~ - ~ ism ABSTRACT THE EFFECTS OF VERBAL CUEING AND INITIAL ETHOS UPON PERCEIVED ORGANIZATION, RETENTION, ATTITUDE CHANGE, AND TERMINAL ETHOS BY James Franklin Weaver This study was designed to answer two overriding questions: First, what are the effects of verbal cueing? Second, what are the effects of initial ethos? Given a well-organized message, are there any advantages to high- lighting or emphasizing the organizational pattern? At the other extreme, might bad cueing prove disadvantageous? Might the credibility of the source produce any interac- tion or main effects. In an effort to avoid any of the effects that oral presentation might have on the results, a written message was employed. Three versions of the speech were devised, exemplifying accurate cueing (good use of the preview, sign post language, internal summaries, and a 1 James Franklin Weaver final summary), no cueing, and inaccurate cueing. Each version of the written message was attributed to a high- credible source and a low-credible source. The 242 subjects, students enrolled in public speaking courses at Iowa State University during the first summer session, 1969, were randomly exposed to one of the six treatments. The effects of verbal cueing and initial credibility upon four dependent variables were measured: perceived organization, retention, attitude change, and terminal ethos. Specifically, sixteen hypotheses were tested: H a: Accurate cueing produces higher perceived organization of the message than does no cueing. H b: Accurate cueing produces higher perceived organization of the message than does inac- curate cueing. H c: No cueing produces higher perceived organiza- tion of the message than does inaccurate cueing. H : High credibility produces higher perceived organization of the message than does low credibility. H a: Accurate cueing produces more retention of the message than does no cueing. H b: ~Accurate cueing produces more retention of the message than does inaccurate cueing. 2 James Franklin Weaver No cueing produces more retention of the mes- sage than does inaccurate cueing. High credibility produces more retention of the message than does low credibility. Accurate cueing produces more positive attitude change than does no cueing. Accurate cueing produces more positive attitude change than does inaccurate cueing. No cueing produces more positive attitude change than does inaccurate cueing. High credibility produces more positive atti- tude change than does no cueing. Accurate cueing produces higher terminal ethos than does no cueing. Accurate cueing produces higher terminal ethos than does inaccurate cueing. No cueing produces higher terminal ethos than does inaccurate cueing. High credibility (initial ethos) produces higher terminal ethos than does low credibility. Results of analysis of variance revealed a signif- icant cueing effect upon perceived organization and re- tention of the message. Analysis of variance revealed a significant credibility effect upon attitude change and terminal ethos, Where analysis of variance showed a sig- nificant F, fiftests were employed to find significant 3 James Franklin Weaver differences between the effects of the versions of the message. Hla: Support was found for the following hypotheses: Accurate cueing produces higher perceived or- ganization (and clarity) of the message than does no cueing. Accurate cueing produces more retention of the message than does inaccurate cueing. No cueing produces more retention of the mes- sage than does inaccurate cueing. High credibility produces more positive atti- tude change than does low credibility. High credibility (initialethos) produces higher terminal ethos than does low credi- bility. THE EFFECTS OF VERBAL CUEING AND INITIAL ETHOS UPON PERCEIVED ORGANIZATION: RETENTION, ATTITUDE CHANGE. AND TERMINAL ETHOS BY James Franklin Weaver A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Speech 1969 g4 /7J‘7 “74.27170 ’\ COpyright James Franklin Weaver 1970 Accepted by the faculty of the Department of Speech, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Doctor of PhilOSOphy degree. Jags. Director of Thesis Guidance Committee: Kenneth G. Hance, Chairman Duane L. Gibson James C. McCroskey David C. Ralph DED ICAT ION —-to my parents who sparked my interest in experimental research when they taught me how to count. ACKNOWLEDGMENTS The writer wishes to thank the following indi- viduals for their contributions to this project: De- baters John Cook, Ludy Edwards, and Velma Wenzlaff for their good imagination and eXpert help in devising the three messages and assembling the experimental packets, Dr. Robert Underhill for his encouragement and for allow- ing Speech students to participate in this study, all my colleagues in English and Speech at Iowa State University who participated in the evaluation of the messages and those who allowed me to visit their busy classrooms, Mr. Paul Kaufmann who kindly and conscientiously helped me to administer the experiment, Drs. Duane sibson and David Ralph who so capably served on my thesis committee, Dr. James McCroskey who provided the extra "push" to get me started and whose experience in experimental research has proved invaluable, and finally, Dr. Kenneth Hance whose patient guidance during the last five years has so greatly influenced my past, present, and future endeavors. TABLE OF CONTENTS ACKNOWLEDGMENTS . . . . . . . . . . . . LIST OF TABLES. . . . . . . . . . . . . LIST OF APPENDICES. . . . . . . . . . . Chapter I. II. INTRODUCTION. . . . . . . . . . Statement of the Problem and Statement of Hypotheses. . . Justification of Project . . Limitations Imposed. . . . . Purpose Review of Experimental Literature. . Summary. . . . . . . . . . . METHOD. . . . . . . . . . . . . EXperimental Design. . . . . Subjects . . . . . . . . . . Procedures of the EXperiment Measurement. . . . . . . . . Statistical Inference. . . . III. RESULTS . . . . . . . . . . . . Preliminary Analyses . . . . Major Statistical Analyses . Summary of Findings. . . . vi Page viii xi 14 16 21 24 42 43 43 47 49 S4 56 67 67 70 81 TABLE OF CONTENTS (Cont.) Chapter Page IV. CONCLUSION. . . . . . . . . . . . . . . . . 84 Discussion of Results. . . . . . . . . . 84 Implications of the Study. . . . . . . . 87 Suggestions for Future Research. . . . . 91 A Final Thought. . . . . . . . . . . . . 94 BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . . 96 APPENDICES. . . . . . . . . . . . . . . . . . . . . 104 vii Table 10. 11. LIST OF TABLES MEAN SCORES FOR PERCEIVED ORGANIZATION AND CLARITY FOR CUEING CONDITIONS . . . . . . MEAN SCORES FOR PERCEIVED ORGANIZATION AND CLARITY FOR CREDIBILITY CONDITIONS. . . . MEAN SCORES FOR RETENTION (10 ITEMS) FOR CUEING CONDITIONS . . . . . . . . . . . . MEAN SCORES FOR RETENTION (10 ITEMS) FOR CREDIBILITY CONDITIONS. . . . . . . . . . MEAN SCORES FOR ATTITUDE FOR CUEING CONDITIONS................ MEAN SCORES FOR ATTITUDE FOR CREDIBILITY CONDITIONS. . . . . . . . . . . . . . . . MEAN RATINGS FOR TERMINAL ETHOS FOR CUEING CONDITIONS. . . . . . . . . . . . . . . . MEAN RATINGS FOR TERMINAL ETHOS FOR INITIAL CREDIBILITY CONDITIONS. . . . . . . . . . MEAN SCORES FOR INTERESTINGNESS AND CONVINC- INGNESS FOR CREDIBILITY CONDITIONS. . . . MEAN SCORES FOR INTERESTINGNESS AND CONVINC- INGNESS FOR CUEING CONDITIONS . . . . . . SUMMARY OF PRELIMINARY CHECK OF ATTITUDE TOWARD TOPIC. . . . . . . . . . . . . . . viii Page 71 73 74 75 76 76 77 78 79 80 158 LIST OF TABLES (Cont.) Table 12. SUMMARY OF PRELIMINARY CHECK OF INITIAL ETHOS O O O O O I 0 O O O O O O O O O O O O 13. SUMMARY OF FACULTY RATINGS FOR ACCURATE CUEING MESSAGE. . . . . . . . . . . . . . . 14. SUMMARY OF FACULTY RATINGS FOR NO CUEING MESSAGE O I C C O O O O O O I O O O O C O O 15. SUMMARY OF FACULTY RATINGS FOR INACCURATE CUEING MESSAGE. . . . . . . . . . . . . . . 16. SUMMARY OF FACULTY RATINGS FOR THE THREE MESSAGES. . . . . . . . . . . . . . . . . 17. SUMMARY OF FACULTY PLACEMENT OF THE THREE MESSAGES. . . . . . . . . . . . . . . . . . 18. ITEM ANALYSIS RESULTS . . . . . . . . . . . . 