit." Psychologische Forschung 33: 254-276. "Standardisierung eines Eindrucksdifferentials." Zsch. exp. angew. Psychol. 12: 22-58. "Weitere Untersuchungen zur Standardisierung eines Eindrucksdifferentials." Zsch. exp. angew. Psychol. 12: 177-208. "Die emotionale Natur des Semantischen Raumes." Psychologische Forschung 28: 1-32. FURNTRATT, E. und A. O. JAGER (1969) Differentieller Wissenstest DWT. Gottingen: Hogrefe. LIENERT, G. A. und T. LEUCHTMANN (1958) "Die Moglichkeiten einer Kurzform des IST-Amthauer." Psychologie und Praxis 2: 177-182. SKOWRONEK, H. (1969) Lehrmittel und Lernleistung. In: H. Roth (Hrsg.) Gutachten und Studien der Billdungskommission, Band 4. Stutgart: Klett. Reinhard Helmreich, psychologist, is a member of the "Zentrum l Bildungsforschung" of the University of Constance, Germany and a free-lance collaborator of the Internationales Zentralinstitut fur das Jugend- und Bildungsfernsehen, Munich.