MEANiNG INTENSITY AS RELM’Efi ‘FO REAQERSHI? OF FOREEGN NEWS Mi: far fits Dog!“ cf Ph. D. MEWEGAN STA‘E‘E UNEV‘ERSITY‘ iohn Y. McNeHy i931 0.169 This is to certify that the thesis entitled MEANING NTLSNSITY AS RELATED TO READERSHIP OF FOREIGN NEWS presented by John T. McNelly has been accepted towards fulfillment of the requirements for Ph.D. degree inCormnunication Arts A 0’ I . . . a /' I , / l ’ . A ;' '-—"' ,.’ I .' l I . x . x , /»r ,1 ‘1 g V" .47, '1/ 4 A « lf‘é/v‘: L‘J'I-‘quo-w r , Ila {nu- nrnfaccnr Date .ié ABSTRACT MEANING INTENSITY AS RELATED TO READERSHIP 0F FOREIBN NEWS by John T. McNelly This investigation grew out of the question of whether mass media messages can influence audience interests. particularly in the field of foreign news. The general question has stirred much speculation among critics. practitioners and researchers of the media. Elpirical evidence in this area. however. is relatively sparse. Psychological and media research- ers have tended to equate interest with |'liking.” This study sought rather to measure interest as a readiness to pay attention and to determine whether it can be affected by foreign news stories. The theoretical rationale for the study centered around the development of meaning. through associational processes. of what Osgood calls "assigns" or "sign-assigns.‘ Osgood's semantic differential has been useful in research on assign development. A nulber of experimen- tal studies were cited which suggest an underlying relationship among polarization on semantic differential scales. meaningfulness. attitude intensity. perceptual facilitation. and expressed interest. An index of meaning intensity was constructed to measure in- terest; it was composed of degree of polarization on evaluative scales of the semantic differential. plus absolute scores on potency and John '1'. McNelly activity scales. . Two experiments were performed. in which the stimulus materials were printed news stories about eight foreign news topics of low famil- iarity. Testing materials included semantic differential scales and a forced-choice headline test. Subjects for the first experiment were 104 high school students; for the second experiment the subjects were 126 university summer session students of a broad age range. Varia- tions were introduced in the design of the two experiments in order to obtain data on number of exposures and on retention. The results supported the hypotheses. (1) Exposure to news stories about the tOpics resulted in increased meaning intensity for then. (2) Increases in meaning intensity were accompanied by increases in expressed interest. (3) Increases in meaning intensity were accom- panied by the probability. as indexed by headline choices. of further readership about the tOpic 3. Data on number of expo sures and retention were somewhat sug- gestive of negatively accelerated curves of learning and forgetting for meaning intensity. The retention data indicated that. while forgetting did set in. some acquired meaning intensity was retained in the days imediately following exposure to the news stories. Meaning intensity levels for exposed groups were not changed significantly by attempted manipulations of evaluative polarization. potency and activity in individual story versions in the first experi- ment. However. separate analysis of data on these three components in the second experiment provided support for their inclusion in the meaning intensity index. Exposure to the news stories resulted in the hypothesized John.T. MbNelly effects for tapics which were unfavorably evaluated. as well as those which were evaluated favorably; I'liking" was not a necessary condition for interest. The results of the experiments were discussed in tenms of their theoretical. practical and social implications. Conclusions were drawn on relationships between meaning intensity. or interest. and exposure to news stories. Further research was suggested. par- ticularly through field experimentation to obtain.evidence on media effects on interests in natural settings. MEANING INTENSITY AS RELATED TO READERSHIP OF FOREIGN NEWS By Jr‘ John TfVMcNelly ATHESIS Suhnitted to Michigan State University in partial fulfillment of the requirements for the degree of ' noc'noa 0F PHIL-0mm Department of General Communication Arts 1961 "111/ 124,9 ‘ /;é.> ‘2 \- 7/247/5 / // ACKNOWS For encouragement and suggestions in the development of this thesis. the writer is indebted to the members of his guidance com.- mittee. Professors Paul J. Deutschmann. Archibald O. Haller. Malcolm S. HacLean Jr.. and Dean Fred 8. Siebert. and also to Professor David K. Berlo. Special thanks are due to Professor Deutschmann. chairman of the committee. for frequent consultation and for making available time. facilities and funds through the Communications Research Center; and to Professor HacLean. for theoretical and methodological advice. Generous cooperation in making exPerimental subjects available was received from the staff of the Communication Arts Institute held at Michigan State University in the summer of 1960. and from faculty members of the Psychology Department at Michigan State. Several staff members of the Commmications Research Center provided ready assistance in the execution and analysis of the experi- ments. Miss Darlene Mieden did the bulk of the tabulating and com- puting e ACKI‘DWLEIBMENTS. . . . LIST 01' LIST OF Chapter I. II. V. TABLES. . . . me...‘ INTRODUCTION . Background and Significance TABLE OF Need for Research Expansion of News Range EMPIRICAL AND THFDREI‘ICAL BACKGIDUND Some Definitions of Interest Media Exposure and Interests Information-Seeking Meaning and Attitude Intensity THEDRETICAL FORMULATIDN. . . . . . . Interest and Meaningfulness Meaning Intensity Application to Foreign News mpothe ses MD 0 O O O O O O O O O O O O O O Pre-Te sting Stimulus Materials Testing Instruments Subjects Procedure Statistical Analysis FIRSTEXPERmENT.......... CONTENTS Characteristics of Sample Meaning Intensity Expressed Interest Headline Choices Number of Exposures Reactions to Experiment iii Page vii 41 Chapter Page VI. RESULTSOFSEDONDEXPERMENT.............. 72 Characteristics of Sample Meaning Intensity impressed Interest Headline Choices Retention Familiarity Reactions to the Experiment VII.DmU$DNOOOOOOOOOeeeeeeeeeeeeeo 90 Theoretical Implications Practical and Social Implications Suggestions for Further Inquiry VIII. SUMMARIANDOONCLUSIONS................. 101 mmcm. O O O O O O O O O O O O O O O O O O O O O O O O O O 10“ mmn O I O O O O O O O C O O O O O O O O O O O O O O O O O 0 no iv LIST 01" TABLE Table Page 1. Distribution of Stories in First Experiment . . . . . . . M 2. Distribution of Stories in Second Experiment. . . . . . . #6 3. Mann-Whitney U Tests for Differences on Meaning Intensity between mosed and Unexposed Groups . . . . 60 4. Mann-Whitney U Test for Difference on Meaning Intensity between Groups Exposed to Positive and Negative Stories on Convention People's Partyothana.................... 61 5. Mann-Whitney U Test for Difference on Meaning Intensity between Groups EXposed to High and Low Evaluative Stories on Tehuantepec Canal......................... 62 6. Mann-Whitney U Test for Difference on Meaning Intensity between Groups hposed to High and Low Potency Stories on Government of H‘itieeeeeeeeeeeeeeeeeeeeeeeeo 63 7. Mama-Whitney U Test for Difference on Meaning Intensity between Groups Exposed to High and Low Activity Stories on Libya. . . . . . . . . . . 63 8. Mann-Whitney U Tests for Differences on Expressed Interest between Exposed and Unexposed Groups. . . . . 65 9. Mann-Whitney U Tests for Differences on Headline Choices between Exposed and Unexposed Groups . . . . . 67 10. Mann-Whitney Test for Differences on Meaning Intensity between No Exposure and One WWOeaaeooooooo...........a 69 ll. Kruskal-Hallis One-Way Analysis of Variance on Meaning Intensity for Three Levels of Exposureothutan.................. 69 12. Kruskal—Hallis One-Way Analysis of Variance on Meaning Intensity for Three Levels of ExposureonBroederbond................ 70 V Table Page 13. Mann-Whitney U Tests for Differences on Meaning Intensity between Exposed and Unexposed Groups........................ 75 14. Mann-Whitney U Tests for Differences on Evalua- tive Polarization between Exposed and Unexposed Groups over Four Topics Combined . . . . . . 76 15. Mann-Whitney U Tests for Differences on Potency between EXposed and Unexposed Groups over FourTopicsCombined................. 77 16. Mann-Whitney U Tests for Differences on Activity between Exposed and UneXposed Groups over FourTOpicsCombined................. 78 17. Mann-Whitney U Tests for Differences on Eacpressed Interest between Exposed and Unexposed Groups. . . . . 79 18. Mann-Whitney U Tests for Differences on Headline Choices between Exposed and Unexposed Groups . . . . . 81 19. Mann-Whitney U Tests for Differences on Meaning Intensity between Exposed and Unexposed Groups over Four Topic 8 Combined in Delayed Tests......................... 82 20. Mann-Whitney U Tests for Differences on Meaning Intensity between Exposed and Unexposed GroupsinDelaysdTest................ 83 21. Mann-Whitney U Test for Differences on Meaning Intensity for Exposed Groups between Original and Delayed Test over Four TOpics Combined . . . . . . 81+ 22. Kruskall-Wallis One-Way Analysis of Variance on Meaning Intensity among Class Groups for Four Imposed Topics Combined in Delayed Test—mperimentalGroupI. . . . . . . . . . . . . 85 23. Kruskal-Wallis One-Way Analysis of Variance on Meaning Intensity among Class Groups for Four Exposed Topics Combined in Delayed TestuExperimentalGroupII. . . . . . . . . . . . . 85 21}. Means on Familiarity Scale for Groups Unexposed toStoriesonEachTopic....... .. ... . .. 88 LIST OF FIGURES Figure Page 1. Distribution of Meaning Intensity Scores forTehuantepecCanal................. 58 2. Meaning Intensity Means on Bhutan and Broederbond for Groups with O. 1. 2 am 3 ExposurOSe O O O O O O O O O O O O O O O O 0 O O 68 3. Distribution of Meaning Intensity Scores for Convention Pe0p1e's Party of Ghana . . . . . . . . 74 1}. Meaning Intensity Means on Four thaosed Topics Combined in Original and Delayed Teas. O C O O O O O O O O O O C O O O O O O O O O O O 8? vii CHAPTER I mmDUCTIQN The question of whether the media of mass communication can influence audience interests or tastes has generated much speculation among media critics. practitioners and researchers. Empirical evidence on this question. however. is relatively scarce (Hovland. 1954. p. 1092; Klapper. 1957-58. p. 2#68). It is the purpose of this thesis to inves- tigate. under eXperimental conditions. some twpothesized relationships between media message equesure and audience interests. The theoretical rationale for this investigation is based largely on the develOpment. through associational processes. of what Osgood (1953) calls 'assigns' or “sign-assigned Two experiments were conducted. in which the stimulus messages were foreign news stories. The present chapter will outline some of the background and significance of the underlying question. The second chapter will sur- vey pertinent literature on the nature of interests. and some scattered empirical evidence on media influences on interests. This will be followed by a consideration of secondary motivation for information- seeking. The chapter will conclude with an examination of some theory and empirical findings in the areas of meaning and perception. With this background. the theoretical rationale and hypotheses for this stuiv will be developed in Chapter III. The methodolog for the two experiments will be explained in Chapter 1v. 1.. 2 Subsequent chapters will present the results. discussion. sumary and conclusions. BACKGROUND AND SIGNIFICANCE The social. practical and theoretical implications of the underlying question are far-reaching. It is important for society to know whether - and if so. how - media messages affect people's inter- ests. Such knowledge would be of value to the media themselves in their day-to-day decisions on content. As will be shown. the question raises theoretical issues ranging from the sociology of knowledge across cultures to twpothetical constructs within the human organism. Possible media effects on interests have been discussed at length since World War II by investigative co-issions on both sides of the Atlantic. and in numerous critical books. articles. editorials. speeches and sermons. This concern is reflected in the ”social re- sponsibility“ theory of the press now merging in the western world (Siebert. Peterson and Schramm. 1956). Two schools of thought are discernible on the broad question- of media influences. One is the 'demand' school. which says media offerings are shaped w what audiences want; the other is the ‘supply' school. which contends that public wants are shaped by what the media make available (cf. M. Lerner. 1957. p. 754). c. Northcote Parkinson. in a recent address before the American Society of Newspaper Editors (Eitor 8: Publisher. April 30. 1960. pp. 21-22). expressed succinctly the latter point of view. shared by law critics of the press: I'Demand is created by supply. not the other way around. The public seldom imagines what is not there." 3 A similar conclusion has been arrived at by Lazarsfeld (1953.; p. 3&5) on the basis of a rich background of research on mass media J audiences: "It is by no means true that people have definite desiresl and pick out what they 'need' from the available supply of programs. magazines. and so on. It seems rather that the supply itself Eggates the demand." On.the other hand. a position perhaps more prevalent among communication.researchers is summed up by Schramm.(1957): 'I stand rather on the side of the people who think that taste shapes the media.“ But Schra. immediately qualifies this: lot wholly. for it is obvious that peOple can choose only from the alternatives they have at hand. They cannot be sure that they would or would not like what they have not seen or read. Therefore. there is unquestionably a certain.amount of con» ditioning of the public taste in.the direction of what the media offer them. (p. 275) ' Or>more explicitly. as thLean and Pinna have suggested. per- haps the media "force' us to learn and care about things we didn't know earlier that we needed or wanted to learn about' (1958e. p. 38). This sort of view has been.prommlgated recently by a number of leading editors. Some of them.conscious1y attempt to put it into practice - particularly in the area of foreign news (McNelly. 1959). ‘Hhile their efforts are not based on an explicit theoretical position. there is an implied notion that readers get more interested in a subject as they learn.more about it. 'William.James suggested this relationship between knowledge and interest seventy years ago (1890. pp. 1106111). The nature of such a relationship seems worthy of investigation. IEED FOR RESEARCH The scarcity of evidence on media influences in the field of interests has been explained by Hovland (195”) in terms of the method- ological difficulties in this research area: Tastes develop slowly over a long period of time and hence the growth.may take a long period of observation. Moreover. to classify the types of tastes on the basis of pgst facto reports on the factors influencing their deve10pment may . . . lead to serious questions of interpretation. (p. 1092) The time problem is one faced in any investigation of the effects of mass communication. The effects often are described as taking place over considerable periods of time rather than on a "one- shot" basis. Schramm.has compared media effects to the almost imper- ceptible. drop-by-drop growth of a stalagmite (19193. p. 397). or to the slow changes wrought by a meandering creek (1957. p. 57). The problem.of Egg; facto interpretation arises particularly in field studies associating media messages with changes in interests. From the waiter of media and inter-personal stimli to which people are exposed over time. it is difficult to single out any one message as a causal factor. But the elusiveness and possible gradualness of the effects do not make them less important. A British observer wrote in the last century that the influence of the press was all the greater because it could not be precisely defined: 'so subtle that we breathe it as we do the air. without being conscious of the minute particles that enter into its composition“ (quoted in Aspinall. 1949. p. h). This. of course. is the nature of a wide variety of influences which.shape our lives. 5 Rather than leave such influences in the realm of speculation. it is up to the researcher to devise means of isolating and measuring them. The question of media influences on interests is one which calls for systematic experimentation (Lazarsfeld. 1953. p. 3146). EXPANSJDN 01" ms RANGE Extending audience interests beyond their present boundaries may be one of the most important social functions of the mass media (liafsiger. HacLean and Engetrom. 1951. P- lI56). In marw cases. audi- ence interests are bounded by geographical considerations. This is reflected in numerous readership studies and in newsroom practices (Hellelly. l959o P. 227). The possibility of creating reader interest concerning lands and peoples abroad is a stimulating one to some editors. The implications for international understanding need no elaboration here (cf. Krech and Crutchfield. 19h8. p. 611). What does the broadening of news interests mean in terms of the individual? Daniel Lerner. in a study of modernization in the Middle East (1958) wrote: "The importance of media exposure. in our theory. is that it enlarges a person's view of the world ('Opinion range') by increasing his capacity to imagine himself in new and strange situations ('empatfi')‘ (p. 96). And further: "hpatlv shows in the 'news range' of an individual because this illustrates his interest in matters beyond his immediate personal and local concerns“ (p. 98). "Hews range' was indexed in the Lemar study by recall of international news items. The expansion of peeple's news range in the field of foreign affairs would appear to lend itself to research more readily than 6 possible media effects in some other areas. Interest in foreign tOpics ordinarily should be easier to trace to particular media messages than interest in domestic topics. A message effect on interest in a foreign topic is likely to be more direct and measurable. less “contaminated" by personal experience or inter—personal communication. Such would be the case particularly with foreign tapics of little or no previous familiarity to the audience. The generation of interest in topics of this nature. which commonly appear in the news media. could play an important part in broadening the news range of media amiences. The present study will deal with such topics in attempting to isolate and measure possible mass communication effects relating to interests. CHAPTER II AND C BAG R0 The lack of a substantial body of evidence on media influences on interests may be due in part to a general vagueness about the con- cept of interest itself. This chapter will examine the concept from several empirical and theoretical points of view. Wgnauamm The literature of mass media research throws little light on the precise nature of interest. Wm: (l956_)_*has_noted__mthat in media research generally the term appears without definition. Interest ....Li' 1" rd“. o. 8.1.998193.."§P£..999§‘721.1¥98 ‘m‘? 1.“ in: Dressage! attengsmpeingiugednrathen widely." and "through context. it can often be established that it is being equated with like' (pp. 12- 13). Interest was defined on a like-dislike continuum also in a wave of studies. particularly in educational psychology. during the 1920's and 1930's (of. Fryer. 1931: Thorndike. 1935a. 1935b: Deutschmann. 1956. p. 6). In a review of the vast literature on interests during this period. Fryer defined interests as "the objects and activities that stimulate pleasant feeling in the individual" (1931. p. 15). He con- trasted interest and aversion: There are . . . two qualities of feeling. pleasantness an! unplea santness. The pleasant feeling is the essential component 8 of the interest experience. The unpleasant feeling is the essen- tial component of the aversion experience. . . . Interests are pleasant feeling experiences with degrees of intensity. Aversions are unpleasant feeling experiences. also with degrees of intensity" (p. I$62). . Dealing with skills and activities in adult education. Thorndike reported. "Interests can be modified. Marinate . . . show that likes and dislikes can be learned as truly as names or dates" (1935b. p. 16). As to what forces can be used to modify interests. Thorndike noted "a diversity of opinion and a regrettable confusion and superficiality" among psychologists. educators and moralists (p. 18). He himself stressed associational learning accompanied by reward: I'To attach in- terest to am situation (aw object. study. activity. etc.) . cause the person in question to have the interest in response to the situation and reward him therefor" (p. 30). writing more than two decades after Fryer. Remmers (195“. p. j 162) accepted the notion of interests as "the reflection of attrac- ; tions and aversions in our behavior. of our feelings of plea santness 1 and unpleasantness. likes and dislikes." A possible objection to this conception of interests was raised but then more or less discarded by i distinction may be made between attitudes and. intemsts, in ithat the latter merely iniicate the degree to which the individual gprefers to hold an object before his consciousness. whether he qacts approvingly or disapprovingly toward the object. while atti- tudes indicate his reaction in terms of its direction. pleasantness or unpleasantness. agreement or disagreement. But since we do not prefer to hold an object before our consciousness unless we agree with it or find it pleasurable. the distinction is a very fine one. and attitudes and interests are for practical purposes identical" (P0 163). But can the distinction be written off so easily? Hovland and associates (19149. p. 110) turned up evidence that it cannot. In testing 9 reactions to films they had subjects give instantaneous like-dislike reactions during the showing by means of a pmgram-analyzer. Afterward the subjects were asked for their paper-and-pencil ratings of parts of the films in terms of "net interesting." "next most interesting." etc. (The results showed discrepancies: Remapgjhujord.-unteresting' " connotes some measure of utility of the material whereas the 'like- rdislike' emphasis in the case of the push-buttons suggested enjoyment 'without regard to utility. " Some support for this view was reported byUeutdi—nm(l956. p. 15). He found that subjects' ratings of head- lines on a "useful" scale correlated almost as highly with "interesting" as did ratings on an "entertaining" scale. These findings are consistent with Dewey's description of the root idea of interest as "being engaged. engrossed. or entirely taken up with some activity because of its recognized worth" (1913. p. 17). On the fact of it. the practice of equating interest with liking seems inadequate .- even naive .- when dealing with media con.- tent. Are mass communicators wrong in assuming that people can dislike but still be interested in objects. ideas or human beings .- and that dislike can be the very reason for the interest? Is it not possible to dislike something but still. to like reading about. listening to or viewing it? This matter will be returned to later. Let it suffice at this M ‘k mewfl'” \ point t_o conceive of interest not in team‘s); liking but in. Emmet a edisposition to pay attention. An attempt will be nude in the next chapter to arrive at an operational definition of this predisposition. ' 10 MEDIA EXPOSURE AND INTERESTS girl“; Eden”. The widespread assumption that the media can generate interest in politics gets some support from research on voting behavior. Berelson. Lazarsfeld and lehee (19511. p. 246) found that media emosure about a presidential campaign was associated with interest in the election and strong feelings about candidates. In field studies. of course. the self-selection factor (e.g. . reading about politics because of a pre-existing interest) obscures direction of causality. On the basis of this and other studies. however. Lane (1959. p. 288) sums up the general effect of media exposure as a "politicizing one" .