19. FACTOR ANALYSIS OF TERMINAL ETHOS RATINGS . . 20. ANALYSIS OF VARIANCE SUMMARY FOR PERCEIVED ORGANIZATION. . . . . . . . . . . . . . . . 21. ANALYSIS OF VARIANCE SUMMARY FOR CLARITY. . 22. ANALYSIS OF VARIANCE SUMMARY FOR 10 RETENTION IT EMS O O C C O I O O C O O O O O O O O O O 23. ANALYSIS OF VARIANCE SUMMARY FOR ATTITUDE SCORES. . . . . . . . . . . . . . . . . . . 24. ANALYSIS OF VARIANCE SUMMARY FOR AUTHORI- TATIVENESS RATINGS. . . . . . . . . . . . . 25. ANALYSIS OF VARIANCE SUMMARY FOR DYNAMISM MT INGS . O O O O O O O O O O I O O O O O 0 ix Page 158 159 159 160 160 161 163 164 166 166 167 167 168 168 LIST OF TABLES (Cont.) Table Page 26. ANALYSIS OF VARIANCE SUMMARY FOR CHARACTER RATINGS . . . . . . . . . . . . . . . . . . 169 27. ANALYSIS OF VARIANCE SUMMARY FOR PERCEIVED INTERESTINGNESS . . . . . . . . . . . . . . 169 28. ANALYSIS OF VARIANCE SUMMARY FOR PERCEIVED CONVINCINGNESS. . . . . . . . . . . . . . . 170 29. RESULTS OF EfTESTS ON ORGANIZATION, CLARITY, RETENTION, AND CONVINCINGNESS FOR CUEING CONDITIONS. . . . . . . . . . . . . . . . . 170 LIST OF APPENDICES Appendix Page A. EXPERIMENTAL MESSAGES . . . . . . . . . . . 104 B. TREATMENT ONE (High-Credible Source-- Accurate Cueing). . . . . . . . . . . . . 128 C. TREATMENT FIVE (Low-Credible Source--No Cueing) . . . . . . . . . . . . . . . . . 143 D. PRELIMINARY CHECKS. . . . . . . . . . . . . 157 E. PRELIMINARY ANALYSES. . . . . . . . . . . . 162 F. MAJOR STATISTICAL ANALYSES. . . . . . . . . 165 xi CHAPTER I INTRODUCTION Statement of the Problem and Purpose From the earliest writing on rhetoric, the impor- tance of the arrangement or organization of a message has been recognized. The development of the first systematic treatment of‘rhetoric is generally attributed to Corax and his pupil Tisias, both Sicilian Greeks. Thonssen and Baird (1948) describe organization as one of the main contributions of this earliest theoretic Greek book. Corax's system of rhetoric had three dis- tinctive features, each of which contributed materially to the modern conception of speech theory: (1) It defined rhetoric . .v. . (2) Arrangement of material received the first formal consideration . . . . (3) Corax showed how probability applied to rhetorical invention . . . (p. 35). According to the classical tradition, as set forth in the §g_Herennium, all rhetoric is divided into five parts: invention, disposition, elocution, memory, and delivery. What the Greeks called taxis, the Latins 1 referred to as dispositio. It has been defined as that constituent of rhetoric which "covers the concept of arrangement, of orderly planning and movement of the whole idea" (Thonssen and Baird, p. 79). Classical references tend to dwell on one main issue: How many parts should a message contain? A fairly faithful reproduction of the more complex classi- cal division is found in the late nineteenth-century Speech plan discussed by Charles COppens. He says that the following parts may appear in an oration: introduction or exordium, narration or explan- ation, proposition, division, proofs or argu- mentation, pathetic (or emotional excitement), and conclusion or peroration (C0ppens, 1885, p. 107). In the Rhetoric Aristotle wrote: A speech has two parts. Necessarily, you state your case, and you prove it. Thus we cannot state a case and omit to prove it, or prove a case without first stating it; one who proves must have something to prove, and one who advances a statement does so for the sake of proving it . . . . Well, then, the indis- pensable consituents are simply the Statement and the ensuing Argument. These are the essential elements of a Speech; at most, the parts cannot exceed four--Proem, Statement, Argument, and Epilogue (CoOper, 1960, p. 200). Over the centuries rhetoricians have suggested that a speech ought to contain from two to seven or eight main parts. At the present time, with the possible excep- tion of those who follow Monroe's Motivated Sequence (Monroe, 1964, p. 208), quite typically, a speech is thought to contain three parts: introduction, body, and conclusion. Plato's advice parallels this modern prac- tice, when in the Phaedrus he has Socrates say: But this at least I think you will allow, that every speech ought to be put together like a living creature, with a body of its own, so as to be neither without head, nor without feet, but to have both a middle and extremities de- scribed prOportionally to each other and to the whole (Thonssen, 1942, p. 29). The classical writers had other advice concerning the arrangement of the message. In Book VII of his In: stitutes of-Oratory, Quintilian wrote: As to what concerns the accuser, I do not altogether dissent from Celsus, who, doubtless following Cicero, persists in maintaining some- what too positively, on this head, that strong arguments should be advanced in the first place, the strongest of all in the last, and the weaker in the middle; because the judge requires to be moved at the beginning, and pressed forcibly at the end. But on the side of the accused, the strongest argument against him must first be attacked, lest the judge, looking to that point, should regard with too little favour our establishment of other points. Yet this order may occasionally be changed, if the lighter points he evidently false, and the refutation of the heaViest charge extremly difficult; so that, after thus detracting from the credit of the accusers, we may proceed to the last point, when the judge is ready to suppose that all the charges may be false (Thonssen, 1942, pp. 133-34). Besides the number of parts a Speech should con- tain, and the best order for the arguments, there were other concerns expressed by the classical writers. Quintilian offered suggestions about the coherence or flow between the parts of a speech. There remains then only the arrangement of part ; and in the parts themselves there must be some one thought first, another second, another third, and so on; and we must take care that these thoughts be not merely placed in a certain order, but that they be also connected one with another, cohering so closely that no joining may appear between them; so that they may form a body, and not a mere collection of members. This object will be attained, if we take care to observe what is suitable for each place, and study to bring together words that will not combat but embrace each other. Thus different things will not seem hurried to- gether from distant parts, all strangers one to another, but will unite themselves, in a sure bond and alliance, with those that precede and those that follow; and our Speech will ap- pear not merely a combination of phrases, but all of a piece (Thonssen, 1942, p. 136). Contemporary public Speaking textbooks present at least a chapter of advice to the speaker on the organiza- tion of a communicative message. Arrangement has main- tained its place of importance through the centuries, good organization being regarded as absolutely essential. "Without reasonable organization, oral messages simply do not come across" (Hance, Ralph, and Wiksell, 1969, p. 179). A The justification for organization is ever pres- ent. Essentially, we organize what we say in Speaking because (1) we want to make sense and (2) we recognize the need within our listeners for a psychological feeling of unity. What we have to say--particu1arly in Speaking at any 1ength--arouses no meaning or at least not our intended meaning within the listener unless we organize the message so that the listener can form within himself a pattern that both rea- sonably satisfies him and reflects the source's meaning. The importance of order or structure in this sense has long been recog- nized by rhetorical theorists (Hance, Ralph, and Wiksell, p. 180). The value of organization exists both for the speaker and for the listener. Strother (1968) listed at least four benefits of organization for the speaker. A sense of order enables the speaker to put all parts of his Speech in their prOper rela— tionship. It enables him to use each idea in its most logical and advantageous place. It enables him to detect the weak spots in devel- Opment of his Speech and to shore them up with additional materials. And at the moment of Speaking, this concern with order helps the Speaker stick with his ideas (p. 127). Monroe (1964) indicated that the listener, too, benefits from a well-organized message. Always arrange the points in your Speech in a systematic sequence so that one idea leads naturally into the next. Not only will such an order make it easier for you to re- member what you planned to say, but it also will enable your audience to follow your thoughts more readily (p. 138). Strother (1968) implied that clear organization is almost an ethical necessity. By the use of good organization, the ef- fective speaker reveals much more than the clarity of his ideas. He reveals both the nature and the sources of his evidence. Further, he indirectly reveals a sense of honesty. He implies: "I make no attempt to send up a smoke screen to hide the issues. I want us to meet them head on" (p. 15). The communicator's job is not completed when the message is cast into outline form. Even the best outline, on paper or in the mind of the speaker, will not guarantee that the receiver will be able to follow clearly the Speaker's ideas. The next, and perhaps most important, task is to present the message in an organized fashion so that the receiver can follow easily the pattern of the Speaker's message. Ideally, the audience should be able to "see" the pattern of organization as the Speaker de- livers the speech. If the speech pattern is clear, it shows them [the listeners] the progression and rela- tive importance of ideas. It should also make clear the general speech purpoSe. Unity and coherence in the speech are perceived by the way the materials are arranged. Much of the persuasive quality of a speech also comes from effective arrangement of ideas . . . . There- fore, we need to plan Speech content so that the order is easily seen (Anderson, Lewis, and .Murray, 1964, p. 243). While some rhetoricians would lead us to believe that a carefully planned outline is enough to assure clarity for the audience, this investigator believes that for maximum clarity more than a well-outlined message is required. During the presentation, certain factors must be included in the message to help guide the listener. Hance, Ralph, and Wiksell (1969) suggested: Remember that the Speech plan is an out- line of what the Speaker intends to say, not a manuscript, and it is frequently necessary to supply transitional statements in order to clarify the intended meaning for the Speaker and to remind him of his obligation to be clear to his listeners (p. 201). Strother corroborates, "It is not enough that the speaker's meaning be clear in his own mind; he must make it clear to his audience or fail to achieve his purpose" (p. 142). There are at least four ways of clarifying the organizational pattern of the message for the audience: initial summary, transitions, internal summaries, and a closing summary. One of the most common methods of clar- ification is the use of the initial summary, preview, or forecast of what is to follow. Hance, Ralph, and Wiksell wrote: Sometimes a Speaker, after acknowledging his hosts or referring to the occasion, tells his listeners what his major points will be . . . (p. 197). McCroskey (1968) offered Similar advice when he stated: "There are three main functions an introduction may serve——to arouse audience attention, to enhance the ethos of the source, to preview the body of the message" (p. 188). Monroe justified the preview when he wrote: The initial summary provides a skeleton around which to group your facts and indicates the relationship each idea bears to the others. It also helps your listeners to follow your pattern of development--to see clearly where your speech is going . . . . The initial sum- mary Should be brief and simple, since a com- plex statement is difficult to remember and needlessly repetitive. Moreover, the order of points which it announces should be fol- lowed exactly in the presentation of the information or the listener will be confused (pp. 187-88). There certainly seems to be ample support for the inclu- sion of this first method of verbal cUeing. Second, the speech textbooks suggest the use of the transition. Typical of the advice which is given is 'that by Dickens (1954) when he wrote: During the delivery of the Sppech the function of your transitions is to show the listeners how each part fits into the scheme of the speech . . . . The relationship among points is clearer to the Speaker than to the listeners. In the first place, the Speaker can see the points, all labeled and indented, as he reads his outline. The listener can only hear them-—he cannot pause to think about the relationship involved or go back and relisten; he must perceive the relationship instantaneously. In the second place, the Speaker has been through the mental processes of preparing the speech and may forget that what seems clear to him now was not clear earlier. The listener has no such advantage . . . . Therefore, transi- tion statements should be scattered liber- ally throughout the speech . . . (p. 305). Monroe stressed the importance of the relationship between the preview and the transition when he said: AS you advance from one point to the next, use connective sentences or phrases to empha- size the transition; if you use an initial summary, relate each new point to the plan you announced there (p. 188). Bryant and Wallace (1960) in discussing transi- tions, advocate use of signpost language. 10 Use signpost devices. Number the main heads: First, Second, Third, and so on. Var- iations of this are: In the firstjplace, The first step, The first matter to be discussed ;§_. . . , Let us first discuss . . . . Al- though such labels may seem obvious and some- what wooden here alone on the page, usually they are not distractingly obvious in a Speech (p. 184). A third method of verbally cueing the organiza- tional pattern is the internal summary. Hance, Ralph, and Wiksell remind us that "the Speech plan, in outline form, does not in itself help the Speaker or the listener to keep in mind what has been said in a previous section of the Speech as they move together to another section." Therefore, an internal summary is necessary (p. 201). Andersch and Staats (1960) offered Similar advice when they wrote: Give your audience time to adjust its thinking to the new idea after filing away the old. You may wish to summarize briefly the point you have just made and then point to the new idea in your transitional statement (p. 128). Soper (1963) concurs with the statement, Restatement in most cases serves as internal summary--a restating of a main head or sub- head after it has been developed through ex- ~position or evidence. Internal summary is vital in most Speeches, particularly if the organization is complex or the subject matter involved or abstract (p. 77). 