- tending to stimulate peOple to be interested in politics. discuss politics. have opinions and information on politics. have firm party and candidate preferences. and vote. Evidence that one medium can stimlate interest in news from another medium has been reported by Lazarsfeld. He found that a rising interest in news transmitted by radio during the decade of the thirties was associated by and large with more newspaper reading than ever be- fore. The best hypothesis. he concluded. was that "listening to radio news increases newspaper reading" (19%. p. 276). In a study of the“ diffusion of major news stories. Deutscknunn and Danielson (1958) also! ; v ”...-— i found some evidence that radio "whats the appetite for news." Early in the development of television as a mass medium. Riley. Cantrell and Ruttiger (19119) reported that television was stimulating new interest in sports. Some evidence of television influence on children's interests emerged in a study by Deutschmann and HoNelly (1958). Attitude test 11 responses by message-exposed children in both field and experimental situations indicated greater interest in the message tOpic. h—H elec- trical clubs. than was the case for unexposed children. Lerner's study (1958) already has been mentioned as pointing to a relation between media exposure and concern with the outside world. "Hoderns" in the Middle East tended to recall international news more than persons who still belonged to the old traditional cul- ture or who were in the process of transition to the new (p. 98). Consistent with this is a finding by Hirabayshi and E1 Khatib (1959) in a study of Egyptian villagers. They reported that use of the mass media. superimposed upon traditional word-of-mouth coamunication. was strongly associated with the emergence of the villagers as a nationally conscious group. In a rural Italian commune. HacLean and Pinna (1958b) found‘j that persons with high interest in distant places and low interest in “(a _ places close to home tended to be younger. more educated. more traveled .Ev, and heavier users of the mass media. - .} Most Americans would be emected to have a greater psychological "life-space" or news range than people in less developed parts of the world. Swanson (1955. pp. 415-1316) found. in analyzing data from readership studies of 130 0.8. newspapers. that local news "does not offer a magic touchstone for reader interest." Local. domestic and Washington. D. 0.. items accounted for 90.9 per cent of all items and 88.5 per cent of total readership - indicating that "psychological proximity' is more important than mere geographical proximity." Swanson suggested. for example. that war news might have stimulated interest in Africa. 12 Hyman and Sheatsley (1947. p. 1415) reported "consistent evi- dence that interest in foreign affairs tends to be generalized. Some people are interested in many or all of the issues: another large group is apathetic toward most or all of them." They found that people who were interested in a particular foreign topic tended more than others to have been emosed to news about it. . "It is probable that the two factors 33 interdependent; as people learn more. their in- terest increases. and as their interest increases they are impelled to learn more" (p. 416). The famous campaign to make Cincinnati conscious of the United Nations and world affairs provided an opportunity to investigate media effects on interest (Star and Hughes. 1950). The campaign was con- ducted through the media . civic organizations. and distribution of .literature and films. It was found that international events "assumed greater importance in the minds of the people" after the campaign than bafore. The investigators attributed this not to the campaign itself but to alarming news about the "cold war" with Russia. On the basis of evidence on the diffusion of major news stories (Deutschmann and Danielson. 1960) it seems likely that the big news breaks on the cold war reached most of the people directly through the media. Animal: [‘9 f the 11.. metastases relayed -tbl‘oufihpersonalnhannels preamble—m- 92!“ th" «magnesium. lmyshhthenediao- Thus news stories would appear to have boosted interest in the cold war. The Cincinnati study indicated that the informational campaign itself reached "those who were least in need of it" and failed to in- terest the apathetic. The investigators blamed the failure of the 13 campaign committee to take into account the principle. "still new to students of public opinion." of the bearing of interest on learning. "The Cincinnati experiment has proved that the creating of interest is the first measure in building public opinion and that only after that will.information.be absorbed" (Star and Hughes. 1950; cf. Hyman and Sheatsley. 1947. p. l‘15). ‘Iet. while communication researchers continue amassing moun- tains of data on Opinion.and attitude change. the amount of evidence available on media effects on interests remains unimpressive. Some editors. on the basis of their workaday impressions and sometimes their private readership surveys. are Operating on the general principle of familiarity breeding interest in foreign.news CHcBelly. 1959). For both.practitioners and researchers. this principle has been.mainly a -matter of speculation. Qgggrglizations. [Despite the scarcity of solid evidence. some general statements about media effects on.interests are available in the literature. Schramm already has been quoted in a previous chapter as holding the view that public tastes shape the media. though he acknow- ledges some conditioning of tastes in the direction of media offerings. Ilapper (1957-58) takes a similarly cautious stand after reviewing the existing evidence: "It has been long known that the media do not seem to determine tastes. but rather to be used in accordance with tastes . otherwise determined." However. in.describing mass communication effects as nediated through reference groups and personality needs. he concedes that "the media serve as contributing agents of reinforcement" (P0 “67). ll} Wiebe (1955) goes a little further. He sees mass comunication as tending to "broaden the areas of counon interest" and to "increase conformity" (p. 167 ). Cultural leadership via mass communications tends. by its very nature. to be moderate and mom-evolutionary. It can "nudge" and "hasten" and "encourage" tendencies for which the audience members are ready. It can initiate mild change or reform so long as such changes do not jeOpardize deep-seated values or biases or so log)“ these changes appear in familiar and satisfying content 13. 1 e Though cultural leadership by the media must be slow and remain close to existing tastes. says Wiebe. its cumulative effect can be "substantial" (p. 185). In connection with their finding of a relationship between media exposure and political interest. Berelson. Lazarsfeld and HoPhee comment (1951!. pp. 245-248): All this relates to an old argument in connection with com- munication. There are those who think it "does no good" to make information widely accessible in which people are not basically interested. If this view prevailed. however. perhaps only a minority of Americans would vote. and even fewer would know. even momentarily. what they were voting for. On the other hand. there are those who hold the equally unrealistic view that if only Americans could be sufficiently flooded with mass-media propaganda. good citizenship could be "sold" like toothpaste. These views are neither right nor wrong: the effective audience for politics today is a mixture of the minority who have reasons for learning about current political events and the majority who do not but who do learn something -- "because it's there." These researchers see the potential for a "spiral build-up" in the fact that "mass-media exposure affects mediating variables like partisanship. interest. and discussion that in turn lead back to mass- media exposure." And they quote the saying. "The appetite grows by what it feeds on." The notion of a spiral build-up of interests is an appealing one from either a common-sense or a theoretical point of view. It is l5 consistent with the scattered research findings cited in this chapter. But few researchers would claim the evidence is conclusive. It is difficult to sort out media cause or effect from a complex of environ- mental processes (cf. Schranm. l9h9h: Lemar. 1958. p. 398; Lipset gt al.. 195‘s p. 1158). I. MENTION-SEEKING Studies of immediate and specific interests might be more fruit- ful if they were based on some theory of the underlying origins of interests in general. Secondary Motivation. Our knowledge of the foundations of human interests has been characterized by Asch (1952. p. 300) as "meager." He sees them as arising along with the development of the processes of perception and cognition: "The capacity to see is at the same time an interest to see. . . . Because we have a capacity for understanding. we have a need and an interest to see into the character of things." And Asch supports the widespread assumption. mentioned in the previous chapter. that as we learn about something we get more in- terested in it: "As knowledge of the surroundings grows. activity am interest gain an increasing reference to them." Hilgard (1956. p. lI67) has noted a recent emphasis in psychol- ogy upon "curiosity" as a motive for learning. Kata and Sarnoff (1951*) assume a need on the part of every person to know about the external , world. "Often this quest for information seems to be a curiosity which is entirely self-contained." Cohen. Stotland and Wolfe (1955) found experimental evidence that a need for cognition may exist as a measur- able characteristic of the organin and that it may operate indepen- dently of other needs. .5 16 In an emperiment with monkeys. Harlow‘and associates (1950) founi evidence for some sort of "exploratory-investigative drive" and wrote that "the curiosity motive frequently attributed to human in- fants and children can also be characterized this way." The experi- menters suggested that drives of this kirxi may be as basic and strong as the homeostatic drives. and as efficient in facilitating learning. "The homeostatic drives have been such convenient drives for the comparative-theoretical psychologist that the potentialities of other drives have been ignored and their role in behavior has been neglected and belittled." Osgood (1953) has noted that most observed behavior. especially in human adults. is not motivated by primary biological drives like hunger. thirst and sex. Secondary drives and reinforcement are handled by Osgood in terms of the self-stimulation released by anticipatory goal reactions. These secondary mechanisms can be learned or extin- guished under specifiable conditions (pp. hMB). A modern adult. of course. can acquire an immense range of secondary motivations. These can be categorized. subdivided and labeled in various ways. Some communication researchers have posited news-seeking as a)! secondary drive. involving a need to know what is going on in the environment (cf. Danielson and Stempel. 1958; Westley and Barrow. 1959). m. There are many possible ways of describing secondary reinforcement. or reward. in connection with the seeking of informa- tion. Schramm's (1918a) concepts of "immediate" and "delayed" reward have been fonmd useful by some communication researchers. Imediate 1? reward news is conceived as rewarding the reader at once. giving him a vicarious experience without the dangers or stresses involved .- e.g. . news of crime. accidents and disasters. sports. recreation. social events. and human interest. The payoff for delayed reward news is seen. on the other hand. as steaming not from the initial reading ex- perience -- it may even alarm or annoy the reader at first -- but from the sense of acquiring information which will be of use in the future. News of public affairs. economics . social problems . science. education and health are placed in this category. The positive relationship found between education and readership of this type of news is held to reflect the learning of interests. is the reader's understanding broadens. Schraam explains. "he perceives more of the causative and repetitive relations of events in society. And thus he substitutes other stimuli for the stimulus which has had a proved reward. and as his horizon broadens comes to see more and more reason for reading news of public affairs" (p. 262). This kind of learning was described by Deway (1913) in terms of . dichoto-w which bears considerable resemblance to the immediate— ' delayed reward notion. He wrote of interests which are direct. as contrasted with "indirect. transferred. or technically speaking. mediated interests. Things indifferent or even repulsive in themselves often become of interest because of assuming relationships and con- nections of which we were previously unaware." As an example. he cited the student who "found mathematical theory. once repellent. lit up by great attractiveness after studying some form of engineering in which this theory was a necessary tool" (p. 22). 18 mi Deny pointed out that direct interest in an object can pass into indirect interest. and vice-versa. His example of the latter is the missr who first seeos gold as the means to an end and later be- comes interested in gold for itself (p. 38). What may be of direct interest to one person may be of indirect interest to another. ‘ So it is with news. Schran- acknowledged that the boundaries between his two classes of news are not stable. The same news story may offer an inediate reward to one reader and a delayed reward to another. And the reader may learn to get both kinds of rewards from the same story. For example. a story read first for public affairs information may take on human interest (in a newsroom sense) as the reader ‘ gets to know about the personalities involved; or a story read first for human interest may come to be seen as offering useful public : affairs information (of. discussion of "consumtory" and "instrumental" comunication in Berlo. 1960. pp. 17-20). The means-end relationship implied by "delayed reward" or "indirect interest" is a significant one for information-seeking. Dewey expressed it simply: "Anything indifferent or repellent becomes of interest when seen as a means to an end already comanding attention" ' (1913. p. 25). This comes close to a concept of attitude structure which is gaining increasing attention (Peak. 1955: Rosenberg. 1956. 1960; Carlson. 1956; Peak and Morrison. 1958: Kata. 196a). Attitmies are seen as composed of cognitive as well as affective characteristics. That is. the direction and strength of an attitude depend upon (1) the strength of an individual 's relevant values or goals and (2) the per- ceived instrnentality of the referent object in attaining or blocking 19 these values or goals. Experimental work has indicated that attitudes can be changed by changing either of these characteristics. This theory of attitude structure can readily be applied to Dewey's case of the engineering student who acquired an.interest in mathematical theory. It can be applied to interest in mass media con- tent. as Carlson.has done in an unpublished study of perceived instru- mentality of news stories. Perceived instrumentality clearly is in- volved in.media efforts to relate their offerings to values. goals or ambitions of their audiences. The notion of affective-cognitive consistency is included in this structural theory of attitudes. A change in.obaect-related values. or in the object's perceived instrumentality as regards these values. is expected to produce an inconsistency between the affective and cog- nitive portions of an attitude. This. in turn. is expected to result in attitude reorganization in which the inconsistency is reduced or eliminated (Rosenberg. 1960). Kate notes that the "problems of cognitive organisation.and the attempts of people to achieve unity and consistency in their own mental worlds" is receiving heavy emphasis in.present-day psychology (1960a. p. i). Osgood. in a recent review of the theories in this area by Raider. Newcomb. Pestinger. and Osgood and Tannenbaum. distills this commonly-accepted principle: "Cognitive modification results from.the psychological stress produced by cognitive inconsistencies" (1960. p. ans; cf. Rokeach. 1960. p. 395). One way of reducing such inconsistencies is to seek new infor- nation (cf. Feetinger. 1957: Osgood. 1960). Thus we have still another approach to reward for informationpseeking -- the reduction of stress ‘ . a ‘9 ‘ . U A . l a ’ . L b: l, ,». . . (I ,e-‘ o . .r‘t ~,, " .l ‘ (I I \" ’ _ -. .4 I” re . ~..- "- I ' J . . JV L t . t , i n x e ‘ I .4 ~ ’ . 1 e e ‘ n e § f' t' '1 £4- 8‘ :‘1 e‘ V . .. ‘._ . .,’~ ‘. . .}\“‘ k.’ “' . ~ -._,. . 1. e? v g) a a a 's .l. f .. t 1 I. , x , .. e.- x.’ 20 or tension caused by cognitive inconsistency. dissonance. incongruity or imbalance (of. Berlo. 1960. pp. 92-97). MEANING AND ATTITUDE INTENSITY DI order to hold out an promise of a reward -- however reward is conceived - a message first must have meaning for the receiver. a news topic which has little or no meaning for a reader will attract little or no attention. Such is the case. as many editors are well aware. with many topics in foreign news. The processes of acquiring and measuring meaning are central to the present stucv. a very large prOportion of the verbal signs used in conunica- tion get their meaning not via direct association with the objects represented but w association with other signs (Osgood. 1953. pp. 697-698). This is particularly true of news communication. in which the source cannot point out the object directly to the receivers. News is full of what Osgood calls m .. their meaning being literally "assigned" to them by association with other signs. Harv news topics. particularly in foreign news. can be termed W. -- signs of assigns. The development of sign-assigns is a complex process. Traced back in Osgood "s terms it involves first the association of various signs with objects. a portion of the behavior originally elicited by an object comes to be elicited by the sign as a representational mediation process. which is the meaning of the sign. Then a number of signs become associated with a new sign. an assign. Portions of the representational behaviors associated with the set of signs transfer to the assign as a "distilled" representational process. which becomes 21 the meaning of this assign (p. 697). A number of assigns later become associated with a sign-assign. and so on. Sign-assigns can be highly abstract. far removed from direct contact with the senses. Appropriate- ly enough for our present purposes. Osgood cites. as an example of a sign-assign. the word "alien." which "has acquired its unsavory signifi- cance through association with assigns which have themselves only a mediate connection with objects and experiences" (p. 698). The formation of signs. assigns and sign-assigns. and the modi- fication of their meanings. can be observed impressionistically in everyday life. It can be observed in the laboratory under conditions ranging from simple stimulus-response to complex word-association studies. Stats and Staats (1959) . for example. have performed experiments in which words have been given meaning by associating them with ptvsical stimuli and with other words. Werner and Kaplan (1950) have studied the development in children of meaning for nonsense assigns which are embedded in meaningful sentences. Using a somewhat similar method. Dodge associated nonsense words with familiar words in the context of stories written in the style of the national Geogaphic again; (Osgood £41.. 1957. pp. 287-289). The develOpment of assign meaniixs. as Osgood and his associates noted. is "replete with tantalizing pos- sibilities for experimental research" (P. 289). Osgood's semantic differential has been useful in such research (e.g.. in the Staats and Dodge studies). Along with its underlying assumptions. the instrument seems to offer a fruitful approach to the study of the generation of. interest in mass media content. The semantic differential is coordinated with the Osgoodian 22 theory of meaning by assuming that the representational mediation pro- cess is made up of a complex pattern of bipolar reactions. That is. his selection of directions in the semantic space -- toward good or bad. active or passive. etc. - is coordinate with E225 reactions are elicited by the sign he is judging. In addition. the polarization or extremeness of his judgments along the scales is coordinate with the may. of the reactions (Osgood 9231.. 1957. p. 319). Intensity. Operationally defined in a variety of ways. has been studied increasingly in connection with attitudes and opinions (Katz. 19M; Green. 1951;. pp. 357-358; Hartley and Hartley. 1955. p. 670). The same relationship repeatedly has been found between extremeness of Judgment and intensity. For example. Cantril. in an early study (19%) defining intensity in terms of a rating on how strongly the respondent held the attitude. showed that "the more extreme an attitude is in its direction. the more intensely it is likely to be held." 7 Guttmmn and Suchman showed how the U- or'J-shaped curve of intensity. in.relation to extremeness of attitude. could be used to provide an obdective cutting point between favorable and unfavorable items in a scale analysis. This point is the lowest on the intensity curve. indicating neutrality or indifference (Suchman. 1950; Guttman. 195k). Guttman has identified the intensity function with the mathe- matical second component of a perfect scale. because both are U-shaped (195“. p. 229). Green. however. calls this similarity "quite accidenp tal" and does not see it as adding in any way to the meaningfulness of intensity (195“. P0 358). {Such-an reported that the use of a Likert scale (strongly agree. agree. undecided. disagree. strongly disagree) made it possible to 23 measure direction and intensity at the same time with one item: in one case the results fitted almost exactly the results obtained when a separate intensity question was asked. However. additional experiments failed to produce the desired intensity curve. so that "for the present it would seem safer to use separate intensity questions" (1950. p. 253; cf. Green. 1951‘. p. 357). On the other hand. Mehling (1959-60) founl that semantic differential scales could measure both direction and intensity satisfactorily at the same time. He had his subjects go back after they had completed the scales and circle a number from O to 9 to indicate "how strongly you feel the way you checked the item." This procedure produced an intensity curve similar to Cantril's and more pronounced than most of those found by Suchman. Hehling's study also gave added weight to the assumption that the middle interval of semantic differential scales represents the neutral point in attitude. Even more direct evidence is available on the assumed relation- ship between polarization on the semantic differential and intensity of mediating reactions. An experiment was performed using concepts flashed on a screen. to which subjects reacted by throwing a lever in the direction of their judgment of each concept on polar adjectives. Later the subjects reacted to the same items in the usual graphic form of the semantic differential. It was found that latency of judgment with the levers tended to be a function of polarization on the graphic scales. Subjects reacted faster to extremely favorable or unfavorable concepts than to "neutral" concepts. As Osgood and associates put it. the results "generally support our basic theoretical assumption that the extremsness of judgment on the semantic differential is a valid _ 21+ measure of the strength with which signs are associated with repre- sentational processes" (1957- p. 159). Results consistent with these were obtained by Postman and Leytham (1950-51) in an experiment based on the Bruner-Postman "Impoth- esis theory." It was found that words rated by subjects as favorable or unfavorable were recognized faster than neutral words. Faster recognition was attributed to a "stronger hypothesis" about the favor- able and unfavorable words. The eXperiment demonstrated that the motivational support for the hypothesis may be of either a pleasant or unpleasant tone. involving either reward or punishment. Directive- state experments have tended to emphasize the role of positive values or "wish fulfillment" in facilitating perception: this study showed that negative reinforcement also can strengthen hypotheses (of. Allport . ‘1955. pp. 309. 