11 A fourth, and final, method of verbal cueing is use of the final summary. This is the last chance for the speaker clearly to impress the organizational pattern upon the listener's mind. Strother indicates that there are several purposes of the conclusion in a speech. The conclusion of a Speech also serves a Specific purpose: to summarize or underscore the main ideas in the speech, to appeal for action, to elicit sympathy or understanding, or to add a note of finality and climax . . . (p. 175). Even though a Speaker has a single controlling motive for presenting his Speech, he will frequently summarize each of the major subdivisions (pp. 179-80). Bryant and Wallace (1960) strongly suggest that the speaker include a final summary in his Speech. Some kind of summary is necessary, and we urge that the beginning Speaker not leave its formulation to the spur of the moment. Any hearer will welcome a summary, because the summary appeals to his sense of order and pro- portion; the Speech as a whole, the multitude of stimulus—ideas he has heard, suddenly are revealed again as orderly and systematic, rather than chaotic (p. 182). Brigance (1961) advocated and justified the use of a final summary when he wrote: If the Speech is complex, or if the Speaker's purpose is to present information, then you had best draw together the impor- tant points in condensed and unified form. "If you can sum up your arguments," said J. H. Gardner, so that listeners "will go off C. O S .1 .1 F as .t S 12 and unconsciously retail your points to their neighbors, you probably have them." Likewise, if you can sum up your information so that listeners can carry it away in condensed and unified form, they likely will remember it (p. 244). Thus we have seen four methods of clarifying the organizational pattern of the message: a preview, tran- sitions, internal summaries, and a final summary. Up to this point the writer has been concerned with four points. First, an attempt has been made to Show that organization or dispositio has been an important element of rhetoric both to classical and contemporary writers. Second, it has been pointed out that organiza- tion has a value to both source and receiver. Third, it should be the goal of the communicator to help the listener "see" the organizational pattern of the message. Finally, there are at least four methods which can be used to assist the listener follow the structure of the mes— sage. The purpose of this study is to examine the ef- fects of verbal cueing and initial ethos of the speaker upon the receiver's perception of the organization of the message, the receiver's retention of the message, the 13 attitude of the receiver, and the terminal ethos of the Speaker. For purposes of this study the following defi- nitions shall be used: Verbal cueing--highlighting or emphasizing the organizational pattern of the message through the use of a preview, transitions, internal sum- maries, and a closing summary. Initial ethos--the attitude or perception of the audience toward the source at the beginning of the communicative act. Perceived organization--the evaluation of the organizational pattern of the message. Retention--the ability to recognize and recall Specific facts presented in a message. Attitude-~the tendency or predisposition to re- Spond favorably or unfavorably toward an object or class of objects. Terminal ethos--the attitude or perception of the audience toward the source of communication at the end of a communicative act. five ite cation review final 5 CUBlng P9rce 8:110 *r! [U 14 The remainder of this chapter will be devoted to five items: a) a statement of hypotheses, b) a justifi- cation of the project, c) the limitations imposed, d) a review of the relevant experimental literature, and e) a final summary. Statement of_§ypotheses The use of two independent variableS--verbal cueing and initial ethos, and four dependent variables-— perceived organization, retention, attitude, and terminal ethos--results in the following hypotheses: H1: Accurate cueing (AC) produces higher perceived organization of the message than does no cueing (NC) or inaccurate cueing (IC) and no cueing produces higher perceived organization and does inaccurate cueing. H a: Accurate cueing produces higher perceived organization of the message than does no cueing. H b: Accurate cueing produces higher perceived organization of the message than does in- accurate cueing. H c: No cueing produces higher perceived,organ- ization of the message than does inaccurate cueing. 15 High credibility produces higher perceived organ- ization of the message than does low credibility. Accurate cueing produces more retention of the message than does no cueing or inaccurate cueing and no cueing produces more retention of the message than does inaccurate cueing. H a: Accurate cueing produces more retention of the message than does no cueing. H b: Accurate cueing produces more retention of the message than does inaccurate cueing. H c: No cueing produces more retention of the message than does inaccurate cueing. High credibility produces more retention of the message than does low credibility. Accurate cueing produces more positive attitude change (in the direction advocated by the Speaker) than does no cueing or inaccurate cueing and no cueing produces more positive attitude change than does inaccurate cueing. H a: Accurate cueing produces more positive attitude change than does no cueing. H b: Accurate cueing produces more positive attitude change than does inaccurate cueing. H c: No cueing produces more positive attitude change than does inaccurate cueing. High credibility produces more positive attitude change (in the direction advocated by the speaker) than does no cueing. Accurate cueing produces higher terminal ethos than does no cueing or inaccurate cueing and no cueing produces higher terminal ethos than does inaccurate cueing. 16 H a: Accurate cueing produces higher terminal ethos than does no cueing. H b: Accurate cueing produces higher terminal ethos than does inaccurate cueing. H c: No cueing produces higher terminal ethos than does.inaccurate cueing. H : High credibility (initial ethos) produces higher terminal ethos than does low credibility. Justification of the Project At this point there should be little doubt in the reader's mind that the arrangement of a Speech is an in- tegral part of rhetorical theory. What was true when the systematic development of rhetoric was beginning, is just as true in 1969. In the preceding pages a rough attempt has been made to justify this project. The purpose of this investigation, as stated above, is to examine the Specific effects of the methods of clarifying the struc— ture of the message (verbal cueing) and initial gth9§_upon the listener. This research project in its present form seems justified for three reasons. First, a good amount of advice is available to the student of public Speaking concerning methods of 17 verbal cueing. Wayne Thompson in his Fundamentals of Communication_(l957) reports that an old time parson once commented, "First I tells 'em where we're going. Then I tells 'em where we is. And then I tells 'em where we's done been!" (p. 262) Other recent texts seem to echo this advice. By following a well-planned outline, the Speaker continually indicates the direction and progress of his thought. Every succeeding idea is related to the previous one, builds on it, and inevitably leads to the next. As he Speaks, his audience is able to follow the develOpment of his ideas, relate them prOperly, and see clearly where the speech is leading (Strother, p. 142). Monroe puts the point more succinctly. Keep the audience constantly aware of where you are, where you have been, and where you are heading (p. 188). One reason for pursuing the present research is to answer the simple question: Is such advice justified? Do the methods of verbal cueing have any effect upon the listener? Second, the present project can be justified as an attempt to affirm or negate the reservations concerning the use of verbal cueing. Up to this point the writer has tried to report accurately the intent of contemporary 18 rhetoricians concerning the use of verbal cueing. There are, however, some reservations lurking between the lines of advice which these writers quoted have suggested. These reservations provide principal justification for the present study. The advice concerning the use of the preview seems to depend upon the purpose of the communicator, whether the intent is informative or persuasive. Hance, Ralph, and Wiksell wrote: "If the purpose of the talk permits, the Speaker should also work a 'preview' of the purpose into the introduction" (p. 198). Based on research by Kiesler and Kiesler (1964), McCroskey (1968) suggests "that the communicator often should not preview the body of the message in his intro- duction if the audience is not yet aware of his persuasive intent" (p. 190). IS it possible that a preview might actually hinder persuasion? After suggesting the use of signpost language (first, second, and third), Soper concludes, "But what you gain in clarity