386. 394). Thus either a favorable or an unfavorable attitude toward a word can strengthen the hypothesis. making its arousal more likely -- or. to use Osgood's terms. strengthening the word's association with representational processes within the individual. Or we can say that words toward which we are favorable or unfavorable - toward which we have attitude intensity -- are more 5.9%. than neutral words. When the semantic differential is used. words judged by checking the mid-points of the scales are said to be relatively meaningless. while those judged further out toward the edges of semantic space are said to be relatively "saturated" with meaning (Osgood my. p. 97). The relationship between meaningfulness and the strength of mediating reactions can be demonstrated ptwsiologically. Bingham (1918) 25 found that the galvanic skin response was greater for words rated highly on meaningfulness. significance and importance to the subject than for words with low ratings. Similar results were reported by Cohen. Silverman and Shmovonian (1959) in a study using "charged" - (meaningful) and "neutral" words. Staats and Staats (1959) used the galvanic skin response in one of their experiments in which words acquired meaning by association with meaningful stimuli. They found a significant correlation between the intensity of a GSR conditioned to a word (by pairing the word with noxious stimulation) and the in- tensity of the "conditioned negative evaluative meaning response" to the word as measured by the semantic differential. Considered together. then. the latency and 6511 studies indicate that speed and intensity of perception are closely related to meaning- fulness (as indexed by attitude intensity or semantic differential polarization). An experiment by Deutsclnann (1956) with headline-type stimuli suggested a similar relationship between expressed interest and perception (his predictions were based on hypothesis theory. with an index of interest used to provide information on hypothesis strength). If meaningfulness and expressed interest are related to per- ception in similar fashion. are they related to each other? Factor analytic work with the semantic differential indicates that judgments of meaningfulness and interest are highly related .. on a receptivity factor (Osgood £31.. pp. 181-187). Some implications of this relationship. in the light of the foregoing theoretical and anpirical considerations. will be developed in the next chapter. CHAPTEIII W The foregoing theoretical and empirical considerations pro- vide the backgrouxxl for the theoretical rationale of the present in- vestigation. The rationale will be developed in general terms in the first part of this chapter. Then it will be applied specifically to the area of foreign news. and the hypotheses and general plan of in- vestigation will be presented. INTEREST AND MEANINGMNESS Interest has been defined provisionally as a predisposition to pay attention. Some scattered pieces of mlpirical evidence have been cited which lend some support to the notion that mass media messages can influence interest. The underlying basis for interest has been discussed in terms of secondary motivation; and reinforcement. in terms of inediate and delayed reward. perceived instrumentality. and cognitive consistency. It has been suggested that in order to hold out any promise of reward. the message must have some meaning for the receiver. The acquisition of meaning was described in terms of associational pro- cesses. The semantic differential has been used to measure these processes. Evidence from a variety of studies. when considered to- gether. suggests an underlying relationship between polarization on 26 2? the semantic differential. attitude intensity. perceptual facilitation. meaningfulness. and expressed interest. Heaningfulness. as indexed by polarization on semantic dif- ferential scales. appears to be positively related to speed and inten- sity of perception. So does expressed interest. In addition. judg- ments of meaningfulness and interest have been shown to be related on a receptivity factor. It makes intuitive sense that interest. as a predisposition to pay attention. should be related to the strength with which signs are associated with representational process - to speed and intensity of perception. Speed and intensity of perception may help determine the predisposition which assigns priorities of attention for a number of competing stimuli perceived in a particular situation. In addition. as Deutschmann (1956. pp. #6407) has suggested. this pre- disposition called interest may be an unconscious factor intervening between the stimulus and the percept itself. Thus interest may be both a cause and an effect of perceptual facilitation. In either case. interest is a hypothetical construct which intervenes between the stimulus (message) and the response (atten- tion) e MEANING INTEJSIT! Interest canonly has been indexed by researchers in terms of ' the individual's own subjective estimate of his internal state .- ex- pressed on a continuum such as interesting-dull or like-dislike. The validity of such an estimate may be open to question in some cases; the individual may‘not be fully aware of his internal state. or he may overestimate. underestimate or conceal his interests for purposes of 28 prestige or privacy. In connection with the like-dislike continuum there is the additional question. raised in the previous chapter. of whether actual interest may not in some cases accomparv dislike. The relationships suamarized in the preceding paragraphs suggest a possible new approach to the measurement of interest. They suggest that interest may be involved in extremity of attitude. in polarization of judylent .- favorable or unfavorable .- about a concept. Such polarization can be indexed through the use of semantic differential scales on the evaluative dimension. There remain to be considered the two other major dimensions of semantic judgment .. potency and activity (Osgood 3111,... 1957. p. 325). It would seem natural that a concept judged as powerful ani active should arouse intensity of feeling. But what about a concept judged as weak and passive? It can be reasoned that the potency and activity dimensions may operate in a different way - that absolute. rather than polarization. scores on these dimensions may be associated with intensity of feeling. In other words. a concept perceived as strong and active may generate more extreme evaluations in either direction than a concept perceived as weak and passive. For example. one might like or dislike communism with extra intensity -- and be more interested in it - if one perceives it as powerful and active. Thus polarization on evaluative scales. plus absolute scores on potency and activity. may add up to a useful index of what might be called mam intensity. This concept may be a valid summation of much of what is involved in the internal state of readiness to pay 29 attention. It seems possible that an index of meaning intensity might be more successful than expressed interest in predicting behavior in some situations. Furthermore. and most importantly for the present study. an index of meaning intensity can be used to relate the process of acquiring interest directly to the process of acquiring meaning. gmloggent of Meaning Qteggity. The acquisition of meaning has been described in the previous chapter as taking place through associational processes. Assigns. or sign-assigns. acquire meaning for individuals through associations with other signs. Dodge was able to show how previously meaningless concepts acquired direction and inten- sity on smaantic differential scales as a function of the intensity and frequency of association with certain adjectives (Osgood my. 1957. pp. 286-289). In complex messages such as news stories. of course. the concept acquires meaning by association not only with adjectives but also with nouns. verbs and other parts of speech in various syntactical combinations. The total process may be virtually impossible to trace accurately in complete detail in a given case. However. the end result of the process has been shown to be measurable with the semantic dif- ferential. A concept which is relatively unfamiliar to an individual. which has little or no meaning for him. would be expected to be judged near the mid-points on semantic differential scales. This is the area of little or no meaning (Osgood at al.. p. 97)."I This concept can acquire 1"There is evidence to suggest that in some cases. mid-point judgments may represent the resolution of response conflict between the polar extremes (Osgood gt g1” pp. 228-229). For the unfamiliar concepts under consideration here. however. the mid-point judgments would seem more likely to represent lack of meaning. 30 meaning for the individual through association. in the context of a message. with other signs which already have meaning -- 1.9.. some polarization on various dimensions of the individual's semantic space. The meaning acquired for the previously unfamiliar concept should re- sult in some polarization of semantic differential judgments of the concept by the individual. The direction and degree of polarization on the various scales would depend upon the context of the message «— upon the direction and degree of polarization of the words or phrases with which the concept has become associated. An index of meaning intensity should reflect this process. An unfamiliar concept should have little or no meaning intensity - that is. Judgments of the concept should lack polarization on evaluative scales and positive absolute scores on potency and activity. If the concept is associated with other signs which have interest. it should acquire evaluative polarization plus potency and activity. Meaning intensity for a particular topic would be expected to be associated with underlying motivational factors. Favorable or un- favorable effect toward a concept. plus a feeling that it is strong and active. make it possible for the receiver to perceive a variety of pos- sible rewards for attending to a message about that concept. The con- cept may offer an innediate reward in terms of a pleasurable emotional sensation. It may offer a delayed reward in terms of information which would be instrumental toward attaining goals. Through relationships it acquires with meaningful concepts. it may hold out some promise of helping the receiver to achieve cognitive consistency. The relation of such motivation (values. needs. etc.) to perceptual facilitation 31 has been indicated in the previously cited studies based on the hy- pothesis theory. As such rewards become more clearly perceived. they may result in increased meaning intensity. Thus meaning intensity may be a pro- duct not only of the initial associations between the concept and other signs. but also of cognitive interaction following the initial associa- tions. The processes involved in acquiring and modifying meaning in- tensity are not limited to any particular stage in a communication cycle (of. discussion of “phases“ of audience reaction in Carter. 1960). Meaning intensity may be evoked and affected as the receiver initially selects. attends to. and reflects upon media messages. goggtativg Ream. It should he noted that the proposed index of meaning intensity involves principally the gonnotatgg meaning for the concept. The semantic differential was designed to measure connotative meanings. Two persons can have similar denotative meaning for an object (e.g. . an apple in the hand) but different connotative (good-bad. beautiful-ugly. etc.) meanings (Osgood 3331.. 1957. p. 322). A person may have denotative meaning for an object (e.g.. he may be able to define it operationally in terms of location. size. etc.) and still have very little connotative meaning for it -- and very little meaning intensity for it. or interest in it. The distinction between connotative and denotative meaning cannot be carried very far. however. Berle (1960. p. 211) makes the point that “denotative meanings are simply connotative meanings on which we all agree.“ Nearly everybody might agree. in making semantic 32 differential judgments about a particular foreign country. that it is large. rich. active and strong. Are these connotative or denotative Judgments? Clearly. denotative elements can enter into connotative meanings. The point is made in other words by Rokeach (1960. p. 8). who suggests that "every emotion has its cognitive counterpart. and every cognition its emotional counterpart." Connotative and denotative meanings are intermingled in his conception of a belief system as repre- senting "all the beliefs. sets. expectancies. or hypotheses. conscious and unconscious. that a person at a given time accepts as true of the world he lives in” (p. 33). m. The theoretical formulation deve10ped thus far can be summarized in terms of an individual who is exposed to a message about a certain tapic. Let us assume that the topic has for him some meaning intensity (evaluative polarization. potency and activity). This meaning intensity has been acquired by associations. within this message and/or previous messages. between the tapic and other concepts which have meaning for the individual. Meaning intensity makes it possible for the receiver to relate the tOpic to secondary motivational factors (e.g.. immediate or delayed reward. achievement of cognitive consistency). These factors. in turn. affect meaning intensity. And meaning intensity leads to attention. This formulation can be diagrammed in this manner: MESSAGE TOPIC —-—-> MEANING INTENSITI -———-> ATTENTION (e.g. . object. : :fi (Readership. idea. person) v I listening. viewing) (SEONDARI MOTIVATIDN) 33 The message topic and the context in which it is presented can be readily manipulated in studying this process. Meaning intensity can be measured in terms of judgments on the appropriate semantic dif- ferential scales. These Judgments reflect acquired meanings for the tepics. and the effects of secondary motivational factors which may interact with these meanings. The meaning intensity index can be used as an operational definition of interest. the predisposition to pay attention. Its behavioral consequence. the act of attention to present or future messages about the tOpic. can be directly observed in terms of selective readership. listening or viewing. APEICATION TO FOREIGN NEWS It should be possible to apply this line of reasoning to the problem of possible media message effects on interest in relatively unfamiliar foreign news topics. a- ...-.0» W-ascuratemccmmt. of . somethinghendidnnotmknov Q1353; (This definition can be operationalized by asking the receiver We can e newg‘agflgflmessageflwhich iswacceptedwbrthe-re- if the message is so accepted; in practice. of course. such acceptance is assumed in the ordinary case of a story written by a trained news- man.) W 3.993.933ng ~53915.1..1‘19d .as..sub3ect mattercperceivedby the meivemamsiatiasmegmée iheikvnnéerimagaa.913.52%!!! ”n- familiarity can be indexed by judgments on a familiar-unfamiliar scale. The following implications can be drawn from the foregoing theoretical fomlation: 1. W..- A previously unfamiliar topic in a news story is presented in context with signs. or assigns. which are 3h meaningful to the receiver. Through associational processes. as pre- viously'deecribed. the tepic acquires meaning intensity for the re- ceiver - i.e.. evaluative polarization. potency and activity. The receiver is able to perceive possible rewards for seeking information about this topic - e.g.. it may help him in an argument about foreign policy. it may be useful in planning a trip abroad. it may help clear up some mental confusion or conflict about a certain country‘s policies. Such.secondary motivation. through cognitive interaction. may increase meaning intensity by making the topic seem.better. worse. stronger. or more active. The individual now should be more predisposed to pay attention to this tepic than he was before exposure to the message. . 2. W. hpressed interest. as well as meaning intensity. would be expected to be low for unfamiliar foreign.news topics. .18 such tapics acquire meaning intensity'for the individual. it is to be expected that his subjective estimate of interest in th. also would tend to rise. The relationship might not be one-to-one: ex- pressed interest might overestimate or underestimate the internal state which.meaning intensity is intended to measure. However. within the limited and relatively low range of familiarity with which this study is concerned. some positive relationship is to be expected between meaning intensity and expressed interest. 3. W. Evidence from a number of studies of reading behavior indicates a positive relationship between antecedent self-expressed interest and actual.media behavior'(waples. 1953; Hyman and Sheatsley. l9h7; Swanson. 1951). This relationship. however. "is not always easy to discover and may be confounded by other factors such 35 as physical or temporal availability of the media message. the emphasis given to the message by the media themselves. and difficulties in measuring interests" (Deutschmann. 1956. p. 20; cf. Naples and Tyler. 1931. p. xx). It was suggested earlier in this chapter that the meaning in- tensity measure of interest may be a more satisfactory predictor of behavior than self-expressions of interest have been. But even if meaning intensity were a perfect index of the internal predisposition to pay attention. prediction still would be far from perfect. is March and Simon have succinctly put it: "The behavior of an organism through a short interval of time is to be accounted for by (1) its internal state at the beginning of the interval. and (2) its environment at the beginning of the interval" (1958. p. 9). Part of the internal state may be evoked under certain environmental conditions. but part of the internal state may not be evoked and therefore not exert aw influence on behavior at the) time (pp. 9-10). 30 it is with the internal state called interest; it may or may not be evoked in a particular situation. Naples (1953) has suggested that accessibility - e.g.. space accorded an item in a given newspaper. or availability of books -- is perhaps the most important influence upon actual reading. Next he places readability. He suggests that "accessibility and readability in combination virtually determine what the general reader reads" (p. 351). Thus if these two factors could be held constant. the correla- tion between subject interest and actual readership should be higher than otherwise (p. 33h). This point can be made in terms of Schramm's "Fraction of 36 Selection" of mass communication offerings (1951+. p. 19): Mtation of reward Effort required If accessibility and readability can be regarded as the major deter- minants of the amount of effort required to read a media message. then holding them constant amounts to an attempt to hold constant the de- nominator of this fraction. The numerator. expectation of reward. can be thought of as interest. as indexed by meaning intensity (the rela- tionship of which to secondary rewards has been indicated). in increase in meaning intensity. then. should increase the Fraction of Selection if the denominator is held constant. It is not difficult to think of mass media situations in which this denominator actually is fairly constant. An example is the common case in which the newspaper reader is confronted with two or more head- lines of roughly equivalent prominence and readability. A headline - or a particular word in it -- serves as a cue. or an index. which can determine the probability of the reader's attending to the news story underneath (Schrann. 1951!. p. 13: Tannenbaum. 1955). The crucial cue may be simply a word or phrase referring to a particular topic in the news; the rest of the headline may be of a fairly standardized nature. with little or no intrinsic interest aside from the crucial word or phrase. For example. one headline might announce a. conference on agronow while another tells of a scheduled speech on marriage and the family. The individual may make his reading decision in such cases primarily on the basis of his interest in the particular topic. It should be possible to reproduce this type of situation under experimental conditions -- to confront the reader with headlines 37 of equal accessibility and roughly equivalent readability. containing crucial cues in the form of references to news tOpics. Choices made . in such a situation could be used to test for effects of interest. as indexed by meaning intensity. upon reading behavior. 1}. Loafing and Retention. If attitude can be defined as a readiness or predisposition to respond to an object in some fashion (of. mport. 1954. p. 1+5). then meaning intensity. as a predisposition to pay attention. can be classed as a variety of attitude. Doob (1947) points out that almost everyone agrees that attitudes are learned. “If this is so. then the learning. retention and decline of an attitude are no different from the learning of a skill. a piece of prose. or a set of nonsense syllables . . ." (p. 135). The acquisition of meaning intensity for a news topic. then. would be eXpected to follow a learning curve. Osgood and associates (1957) postulate |'an ordinary growth curve of learning" for the develop- ment of assigns. "The terminal degree of polarization of the assign on. any one dimension of space will be a negatively accelerated positive function of the frequency of association between sign and assign and a linear function of the initial degree of polarization of the sign" (p. 288). Direct experimental evidence of this negatively accelerated function is not cited by these investigators. though they do cite evidence of a positive relationship between number of exposures and degree of polarization (pp. 288-289; cf. pp. 208-209). In a news context. this negatively accelerated function would be expected to operate only over a limited range. Eventually. through contextual associations. an assign (or sign-assign) may acquire about 38 as much meaning intensity as is possessed by the signs with which it becomes associated -- and perhaps even more than some. Such could be the case with some tapics in the news; for persons to whom such tOpics are familiar. repeated exposures might not increase meaning intensity at all. In fact. if a tepic of high meaning intensity becomes repeated- ly associated with other tapics (or adjectives. verbs. etc.) of lower meaning intensity. it may actually lose some of its meaning intensity. Predictions in such situations could be made on the basis of congruity theory (Osgood 3131.. 1957. pp. 199-216). The present investigation is limited to testing for increases in meaning intensity for news topics of low initial familiarity. If the acquisition of meaning intensity for a news tepio follows a learning curve. its retention should be subject to a curve of for- getting. In the case of relatively unfamiliar foreign news tOpics. initial exposure could well be followed by a period of many days without additional earposure through the mass media or personal conversation. The forgetting curve of barely or partially learned material begins with a sharp drop and gradually flattens out (Woodward and Schlosberg. 1951:. p. 786; Osgood. 1953. p. 551+). HIPOTHESES The foregoing considerations provided the basis for the follow- ing hypotheses of the present investigation: 1. Readership of a news story about a previously unfamiliar topic will result in an increase in meaning intensity for that tepic - as indexed by evaluative polarization. potency and activity. 39 2. Accompaming an increase in meaning intensity for the news topic will be an increase in expressed interest in the topic. 3. Accompanying an increase in meaning intensity for the news tepic will be an increase in the probability of selec- tive readership of further news about the tepic -- this probability to be indexed by headline choices. Meaning intensity acquired for the news tOpic would be expected to increase. in the initial stages. as a negatively accelerated function of the number of exposures to stories about the topic. Following ex- posure. meaning intensity would be expected to decrease as a negatively accelerated function of time. finpirical learning and retention curves. even for such relatively simple subject matter as nonsense syllables. vary greatly depending in large part on the particular experimental and computational procedures used (Osgood. 1953. pp. 330. 557-558). For such complex materials as news stories. the empirical curves could be affected by a profuse variety of factors. However. data of an explora- tory nature relating to learning and retention curves will be considered in this study. The general plan of this study was to expose subjects. under experimental conditions. to news stories about previously unfamiliar foreign tepics and then to test for the twpothesized effects. Semantic differential scales were prepared to measure meaning intensity and expressed interest. A forced-choice headline test was used as an index of the probability of selective readership of further news about each topic. Two experiments were designed. with variations which made it 1 40 possible to obtain supplementary data on learning and retention. The methodology will be explained in detail in the next chapter. CHAPTER IV METHOD As was pointed out in Chapter II. it is difficult in field studies. which have produced most of the evidence of media influences on interests. to sort out media cause and effect from other environ- mental processes. Laboratory studies. on the other hand. are con- fronted with the problem of generalizability to real-life situations. The aim in the present study was to test the hypotheses under controlled conditions without sacrificing some essential components of an actual media behavior situation. I The methodology of this study will be described in terms of (l) pro-testing. (2) stimulus materials. (3) testing instruments. (‘4) subjects. (5) procedure. and (6) statistical analysis. PBS-TESTED Pro-testing for this study was done in three phases. First. the names of eight countries were used as concepts in a semantic dif- ferential test administered to twenty-one Michigan State University students. On the basis of the findings. an expanded list of foreip news topics was drawn up to be judged on a refined list of scales by titenty additional university students . The third pha so was a pilot study in which factual news stories. the semantic differential. a headline-choice test. and a brief questionnaire were used; the subjects were twenty-nine high school students attending an institute at Michigan 1&1 1+2 State University. In each phase. some informal interviewing was done with subjects in order to check on their understanding of and reactions to the materials. Analysis of the pilot study results led to various modifications in design. stimulus and testing materials. and procedures for the final two experiments. STIMULUS MATERIALS In studies of communications effects it often has been founi convenient to use variations of a single message on a single topic. While this solves some design and control problems. it may limit the extent to which generalizations can be made about different messages on different tepics (cf. Hovland. 1951;. p. 1080). It was decided. therefore. in the present study. to use news stories on eight dif- ferent foreign news topics. The topics were genuine. The stories were written in normal newspaper fashion. leaning toward the interpretive. Story organization and stylistic factors were allowed to vary according to the require- ments of the subject matter. An effort was made throughout to maintain a high level of authenticity. All of the stories were based on factual infomation from newspaper or magazine articles. Two stories were taken verbatim from newspapers -- one in its entirety and the other with some paragraphs deleted. None of the stories was more than 200 words long. The stories were presented in printed form. with headlines. on narrow separate sheets as advance proofs for a Sunday page on world affairs. They are reproduced in the appendix. The tapics are listed ‘43 below. along with a brief explanation of the story or stories about each : l. 2. 3. it. 5. 6. 7. 8. _B_t_1_u_t_a_r_1. Three versions. A. B and C. which are merely I'rewrites'l of the same factual material (as in case of several news agency or newspaper stories on same event). No significant differences among versions in reading ease ratings by pilot study subjects. Breederpggg. (Three versions. A. B and C. equivalent as above in factual content and reading ease. Tehuantgpgg Gang}. . Two versions. identical except for a difference in the final paragraph in degree of favorability of evaluation of proposed canal. Version A high in favor- able evaluation; version B. low. W. Two versions. identical except for potency of tapic in final paragraph. Topic pictured as strong in version A. weak in version B. m. Two versions. almost identical except for activity. Version A pictures Libya as active; version B. passive. Convention People '3 Party of 995m .. Two versions. dif- fering in evaluative direction. Version A emphasizes positive evaluation of topic; B. negative evaluation. mtg, One version. a verbatim cepy of a United Press International dispatch from the Lansing (Mich.) M Journal. W. One version. a verbatim cepy of a New log me dispatch except for the deletion of some paragraphs for the sake of brevity. m. In the first eXperiment. the stories were presented to four experimantal groups (to which subjects were randomly assigned) according to the distribution shown in Table 1. TABLE 1 DISTRIBUTION OF STORIES IN FIRST EXPERD’ENT Order Presented Group I Group II Group III Group I! let Malta Malta Mozambique Ho zambique 2nd Bhutan Bhutan Broederbond Breederbond (A) (B) (A) (A) 3rd Canal Canal Libya Libya (high eval.) (low eval.) (passive) (active) nth Bhutan Bhutan Brosderbond Breederbond (B) (B or C) (B or C) (B) 5th Haiti Broederbond Bhutan Party (strong) (a. B or c) (a. a or 0) (p08. eval.) 6th Bhutan Haiti Party Brosderbond (C) (weak) (neg. eval.) (C) NOTE: Where alternative versions on Bhutan or Breederbond are listed (e.g.. A. B or C). the respective versions are distributed equally and randomly within the group. This design gave every subject six stories to read. Each group served as an experimental group for the topics to which it was exposed. and as a control group for the topics to which it was not exposed. Thus. in effect. several experiments were being carried on with the same subjects. but with different stories. fatigue. etc.. with which they approached the post-test would be eXpected Each group went through a similar reading eXperience. The general frame of reference. degree of #5 to be rather similar. Differences found between experimental and con. trol groups would be attributable to exposure or non-exposure to partic- ular stories - rather than to a difference in over-all I'set." which might have been produced by the use of a control group receiving no stories at all. The use of four groups in the first experiment made it possible to test for effects of different numbers of eXposures on Bhutan and Breederbond. Table 1 shows that there were four levels of exposure -- O. l. 2 and 3 stories -- on each of these two topics. Table 1 also shows how alternative story versions were admin- istered on the Canal.‘Libya. Government of Haiti. and the Convention Peeple's Party. This made it possible to obtain supplementary data on the components of meaning intensity - evaluative polarization. potency and activity. In the second experiment the stories were presented to two experimental groups as shown in Table 2. It can be seen in Table 2 that all eight topics are involved. with each group reading five stories. Number of exposures was not varied in the second experiment. but delayed tests were given in order to obtain retention data. Table 2 shows also that alternative verb sions were not presented for the Canal.‘Libya. Haiti's government. and the Convention.Pe0ple's Party. TESTING INSTRUMENTS All subjects in both experiments took the same post-test imme- diately after reading the news stories and rating them on reading ease. The testing instruments used were the semantic differential and a forced-choice headline test. 146 TABLEZ DISTRIBUTION OF STORIES IN SECOND EXPERDIENT Order Presented Group I Group II 1st Malta Mozambique 2nd Bhutan Broaderbond (A) (A) 3rd Canal Libya (high eval.) (active) hth Bhutan Breederbond (B or C) (B or 0) 5th Haiti Party (strong) (pos. eval.) NOTE: Where alternative versions on Bhutan and Breederbond are listed (B or C) . the respective versions are distributed equally and randomly within the group. The use of an 'after-only' design. in which testing materials were administered following exposure to the message. offered several advantages in this study over the I'before-after" design (of. 110le 1L5}... 1949. pp. 308-328): (1) The after-only design made it more plausible to present the stimulus and testing materials as part of a “survey” of reading preferences. reducing the likelihood of "guinea- pig'I or 'test' reactions. (2) "Sensitization" to the experimental variables before exposure to the messages was avoided. (3) The pos- sibility of a I'consistency" reaction. a tendency to give the same responses after as before exposure. was avoided. (h) The after-only design appeared to involve less risk of producing fatigue and boredom among subjects. These advantages. of course. were obtained at the 47 I expense of some advantages of the before-after design. which provides a record of each individual's initial and subsequent responses (Hovland et ai... 1919. p. 311»). §emantic‘Differentigl. Each subject was asked to check semantic differential scales on all eight tepics involved in the experiment. plus Canada. Canada was used for some comparison purposes. There were three evaluative scales - good-bad. harmful-benefi- cial. and unpleasant-pleasant; a potency scale. strong-weak; and an activity scale. active-passive. For the index of meaning intensity. the polarization (folded-over) scores on the three evaluative scales were combined with the absolute scores on potency and activity. Thus the highest possible meaning intensity score would be 15 -- made up of the highest polarization scores (in either direction) for a total of 9 on evaluation. plus a plus 3 each on potency and activity. The lowest possible score would be a minus 6 - made up of three zeros on evalua- tive polarization plus a minus 3 each on potency and activity. This index. developed in the pre-test and pilot study stages. gives heaviest weight to evaluative polarization; evaluation repeatedly has been found to be the dominant factor in semantic judgments (Osgood et al.. 1957. P0 325)e Combined absolute scores on two scales - interesting-boring and would read about-would ggt’read about - were used as the index of expressed interest. The scores have a possible range from.minus 6 to plus 6. . For the second experiment. one additional scale was added - familiar-unfamiliar - but the indexes of meaning intensity and interest remained the same. The evaluative scales were intermingled with the other scales. The direction of "favorability" on the scales was varied from.right to left (see appendix). 1.3 The reliability and validity of the semantic differential have been demonstrated in a variety of studies (Osgood M... 1957. pp. 125-166. 192). Test-retest reliability coefficients of .85 and more have been reported. In addition to its face validity. in terms of agreement with other kinds of judynents made by subjects. the differ- ential has produced data which agree with outside criteria. such as political preference. The relationships already cited. in Chapter II. between polarization on the differential and latency of judgmental reaction isoffered by Osgood and associates as evidence of the validity of the differential as an index of representational mediation processes (1957. p. 155). 331%.. Following completion of the semantic differential scales. each subject was given a forced-choice headline test. The test. which is reproduced in the appendix. was made up of pairs of printed headlines. The subject was to circle the one from each pair which he would choose to read first if he saw the stories in a newspaper. A headline was written about each topic . in such a manner as to provide little or no intrinsic interest or information aside from the word or phrase referring to the topic. Bach headline appeared four times. paired with a headline about another tepic. The pairings were arranged in such a way as to mini- mize ”competition" between topics to which individuals had been ex- posed through the news stories. For example. the Malta headline was paired with headlines about Mozambique. Libya . the Convention Peeple 's Party. and Broederbond. It can be seen in Table 1 that subjects exposed in the first experiment to the Malta story were not exposed to Mozambique. Libya or the Peeple's Party - and that only one of the two Malta groups 1.9 was exposed to a Brosderbond story. In the second experiment. as Table 2 shows. the group which read about Malta did not receive stories on any of the other four topics. In all. then. there were 16 pairs of headlines on the experimental topics (plus two pairs of headlines on extraneous topics). Since each headline had four chances of being chosen. the possible scores for each tapic ranged from 0 to h. The forced-choice headline test was developed in order to pro- vide an approach. within the limitations of experimental conditions. to an actual reading behavior situation. As was pointed out in the pre- vious chapter. newspaper readers often are confronted with two or more headlines of roughly equivalent prominence and readability. from.which they must choose one to read first. The choice may well be made on the basis of a word or phrase referring to a topic in the news. Ratings of headlines on an "interesting-dull" ani "would read- would not read" scales were used by Deutschmann (1956. pp. 66-69) with results suggesting relatively satisfactory reliability and validity. In a study of pre-testing editorial items for a mass-circulation maga- zine. Haskins (l960a)used ratings of titles on a thermometer-type scale of interest. The scale ranged from "extremely sure not to read" to "extremely'sure to read." The title-ratings were validated in a national survey in terms of actual readership of the editorial items; predictions were within 6 per cent of actual item readership. on the average. Questionnaireg. After completing the semantic differential and headline tests. each subject filled out a brief questionnaire. The questionnaires. reproduced in the appendix. provided demographic infor- mation about the subjects and some indication of their prior level of 50 news-seeking. In the second experiment the questionnaire information made it possible to match.delayed tests with original tests for analysis without obtaining the names of the subjects. SUBJECTS zirst gaperiment. The subjects for the first experiment were 104 high school students attending a two-week Communication Arts In- stitute at Michigan State University in July. 1960. The institute was divided into sections on high school publications. forensics and theatre. The students. largely seniors and juniors. were selected by their schools to attend the institute and in general represented the upper levels of their classes. Upper-level high school students are members of an important target audience for newspapers. Schramm. Lyle and Parker (1960) found that about two-thirds of high school seniors in San Francisco read a newspaper nearly every day. and that newspaper reading increased with intellectual ability. In a large-scale survey in Texas. 99 per cent of high school teenagers said they "read newspapers at some time"; 60 per cent had had school assignments requiring readership of newspapers (Davis and watkins. 1960). Since their interests presumably are in a formative stage. high school students are appropriate subjects for a study of possible media message effects on interests. §gggp§_§xpgzigggt. Subjects for the second experiment were 126 students in psychology classes at Michigan State university in August. 1960. Two of the classes were in introductory psychology; the third. in social psychology. 51 The second experiment provided an Opportunity for a partial replication of the first experiment. with the addition of the delayed test. on.an.older group of subjects. Among them.were adult school teachers taking summer course work. The age range. therefore. while not equaling that of the general newspaper audience. was considerably broader than that of most college groups. PROCEDURE The subjects for both experiments were made available in intact class groups. varying in size. For the first experiment there were six class groups. The experiment was administered to five classes on the same afternoon. in different rooms. and to the sixth class two afternoons later. Responses to a brief supplementary questionnaire administered to the sixth class gave no indication that the two-day delay had seriously affected the subjects' performance; only two of the seventeen subjects reported having heard of the experiment from.students in other classes. There were three intact class groups for the second experi- ment. The main.test. in which complete stimulus and testing materials were presented. was administered to two classes. in different rooms. on one morning: and to the third class two mornings later. The delayed test. in which only the semantic differential and the questionnaire were administered. took place on subsequent days. For the first class the delay was one day; for the second. two days; for the third. four days. Responses to a brief supplementary questionnaire did not inn dicate that outside discussion of the experiment following the original test had seriously affected results on the delayed test; only nineteen 52 subjects reported having talked with classmates about the experiment and of these only three said it affected their reaponses (they did not specify in what way). For both experiments. the instructor in charge of each class turned the classroom over to the experimenter for the duration of the experiment. The experimenter (or. in several cases. an assistant) in. troduced himself as a staff member of the Communications Research Center at Michigan State University and explained'briefly that the purpose was to get the students' personal reactions to some news stories and news topics. (In the second experiment. the subjects were told that the experimenter would return at a later date to present some of the results of the 'survey.') He then passed out the stimulus and testing materials. clipped together with complete instructions (see appendix). The materials were passed out in random order. but the alloca- tions for each class were such as to divide its members into experi- mental groups of equal size. There were four experimental groups for . the first experiment and two experimental groups for the second experi- ment. Only the stimulus materials (news stories) differed from.one experimental group to another: the tests and questionnaires were the same. The instructions on the first page. preceding the 'proofs" of the news stories. told the subject to read each story just as he might if he happened to see it in his newspaper. and then to write on the proof sheet his numerical rating of its level of reading ease (or dif- ficulty). The purpose of this task was to encourage reading of each story but to minimise "guinea pig" or 'test' reactions. The instructions pointed out that this was not an examination. that it was not necessary 53 for the subject to sign his name. and that his honest answers would help achieve the purpose of determining how better to serve readers of various age groups. After completing the reading ease ratings. subjects went on through the semantic differential and headline tests and finally the questionnaire. The entire procedure took about half of a Schminute class period. Two subjects were eliminated from.each experiment because of failure to complete a major portion of the testing materials. The delayed tests for the second experiment were conducted in a similar’manner. The experimenter told the subjects that he would like to take a few minutes to get one more brief look at their reactions to some news topics. Then just the semantic differential test and questionnaires were handed out. clipped together. They were the same for both experimental groups. Return visits were paid by the experimenter to two classrooms of subjects for the first experiment. and to all three classrooms of the second experiment subjects. The visits provided an Opportunity'to present a few of the preliminary findings of the study and. at the same time. to get informal group reactions to the news stories. topics. testing materials. and the study in general. STATISTICAL ANALKSIS Analysis of pro-test and pilot study results indicated that the semantic differential and headline tests. as used in this study. tended to produce data which did not meet two assumptions underlying the use of parametric tests of significance. These assumptions are 5h normality of distribution and homogeneity of variance. Some of the distributions which appeared in the data were markedly skewed. with a few approaching a J-shape and some tending to- ward flatness. Drastic differences in variances were evident from in- spection of the data. Such departures from the assumptions could re- duce the accuracy of probability levels for parametric tests used on the data (Siegel. 1956. p. 20; Lindquist. 1953. pp. 81-83). a third assumption of parametric tests. that of equal inter- vals. may be reasonably well justified for the semantic differential data (Osgood 92.91:." 1957. p. 152). Means as well as medians were com- puted. and the means appeared generally to be a more stable and infor- mative measure of central tendency (cf. Sanders. 1958. pp. 161-163). Because of the non-normality and the heterogeneity of variance. it was decided to make use of nonparametric tests -- theiHannAUhitney U test (Mann.and Whitney. 19“?) and the KruskalAWallis onedway analysis of variance. Compared with the parametric tests (t and F) under conditions where the assumptions for the parametric tests are met. these ranking tests have a power-efficiency approachinz 95.5 per cent. A nonparametric test may be used without any loss of power simply by drawing a somewhat larger 1! (Siegel. 1956. pp. 21. 126. 192-193). For some population dis- tributions. a ranking test may actually be more powerful than a para- metric test (Kruskal and'wallis. 