by this method you may lose in in- terest" (p. 203). In an earlier chapter SOper writes, "The repetitive use of first, second, and in conclusion 19 is appropriate when enumerative precision is desired. But greater conversational informality is ordinarily prefer- able" (p. 80). Is it possible that such Signpost language might make the message less interesting? Based on the advice of contemporary textbooks and the reservations expressed above, it seems justified to test under controlled conditions the effects of verbal cueing in a persuasive message. The third justification for the project comes from the use of initial ethos as a second independent variable. Since the time when Aristotle wrote It is not true, as some writers on the art main- tain, that the probity of the Speaker contri- butes nothing to his persuasiveness; on the con- trary, we might almost affirm that his character (ethos) is the most potent of all the means to persuasion (Cooper, 1960, p. 9), rhetoricians have SUSpected the power of source credi? bility in persuasion. Anderson and Clevenger (1963, p. 77) were able to conclude that the ethos of the source is related in some way to the impact of the message . . . . Research shows that expert opinion may be about as in— fluential as majority opinion in inducing atti- tude change. Likewise, McCroskey and Dunham (1966, p. 463) concluded, "The research and analyses reported here lend support to u" L, -'V‘ 20 the theory that ethos is a very important factor in per- suasive communication." But more important is the state- ment of McCroskey and Dunham that ethos is a confounding variable. Concerning "confounding" variables, Kerlinger (1965) wrote: The term "confounding," often used in treatises of statistical research design, means the "mixing" of the variance of one or more in- dependent variables, usually extraneous to the research purpose, with the independent variable or variables of the research problem. As a re- sult it cannot be clearly said that the rela- tion found is between the independent variables and the dependent variable of the research, or between the extraneous independent variables and the dependent variable, or both (p. 308). McCroskey and Dunham continued: If ethos can confound research on evidence, why can't it also confound research on other variables? And is there any theoretical ground for supposing it would not do so in all cases where ethos of any sort could operate? . . . On the basis of our own experience with dis- crepant findings, we would be very hesitant to accept any findings of research in persuasive communication unless the ethos level were mea- sured (p. 463). These conclusions seem to supply a third reason for the present research. In the belief that ethos might be a confounding factor, it was included as an independent variable in the study. Kerlinger suggested: 21 An extraneous variable can be controlled by building it into the research design as an assigned variable, thus achieving control and yielding additional research information about the effect of the variable on the dependent variable and about its possible interaction with other independent variables (p. 285). The inclusion of ethos as a second independent variable not only avoids the effect of source credibility as a confounding factor, but supplies the answer for two questions: What effect does credibility have on the four dependent variables? Does credibility interact with verbal cueing? Limitations Imposed At least two limitations were imposed in the 'present study. The first was the decision to use a written message.’ There are some good reasons for pro— ceeding with a written message. In the first place, this investigator was convinced that a tighter study could be performed as described. Although some might argue that verbal cueing would be more important in an oral message than in a written message, investigation of this variable has to begin somewhere. Given the time and resources 22 available, it was believed that more sound conclusions could be drawn from the present study than from one in- corporating the use of an oral message. Second, because verbal cueing (emphasizing the organizational pattern of the message) is so directly re- lated to even minor aspects of delivery, reliance upon an oral message could easily lead to false conclusions. A gesture, a facial expression, a tone of voice, any number of delivery factors, could cause one to reject (or accept) an hypothesis in error. It was hOped that the use of a written message would avoid this pitfall, and would come closer to measuring the effect of verbal cueing, as defined above. A second general area of limitation has to do with devising the messages with various levels of verbal cue- ing. In terms of possibilities for cueing, at least three broad areas exist: preview, develOpment, and summary. In other words, there are opportunities for cueing in the introduction, in the body, and in the conclusion of the message. II. III. 23 Possibilities Preview A. General 1. None 2. Inaccurate B. Specific 1. None 2. Inaccurate a. More or less b. Out of order Development A. Internal summaries 1. None 2. Weak or inaccurate B. Transitions 1. None 2. Weak or inaccurate Final Summary A. General summary 1. None 2. Inaccurate B. Specific l. Incomplete 2. Out of order Examples ("I have three points.") (No preview given) (Preview of four points, a different number 1 developed) , Q: ("My three points are: (1) . . . .") (Preview of four points, more or less developed) (In relation to how points are develOped) 24 With all of these choices available, an almost endless number of messages could be devised. Therefore, some limitations had to be imposed. It was decided that the main effects of cueing could be measured with the use of three messages: an accurate cueing version, a no cue- ing version, and an inaccurate cueing version. The preceding pages have included what might be considered an informal review of the literature, with both classical and contemporary references to organiza- tion of the message. Before proceeding to a description of the method used in the present experiment, it seems apprOpriate to take a more formal and detailed look at the available experimental literature relevant to this study. Review of Experimental Literature We are concerned with the effect of two indepen- dent variables (verbal cueing and initial ethos) upon four dependent variables (perceived organization, retention, attitude change, and terminal ethos). There is no one <31ear-cut area of research to which we can turn for all .‘ 2‘. I -, ”a of the we she] and sec most c] :3 effects the nif cat nes jut cat ere 0f die gre con 25 of the-answers. In an effort to achieve maximum clarity, we shall examine first, the effects of source credibility, and second, the effects of those areas of communication most closely related to verbal cueing. Effects of Source Credibility Hovland, Janis, and Kelley in Communication and Persuasion (1953), one of the earliest summaries of the effects of source credibility, wrote: In summary, the research evidence indicates that the reactions to a communication are Sig- nificantly affected by cues as to the communi- cator's intentions, expertness, and trustworthi- ness. The very same presentation tends to be judged more favorably when made by a communi— cator of high credibility than by one of low credibility. Furthermore, in the case of two of the three studies on credibility, the imme- diate acceptance of the recommended opinion was greater when presented by a highly credible communicator (p. 35). With notable reservation, Anderson and Clevenger (1963) indicated: "The finding is almost universal that the ethos of the source is related in some way to the impact of the message" (p. 77). A year later, Arthur R. Cohen, in his book 55;;- tude Change and Social Influence (1964), began his second 4c 26 chapter on the characteristics of the communicator, with these statements: Who says something is as important as wha£_ is said in understanding the effect of a com- munication on an attitude. How the listener perceives the communicator can affect attitude change in numerous ways (p. 23). In light of these three statements, one of the most unexpected conclusions concerning gghgg is that of Wayne N. Thompson. In Quantitative Research in Public Address and Communication (1967), he wrote: "The avail- able evidence does not support the position that ethos increases persuasiveness in speech situations" (p. 59). Thompson is making two points. First, most of the studies concerned with the effects of gghgg upon persuasion did not use oral communication. Second, results from the few studies using oral communication are inconclusive. Results from the Hovland and Mandell (1952) study showed that there appeared to be no differences in atti- tude between subjects (SS) who heard a Speech by a highly credible source and those who heard the speech by the low credible source. Gilkinson, Paulson, and Sikkink (1954) found that Speeches in which quotations were attributed to h~s-—.__--q—. 27 well-recognized authorities and those in which the attri- butions were removed did not differ significantly on atti- tude shift, retention, or convincingness ratings. Sikkink's study (1956) Showed that authority and nonauthority presentations did not differ significantly in attitude shifts or in ratings of persuasiveness. Ludlum (1958) found that attempts at increasing gthgg actually backfired. However, he did not validate his variable by measuring gphgg as received by the listeners. There are at least four studies which lead to the opposite conclusion. In 1949 Haiman found that variations in ethical appeal significantly affected persuasiveness. Wayne Thompson (1967) wrote: It is worth considering that the one study [Haiman, 1949] that found the effects of ethos to be Significant was very carefully conducted, but the failure of other experimenters to con- firm this result also is impressive (p. 60). Another study reporting positive results, not men- tioned by Thompson, was the research by Hovland and Weiss (1951). Change of attitude in the direction advocated by the communication was greater when it origi- nated from the highly credible source than when it came from the low one (Cohen, p. 24). 28 In 1953 Kelman and Hovland reported: The data on attitude are consistent with the perceptions of the source: the group hearing the communication from the positive source favored more lenient treatment than those hearing it from the negative source (Cohen, p. 25). In a study by Paulson (1954) ephpg was signifi- cantly related to Shift of opinion for men, but not for women. Granted, the results of these studies are not totally conclusive. The evidence for the effects of credibility upon attitude change with a written message seems stronger than with an oral message. Nevertheless, Cohen concluded: The data from these different experiments Show with reasonably good agreement that vari- ations in the credibility of the communicator do indeed determine variations in attitude change (p. 29). The relationship between source credibility and retention seems more conclusive. One might hypothesize that a receiver would be more inclined to understand and remember a greater amount of a message presented by a ihighly credible source than by a low credible source. Ikmvever, experimental evidence fails to confirm this. _f' 29 Wayne Thompson (1967) wrote, The most striking of the six generalizations . . . is the statement that ethos does not affect learning . . . . Factors such as mo- tivation, selection and arrangement of mate- rials, and clarity all seem to have much more to do with retention than does the ethos of the Speaker (p. 58). Charles Petrie (1963) reached the same conclusion when he wrote, "There is little experimental support for the assumption that source credibility or source sincerity influences the amount of information learned and retained from an informative Speech" (p. 81). Effects of Verbal Cueing The second independent variable of major concern is verbal cueing. To predict accurately the effects of verbal cueing, we Should consider the research performed in five areas: forewarning, verbal emphasis, conclusion- drawing, organization, and three studies directly measur- ing the effects of verbal cueing. 30 Forewarning One aspect of verbal cueing is a preview or ini- tial summary of what is to follow in the body of the :* Speech. Those rhetoricians who offer reservations con- cerning the use of the preview base their advice primarily on research concerning forewarning the audience of the intent of the communicator. Their reasoning is as follows. If the audience is hostile to the position of the commun- icator, then it would be better to "sneak up" or use an inductive approach. The use of a Specific preview would be counter to this advice. Allyn and Festinger (1961) found that SS who were forewarned of the nature of the communication changed their opinion less and rejected the communicator as biased to a greater degree than unprepared SS. Kiesler and Keisler (1964) confirmed the results obtained in other experiments that warning the subject about the intent of a communication has the effect of rnallifying the persuasive influence of a communicator. 31 Verbal Emphasis Another area somewhat closely related to verbal cueing is the work done with verbal emphasis. Arthur Jersild in a study reported in 1928 and 1929 presented 70 statements of fact to some 253 SS. The investigation was designed to test the relative effectiveness of var— ious degrees and forms of primacy, recency, frequency, and vividness on the immediate memory of meaningful mate— rial. The material consisted of 70 narrative statements presenting a biographical sketch of a victitious char- acter. The most effective devices included: "five, four, and three distributed repetitions, first degree primacy, proactive verbal emphasis ('Now get this'), and two widely distributed repetitions" (Jersild, 1929, p. 69). Ray Ehrensberger (1945) attempted to relate the Jersild finding to a more realistic public speaking situ- ation. His five most significant findings are as follows: 1. The most effective means of assuring reten- tion of a statement is to precede the state- ment by pro-active emphasis such as "Now get this." 2. The immediate repetition of a statement early in the discourse assures high retention. 32 3. The immediate repetition of a statement late in the discourse assures high retention. 4. The difference between immediate repetition early in the discourse and immediate repe- tition late inrthe discourSe is not signif- icant. 5. Three scattered repetitions are very effec- tive and more significant than two or four (p. 111). It is the belief of this investigator that forms of verbal cueing (the preview, internal summaries, tran- sitions with signpost language, and the final summary) are similar to the successful forms of verbal emphasis. If these modes of emphasis are successful in producing reten- tion, then certain aspects of verbal cueing should also produce higher retention than would a message without these elements. Therefore, a higher amount of retention can be predicted for the message incorporating the use of accurate cueing. Qonclusion:prawing Several researchers have been concerned with the question: kShould the speaker draw a conclusion or let the audience draw its own conclusion? 33 Hovland and Mandell (1952) reported that over twice as many SS changed their Opinions in the direction of the position advocated by the communication when the Speaker drew the appropriate conclusion than when drawing of the conclusion was left to the audience (p. 588). In contrast to the Hovland and Mandell study, Thistlethwaite, De Hann, and Kamenetzky found that "conclusion-drawing and no conclusion-drawing by the Speaker were not differentially effective in changing attitudes" (p. 112). However, conclusion-drawing was more effective in producing comprehension of the intended conclusion. One final study by Tubbs (1968) adds support to the Hovland and Mandell finding. The results of the Tubbs study indicate "that regardless of the audience's degree of commitment, the explicit conclusion was more effective in eliciting attitude change" (p. 117). Believing that use of a Specific final summary is similar to conclusion-drawing, again this investigator has reason to predict more attitude change with an accurate cueing message, where the conclusion is specifically sum- marized. 