1952. p. 585); there was some sugges- tion from.pilot study analysis. that such.might be the case for some of the data in the present investigation. In both experiments. the ManneWhitney test was used to test for differences in central tendency between two distributions; for more .V‘ .' ..v 55 than two distributions. the Kruskalawallis one-way analysis of variance was used. The standard of statistical significance used in this study was the .05 level. It was decided to use two-tail tests. which. in addition to being more conservative than.one-tai1 tests. provide a basis for statements of significance in the event of results in the unexpected direction. CHAPTER V W The results of this experiment will be reported in order of the hypotheses to which they relate: (1) meaning intensity. (2) ex- pressed interest. (3) headline choices. A subsequent section will present some data relating to number of exposures. Some supplementary analysis of subjects' reactions to the experiment will be presented at the end of the_chapter. CHARACTERISTICS OF SAMPLE The 104 subjects for the first experiment ranged in age from 14 to 18. with a mean of 16.2. Seventy-one per cent were girls. The subjects' self-ratings of their newspaper reading during the school year were similar in general to those of high school students in the San Francisco and Texas studies cited in the previous chapter. Nearly two-thirds of this sample reported reading the front page of a daily newspaper every day during the school.year. Only h per cent reported . never reading the front page of a daily during the school year. A more objective test of prior news-seeking was provided by the subjects' recall of the names of five world news figures -- the heads of state of Russia. Cuba. India. Britain and France. All but 8 per cent of the subjects were able to name two or more of these persons; more than half of the subjects named at least four correctly (aside from spelling). Boys were somewhat ahead of girls on both newspaper reading 56 ‘ 57 and recall of the heads of state. In the main. the subjects of both sexes can.be classed as members of the news-reading public - available at least some of the time for exposure to news stories under natural conditions. 'MEaNING INTENSITY The first hypothesis was tested by comparing the meaning in- tensity CHI) scores for subjects who were exposed to a story on each topic with the scores for those who were not exposed to a story on that tepic. It should be noted that the exposed subjects for each topic did not all receive the same stories on the topic - except in the cases of Malta and Hbzambique (see Table l in the previous chapter). For example. the exposed groups forfiLibya were III and IV; group III got the “passive“ version while group IV got the “active“ version. In the cases of Bhutan and Breederbond there were three exposed groups. but the numbers of stories to which they were exposed differed from group to group. For the present test. groups receiving some exposure on each tepic. whatever the type or degree of exposure. were lumped together to comprise an exposed group. The group or groups which re- ceived no story at all on the topic comprised the unexposed group for that topic. ‘ TheIdistributions of meaning intensity scores for unexposed and exposed groups on one topic. the Tehuantepec Canal. are presented visually in.Figure 1. It can be seen that the distributions for the two groups on this topic. the Tehuantepec Canal. are rather skewed. with sizable variances. The mean for the unexposed group is 2.10; for 58 £232.11 DISTRIBUTION OF MEANING INTDJSITI SCORES FOR TEHUANTEPEC CANAL HI Score Unexposed (n-50) Exposed (n—52) 15 x 1h x 13 xxx 12 x 11 m 10 x xx 9 mom 8 xxx Jason 7 xx *xx " 6 xx :30on . 5 m m 1! xxxx . 3 xxxxx 2 xxxxx xxxxx 1 xxxx xxxx 0 xxxxxxx xx -1 xxxam xx -2 xx .3 xxx .1} x -5 .6 x 59 the exposed group. 6.19. The scores for the two groups combined cover the full possible range of meaning intensity scores from minus 6 to plus 15 (the HI scores being made up of the combined polarisation scores on three evaluative scales plus the absolute scores on the "strong-weak" and I'active--passive" scales). The mean scores on meaning intensity for the unexposed and exposed groups on each topic are given in Table 3. In each case the mean MI for the exposed group is higher than for the uneXposed group. The Mann-Whitney U tests show that the differences are statistically significant for all topics except Malta and Libya. The first hypothesis. then. is supported for six of the eight topics. Readership of a news story about each of these six tepics did result in an increase in meaning intensity for the topic. Dgferences Between Stories. The news stories differed within and among topics in several cases on individual dimensions of meaning intensity .. evaluation. evaluative polarization. potency. and activity. The remainder of this section will deal with some data on effects of these differences. It can be seen in Table 3 that the meaning intensity hypothesis was confirmed for two tapics. Broederbond and the Government of Haiti. the stories about which contained information which would be expected to result in unfavorable evaluation. Inspection of the data showed that the absolute evaluative scores for both of these topics were negative for the exposed groups. and nearly neutral or slightly favor- able for the unexposed groups. Thus. in order to produce increased meaning intensity for a tepic it was not necessary for a news story to produce 'liking' for the topic. 60 TABLE 3 MARI-WHITNEY U TESTS FOR DIFFERENCES ON MEANING INTBISITI BETWEEN EXPOSED AND UNEXPOSED GROUPS Mean Mean Sum of Sum or HI ‘MI Ranks Ranks Tepic Unexp. Exp. Unexp. iExp. U a Bhutan, 1.73 n.2u 981.0 h069.0 630.0 2.61" . (xx-26) (rs-74) Broederbond 1.62 7.12 931.0 h220.0 631.0 2.3hf (II-24) (II-77) Canal 2.10 6.19 1901.5 3351.5 626.5 n.51?* (nw50) (DP52) Haiti 2.65 u.u2 2298.5 3057.5 972.5 2.33‘ (rt-51) (11-52) Libya 3.36 n.u5 2076.5 277n.5 948.5 1.78 (Ix-#7) (11-51) Party 3.62 6.20 197n.5 2975.5 798.5 2.96** (II-1+8) (In-51) Malta 3.2a 4.08 2283.5 2766.5 1058.5 1.32 (II-“9) (rt-51) Mozambique 2.62 n.44 20u8.0 3002.0 872.0 2.59“ (n.u8) (n-52) 'Significant at the .05 level (two-tail test) “Significant at the .01 level (two-tail test) NOTES: The unexposed and exposed groups differ among tepics .. e.g.. the unexposed group for one tepic may be the exposed group for another (See Table 1). Slight differences in total N among topics resulted from skipping by a few subjects of semantic differential scales on particular topics. The a values in this and subsequent tables on the Mann- Whitney U test are slightly "conservative.“ because a correction for ties was not used. A correction for ties slightly increases the value of 2; this effect usually is negligible (Siegel. 1956. p. 125). 61 This was evident also in the results on the Convention Peeple's Party of Ghana. The story for one experimental group emphasized posi- tive evaluation of the tepic . while the story for the other group emphasised negative evaluation. The meaning intensity scores for both groups combined were. as shown in Table 3. significantly higher than for the unexposed groups. As eXpected. the absolute evaluation score for the group exposed to the positive story was in a positive direction. while that for the group exposed to the negative story was in a nega- tive direction. Table 1% shows that the difference in meaning intensity for the groups receiving the positive and negative stories was not statistically significant. TABLE“ MANN-WHITNEY U TEST FOR DIFFERENCE ON MEANING INTENSITY BETWEEN GROUPS EXPOSED TO POSITIVE AND NMTIVE STORIES ON CONVENTION PEOPLE'S PART! OF GHANA Mean HI Sum of Ranks (n) Group 17 (Positive 7.00 637.5 (27) Evaluation) Group III (negati'. 5e29 688e5 (2“) Evaluation) 0 C 259.5 s - 1.21. not significant Absolute evaluation scores were favorable for the remaining five topics in the experiment - Bhutan. Libya . Halts . Hoaambique and the Tehuantepec Canal. There were two versions of the Canal story. identical except for a difference in the final paragraph in degree of evaluative ..., 14-” 62 favorability as attributed to 'astute observers.“ The results were in the opposite direction to that expected; the obtained mean meaning in- tensity score for the group receiving the "low evaluation" story was higher than that for the group which read the "high evaluation" story. However. Table 5 shows that the difference between the two groups was not statistically significant. TABLE 5 MANNJWHITNEY U TEST FOR.DIFFERENCE ON MEANING INTENSITY BETWEEN GROUPS EXPOSED‘TO HIGH AND LOW EVALUATIVE STORIES ON TEHUANTEPEC CANAL Moan HI Sun of Ranks (n) Group I (High Eval.) 5.69 722.0 (26) Group II (Low Eval.) 6.69 656.0 (26) U - 305.0 a - .58. not significant An.attempt to manipulate the potency dimension was made on the two story versions about Haiti's government. The versions were idenp tical except for the final paragraph. The government was pictured as weak in one version. strong in the other. This variation failed. hows ever. to produce the expected judgments on strong-weak scale judgments; in fact. the results were in the Opposite direction. The manifest leaning intensity scores for the group receiving the 'weak' version were higher than for the group receiving the 'strong' version. as Table 6 shows: however. the difference did not reach statistical signif- icance. 63 TABLE 6 MANIQWHITNEI U TEST FOR DIFFERENCE ON MEANING INTENSITY BETWEEN GROUPS EXPOSED TO HIGH AND LOW POTENCI STORIES ON GOVERNMENT OF HAITI Mean HI Sum of Ranks (n) Group I (High Potency) 3.89 735.5 (26) Group II (Low Potency) “.96 642.5 (26) U - 291.5 3 II .85. not significant Activity was varied in the two versions on Libya. One group received a story which stressed activity in Libya. while the story for the other group was nearly identical except that it pictured Libya as passive. The results were in the expected direction on the active- passive scale and on the meaning intensity scores. However. as Table 7 shows. the difference on meaning intensity between the two groups was not significant at the .05 level. TABLE 7 HAM-WHITNEY U TEST FOR DIFFERENCE ON MEANING INTENSITY BETWEEN GROUPS EXPOSED TO HIGH AND LOW ACTIVITY 81'0le 0N LIBIA Mean HI Sum of Banks (n) Group IV (High Activity) 5A0 632.0 (27) Group III (Low Activity) 3.37 69h.0 (21+) U ' 255.0 s - 1.32. not significant 61. The above results show that the individual variations on activ- ity. potency and degree of favorable evaluation. through.different story versions on the same tepics. did not produce significant dif- ferences in meaning intensity. Manipulation of direction of evalua- tion. as in the case of the Convention PeOple's Party of Ghana. appar- ently did produce evaluative judgments in the eXpected directions. Because of limited Na in the individual groups. significance tests were not applied to comparisons between groups on separate components of meaning intensity. Subsequent analysis of expressed interest and headline test data will not be broken down on the basis of different story versions on the same topic. The analysis. as in the test of the meaning inten- sity hypothesis (Table 3). will deal with differences between groups receiving some exposure on each topic. and groups receiving no exposure. EXPRESSED INTEREST Results on expressed interest (E1) for the eight topics are presented in Table 8. (Possible 21 scores. based on the "interesting- boring" and 'would read aboutawould not read about" scales combined. range from.minus 6 to plus 6.) The differences between the unexposed and the exposed groups are in the predicted direction in seven of the eight cases. The differences are significant beyond the .01 level for four tapics -- Breederbond. Libya. Convention.People's Partyw and Mozambique. The other differences are not significant at the .05 level. For Malta the manifest difference is in the Opposite direc- tion to that predicted. though not significantly so. 65 TABLES MANN-WHITNEY U TESTS FOR DIFFERENCES ON EXPRESSED INTEREST BETWEEN EXPOSED AND UNEXPOSED GROUPS Mean Mean Sum of Sun of E1 E1 Ranks Ranks Tepic Unexp. Exp. Unexp. Exp. U 2 Bhutan e76 1050 1173 e5 387605 822e5 1009 (ta-26) (11-74) Breederbond .. 50 2.22 sumo 11307.0 sumo 3 .013" (In-24) (rt-77) Canal -.50 03,-" 2380.0 2873 .0 1105.0 1.31 (nPSO) (n~52) Haiti e51 1483 2381e0 297500 105500 1e79 (nPSI) (n-52) Libya .00 2.01 1820.0 3031.0 692.0 3.60" (n~47) (n-Sl) Party -.69 2.05 18h8.5 3101.5 672.5 3.84“ (rt-48) (11-51) “3 lta 10 06 e 59 2571 a 0 2,479 e 0 1153 o 0 e 66 (nvh9) (n~51> Mozambique -.38 1.63 2020.0 3030.0 euu.0 2.79” (n48) (rt-52) ”Significant at .01 level (two-tail test) It can be seen that the expressed interest difference between unexposed and exposed on Libya is highly significant. although on meaning intensity (Table 3) the difference missed significance at the .05 level. The above results provide only partial confirmation of the second hypothesis. A significant increase in expressed interest accom- panied an increase in meaning intensity for some topics. but not for others. HEADLINE chcss Results on the headline test. as shown in Table 9. were in the predicted direction for all eight topics. The mean scores on headline choices. which for each tepic had a range from 0 to 4. were higher for the exposed group than.for the unexposed group in every case. For only one topic. Haiti. did the difference fail to reach the .05 significance level. For two tepics. Libya and Halts. the headline test differences between exposed and unexposed are highly significant even though the differences on.meaning intensity (Table 3) had not reached the .05 significance level. The data in Table 9 generally support the third hypothesis. Increases in.meaning intensity were accompanied by increases in prob- ability of selective readership of further news about the topics. as indexed by headline choices. NUMBER OF EXPOSURES ‘ The means on.meaning intensity for each level of exposure on Bhutan.and Brosderbond are plotted in Figure 2. One exposure is de- fined as readership of one news story about the topic. (The number of exposures on each tepic for each of the four experimental groups is shown in Table 1 inthe.previous chapter.) The stories on each topic. being "rewrites'l of the same factual material. were rated similarly on reading ease by the subjects in this experiment. as in the pilot study} the ratings averaged around the “somewhat easy“ mark. If. as would be expected meaning intensity increases as a negatively accelerated function of the number of exposures. then it should increase more on the first exposure than on the second exposure. 67 and more on the second than on the third. and so on. This does not specify the precise shape of the learning curve in terms of sharpness of initial rise. area of inflection. etc. Such characteristics depend on a variety of factors. including the index of learning used. the definition of exposure. complexity of material. and amount of reinforce- ment (Cf. 03g00d. 1953s PP. 328 ffe)e TABLE 9 MANN-WRITNEY U TESTS FOR DIFFERENCES ON HEADLINE CHOICES BETWEEN EXPOSED AND UNEXPOSED GROUPS Mean Mean Sum of Sumof Score Score Ranks Ranks TOPICS Unexp. Exp. UneJCp. DIP. U 3 Bhutan 1.52 2.26 1073.5 0282.5 695.5 2.23" (In-27) (...-76) Brosderbond .96 2.63 607.5 0708.5 322.5 5.15" (II-25) (rt-78) Canal 1.00 2.01 2156.0 3200.0 778.0 3.69" (In-52) (In-51) Haiti 2.35 2.76 2053.0 2903.0 1075.0 1.71 (rt-52) (rt-51) Libya 1.67 2.38 2201.0 3155.0 875.0 3.05" (rt-51) (In-52) Party 1.27 2.15 2105.0 3211.0 819.0 3.03” (11-51) (In-52) 11.11;. 1.23 2.20 2085.5 3270.5 707.5 0.23” ' (rt-52) (II-51) Mozambique 1.35 2.08 2207.5 3108.5 881.5 3.01” (11-51) (rt-52) 'Significant at .05 level (two-tail test) "Significant at .01 level (two-tail test) 68 3.239.}. MEANING INTENSITY MEANS 0N BHUTAN AND BROEDERBOND FOR GROUPS WITH 0.1. 2AND3 EXPOSURES 7 - Broederbond Meaning Intensity Number of Exposures It has already been shown.(Table 3) that meaning intensity scores on both Bhutan and Broederbond were significantly higher for subjects receiving some exposure (1. 2 or 3 stories) than for those who were unexposed. Table 10 also shows that those who read just one story had significantly higher meaning intensity for each tepio than those who read none. 69 TABLE 10 HANN-MHITNEI TEST FOR DIFFERENCES ON MEANING INTENSITY BETWEEN NO EXPOSURE AND ONE EXPOSURE Mean Mean Sum of Sum of HI IMI Ranks Ranks Topic Uhexp. 1 Exp. Unexp. 1 Exp. U 3 Bhutan 1.73 3.95 793.0 052.0 182.0 2. 52" (n.26) (n.2h) Broederbond 1.62 6.72 773.0 052.0 127.0 3.06” (II-21+) (In-25) I’Significant at .05 level (two-tail test) ”Significant at .01 level (two-tail test) There are no significant differences in meaning intensity. hows ever. among the remaining levels of exposure -- 1. 2 and 3. Tables 11 and 12 give the results of Kruskal4Wallis onesway analysis of variance tests on these scores. TABLE 11 KRUSKAL-WAILIS ONE-WAY ANALYSIS OF VARIANCE ON MEANING INTDJSITY FOR THREE LEVELS OF EXPOSURE ON BHUTAN 1 EXP. 2 EXP. 3 Exp. Kean 3.95 4.80 3.92 S“. of Rinks 850e0 1058e0 867.0 (n.24) (n926) (n-Zh) H - .67. df - 2; not significant 70 TABLE 12 KRUSKAL-WAILIS ON E-WAI ANALYSIS OF VARIANCE ON MEANING INTDISITI FOR THREE LEVELS OF EXPOSURE'ON BROEDERBOND 1 EXP. 2 Exp. 3 Exp. “can 6e72 6e72 7e85 Sum.of Ranks 935.0 906.5 1161.5 (11-25) (rt-25) (11-27) H ' 103?. df - 2; not significant The fact that there was a significant increase between 0 and 1 exposure on both topics. but no significant increase between 1. 2 and 3 exposures. is suggestive in itself of a negatively accelerated curve within a limited range. That is. the meaning intensity scores would appear,to increase substantially on the first exposure and then to level off. Visually. the points for Bhutan in Figure 2 suggest negatively accelerated increases up through the second exposure and then a de- cline on the third exposure. For'Broederbond the means suggest a leveling off on the second exposure and then a slight rise for the third. The combined means for Bhutan and Brosderbond (connected by broken line) give some suggestion of a negatively accelerated curve. REACTIONS TO EXPERIMENT Some general impressions of subjects' reactions to the experi- ment were obtained during return visits to two classrooms of students who participated. The subjects appeared to perceive the news stories as genuine. understandable and interesting. The procedures involved ... .uYWJEEFNau 71 in the semantic differential and headline tests appeared to have been generally well understood. The informal questions and answers of the subjects were suggestive of a general perception of the experiment as a survey of reader preferences. in which their own candid reactions were obtained. These impressions. while not conclusive in themselves. were supported by some supplementary analysis of data from.the experiment: 1. Meaning intensity and expressed interest scores for Canada furnished some check on the validity of the scores for other tepics. No news stories were presented on Canada. but it was expected that MI and ET for that topic should be relatively high. if subjects' judgments were honest. because of heavy prior exposure through media and interb personal channels. The results confirmed this expectation. The mean IMI score for all subjects on Canada was 8.13; the mean.EI score was 2.88. These were higher (though not. in every case. to a statisti- cally significant degree) than the corresponding means for exposed or unexposed groups on any of the eight topics about which news stories were presented. 2. Subjects' ratings indicated that the news stories were readable at the high school level. None was rated by a majority of readers as being somewhat or extremely difficult; all mean readability iratings were between "extremely easy“ and ”so-so.“ CHAPTER'VI RESULTS OF SECOND EXPERIMENT As in the previous chapter. the results of the experiment will be presented in order of the relevant hypotheses: (1) meaning intensity. (2) eXpressed interest. (3) headline choices. Supplementary data. in- cluding results on retention. will be analyzed in subsequent sections. CHARACTERISTICS OF SAMPLE The age range for the 126 subjects in the second experiment was 18 to #6. with a mean of 2U.1. Sixtyhfour per cent of the subjects were men. Only A per cent of the subjects reported never reading the front page of a daily newspaper during the school year - the same figure as for the high school students in the first experiment. Front page exposure every day was reported by 56 per cent. not quite so high a percentage as that obtained for the high school subjects. The university group was slightly ahead of the high school students in knowledge of news figures; nearly two-thirds of the university sub- jects were able to name at least four heads of state for the five countries listed (Russia. Cuba. India. Great Britain and France). Only'10'per cent of the second experiment subjects were unable to name at least two of the national leaders. JMen tended to be somewhat ahead of women on both newspaper reading and recall of the heads of state. The subjects for this experiment. like those for the previous experiment. can be classed in the main as members of the news-reading 72 73 public ... available at least some of the time for exposure to news stories such as those used in the experiment. MEANING INTENSITY As shown in Table 2. Chapter IV . two experimental groups were used in the second experiment. Number of exposures on Bhutan and Brosderbond did not vary. nor were alternative story versions presented on other topics. An example of the distribution of meaning intensity scores for individual subjects is presented visually in Figure 3. These scores. on the Convention People's Party of Ghana. cover nearly the full possi- ble range of minus 6 to plus 15 (comprised of the combined polarization scores on three evaluative scales plus absolute scores on the I'strong-- weak'I and “active-passive" scales). The mean for the unexposed group is 2.610; for the exposed group. 5.59. The mean meaning intensity scores for all tapics are presented in Table 13.. For each tOpic the results are in the predicted direction: the eXposed group. has a higher mean score on meaning intensity than the unexposed group. The Mann-Whitney tests show that the differences are highly significant in every case but that of Mozambique. The first hypothesis is supported by these data. Readership of news stories about the tepics did result in an increase in meaning intensity for them. Results on Mividual Cownents. The effects of eXposure to the news stories were investigated for each component of meaning in- tensity -- evaluative polarization. potency and activity. This was done by suming each subject's scores on each component 74 £123.22}. DISTRIBUTION OF MEANING INTENSITY SCORES FOR. COWTION PEDPLE 'S PARTY OF GHANA HI Score Unexposed (n-56) Exposed (11-64) 15 x n 5 110 x 13 x x 12 x noocx 11 x 10 x xx 9 xx xxxx 8 xxx m 7 x xxxx 6 xxx momma: 5 m h x mom 3 ram moocxx 2 momma: mo: 1 ma xxxx 0 monsoon xxxx .1 xx -2 xx xxx 6. fl 75 TABLE 13 MANN-WHITNEY U TESTS FOR DIFFERENCES ON MEANING INTEJSITI BENEEN EXPOSED AND UNEXPOSED GROUPS Mean Mean mm of Sun of MI MI Ranks Ranks Tonic Unexp. Exp. Unexp. mp. U a Bhutan .66 0.38 2896.5 0978.5 816.5 5.61“ (n-60) (ii-61) Broederbond .05 6.10 2219.0 5001.0 623.0 6.15" (n-56) (n-6ll) Canal 1.06 6.78 2595.0 5155.0 579.0 6.71” (ts-63) (xx-61) Haiti 1.89 6.23 3105.0 0730.0 1065.0 0.38” (n-6l+) (n-61) Libya 2.00 0.23 2952.0 0026.0 1201.0 3.15“ (11-58) (Ix-64) Party 2.60 5.59 2602.0 0658.0 1006.0 0.13” (n-56) (n-6lt) Malta 2.01 0.30 3369.5 0380.5 1353.5 2.80“ (xi-63) (n-61) Mozambique 1.17 2.33 3230.0 l0273.0 1519.0 1.73 (n-58) (rt-61*) "Significant at .01 level (two-tail test) over all four topics to which he was exposed by news stories. and over all four topics to which he was uneXposed. As shown in Table 2. Chapter IV. subjects in experimental Group I were exposed to Malta. Bhutan. the Tehuantepec Canal and Haiti's government: Group II sub- jects were exposed to Mozambique. Broederbond. Libya and the Convention People's Party of Ghana. versions were not used for evaluation. potency and activity: (Table 2 also shows that alternative story only the Cl? 0‘ .' e'a‘l 76 high evaluation story was used for the Canal. the positive evaluation version for the Party. the strong version for Haiti 's government. and the active version for Libya. It will be recalled that in the first experiment the alternative versions on these topics did not produce significantly different meaning intensity scores.) Table 14 shows that for each group of four topics. the exposed subjects had higher mean combined evaluative polarization scores. (The score for an individual subject could range from 0 to 36 -- the highest score being made up of a polarization score of 3 on each of the three evaluative scales for each of the four topics.) The differences between exposed and unexposed grOUps were shown by the Mann-Whitney tests to be highly significant. TABLE 14 MANN-WEI U TESTS FOR DIFFERENCES ON EVALUATIVE POLARIZATION BETWEEN EXPOSED AND UNEXPOSED GROUPS OVER FOUR TOPICS COMBINED Mean, Mean Sum of Sum of Be POle Be P014 Ranks Ranks TOpics Unexp. Exp. Unexp. EXp. U 3 Malta. Bhutan. 