34 Organization Befbre examining the three experiments which have attempted to measure the effects of verbal cueing, it seems apprOpriate to look at the large body of research which most nearly relates to the present study--organiza- tion. There are at least two areas of research which are closely related to organization which are not considered relevant to the present study by this investigator: the work with particular orders (climax, anti-climax, etc.), and the examination of primacy vs. recency. The most useful work to our present purpose concerns the "quality" of organization rather than the "type" of organization. This review of research on organization will progress from what is considered to be the largest distortion of organ- ization by the researcher, up to the most subtle changes of structure of the message. Donald Darnell (1963) chose the sentence as the basic variable unit. The same sentences compose each message, but their order varies. The sentence was chosen as the unit to be manipulated because the effect of order at letter level and the word level seems obvious and because at least one study at the paragraph level had been reported (p. 97). 35 Seven versions of the message were devised. Cloze procedure was chosen as the measuring instrument. Words were deleted from the message according to some fixed system--from a random start, every fifth word. Blanks of uniform size replaced the deleted words: and the muti- lated message was presented to a group of SS, who attempted to reproduce the missing words. The number of deleted words which a subject replaced was his "cloze score." "Cloze procedure is a highly reliable index of the com- prehension of a written message" (p. 97). Darnell concluded that "the seven forms of the message were significantly different and the differences were in the expected direction" (pp. 99-100). Darnell's study showed "that disorder can affect comprehension ad- versely and that the amount of loss of clarity becomes greater as the degree of disorganization occurs" (p. 100). In terms of organization and disorganization, it seems apprOpriate to consider next the work of K. C. Beighley, who used the paragraph as the basic unit. Beighley (1952b) randomly (with controls) rearranged paragraphs from two speeches, so that he had an organized and disorganized version of each. Comprehension was the . I 36 main dependent variable. Beighley concluded: "It is apparent from these comparisons that organization shows no statistically dependable superiority for comprehension over the kind of disorganization Used in this study" (p. 255). It should be pointed out that at least three dif- ferences exist between the work of Darnell and that of Beighley, besides different messages with different topics. Beighley's messages were oral; he used the random para- graph as disorganization; and he used a different measure for comprehension. It appears that any one of these dif- ferences could account for the different results. Beighley (1954) replicated the original eXperiment and found that with a different pOpulation, again "organ- ization shows no statistically dependable superiority for comprehension over the kind of disorganization used" (p. 253). In the second study it was also hypothesized "that if it were true that an audience achieves about equal com— prehension from the disorganized material because it works harder to make sense of the material, then that audience Inight possibly remember it longer" (p. 253). No support *waS found for this second hypothesis. 37 Smith (1951) devised nine messages by tranSposing the main parts of a well-arranged Speech. He concluded that "the transposition of a single main part of a speech to a position in the sequence other than its normal one does not affect the persuasive outcome of the Speech" (p. 299). When two parts were transposed, however, the differences between the effectiveness of the resultant speeches and that of the respective normal order speech did prOve (in Experiment II) to be significant. .When all of the parts were transposed, with complete randomization of the main part order, the second experiment actually yielded a negative audience reaction. Thus the evidence indicates that with cer- tain tOpics and/or methods of total organiza- tion in the compositional pattern of the ori- ginal Speech, over-all organization per SQ, is an extremely important factor in persuasion (p. 299). These messages devised by Ernest Thompson (1960) have been used in several studies of organization. In order to achieve variation of the struc- ture, each sentence of Communication I was ar- bitrarily treated as a unit. A second communi- cation (Communication II) was prepared by ran— domly re-arranging the units (sentences) of Communication I within each of the eight main points presented in the communication. Thus, while the original order of the main points was 38 preserved, point one preceded point two, etc., the arrangement of the units within each main point was a matter of chance. A third commun- ication (Communication III) was prepared by randomly re-arranging the units of Communica- tion I within each of the three divisions of the communication: introduction, body, and conclusion (p. 62). Results of the Thompson study (1960) were as fol- lows: Subjects who listened to a better struc- tured communication consistently and signifi- cantly scored higher on the immediate retention test than did Ss who listened to a less well- structured communication. The same trend was observed in delayed retention results as well, but the differences were not significant . .l. . The shift of opinion results were inconclusive; no Significant differences were observed (p. 69). A second study using two of the three speeches by Thompson was reported by Sharp and McClung (1966). Sharp and McClung were concerned with the effects of organiza- tion on the Speaker's ethos. The researchers found that students exposed to a disorganized speech thought less of the Speaker after hearing his talk than before he Spoke. In contrast, students liStens ing to the organized address shifted very little in their attitude toward the Speaker (p. 183). A third study employing these same Speeches by Thompson is reported by McCroskey and Mehrley (1969) . These researchers were interested in the effects of 39 nonfluency and disorganization upon attitude change and source credibility. They found that "the well-organized, fluently presented message produced significantly more attitude change than another condition, and the other three conditions were not significantly different from each other" (pp. 18-19). In other words, the presence of either serious disorganization or extensive nonfluencies was sufficient to significantly reduce the amount of attitude change produced by a Speaker; but the presence of both of these detracting elements produced no greater reduction of attitude change than did the presence of either one alone (p. 20). Eldon Baker (1965) devised a Speech which advo- cated the legality and social acceptance of euthanasia. The only variation in content between the two forms of the message was the inclusion of eighteen Speaker dis- organization cues. These cues, or verbal/non-verbal suggestions of the Speakers' difficulty in presenting his ideas systematically, included apologies for not remembering the next point, apologies for including an idea out of sequence, explanations for not being able to place materials in prOper sequence, and long Silent pauses (p. 150). Baker found that perceived Speaker disorganization cues did p93 Significantly reduce the credibility of the Speaker . . . . There wa§_a Significant in— crease in Speaker credibility