7.05 16.38 5287.0 2588.0 697.0 6.20” Canal. Haiti (xi-61+) (n-6l) Mozambique. 7.05 11.92 0008.5 3050.5 970.5 0.52“ Broederbond. (n-58) (n-6h) 1.0in e an, ”Significant at .01 level (two-tail test) Table 15 shows that the exposed subjects had higher potency scores than did unexposed subjects for each group of four tapics. (The score for an individual subject could range from a minus 12 to a plus 77 12 - being composed of his absolute score on the strong-weak scale for each of the four topics.) The Mann-Whitney tests show that the dif- ferences on potency between exposed and unexposed groups were highly significant. TABLE 15 MANN-WHITNEY U TESTS FOR DIFFERENCES ON POTENCI BETWEEN EXPOSED AND UNEXPOSED GROUPS OVER FOUR TOPICS COMBINED Mean Mean Sum of Sum of Potency Potency Ranks Ranks Topics Unexp. Exp. Unexp. Exp. U z Malta. Bhutan. -l.88 1.28 5028.0 2807.0 956.0 0.92“ Canal. Haiti (n-6h) (n-6l) Mozambique. -.88 2.00 0072.0 3031.0 951.0 0.60“ Broederbond . (Ii-58) (II-64) Libya. Party "Significant at .01 level (two-tail test) The results were similar for the activity component. as shown in Table 16. (Possible score for an individual subject could range from minus 12 to plus 12 -- being comised of his absolute score on the active-passive scale for each of the four tapics.) The difference on activity scores between exposed and unexposed are shown by the Mann-Whitney tests to be highly significant. Rank correlations provided additional evidence of the inter- relationship among the three components of meaning intensity. Means for each of the three components were computed for a single experimental group so that the subjects involved all received the same exposure on each topic. The rank correlation between the evaluative pole rization and potency means on the eight experimental tepics plus Canada for Group I was .66. activity was .83. correlated .70 with each other. the .05 level. The correlation between evaluative polarization and Potency and activity means on the nine topics were TABLE 16 These correlations are significant at MANN-WHITNEY U TESTS FOR DIFFERENCES ON ACTIVITY BETWEEN EXPOSED AND UNEXPOSED GROUPS OVER FOUR TOPICS COMBINED WW 3 Mean Mean Sum of Sum of Activ. Activ. Ranks Ranks Tepics Unexp. Exp. Unexp. mp. U z Malta. Bhutan. .88 0.23 5069.0 2806.0 915.0 5.12“ Canal. Haiti (nu-64) (rt-61) Mozambique. .50 4.39 4604.0 2899.0 819.0 5.32MK Broederbond . (n-58) (n-64) Libya. Party *‘Significant at .01 level (two-tail test) EXPRESSED INTEREST The results on messed interest ratings for the eight topics are shown in Table 17. (The possible range on E1 scores. made up of the combined scores on the ”interesting-boring" and I‘would read about- would not read about" scales. was from minus 6 to plus 6.) For all eight tepics the results are in the predicted direction. Six of the differences beth exposed and unexposed groups are sta- ti stically sigiificant. Party. the .05 significance level was not reached. For Mozambique and the Convention Peeple ' s In the case of Mozambique. the mean EI scores were almost equal 79 for the exposed and unexposed groups. Mozambique also was the only topic for which exposure to a news story did not produce a significant increase in meaning intensity (Table 13). TABLE 17 MANN-WHITNEY U TESTS FOR DIFFERENCES ON EXPRESSED INTEREST BETWEEN] EXPOSED AND UNEXPOSED GROUPS Mean Mean Sum of Sum of El EI Ranks Ranks Topic Unexp. EXp. Unexp. Exp. U z Bhutan .02 2.07 3177.0 0698.0 1097.0 0.22?* (xi-64) (xi—61) Broederbond. -.11 2.30 2506.0 0710.0 950.0 0.03‘* (n-56) (In-64) Canal .71 1.87 3025.5 0320.5 1009.5 2.56* . (n-63) (n-61) Haiti .80 2.23 3039.5 0035.5 1359.5 2.93" (ii-64) (n-61) Libya .97 2.19 3077.0 0026.0 1366.0 2.51“ (rt-58) (n-64) ~ Party .89 1.58 3105.5 0110.5 1509.5 1.28 (n-56) (n-64) lMalta 1.08 2.08 3301.0 0009.0 1285.0 3.183‘ (It-63) (xx-61 lezambique .70 .81. 3595.0 3908.0 1828.0 .10 (n~58) (ne64) I"Significant at .05 level (two-tail test) "Significant at .01 level (two-tail test) The second ”pothesis is supported by the results on six of the eight topics. Increases in meaning intensity were accompanied by significant increases in expressed interest in these tapics. 80 Further evidence of the relationship between meaning intensity and expressed interest is a rank correlation of .83 (p<.05) between MI and EI means on the eight experimental tOpics plus Canada for ex- perimntal Group I (second experiment). HEADLINE CHOICES Results on the headline test are presented in Table 18. Sub- jects chose headlines more frequently about the tepics on which they had read news stories than about topics on which they were unexposed by news stories. Only in the case of the Convention Peeple's Party did the difference between the uneXposed and exposed groups fail to reach at least .05 significance. (Since subjects had four opportunities to choose a headline on each topic. in preference to a headline on another topic. the possible headline score range was from 0 to 4.) The Convention PeOple 's Party was one of the two tepics for which exposure to a news story also had failed to produce significantly greater expressed interest (Table 17). For Mozambique. on the other hand. the headline score difference between unexposed and exposed groups was highly significant. though the required significance levels had not been reached for either meaning intensity or expressed interest on that topic. For the other six topics. differences between exposed and unexposed groups were statistically significant. in the hypothesized directions. on all three tests - meaning intensity. expressed interest. and headline choices. These included two tOpics. Broederbond and Haiti. for which exposure to the news stories produced dislike (nega- tive scores on the evaluative scales) on the part of most subjects. The headline test results support the third hypothesis. In- creases in meaning intensity were accompanied by increases in 81 probability of selective readership of further news about the topics -- as indexed by headline choices. TABLE 18 MANN-WEETNEI U TESTS FOR DIMCES ON HEADLINE CI'DICES BETWEEN EXPOSED AND UNEIG’OSED GROUPS Tapic Unexp. hp. Unexp. mp. U 3 Bhutan A 1.28 2.25 3310.5 0690.5 1165.5 0.12” (n-65) (xi-61) Broederbond 1.07 2.78 2598.0 5003.0 907.0 5.383* (n-61) (ti-65) Canal 1.02 2.89 2986.0 5015.0 801.0 5.70“ (nw65) (nw61) Haiti 2.17 2.66 3721.5 0279.5 1576.5 2.00? (n~65) (n-61) ' Libya 1.80 2.62 3200.5 0800.5 1509.5 2.37‘ (n-6l) (n-65) Party 1.01 2.11 3303.0 0698.0 1612.0 1.85 (Ix-61) (n—65) Malta 1.18 2.51 3061.5 4939.5 916.5 5.30" (rt-65) (n-61) Mozambique 1.33 2.17 3153.0 0808.0 1062.0 2.61?‘ (n-61) (II-65) ‘Signiricant at .05 level (two-tail test) "Significant at .01 level (two-tail test) Further evidence of the predictive power of meaning intensity in relation to headline choices is a rank correlation of .95 (P< .01) found between MI and headline test means on the eight topics for experi- mental Group I (second experiment). Expressed interest means for the 82 same group on the same topics correlated .83 (p<.05) with the headline test means. RETENTION Was increased meaning intensity measurable only innnediately following exPosure to a news story about a tOpic. or did it endure over a period of days? Results of a delayed test. shown in Table 19. show significant differences between exposed and unexposed groups on meaning intensity over four tOpics combined. or four days following the original experiment. TABLE 19 The delayed test was administered one. two MANN-WHITNEY U TESTS FOR DIFFERENCES ON MEANING INTENSITI BETWEEN EXPOSED AND UNEXPOSED GROUPS OVER FOUR TOPICS COMBINED IN DELAYED TEST Mean Mean Sum of Sum of MI MI Ranks Ranks Topics Unexp. Exp. Unexp. Exp. U 2 Malta. Bhutan. 8.71 15.11 3060.0 2396.0 965.0 2.51? Canal. Haiti (n-51) (0'53) Mozambique. 8.06 13.90 3078.0 2073.0 707.0 3.59** Broederbond . (n-50) (n-51) Libya. Party *Significant at .05 level (two-tail test) ”Significant at .01 level (two-tail test) I Table 20e expected direction for all eight topics. Results for individual topics on the delayed test are given in The differences in means on meaning intensity are in the For Broederbond . the 83 Tehuantepec Canal and the Convention People '3 Party the differences are highly significant. Mann-Whitney test do not reach the .05 significance level. TABLE 20 For the other five twice the z-scores from the MANN-MIITNEY U TESTS FOR DIFFERENCES ON MEANING INTENSITY ESTHER] EXPOSED AND UNEKPOSED GROUPS IN DELAYED TEST Mean Mean Sum of Sun of MI MI Ranks Ranks Topic Unexp. hp. Unexp. Exp. 0 5 Bhutan 1e22 2030 2141705 304205 109105 1069 (In-51) (In-53) Broederbond 1.146 £1.69 20111.0 3110.0 766.0 3A6" (rm-50) (II-51) Canal 2.25 5.34 2130.5 3329.5 809.5 3.56" (n-51) (21.53) Haiti 2.29 3.70 2M7.5 3012.5 1121.5 1.50 (13.51) (II-53) Libya 2.148 3.06 235200 279900 107700 103“ (II-50) (rt-51) Party 2.62 n.u5 2088.0 3063.0 813.0 3.1u’* (II-50) (In-51) Halts 2.98 3.71; 2705.0 2755.0 13211.0 .18 ' (II-51) (10.53) Mozambique 1.50 1.78 2143305 271705 115805 079 (In-50) (xx-51) “Significant at .01 level (two-tail test) The significance levels in the delayed testmuld be expected to be lower than in the original test (Table 13). First. smaller Rs are involved in the delayed test .. some of the original subjects were V v 0. luv 814 not included in the delayed test due to absenteeism or to difficulties in matching delayed with original tests. Second. and more importantly. a forgetting curve would be expected to set in. Table 21 shows that meaning intensity scores. sununed over the four t0pics to which each experimental group was exposed. were signif- icantly higher in the original test than in the delayed test. Thus. acquired meaning intensity did decrease in the period of days following exposure. TABLE 21 MANN-WHITNEY U TEST FOR DIFFERENCES ON MEANING INTENSITY FOR EXPOSED GROUPS BETWEEN ORIGINAL AND DELAYED TF3! OVER FOUR IOPICS COMBINED Mean Mean Sum of Sun of MI MI Ranks Ranks Tapica Orig. D01. Orig. Del. U z Malta. Bhutan. 20.92 15.11 2h21.5 32A9.5 990.5 2.62“ Canal . Haiti (ll-53) (ll-53) Mozambique. 18.37 13.90 2320.5 2932.5 99u.5 2.o5* Broederbond. (n-Sl) (n-Sl) Libya . Party *Signiricant at .05 level (two-tail test) "Si¢nirieant at .01 level (two-tail test) The number of days between tests differed for each of the three class groups used in the experiment. Class A received the delayed test one day after the original test; for Class B the delay was two days; and for Class C. four days. Tables 22 and 23 show that variations in number of days between original smears and the delayed test did not produce significant 85 differences in meaning intensity -- using the combined scores for the tepics to which each eXperimental group was exposed. TABLE 22 KRUSKAL—WALLIS ONE—WAY ANALYSIS OF VARIANCE ON MEANING INTENSITY AMONG CLASS GROUPS FOR.FOUR EXPOSED TOPICS COMBINED IN DELAYED TEST-—-EXPERIMENTAL GROUP I Class A Class B Class C (lpday delay) (2-day delay) (fl-day delay) Hean 13.13 28.25 12.00 Sun or Ranks 87200 1160‘) M300 (n931) (n-8) (Dela) H - 2.9”; df - 2: not significant TABLE 23 KRUSKAL-WAILIS ONE-WAY ANALYSIS OF VARIANCE ON MEANIN} INTENSITY AMONG CLASS GROUPS FOR FOUR EXPOSED TOPICS COMBINED IN DELAYED TEST -- EXPERIMENTAL GROUP II Class A Class B Class C (lpday'delay) (ZLday delay) (h—day delay) Mean 15.08 16.13 11.06 Sum of Ranks 617.0 170.5 534.5 (n-26) (n98) (Del?) H I 2.6#: df - 2; not significant Thus. though there was a significant over-all difference in meaning intensity scores between the original and delayed tests (Table 21). the length of delay did not produce significant differences (Tables 22 and 23). This would suggest that meaning intensity may 86 tend to level off after an initial drop. The means for the original and delayed tests for each experi- mental group are plotted in.Figure #. The combined means for both experimental groups (broken line) do not give a fully consistent picture. The high score for Group I in the two-day delay test (Class B) makes the combined mean for that test higher than that for the orig— inal test. This could easily be a chance inconsistency. however. due to the great variability among individual subjects' scores within groups and the fact that Class B was the smallest of the three classes (the Group I mean for that class being based on an.N of 8). In summary. the combined scores of the delayed test are lower than the combined scores of the original test (Table 21). The fact that differences in number of days of delay did not result in signifi- cant differences in meaning intensity (Tables 22 and 23) suggests a possible tendency to level off after an initial decline. This would be consistent with the expectation of a negatively accelerated down- ward curve of forgetting for meaning intensity. FAMILIARITI The use of a 'familiarbunfamiliar' scale in the semantic dif- ferential test for this eXperiment made it possible to check on the assumption of unfamiliarity of topics. Table 2# gives the mean scores on the familiarity scale for the unexposed groups on each of the eight topics. The familiarity means are all negative: that is. for each tepic. the average subject checked closer to the “unfamiliar" than to the 'famdliar' side of the scale. 0n the other hand. the mean.for all [iggge ’4 MEANING INTENSITY PEANS ON FOUR EXPOSED TOPICS COMBINED IN ORIGINAL AND DELAYED TESTS 30 r Group I 25 ‘ /\ . // \ 20 k / \\ “93111-38 \\ [/Combined \\ . \ Intensity \ // \\ / \ / \ \ 15 F Group II \ . \ \ \. \ \ 10 " 5 L- 0 I l l J 1 2 3 4 Number of Days Delay 88 subjects on Canada. an obviously familiar tepic on which no news story was presented. was a positive 1.16. TABLE 24 MEANS 0N FAMILIARITI SCALE FUR GIDUPS UNEXPOSED T0 STORIES ON EACH TOPIC o lo M Bhutan -1.78 Broederbond -l.76 Canal al.89 Haiti -l.47 Libya .. .72 Party -l.50 Malta - .63 Mozambique -1.36 Thus. the topics about which news stories were presented in this experiment were previously unfamiliar to the subjects. REACTIONS TO THE EXPERIMENT Subjects' informal questions and answers during the return visits by the experimenter to each of the three classrooms gave some indication. as in the previous experiment with high school students. of a general perception of the news stories as genuine. understandable and interesting. and of the study itself as an authentic survey of reader preferences. Relevant to these impressions were some supplementary data from the experiment: 1. Meaning intensity ani eXpressed interest for Canada -- 89 which with its familiarity and physical proximity would be expected to be relatively high if subjects were reacting candidly - were higher than for am of the tapics on which stories were presented. The MI mean for all subjects on Canada in the original test was 7.91. Since meaning intensity for Canada was developed over a long period of time and not by an emperimental news story. it would be expected to remain stable. rather than to decline. between the original and delayed tests. This expectation was borne out; the MI mean on Canada in the delayed test was 8.28. very close to the mean in the original test. 2. As in the previous experiment. mean readability ratings of the news stories by the subjects were between 'extremely'easy' and “so-so.' None of the stories was rated as difficult reading by the university subjects. CHAPTEI VII D USS The first section of this chapter will deal with theoretical implications of the results of the two eXperiments. In the second section. practical and social implications will be discussed. Pos- sible lines of further inquiry will be suggested in the final section. TiflDRETICAL IMPLICATIONS This study was designed to study possible media message effects on audience interests. specifically in the area of foreign news. The underlying basis for interest. as a readiness to pay attention. was discussed in terms of secondary motivation. Reinforcement was discussed in terms of immediate and delayed reward. perceived instrumentality. and cagnitive consistency. It was suggested that in order to hold out aw promise of reward. a message must have some meaning for the receiver. The acquisition of meaning was described in terms of associational pro- cesses resulting in the develOpment of what Osgood calls I'assigns" or 'sign-assigns." Evidence from a variety of studies was cited which. when considered together. suggested an underlying relationship among meaningfulness of assigns. polarization on the semantic differential. attitude intensity. perceptual facilitation. and eXpressed interest. In the light of these studies. an index of meaning intensity was constructed as a new approach to the measurement of interest. The 90 91 index was made up of polarization on evaluative scales of the semantic differential. plus absolute scores on potency and activity scales. The use of this index made it possible to relate the acquisition of interest to the development of meaning for assigns. According to the theoretical formulation. a previously unfa- miliar topic could acquire meaning intensity through associations. within the context of a news story. with other concepts which already had meaning for the reader. This meaning intensity would make it pos- sible for the reader to relate the topic to secondary motivational factors which in turn might affect meaning intensity. It was reasoned that the acquisition of meaning intensity for a previously unfamiliar topic should be accompanied by an increase in the reader's subjective estimate of his interest in it. More impor- tantly. increased meaning intensity should be accompanied by a higher probability of further selective self-exposure to news about the topic. The two experiments provided support for the theoretical scheme. The results. in general. were as hypothesized: (l) mtposure to news stories about the previously unfamiliar foreign topics resulted in in- creased meaning intensity for them on the part of the experimental subjects. (2) Increases in meaning intensity were accompanied by in- creases in expressed interest. (3) Increases in meaning intensity were accompanied by increases in the probability. as indexed by head- line choices. of further readership about the topics. The results supported the use of the meaning intensity index as an Operational definition of the internal state of readiness to pay attention. Support for the inclusion of evaluative (polarization. potency and activity in the index was provided by the data on these .a 92 individual components in the second eXperinent. The data on meaning intensity. expressed interest and headline choices clearly were incompatible with the notion. widespread in psy- chological and media research literature. that interest is equivalent to liking. It was found possible to develop interest along with -- in large part as a consequence of -- either favorable or unfavorable evaluation. The evidence supported the assumption. involved in the meaning intensity index. of a curvilinear relationship between judgments of evaluation and judgments of potency and activity. That is. strong feelings of either like or dislike toward a tepic were associated with feelings that it was strong and active. as well as interesting. Confirmation of the hypothesis of this study does not imply un- qualified support for the notion. alluded to in the first chapter. that knowledge breeds interest. For the previously unfamiliar topics used in the experiments. increases in knowledge resulting from readership of single news stories clearly were associated with increases in in- terest. But this does not justify the conclusion that further increases in knowledge about the tOpics would always result in further increases in interest. Repetition can weaken (or extinguish) as well as strength- en habits (Osgood. 1953. p. 328). One may tire of reading about a particular topic in the news. even when new information is presented about it. Berelson. Lazarsfeld and McPhee (1951+) saw this as a limiting factor in the ”spiral build-up" of interest in politics: There is probably a built-in check of satiation - a kind of psychological regression effect whereby more and more political 93 exposure at one point becomes too much and the more interest builds up. the harder it becomes to build it further. In these ‘ways. the spiral effect of exposure and interest on each other must be dampened . . . (p. 2h8). The I'satiation" or "dampening" effect may be explained in terms of decreases in reinforcement. Additional information about a political issue or personality may offer the reader progressively less in the way of immediate or delayed reward. of serving his goals. of achieving a more balanced picture of the world. And such secondary .motivational.factors may even be reduced by overbexposure to the tepic. The reader may become apathetic about it; he may have less meaning in- tensity for it. This process could be expected to set in after multiple exe posures to the tapics used in this study. Since the tOpics were pre- viously unfamiliar to the subjects. satiation was unlikely at first unless - as may have been the case for some individuals - there was a satiation with news in general. For the exposed groups as a whole in this study. meaning intensity'for’a topic increased on one exposure to a news story about it. The data on number of exposures to Bhutan and Broederbond (Figure 2) suggested negatively accelerated increases over a limited range. If data had been collected for’more than three eXposures. a decline might have set in: this might be the explanation for the lower mean obtained for Bhutan on the third exposure. It should be noted. however. that the second and third exposures on Bhutan and Broederbond involved no new factual material; the succes- sive stories were rewrites of the same information. .Exposure to two or more rewrites of the same news can occur in a natural situation (e.g.. reading more than one newspaper or hearing successive broadcast 94 "rehashes' of the same event). but it probably is more common for successive exposures to provide additional news about a topic. In the latter case. the meaning intensity curve would be expected to continue longer on the rise. Even then. the rise could not be expected to continue indefi- nitely. The reader's meaning intensity for the tepic eventually may become as great as for most of the concepts with which the topic is likely to be associated in successive messages. Leveling off. or stabilization. would then be expected. Such would appear to be the case for Canada. a tOpic to which the subjects of these experiments must have been exposed thousands of times in news stories and through personal contacts or experiences. The meaning intensity means for Canada (see pp. 71 and 89). while higher than those for the experi- mental tOpics. were well below the upper possible limit. Exposure to sensational news about Canada might raise its meaning intensity meas- urably for these subjects. and extremely dull news might conceivably reduce it. but general stability would seem likely in the normal course of events. Some evidence of such stability was found (p. 89). In sumnary. for previously unfamiliar topics such as those used in (this study. the evidence does lend support to the notion that knowledge breeds interest. As tepics achieve familiarity. additional knowledge may produce successively smaller increments of interest: eventually the level of interest may become relatively stable. with possibilities of further increases or declines resulting from excep- tional messages. PRACTICAL AND SOCIAL IMPLICATIONS Aw generalizations from this investigation to real-life 95 situations.must be hedged with the usual qualifications applying to controlled experiments. .A laboratory'situation.cannot precisely re- produce natural.conditions. Even if a subject responds to a message in the laboratory in approximately the way he would have responded in his home. it is possible that he might not be exposed to such a mes- sage in his heme in the ordinary course of events. Controlled experi- ments thus tend to over-estimate the effects of communication; the self-selection of audiences may cancel out effects found in the labora- tory (Rovland. 195“. p. 1158). On the other hand. several aspects of this study can be comp sidered as justifying some degree of generalization to situations out- side the laboratory. A variety of news stories was used in the experi- ment. The stories were realistic and based on factual material; two of them were taken directly from newspapers. Results followed a close- ly similar pattern over the two experiments. embracing subjects of both sexes and a relatively large age range. The subjects in both experiments can be classed generally as members of the news-reading public. available at least same of the time for exposure to news stories such as those used experimentally. It seems highly probable that results similar*to these pro- duced in the experiments could be produced under natural conditions. Increases in meaning intensity and expressed interest. and changes in reading behavior. might be small and difficult to measure. The large variability among subjects on.meaning intensity in this study (see Figures 1 and 3) indicates that there would be many individual excep- tions to the general trend of results in the field. Numerous social and personality factors could interact to influence and obscure the 96 results. Yet. as was suggested in the introductory chapter. the elu- siveness and gradualness of media effects does not make them less im- portant. ngggg.ww “tenets—frondayimfldaxngte in long-range changes of great significancg‘tojhe media‘and...toq.aocie~t;. —~—- “....- “IO‘MN—a‘u ‘ a ,, For unfamiliar topics such as those used in this study. the evidence indicates that measurable increments of interest can result from readership of a single news story. The evidence also indicates that some of this acquired interest can survive at least over a period of several days following exposure. Additional exposure during the period of partial retention might well sustain or increase the interest previously generated. Such may be the effects produced by series of news or inter- pretive articles on unfamiliar tapics which editors feel deserve atten- tion. Series of this kind. printed in the New York Tgeg during the early sumner of 1960. provided the Mozambique story and the information for the Bhutan stories in the present stuck. It should be emphasized that the effects produced in these ex.- periments should not necessarily be expected for news topics which already are familiar to readers. Readers may already have high meaning intensity for such tepics. Additional exposure may result in the 'satiation' or stabilization effects discussed previously. Are any implications to be drawn from this study as to. what kinds of news stories may be most effective in generating interest in unfamiliar topics? A general principle is implied by the theoretical rationale for the acquisition of meaning intensity. It is that a topic will acquire more meaning intensity (or interest) to the extent that it is associated through the context of the message with other 97 signs or assigns which themselves have more meaning intensity for the reader -- to the extent that he feels they are good or bad. strong and active. The ways in which this general principle can be applied depend. of course. upon the special circumstances involved in the case of the individual news tepic. In this study. Broederbond. which ranked high in acquired meaning intensity for exposed subjects. happened to be associated in the news story context with concepts of presumably high meaning intensity such as 'anti—Catholic.‘ I'i‘anatic" and "anti-Jewish.” There remains a question. however. about the extent to which editors can engage in generating new interests without running the risk of alienating their readers. This type of risk often is cited w media practitioners who are accused of pandering to low public tastes. Iet. while no realistic editor would attempt a wholesale refor- mation of his readers' tastes. the media do appear to have some depee of latitude within which they can vary their offerings without undue risk of losing their public. 9 To some extent people seem to accept passively what is put in front of them; Berelson found that for may people. newspaper reading was a 'ceremonial or ritualistic or near- compulsive act for may people" (1919. p. 129). A researcher for a mass-circulation magazine reports that certain "basic editorial mixes'I can assure the editor of a certain satisfaction threshold for his audience. with a fair amount of "free" space remaining within which the editor can exercise “creativity and social responsibility" (Haskins. 19606. Control of space in his publication gives the editor some measure of control over what peOple read. Waples assigns a pro-endnent role to the space factor in determining newspaper reading behavior 98 (1953-! PP- 318-351). Thus some Icreative editing" to influence audience interests l would appear to be possible without undue economic risk. As Max Lerner (1957. p. 767) has put it. “there is at any time a frame of permis- siveness within which the technicians of the Big~Audience media can work either creatively or destructively. either to degrade and cor— rupt what they find or to evoke its potential.‘ In the latter spirit. some editors. as was pointed out in the first chapter. are actively engaged in efforts to enlarge their readers' ‘news range" in the international field. They are using such editorial 'tools as headlines. display and “local angles" to win readership for foreign news (McNelly. 1959). The results of the present study lend support to the position that such efforts can.bear fruit - that public demand to some extent can be shaped. or conditioned. by messages supplied by the media. SUGGESTIONS FOR.FURTHER INQUIRY Further research seems warranted in the area of media effects on interests. which are basic to the informational and persuasive aspects of mass communication. Some possible lines of inquiry are these: 1. Correlational analysis might lead to refinement of the meaning intensity index in terms of components and weights. Such analysis could be based on semantic differential judgments of a large number of news tepics. selected from newspapers by a sampling method. It might be possible to improve predictions to reading behavior'by varying weights assigned to components in order to produce special " 99 indices for use with certain categories of news topics (e.g.. persons. countries. events. policies. etc.). 2. Studies could be made of content. stylistic and structural variables in news stories which may affect meaning intensity. Numerous control and interpretation problems would arise in such studies. In the first experiment of the present study. attempted manipulations of evaluative polarization. potency and activity failed to produce signif- icant differences in meaning intensity for individual tepic s. The effects of individual variables may be offset. or neutralized. in the complex interactions among lexical and syntactical elements by which meaning intensity is acquired. 3. A problem which might be encountered in using the meaning intensity index for familiar tepics is that of the meaning of mid- point judgments on the semantic differential (see footnote. p. 29). With the unfamiliar tepics used in the present study. the assumption that judgments around the mid-point represent little or no meaning seems justified; but mid-point judgments on familiar topics might be worth further scrutiny. In an experimental study of oral communica- tion. Bettinghaus reported finding some support for ”the notion that neutral attitudes may be intense ones. and not easily changed' (1959. p. 72). Cases of this nature might be indicated by mid-point judguents on evaluative scales accompanied by judpents of high potency and activity. ~ I - hr. A special study might be made of individuals like some in the present investigation whose meaning intensity for a topic remains at the zero point or below following exposure to a news story about the topic. These “deviant cases" might differ from others on on ... ‘V‘ 0.. 100 demographic. personality or sociometric variables. 5. Perhaps the most fruitful type of further research in this area can be done under field-experimental conditions. Split-run tech- niques could be used to test effects of different kinds of newspaper stories in natural settings. with the self-selection factor controlled. It would be possible in a study extending over a period of time to study the effect of news story exposure on meaning intensity (or inter- est). and the effect of acquired meaning intensity on further reading bOh8V1OI‘e CHAPTER VIII §Qflfl§fil AND GONQLUSIQES Two exPeriments were conducted in an investigation of possible media effects on audience interests. specifically in the area of foreign news. . Backgrgund.” .Some observers and practitioners have taken the position that such effects are possible. that public demand can be shaped or conditioned by what the media supply. Others. including some re- searchers. have tended toward the opposite position. In much of the psychological and media research literature. in- terest has been equated with '11king.‘ In the present study. interest was conceived as a readiness to pay attention. whatever the direction of affect. Some scattered research evidence was cited relating to media influence on interests. 1222.2- The theoretical scheme for this study centered around the development of meaning for what Osgood calls "assigns“ or "sign- assigns.‘ In the light of a number of studies involving the semantic differential. attitude intensity. perceptual facilitation. and expressed interest. an index of meaning intensity was constructed. The index was composed of polarization on evaluative scales of the semantic differ- ential. plus absolute scores on potency and activity scales. Use of this index made it possible to relate the acquisition of interest to assign development. 101 102 m. In order to test hypotheses based on the theoretical scheme. experimental subjects were exposed to news stories about fore eign topics of low familiarity. Eight topics were used. Testing materials included semantic differential scales and a forced-choice headline test. Subjects for the first experiment were 10# high school students: for the second. 126 university summer session students of a large age range. Egsultg.~ The results. in general. were as hypothesized: (1) Exposure to news stories about the tapics resulted in increased meaning intensity for them. (2) Increases in meaning intensity were accomp panied by increases in expressed interest. (3) Increases in.meaning intensity were accompanied by increases in the probability. as indexed by headline choices. of further readership about the topics. Exposure to news stories resulted in the hypothesized effects for both favorably and unfavorably evaluated tepics. Attempted manipulations of evaluative polarization. potency and activity in individual story versions in the first experiment failed to produce significant differences in meaning intensity. How-- ever. separate analysis of data on these three components in the second experiment provided support for their inclusion in the meaning inten- sity index. Data on number of exposures and retention were somewhat sug- gestive of negatively accelerated curves of learning and forgetting for meaning intensity. The retention data indicated that. while forgetting did set in. some acquired meaning intensity was retained in the days immediately following exposure to the news stories. Conclusions. Some theoretical. practical and social implications 103 were drawn from the results. Suggestions were made for further re— search in this area. The following conclusions are based on the findings of this investigation: 1. Interest. in the sense of readiness to pay attention. can be acquired for previously unfamiliar tapics through exposure to news stories. 2. The index of meaning intensity used in these experiments can be a useful measure of interest. 3. Meaning intensity acquired by exposure to news stories is not merely a momentary phenomenon: it was found to endure to a mea sur- able extent. and thus be a potential influence on reading behavior. in the days immediately following exposure. 4. The evidence suggests that meaning intensity acquired on one exposure to a news story about an unfamiliar topic may tend to level off. or stabilize. as the topic gains familiarity on exposure to subsequent news stories. 5. I'Liking" is not a necessary condition for interest (readi— ness to pay attention). Either favorable or unfavorable evaluation may be associated with the generation of interest in a news tepic. 6. Further research along the lines of this study should be carried out through field eaqaerimentation. in order to obtain evidence on media message effects on interests in natural settings. REFERENCES Allport. F. H. Theories of perception and the concept of structure. New York: Wiley. 1955. Allport. G. H. The historical background of modern social psycholog. In Lindzey. G. (ed.). Handbook of social ppychologz. Vol. I. Reading. Mass.: Addison-Wesley. 1951+. Pp. 3-56. Asch. S. E. Social psychology. New York: Prentice-Hall. 1952. Aspinall. A. Polipics and thg press mpg-$559. London: Home and Van T1181. 19%. Berelson. B. What 'missing the newspaper' means. In Lazarsfeld. P. F.. and Stanton. F. N. (eds.). Communications research l2’48-l232. New York: Harper. 1949. Pp. 111-129. Berelson. B. R.. Lazarsfeld. P. F.. and McPhee. W. N. 221.jpg. Chicago: University of Chicago Press. 1954. Berle. D. K. The process of communicatiop. New York: Holt. Rinehart and Winston. 1960. Bettinghaus. E. P. Jr. The Operation of congruity in the oral com- munication situation. Unpublished doctoral dissertation. University of Illinois. 1959. Bingham. W. E. A study of the relations which the galvanic skin re- sponse and sensory reference bear to judgments of the meaningfulness. significance. and importance of 72 words. slourna]= of thologz. 19143. lé. 7.1-3“. Cantril. H. The intensity of an attitude. Journal of Abnormal and 30013]. P halo . 19%. £9 129-136. Carlson. E. R. Attitude change and attitude structure. Journa o Abnormal and Social Eacholog. 1956. 1;. 256-261. 10“ 105 Carter. R. P. On reactions to mass media content. Audiovisual Com.- munication Review. 1960. Q. 210-213. Cohen. A. R.. Stotland. E.. and Wolfe. D. M. An eXperimental investi- gation of need for cognition. Journal of Abnormal and Social W. 1955. 51. 291-294. Cohen. S. I.. Silverman. A. J.. and Shmavonian. B. M. The measurement of human adaptation to stressful environments. Genera; astems; yearbook of the Society for Geperal flaws gesearch. 1959. 3. 231- 241. Danielson. W. A.. and G. H. Stempel. News-seeking behavior on two college campuses. Paper presented at Association for Education in Journalism convention. University of Missouri. August. 1958. Davis. N. G.. and Watkins. S. Teen-age newspaper reading. Editor and Publisher} JUly 23’ 1960. pe 13. Deutschmann. P. J. The relationship between interest and perception of headline-type stimuli. Unpublished doctoral dissertation. Stanford University. 1956. Deutschmann. P. J.. and Danielson. W. Diffusion of the major news story. Paper presented before Association for Education in Journal- ism. University of Missouri. August. 1958. Deutschmann. P. J .. and Danielson. W. A. Diffusion of knowledge of the major news story. Journalism Quarterlz.__1960. 31., 395-355. Deutschmann. P. J.. and McNelly. J. T. Audience and effects of the #41 TV Electrical Club series. garter; Bullptin (Michigan Agricultural Experiment Station. Michigan State University). 1958. A. 395-409. Dewey. J. Interefi adp 93:0 gt ip gdpcation. Cambridge. Mass.: River- side Press. 1913. Doob. L. W. The behavior of attitudes. W. 19W. 52. 135.156. Festinger. L. A thpon of cognitivg diggnapcg.,,_ Evanston. 111.: Row. Peterson and Comparv. 1957. Fryer. D. The measurement of into sts. New York: Holt. 1931. Green. B. F. Attitude measurement. In Lindzey. G. (ed.). Handbook of social pmhologz. Reading. Mass.: Addison-Wesley. 1954. PP- 335-369. Guttman. L. The principal components of scalable attitudes. In Lazarsfeld. P. F. (ed.). Mathematical thinking in the social, scienceg. Glencoe. 131.: Free Press. 1951!. Pp. 216-257. a. t O a L a . o 0 ¢ 0 I I o. ( I v. e E e t I e I a a I e o e. a . u n .. 106 Harlow. H. F.. Harlow. M. 1.. and Meyer. D. Learning motivated by a manipulation drive. Journal of Mgrmpta; thologz. 1950. 59. 228-234. Hartley. E. I... and Hartley. R. E. Fundamentals of social pmpologz. New York: Knapf. 1955. Baskins. J. B. Pre-testing editorial items and ideas for reader in- terest. Journalim Quarterypl960a. 31. 2204330. Haskins. J. B. The editorial mix: a holistic approach to the analysis of a publication's effectiveness. Paper delivered at convention of Association for Education in Journalism. Pennsylvania State Univer- sity. August. 1960b. Hilgard. E. R. Theories of 1earni__ng. New York: Appleton-Century- Crofts. 1956. Hirabayashi. G. K.. and El Khatib. M. 1“. Communication and political awareness in the villages of Egypt. Public OpiniOp Quarteru. 1959. £39 357-3530 7 Hovland. C. I. Effects of the mass media of communication. In Lindzey. G. (ed.). Handbook of social cholo . Vol. II. Reading. Mass.: Addison-Wesley. 1951}. Pp. 16g2-1103. 30"th C. Io. Lmaineg A. A.. and Sheffield. F. D. W on mass conununication. Princeton. N.J.: Princeton University Press. 1949. Hyman. H. 3.. and Sheatsley. P. B. Some reasons why information campaigns fail. Public Opinion garterll.” 191W. _l;._. 1:13-1:23. James. W. The principles of psycholog.~ ,Vol. II. New York: Holt. 1890. Kata. D. The measurement of intensity. In H. Cantril (ed.). M public Owen.” Princeton. N. J. : Princeton University Press. 19Me Pp. 51-65. Kata. D. Preface. Public Opinion Quartem.,, -1960a. 3.1!.- i-iii. Kata. D. The functional approach to the study of attitudes. Public; Opipion ngrterlz. 7,1960b. 311. 163-20“. Kata. D.. and Sarnoff. I. Motivational bases of attitude change. 10ml of “norms; gpd Social tholog. .1951}. 32. 115-12“. flapper. J. T. What we know about the effects of mass coqunication: the brink of hope. blic o o ,_1957-,-58.. g1. “SB-#71». Krech. D.. and Crutchfield. R. S. Iheon and problems of social cholo .‘ Res York: McGraw-Hill. 191+8. 107 Kruskal. W. B.. and Wallis. W. A. Use of ranks in one-criterion variance analysis. Journal of the American Statistical Assogiatigp. 1952. 51. 583-621. Lane. R. E. Political life; El people get involved in mlitics. Glencoe. 111.: The Free Press. 1959. Lazarsfeld. P. F. Radio and the printed page. New York: Duell. Sloan and Pearce. 19140. Lazarsfeld. P. F. Audience research. In Berelson. B.. and Janowitz. M. (eds.) Reader in public Opinion app pomnunicationt, Glencoe. 111.: Free Press. 1953. Pp. 337-346. Lerner. D. The Passing of Traditional Socigty; ModegLizing the Middle East. Glencoe. Ill.: Free Press. 1958. Lerner. M. America as a civilization. New York: Simon and Schuster. 1957. Lindquist . E. F. Desi n and anal sis o e ent he 0 education, Boston: Roughton Mifflin. 1953. ‘ LiPBCt. Se He. Lazarefeld. Fe Fe. Barton. ‘0 Ho. 8m 111112». J. Th9 psychology of voting; an analysis of political behavior. In Lindzey. G. (ed.). Handbook 0; social pmhologx. _,Vol. II. Reading. Mass.: Addison-Wesley. 1954. Pp. 11214-1175. MacLean. M. 3.. and Pinna. L. Distance and news interest: Scarperim. Italy. Journalism QparterlyLl958a. 31. 36-198. MacLean.. M. S.. and L. Pinna. News interest. movement. and mass media use of the Scarperiesi. Paper presented at Association for Education in Journalism convention. University of Missouri. August. 1958b. Mann. H. B.. and Whitney. D. R. On a test of whether one or two random variables is stochastical];r larger than the other. M}; of Mathematical Statistics.__l9l+7. l_§. 50-60. March. J. G.‘. and Simon. R. A. Organizations. New York: Wiley. 1958. McNelly. J. T. Press responsibility and readership of foreign news. Qazgt‘te. .1959. 5,. 225-233- Mehling. R. A simple test for measuring intensity of attitudes. Public Opinion Qu_arterlz._ 1959-60. 31. pp. 576-578. Nafziger. R. 0.. MacLean. M.. and Engstron. W. Useful tools for inter- preting readership data. Journalism Quarterly._l951. Q. tum-#56. Osgood. C. E. Method and theog in egerimental pmholog. New York: Oxford University Press. 1953. 108 Osgood. C. E. COgnitive dynamics in the conduct of human affairs. Public Opigon Quarterly. .1960. 2.3. 341-365. Osgood. C. E.. moi. G. J.. and Tannenbaum. P. H. The Measurement of WW Urbana. 111.: University of Illinois Press. 1957. Peak. H. Attitude and motivation. In Jones. M. (ed.). Nebraska sium on motivation l ... , Lincoln. Neb.: University of Nebraska Press. 1955. Pp. 149-189. Peak. H.. and Morrison. H. W. The acceptance of information into attitude structure. Journal of Abnormal and Social P cholo . 1958. 51. 127-135. Postman. L.. and Leytham. G. Perceptual sensitivity and ambivalence of stimuli. Journal of Personality. 1950-51. 1.2. 390405. Remers. H. H. I_ntroduction to Opinion and attippge mggpprement. New York: Harper. 1951!. , Riley. J. W.. Cantwell. F. V.. and Ruttiger. K. F. Some observations on the social effect of television. Public Opinion Qpprterlz. 1949. 1.3... 223-234. Rokeach. M. The Open and closed mind: Investigations into the nature of bel e stems and rsonalit stems. . New York: Basic Books. 1930. Rosenberg. M. J. Cognitive structure and attitudinal affect. 99ml of Abnormal and Social chhologz. 1956. 51. 367-372. Rosenberg. M. J. A structural theory of attitude dynamics. 2591;; Opinion Quarterliz.,1960. a. 319-3%. Schramm. W. The nature of news. WW. 191+9a. _2_6_. 259-269. Schramm. W. The effects of mass conmunications: a review. Journal}; mus. 19l49bo 3.6.0 397409. Schramm. W. (ed.) E e process and effects of mass comications. Urbana. 111.: University of Illinois Press. 1954. Schramm. W. Respgnsibility Q mass commuicatiop... New York: Harper. 1957. Schramm. W.. Lyle. J.. and Parker. E. B. Patterns in children's reading of newspapers. Journalism Quprterly..1960. 31. 3540. Senders. V. L. Measgpgp ant and stgtistics. New York: Oxford Univer- sity Press. 1958. O I Q I. m a e v. G C I . . e t e . I 0 V e _ t u 1 .. Q 109 Siebert. F. S.. T. Peterson. W. Schranm. Four theories of the ress. Urbana: University of Illinois Press. 1953. Siegel. S. Nepparametric statistics for the pehavioral sciences. New York: McGraw-Hill. 1956. Staats. A. W.. and Staats. C. K. Meaning and p: correlated but separated. chhological Revigw. 1959 . g. 136-144. Star. 5. A.. and Hughes. H. M. Report on an educational campaigm the Cincinnati plan for the United Nations. Amepican Journal of Sociology. 1950. 55. 389-400. Suchman. E. A. The intensity component in attitude and Opinion re- search. In S. A. Stouffer et al.. Measurement and prediction. Princeton. N. J .: Princeton University Press. 1950. Pp. 213-276. Swanson. C. E. Predicting who learns factual information from the mass media. In Guetzkow. H. (ed.). Greys. leadersm and map; research in human pelations. Pittsburg. Penna Carnegie Press. 1951. Pp. 218-233. Swanson. C. E. What they read in 130 daily newspapers. Journalism ngrterlz. 1955. 3_2_. 411-421. Tannenbaum. P. H. The indexing process in communication. Pubgc 021111011 Quartery p 1955. 1.2. 292-302. Thorndike. E. L. The pgcholog of wants. interests and attitudes. New York: Appleton-Century. 1935a. - Thorndike. E. L. Adult interests. New York: MacMillan. 1935b. Waples. D. The relation of subject interest to actual reading. In Berelson. B.. and Janowitz. M. (eds.). Reader 1;; Public Opipion ard Communication. Glencoe. Ill.: Free Press. 1953. p. 347-351. Waples. D.. and Tyler. R. W. What peeplg want to read apgup. Chicago: University of Chicago Press. 1931. Werner. H.. and Kaplan. E. DeveIOpment Of word meaning through verbal context: an experimental study. Journal of Psycholog.,_l950. g2. 251-257. Westley. B. B.. and Barrow. L. D. Jr. An investigation of news- seeking behavior. Journalism Mrterlz...l959. 1Q. 431-438. Wiebe. G. D. Mass communications. In E. L. Hartley and R. E. Hartley. Fundamentals of social psycholog.,_ New York: Alfred A. KnOpf. 1955. PP. 159-195. Woodward. R. S.. and Schlosberg. H. memental pgcholog. New York: H0113. 1951+. . a. . . . . r v . I a C . a A. Q .. v . I O u e .. e I e e t a. t e a I e l . e e v. . e e e K o . q .. . . u . p . t . I I . _ Q l e . I a . a I u e l . n t v I . Y t . . a r e . . .. . I e I a e a. ~ 7 .- .. t b AETEEQI! STIMULUS AND TESTING MATERIALS WITH INSTRUCTIONS 0n the following pages are the stimulus and testing materials for this study. along with the instructions. The testing materials were the same for both experiments. except for the addition of a 'familiar-unfamiliar' scale in the second experiment. The news stories. reproduced here photographically. were clipped together in the order (for subjects in each experimental group) shown in Tables 1 and 2. The headline test is reproduced photographically about 70 per cent its original size. 110 111 Communications Research Center Michigan State University NOTE TO READER: We want to get your reactions to a few news stories and some topics intended for use in a Sunday page feature on ''The World Today." The news stories you are about to read are edited from current news agency dispatches. They are advance proofs. We would like you to read them just as you might if you happened to see them in your daily newspaper. It probably will take you less than five minutes to read all six stories. After you read each story. we would like you to give us your estimate Of the level of reading ease (or difficulty) for that story {21; 333. You can do this simply by writing a number at the top of the story. right on the proof sheet. The appropriate numbers are: luextremely easy 2-somewhat easy 3-So-So 4—-somewhat difficult 5--extremely difficult Note: You may get more than one story on the same tepic. But it will not be exactly the same story. We'd still like you to read it and give us your reading ease estimate on it. This is pp; an examination. You need not sign your name. We just want your own personal. honest reactions after you read these stories in your regular way. Then we would like you to go on to the next instruction sheet to give us some of your reactions on news topics. The purpose of this is to determine how better to serve readers of various age groups. Your honest answers will help us to achieve this purpose. III“ 09:39.0 030300 .u>x0. w9fiejmvs0 O. 90310.0 .000. .995. 00:20.00 9.0 906.: 0 @000. 00.0.8299 9.000000 0» 9.0 2.30.0.0: .05: 02: 0. 2.50: 0000.2 90. ‘09.: 0 not «.0000 900 .m9 00:232....000- 0.0: S: 00<0900 2<0 00:30.00: 50...:m 90 00:30.. 0 000000020 503000 0. 90 30000: £010. wcn 0203030. 0...... 8 90 .609 9000.090 w950:.0 <03. 00000000.. 4.9.0 00.000 90 0:023: on 9090.. 9000 a... 90 :90 8 3090 90 00- <0=000 00 0 .000 9030:. .OE&:m w9:~0:.0 .0005: 020.00 .0 .:&0. 000000 .90 900000 .02: m9:~0:.0 00:92.: ..::m_0. W950: .0 .0055 0<00 30900. 90.0300. 3- £000 90 "00990.05. 0:0 08:03.0 09.500000 0. 9.0000 0:0 >30100. > .2 30059000. . . 30:00 8 .920: 00003:. 00 003 on 90 00:23.0 0:10.835 8 90 ~99 .0983: .n .0 0 .0:0. 93. 0000300. on .00....0 50:00.0. 30.200 959.0900. £20 m00m00 0:0 .::n.00 a: 0. m030. 300. o. 90 0000.0 0. 9.0 02:20 00:200. 000 0. 0.9000: 00.- «.9. .0: 0:0 3:02:00. .090 000:..- :0: .0 090:. 08.80.: 0: 0000 090:— 90 0.00 0. 2.000.009.0030 0:0 <00: 30:. .3090... 2.0.00 09.0.00902 .: m9:..:.0 0.00.8290 .0 90 00:20:00.0: 0. :0 :02 0209050 00003.02: 90 _:n 0.0: “002.00 .0 1000. 90 000:0? 09:30.. 0300000 .028. 2.5.7.5... .0520 xiii? 0:602: .0 00.0.0.8 8 29.0000 .20 00:80.00 $99 90 :05 net «.02... 0:0 2 90 0.00.0.0 .002 1.95. 00:30.00. 2.20: 9.0 . 59 00:- 20.. 05:00:00. :00 .000000 0..: 00 5.90900 003000000000... 32:. 900 2 90 89 003:3. 0.00.0. 9: 5:0 .0 0. 90 0000000. 02:- 3:0... 09.3 8 90 :29 90020:. 0909:... 80002.00 00 0 0.0.0:. .00... 92600.8 9.52... 02.9: 00: .260. 9.000 E..—0.0.0 “can: 2.0.00. .20.... 90.0200. 2.5.09.0 909:. BER. ... .220 500002 ... 90 30900.0: 0:0 0000230 0.0.5. 2 @2000 0:0 90 c380 9.000. ”00000.0. 92... 02:50: 0 .0! 300.0000... 50.00.90 >350... 901080... .0 00000 900 01.90....“ .0 90 98000.: .00. 4.9000 30.800 00! 0 500.0,... ...:0 2 .905 08.80 0000.0. 3000 2 90.: 8... 5:09:00 0.00 2 ....- 9000: 000002. ...90 0000.0 <53... 003000 959.900. 259.2003. «00": 2.0 .020...“ €3.00 2 $9:- 2.. .00 .0 ”00.20.008.359. 0.00 2 2.00.09.30.00 0:0 <32.» 02:30.... 09:00.... :02 050302.. 00.0! .003 90 .00.... 2.2.0.00 .0 0.00. 90 0.0.0.7-.. :0! ...:000 02.- 20:000.. 8 .000 90 t... 8 90 00.5000... 0.00.823. 09:39.0 0300000 0>xo. m90..:l> N002 5290:: .:m 9.0 909—: ... 9.0 1.50.0.0: 5:002... 0:0 0. 90 0.00.0.0 .002 .3005 00:30.00. $959.0 .000000 0000.2 9.. ‘09.: 0 .0... 0.0000 900 7.9 00:. 2.00. 0.300000: t... 09320 3.0 02.213.350.05 90 00:30.. 0 00000035 505900 2 90 30000: $00.0. .....0 0:00:09 .0 90900 090 t... 90.0 9000 .2 0.0.00 8 00 ... 9...: "09.0 <05. 0.020900 .0 0 90:0: .... 90022.00 ...? 0033.503 0...... .0 90 9009. 9:00: .0000 900:9 900 00:92.: .25... 3:300 .90... 09.09 as m9S0:.0 .0003: 020.00. mi 9E: 00: .0 .85—5 0<0: 000907100300 90 ”009.00.. 05:0 0:0 00263.0 0.0.5. on m:0000 0:0 90 €3.30 0.0.00. ... 00000 8 00000 900 0.009205 3 90 ~09 00:83.. w920: 00002- E 000:00 :0 .00 9000000 3 0 001 300.8000 l 32:25 3:00.00: 92090:. .093. 3:90 0: 30000.9... 50.200 90 0000.0 000 .0... 5:02:00 .00 0. 0.900.: 005.... 00:20:28: 2. 90 005.00.00.00» 022:3... 00.01002: 90 50.0: 2.03.00 8 0.00. 90 000.20.»-.. 90:. .5000 i: 00 90 50.00 09.0.- 0030... ... W952... 01.90.08. Z0500 0032.00 r0282! 9209.0 020.0030: 00: 2 00.020. 900000900 .0 92:0 0000000 :0 9.. :0!.:0000:00s0000u 30:00 2.2: 0.310202000300902... .090 330000 .2090 00.00... (9.09 :3: 90 ::0.0.0 000r00 0N0 90.090 00.0900 900. 8 90 00:00.. 2007 "01.0909. .0 0.050005 $0.: .0 .2: 90 0£0...:m 00:90 0. 30500 9.509 000080.90 £9.09 93.0 90: 02:0 900000002 9.00:0. wE E..—2:. £9.09 900 00.00 90 0.000 ~00 Io «0000 0:0 (9.09 0? 0000.00 .n :09 90 00.0000 00030 00000 .00 0.0.5.:m :0 8 1500.0 0.<0 9039000 902.0905 $00.0 $.00 ... 90:90 0500037.. .....0 mung «0.1009502 0000 :0 30:00 .0 0: 50003002 0.0.2. ~00 20.2.0 “598° 0000.00.09 002:: 500 .30 0.. .00. on .000 9.: .8 04.000 5:00. 99.00: 50:90 .no 00.00.: 0:0: 000000 20000.0 00:05:03: 0:0 39000! .9500 .302... 00:- .3303 9000:00 90 3.000.. 00:02.... 003000 3 080002059 900. 2:00 90: 90 3.0:: 900 000:5... 3. 90 E059 «9.00900. 10.0 mag...N ..01>221me10. m0:9 >.0.00| > 1.309905 9:. 00.09. 0.0.9.0.: ..0: :00 909.90 .90 0009-190 00.00- 6.92.0: 2 m0:9 >500... (9:0 m0<009509. .0 9000 9.009 09000.0... 00 0000.0 .9 .90... 0.0000. 49.0 0309.02.09 .0 90100000... 9090. .. 50.90 ..9090 2 9009000: .9 .90 9.99.50 2 .90 £990 >.07 9.9000 90 0:: 90 00:20... > .0929 00000. 000.0... .90 10000009090 .0 0:000000 .0 90<0 900: 35000 00 .9 09.710509 N005. .9 $00.0 $00 .. 20900.. .0 0:00 090:. ..0 0.00. 9:. 90 100000..- 9090 000909... 900 090... 90850...- 9000. ...90 10000009090.0 009.00 .0 .9: 0.00.00 9.. .90 .00. 90. 500. 2 90 0.9.90. 5.9.0.000 .9 90 20:09....0. 0...... 09.009509. 90.09m .0 ... .9150 3.9.0.00 10:00.09 <00So000 .0 0 50.9900. 000.0990 .9 300.0200. 09.0 .00 >....9.90..0. 90 10000009090 .0 0.: 500. .0 09.70.005.009. 09.7009- 010 .90 09.7.0509 .0 .. .0 09.7 9...<0. .090 9.0090. £90 90:. 0.3mm? .00 9:50: 059.0 .9 90 .000 2 09.009509. 5009.90 m::0. 0:? 929900 90 20.9.9000 27.0 .0 0:0. 1 1 o m U M 1 1 o 2 U 10.0 «0.0.0? .gzzmmmcxo. m0:9 >....00I. >9 03.9.0020: :09 .....0 .050 9:. 09:09 03.00 90090 :0 .90 0.09.. 00. 8.5.90.8: 2 mo:.: >.0.00.0 (9:0 m0<009509. .0 9000 90 9..09 .90 00.0000 0000.0 .9 .90... 0.0000. .. .0 0.0:...0 00000. 000.0... 990...: .0 9010000009090...0 5059000.? 09.00 9.00. 2 .90 009.90. 5.9.0.000 .9 90 22890.5. 000... m0<009509.. .9 90 ..9m:.m0 2 .90 (9:0 30.9.9000 £90 0:: 90 00:92.... 10000009090 50.90 :9090 2 9009000.: w:000000 .0 90<0 900: 35000 .0 .9 09.710509 m00:0u .9m .9 $00.0 $00 .. .90 1000000: 9090 90... 0009.9... 900 090:. 980 5059000. 9.000 000...: 909.90 505900- 09.0 .0... .0 90 00.40.. 2 90 10000009090. 090 2 ..0 5059000 .0 00.50 2:50.00 10900.09 <00£0000. ....90 10000009090 .0 0.500. 00 09.7000050009. 09.70.9010 .90 09.7.9509 .0 .. .0 09.790..<0. .. 00.400 00 00509.5 2 .9 .0028: .00. 09.0 .00 .90 10.909000. 0:.- 9.9990000 ....0 .0 090 9.. 90 9.0090 $.90 0.00%.0 .00 9:50: 0.9.0 .9 90 .000 2 N03550:. 5.09.90 9.90. 10.0 mono... .gzgmmcxn. m0:.9 >300 nu 109.9090 0.0:. 00.0.5.9...0: 2 00:9 >....0..0 ......0 03009509. .0 9000 9.09 .90 00.0000 0000.0 .9 90... 0.0000 .0 0 9.209905 9:. 00.0009. 0.0.9.0239. _. .0 90 10000009090. 0 2.90.: 00000. 000.0... .0 £9.09 90.090 500. 2 90 009.92 5.9.0.000 .9 90 z.- ..09...0. 0...... «3009509.. 10000009090. .9 90 $92.00 2 .90 5.900.... (9:0 30.909000 £90 0..: 90 095...... 50.90 :9090 2 9009000... .. .0 2.000000 .0 92.0 9009 22.9000 00 0: 5.710.109 .9- 0.02909. .9 $00.0 $0.. .. 20¢. 90 10000009090 000909... 9299000 0- 00:90 98° 5059000190903 9900.0 .2. 0:00. $90. .0 000...: .0 .90 000.00 2 90 1300009090. 00.50 2.9.0.0.. 10900.09 <00to000 .0 . 9.05900. >500. .0 09.70.005.009. .9..- 0092.0 .90 09.7.0209 00 .. .0 02790.30. .90 10000009090 0..? ...000 0: 3023.9. 09.0 .2. .90 20.9.9000. .090... 000 9......590000 2<0 .0 2.0 9< 90 9.0090. (90 90¢. macaw—0 .00 9:50: 0.98 .9 90 .000 2.0205509. 5009.90 n50. KON>21HDCN :3. 3000.993... 9.811200 §>10C11. 9.00.05: ...:0'09000 .0900 090 2.0 9.....27 (90 5.590.. 9.89.95 . 00. 2 90.9 0. 90 .9000 090 2 0 000.0. m050 00007.0 .00... 92.0: 3025.00.50 50.0.. 000030 000 2.09. 2m... 090 0.090.... ::.0. m0..- .00 003.000 .0 .9 09 2.0... .0 00.09 :0 .0 0050 .0m000 :09 ..0 90.099000. 9.0005: 1 000.50. .0 .0090. ....90 000. 9.0 0050 , 005 200.5935. ammd>zammwo o>z>h b .9pm9 oduwauowmu. 4090020000 G090. 2.3200 00.41090 0000.: 2 0 :00 00.000... .00.: 90 2.3.0 .0 90 00050 900 9009 00.0.50 9.. 20.000. C9000 «2009509. 0:00.000. 09: «90000 9.20 900:: 0900.90 90 8.0.9.9... 0. 0 08.0 000000 28.: .00.0 99530000 .0935. ...90 00000000 0959.008 0090. .00—00 90090:. ..80 .998 9009.000. 0. .90 0.080.. 0.00990 0050 .00 >..09..0:000...0 09.00.:«390 .009: 0.90 0090.. ...90 00.90 0. 90 0959.008 0090. 20:... 3:0... 0 1.6.. .00 090:. 90 .....00. ...90 000. 290 02.0 .0020 92:00 0 000.00 2 00.2.0.0. .0900 0.5. 0 0:. 00 .::.,.0....9009..00 00:... .2050 90 .8 5:8. >50100: :90 20.0009 09“.:0000 , ..00. 080.00 .90 .000 0. 0.09:0? .0000 00:0. 09:00. 0 0039.... 000. 28000.0 80905.0 900... .9 900-0. 9000 00000 900 00<.<00 .9808. .9 ... .. .0 80.90.00 90... 2990.090 0... 00000 .0 92.0 90 0090. 0.0000. 0. 0 90.79.50: 02.000 00 .9000. 209.. 008.0 09002000 0.00 00:: ...:000 90. .90 000. . 2 90 09.39.0000 00:0. 20:.0 90 00.. 09.00.:m 90.9.00: 90 3.03.0 090 000...? 1 .wo: 040.900.09. 4099020000 n0:0_ 2.3200 03'90 0000.: 2 0 90.0 9.0.00.0... .003 .90 3.03.0 .0 90 00050 900 900: 0000.30 9.. 20.000. C9000 «2.00.509. 0:00.08. 09: 0.90000 92.0 900:: 020290 90 .000.9...... 2 .0 00.90 000000 20.? .00.0 09530000 .0959? .....0 00000000 0953008 0090. 9.0:... 90 90:. ..8o 5:8 9009: .000. 0. 900.000.... 0.00090 00.90.00 >..03.0:.u:0...0 09.00.:0'90 009: 0.9- 0090.. ...90 00:8 2 9009:0388 009: 9000 ..0000 900 00.0000 9.0000. .9 ... .. .0 00030.00 .90. .. £900 .090 0... ..0000 .0 95.0 90 0090. 0. 0 000. 0. .0 90.79.50: 02.000 00 .9000. 209.. 008.0 090000000 000 90. 009. .00 090.90.. 09.00.90 0050 90: .0009 .90 2.0.9.0 09.. 0.0.90. oedmmalmze ow mbHeH > 10:... 0033309. $2.0 0:90 50900 90 9.0 9000: :98. 0.... 090900 00.9.9920. 900099000 0: 2:00.006 00000.00 .0 0800.00 .9 .90 309.90 .9 09: 00.00 000.9900: 00:300. 0900.....00 00... 0 29.0000 0.0.2.07. 09.0.: :920900 00.90 .0 0022. 0909.. 5.0.. .3072901009 000000030: ....0 2 90 >.....::0.0. 9009250 2 90 90! 10...... 09.00503 92.0 900: 098.009.? 900000. 009.09 .30 0.0.0 00 90.7 02992.0. 920...... 900 g .98 9... 09.00.1909. 90. ”0.9.5.9 00000.98 00 :0 «090 2 03:08:: 85.0. 00.00000 (90 000 0.9..“ 9.... 0.0000 .490 0.0000 .0 90 09: 3:20. ..90... .090 «008.3903 0. 10... 9.0.0 90‘: ..0 0.0090... .9 000.80 1000. 903: 10:... 003.930... $9.5 0:90 9.0900 90 In 9000: :98. 0.... 090900 00:99:20. 900099000 0: 2900.006 00000.9 09005.8. 00.... 0 ..2800 0.0.0.07. gus- ..920..00 0390 .0 00100 0908.. 115 >0..<..0 .9 ...900 00.00.... ....900::09.0 0009i..9m 00000. 9.:m003 0: .90 9009 0000. 0. >....00 .0 90003.90 0 .090 0. 00x00 090 9.9030 00009.0. >30...009 «3.0m. 0.0. 00...000 090 003009.. 0.800.300 90<0 000000.98 .90 9:00 00:30.. .0 008.00 90 0.09 0.. ..0.00 0.9.0... ..0 909009 90 00900. ...900 ..90 .9 ...900 0 90.8.0: 090.80 009.000. 90.0.00: 90 0.0 09.. 9090.... 2:09 0. .90 00050.89 030.08 0. 7.00.03 3.900309 .90 9000 0030.0 000000 90 90. 00000.. m:00030 0:.00 0. ....900 .0 ..o: 00070.0 ...:m .010. 9:. 99.90 .90 0.00.0 0000. .00000 .:09 00030.8: 090 020.00 03090 .00000 00.50.: 090. 090.. 9::m00 0.80 .90 0.09 00.000 .9 0.000 .00 9000 £90 ...... 90. .9 00.20.. .090: .90 0.. 30900 00300. 090 .0<00.09 0000 0. 5900.0 00: 8.00309. .0 0000000 90 >3010090 .090 000 90.0.90 95... 00.000 0.0.90. 00000 090 90.300. 000.300 .0 9000.90 090 0. 90.7.0 .00....9m 0..:0000:0.9m 00000 .9 90 ...900 .0 0009.90 ....9 w .Hoi 000»<»00. O I I 0.. .9 ...900 00.09.... .:.9004:...9.0 0000.2..9m 0000... 9.:m003 0: 90 90.9 0000. 0. >.0.00 .0 90003.5 0 .090 0. 00x00 090 09.09030 00009.0. >30...009 m8.0m.0.0. 01.800 090 003009.. 0900:2200 90<0 00:00.. .98 .90 003.00 .0 00.0.00 .90 0.09 0.. 90.00 9.09 :0 909009 .90 0:09. 00:00. 090.. ..90 .9 9.900 0 0.009m0 00:: .000. 90.9.00: .90 0... 090 90 90.... 2:09 0. 90 00052.09 890.08 0. 300.03 2.90.0309 £90 9000 0030.0 00.000 .90 0000.... m:00030 3.00 0. ....900 .0 0o: 00070.0 .95 9.0.0. 9:. 99.90 .90 0.00.0 0000. .00000 .:09 000.0000: 0:.. 0.9.00 0309m .0000: 00.: 9.0.090. .0900 9.3000 0.80 90 0.09 00.~00 .9 0.000 .00 9000 90...... 90 .9 00.20.. .90: .90 0.. 3090000300. 5900.0 030.0039. .0 9.000 9.. >300.00:0 £90 000 90.0.:m 99.0 03.00 0.09.0. 00000 090 90.0000. 090:9 000090.. .0 900030 090 0. .90 ..<0 .0095 0..:030:0.9m 00000 .9 .90 0.00.0. ....900 0.... 3.00.0090: .0 .9 0 0.90.0... 3000 00007.0 000. > 00090.40 040.000.09 09090.... .0000... >000? 09090l090 009009.89 0000.0.0 000.0 .0 800.8 90 000.5 90.8: 0. 09090 .98 90 2090.09 009.900. .090 0.0.3.00 0. 90 000.0 090 90 90.8: 000000 .0 90 ..9900 0800.0 8090.. .00 0 .09m ..30 .0 0030. >..00 900.0 00000 0. 0053.0 0009.00 9.. 0000. .w0..0.9. 0909090: 0030 0 000:9..0 .59.: 90 9.509 003.909.0009 ..:.0 .. >90 0. .90 90.3 0030.90 .9330 29309. 9000 0. 90 00309.8: 0000.0.0 000.0. 0930.090 930 00.00 2000.90.09: .9m.0 .00 90.0 :0... 009008.89 090 920 0900000.. 29309.0 0.09 .0 0039.90 90 .090 0. 0000.009. 090 00.30 3.9.0.00 .98 090. 09.0 09.0. 0.80:..00 .0 29309. 0.39 90...... 00 ..>....00.0 309 0. 000090... C9000 90 00989.89 080.00 00...... 09090 900 9:... :0 0 9090.: 00.90 .000050 0:00.:0. 09.. 90 90: ..090. 00.. 0.90.000... ...90 00.00 0. 00000. 30.: 90.890. 00030.. 900 90.0 :0 £0... >30...00: .90:0....0.: .0.0 000 300.9200 :99 09090.0 900.00.00.10 00.0920. 090 900.08 00000.9. 29309.0 3.9.0.000 900909000 00900 .0 9000 0 0.39m. 09.0 00000 0. 99.89 002.0000 090 899.0890 09 .90 .09. ...90... 0030.95.09 .0 09.009.90.90. 030.0900 .0 09090.. .0080 .9 90 1.00.30 0. 90 00:: <0:..09 0000.0 0 000.... oozdmzeHOI 0mo00m.m 0>w0H o0 om>z> m .90m00940 odwwsnapou 09090.. 900.30 >000? 09090|090 :0... 90..0: .. 09090 0000000 .00900 09.00 90: . 00 .9 90 «0.0 0. 090.000... 050:: 0. 0. .90 00303.0: 0000.0... .00.... >..00 9000 00000 0. 00..:m0<009: 0:. «009.00 9.. 0000. $1.93. 9090 900030 0 00009.8 .59.: 90 "0...09 003.909.0019 0: ..:.0 .. 030.35 0. 90 90.3. 0. 00:000. 0 .9830 29.039. 9000 0. 90 0:<0:..0: 0000.0 0 00...... ...90 000.0 0890.300 90 00: ..0...0: .9 0: 0.02.09 09 90 :0... 090.3309. 309.:m .. 00009.0 .00 90:39 .0 .:0: 90 000030900 .98 . 9.00.30 .09 .00 9.300... .090 90000000 00000...0: 000.0 .0 00000 9.. U0. .. w. 0:90:09. 90 00090 0.0 30: 0. 0909.0: .009: .00. :0 090 9.0 9009000 90<0 m:09 80:00 90<0 000.300.3903 003 0030000..0:.:.9000 0.30:.0 .. 0.0880098? 116 BE SURE YOU HAVE: 1. Read each news 'story. 2. Written at the top of each story a number representing your estimate of its level of reading ease £3; mo WHEN YOU HAVE FINISHED. TURN THE PAGE.... 117 Now we would like your 9111019 reactions to some topics in the news. On the following pages. you will find a word or a phrase followed by a series of rating scales. Each scale is composed of Opposite meaning adjectives. like this: active : : : : : pa ssive Let's say the news tOpic you are asked to judge is MICHIGAN: If you feel that MICHIGAN is very closely related-to one or the other adjective. you should place an "10' mark in one of the extreme positions. as follows: fair : : : : : : unfair * “ ~_‘* If you feel that the topic is Quite closely related to one side as Opposed to the other. then. you should check on one of the spaces one step in from the extreme: If the tOpic seems only slightly related to one side as Opposed to the other. then you should check one of the spaces on either side of the middle space: strong ___:____: A :___:___: : weak or strong __ :____: :___:_X.:___:__ weak If you consider both sides of the scale an all! associated with the topic. or if the scale is comletey irrelevant. then you should place your checkmark in the middle space: harmful : : :X: :___:___ beneficial WEBER: 1. Work fast. Don't puzzle over items. Give your first reactions. 2. Be sure you put one '1' on eveg. scale—don't skip am. 3. Put only one 'X' on an scale: and put it in the middle of spaces (as above). not on the boundaries. . I . r e e I . t . n t u I . . . . . . n I. a _ t . . 1 . . I — . .. .. .- . . .. . . . . . Q o . .- .. . . .. . . . .. Z a. .. .0 cs _ .. l .. .. .. . . v . I .. . _ . .. u w . . . . . . . . .. C 118 CANADA weak : : : : : : strong godd : : : :__:___:___bad boring___:___:___:___:____:__:__interesting harmful___:___:__:___:___:_:_benefic ial active___:___:___:_:_:____:___passive unplea sant__: : : : : : pleasant would read about : : : : : : would at read about familiar : : : : : : unfamiliar GOVERNMENT OF HAITI weak : : : : : : strong good : : : : : : bad boring__:___: : :_:___:____intere sting harmful : : : : : : beneficial active__: : : : : : passive unpleasant__: : : :___:__: pleasant would read about____: : : : : : would 92; read about familiar : : : : : :' unfamiliar ~ _— —*_ —.*. TEHUANTEPEE CANAL harmful : : : : : : beneficial active : : : : : : passive unpleasant__: : : : : : pleasant would read about : : : : : : would 3.195. read about familiar::::::unfamiliar 119 CONVENTION mom's PARTY OF GHANA weak___: : : : : : strong good::::::bad boring__:__: :___:__: :__,interesting harmful : : : : : beneficial active____ 2 :___:___: passive unplea sant__: : : :____: pleasant would read about__: : : : : would. not read about boring__:__:____:___:___.___:___int eresting harmful : : : : : :___beneficial active : : : : : : passive unplea sant__: : : : :___: pleasant would read about : : x : : : would not read about harmful : : : : : : beneficial active___: : : : : : passive unplea sant___:___: : : :___: pleasant would read about__: : : : : :___would not read about “-__‘_’ familiar___:___: : : : :___unfamiliar good___:__:__:___:__:___:___bad boring__:_:__°_: :___:__interesting harmful : : : : : : beneficial : passive active : : : : unpleasant___:__:___:__:___:___:__plea sant would read about_:__:_:__:,___:____:___would M read about familiar___:_____:__:___:___:____:__unfamiliar MALTA weak__:__:__:___:___:___:___strong good___:___:__:___:____:_:___bad boring : : : : : : interesting : : beneficial harmful:: O. 0. active___: : : : : : passive unpleasant : : : : : : pleasant would read about__:___: : :_: :___would 9.2 read about familiar____: : : : : : unfamiliar MOZAMBIQUE weak::::::strong #——__'__ good__: : : : : : bad boring__:___:__:___: : :__J.nteresting harmful__: : : : : : beneficial active : : : : : : passive unplea sant___: : : :__:__: pleasant would read about : : : ° : : wouldggt read about familiar“: : : : :___:__un1'amiliar 121 AS A FINAL INDICATION OF READING PREFERENCES. we would like you now to imagine yourself reading a newspaper. When you do that you have to decide Often which of two or more stories to read first. You often choose very quickly just by looking at the headlines. They give you some idea of what to expect in the story. Printed on the next sheet are some of pairs of headlines. side by side. Each will give you some idea of what the news story with it would be about. On this basis. circle the story in each pair which you would read 9551. Don't spend am more time on each decision than you would in actually reading a newspaper. For example. here's one pair of headlines. Let's say you happen to think the one on the right would capture your interest first. SO you circle it.... Crowded Month Lies Ahead Problems of South America For United Nations Group e Aired at Convention I D I C 1.22 Arabian Diplomats Divided On Mediterranean Proposal Earthquake Study is Planned By 0.5. Geologists in Chile Malta's Aspirations Realistic. New Foreign Minister Declares Envoy Sees New Factors In Mozambique's Future New Growth for Broederbond ls Predicted by Err-Diplomat Bhutan leader Visits Capital For Talks on Development Plan Tehuantepec Canal Backers Given New Encouragement Changes in Libya's Prospects Seen During Corning Decade Situation in Haiti Placed On Agenda for Conference Ghana Official Clarifies Role Of Convention People's Party Ghana Official Clarifies Role Of Convention People's Party Malta's Aspirations Realistic, New Foreign Minister Declares Bhutan Leader Visits Capitol For Talks on Development Plan Changes in libya's Prospects Seen During Coming Decade New Growth for Broederbond ls Predicted by Ell-Diplomat Tehuantepec Canal Backers Given New Encouragement Envoy Sees New Factors In Mozambique’s Future Situation in Haiti Placed On Agenda for Conference Britain's Foraign Policy Criticized by Europeans Economist Sees New Hope For Prosperity in Asia Malta's Aspirations Realistic. New Foreign Minister Declares New Growth for Broederbond ls Predicted by Ext-Diplomat Bhutan leader Visits-Capital For Talks on DeveIOpment Plan Envoy Sees New Factors In Mozambique's Future Envoy Sees New Factors In Mozambique's Future Tehuantepec Canal Backers Given New Encouragement Situation in Haiti Placed On Agenda for Conference New Growth for Broederbond ls Predicted by Err-Diplomat Changes in libya's Prospects Seen During Coming Decade Malta's Aspirations Realistic, New Foreign Minister Declares Ghana Official Clarifies Role Of Convention People's Party Bhutan leader Visits Capital For Talks on Development Plan Tehuontepec Canal Backers Given New Encouragement Ghana Official Clarifies Role Of Convention People's Party Changes in Libya's Prospects Seen During Coming Decade Situation in Haiti Placed On Agenda for Conference 123 Your Date of Birth (Day) (Month) (Year) Sex (circle): Male Female Class (circle): Freshman Sophomore Junior Senior Graduate Major How often do you read the front page of a daily newspaper during the school year? fiery day Circle one: Sometimes Never In order to give us some idea of background knowledge on foreign news for readers of your age group. would you fill in as marw of the names below 88 you can? (Last name only....and go ahead and write it even if you're not sure about spelling.) Prime Minister of Russia Prime Minister of Cuba Prime Minister of India Prime Minister of Great Britain President of France Hm THANKS FOR YOUR COOPERATION AUG i4~19a53 “iii AW W ’ A e ' - A.., f I‘ ‘ , -' ‘ a. ..- -: J . * ~ g ' J : e - _- I }' 4' V" t, M." L‘ n 5 fx‘wT‘} l‘xi" Wfiam 3‘- k (I ‘If some "ITllllllllll‘l‘lllTIES