in the absence 40 of Speaker disorganization cues . . . . Speaker disorganization cues did not sig- nificantly or differentially affect audi- ence attitude change (pp. 160-161). In summary, the studies dealing with the organ- ization of messages indicate that there is rather strong support that organization and disorganization can affect the amount of retention of the message by the receiver. There is only moderate support that organization has an effect on Speaker credibility or attitude change. Verbal Cueing We shall complete our review of the literature by looking at three studies which have been concerned with verbal cueing as defined in the present study. Thistlethwaite, De Haan, and Kamenetzky (1955) first concerned themselves with methods of verbal cueing as they hypothesized that well organized communications (those with liberal Sign posts and guides for orientation) would aid the listener in "getting the point," and thereby increase the effect of the commun- ication in changing attitude (p. 107). Two conclusions were reached: 41 The programs with well—defined organizations of content similarly produced greater comprehen- sion than the programs having poorly defined organizations of content . . . . There were no significant over-all differences between exper- imental treatments in effects upon attitude scores (p. 113). Parker (1962) probed the effectiveness of four widely used rhetorical elements: tOpic sentence, cross headings, the beginning summary, and the closing summary. The use of tOpic sentences increased comprehension when the SS were tested immediately. No effect was found for delayed recall. College freshmen generally were unaware of the presence or absence of the organizational devices included in the study. Over—all, the more devices used, the more interesting the readers found the material. "Some readers, however, noted the presence of both begin- ning and concluding summaries and found the use of both objectionable" (p. 29). The most recent study concerning the use of verbal cueing was performed by Ernest Thompson (1967). This is the fourth experiment using the Thompson speeches devised in 1960, and referred to above. Thompson concluded that the addition of statements to highlight relationships among units in a speech can enhance comprehension (p. 56). 42 W A careful review of the literature can be just as frustrating as it is valuable. There are few rhetorical factors which can be counted on always to make a differ- ence. Whereas most contemporary public Speaking textbooks would lead us to believe that in most cases the use of verbal cueing is well advised, experimental literature would lead us to conclude that in many cases verbal cue- ing would have little effect. In light of this review of the experimental liter— ature, one would expect that high credibility would pro- duce a significant increase in attitude change, but onld most likely not affect retention of the message. On the other hand, verbal cueing is likely to produce greater retention, but most likely will not produce a significant increase in attitude change. The effects of credibility and cueing on the other dependent variables is difficult to predict. CHAPTER II METHOD This chapter includes a full discussion of the following: a) design of the study, b) subjects for the experiment, c) procedures of the study, d) measurement, and e) methods of statistical inference. Experimental Design This study employed a 2 X 3 factorial analysis of variance design. Lindquist (1953) described factorial designs as "those in which there are two or more cross- classifications of treatments, or in which the effects and interactions of two or more experimental variables are simultaneously observed" (p. 8). Use of this fac- torial design enabled the experimenter (E) to observe the possible main effects for two levels of credibility and three levels of cueing, as well as the possible interac- tion effects between these variables. 43 44 In the research study reported by this thesis, two variables were manipulated (cueing and credibility) and four variables were observed (perceived organization, retention, attitude change, and terminal ethgs). The following section will offer an Operational definition of each of the variables of the study. Verbal cueing was defined in Chapter I. For this experiment three levels or degrees of cueing were created: accurate cueing, no cueing, and inaccurate cueing. l. Accurate cueing includes: the use of a specific preview ("Specifically, I would offer three indictments . . . .") the use of signpost language (First, U.S. aid has been . . . . "A second area of concern is . . . . "The third indict- ment of U.S. aid . . . .") the use of internal summaries ("Because of lack of careful plan- ning, and . . .‘. "Second, then, I would contend "In this case U.S. aid actually backfired . . . .") 45 the use of a specific final summary ("In summary, I have suggested three problems.") No cueing--the omission of all devices used in the accurate cueing version. Inaccurate cueing includes: the use of an inaccurate preview the use of some Signpost language (Specifically, I would offer four indictments . . . ." Four indictments were previewed out of order, and only three were ac- tually develOped.) ("First, United States foreign aid . . . ." The second indictment was introduced but never devel» oped. The third indictment was not highlighted until the internal summary. The fourth point was introduced by "final" rather than by "fourth." The last internal summary included the word "final" rather than "fourth." The closing summary was general--"I 46 have suggested several problems . . . ." Only two points were mentioned in the final summary. Four points had been previewed and three points had been devel- oped.) High initial ethos—-attributing the message to the former U.S. Ambassador to Brazil, Thomas A. Wilson. Low initial ethos--attributing the message to the Red Chinese Ambassador to Brazil, Lin Tai. Perceived organization-~the degree of agreement with the statements: "The Ambassador's Speech was very well organiZed," and "The main points develOped by the Speaker stood out very clearly." Retention-~the score on a l3-item multiple choice examination (subsequently reduced to 10 items as a result of item analysis) administered after the message was presented. Attitude toward the tOpic (Attitude Change)--the score on a set of semantic differential scales 47 measuring attitude toward the statement "The U.S. has a well administered foreign aid program for Brazil." 9. Terminal ethos-—the score on six semantic differ- \ ential scales for each of three dimensions of .....—-‘:' ethos: authoritativeness, dynamism, and character. ) Subjects The purpose in choosing subjects (SS) for the ex- periment was to select subjects, as nearly as possible, who were similar to those persons in the "real“ world. With one eye on generalizability, with a second eye on practicality, and a third eye, figuratively speaking, on the factors which might affect the main effects of the study, an effort was made to secure 83 with a minimum amount of Speech training. This is not meant to imply that a majority of persons in the "real" world have not had some Speech training, but at least that training has not come in the last several months, or even years. In the belief that such recently trained Speech students 48 might be sensitized to the manipulated and observed var- iables, beginning Speech students were sought. The Ss for the experiment were students enrolled: in five different Speech classes1 during the first six- week session in the 1969 summer quarter at Iowa State University. Some 257 students, 140 males and 117 females, participated in the experiment. The vast majority (73%) of these SS were enrolled in one of ten sections of public Speaking (Speech 211). Twenty-three students were en- rolled in another beginning class, Speech 30. The re- maining 47 SS were enrolled in advanced Speech courses: 305, 312A, and 336A. The SS were randomly assigned to one of the six treatments of the experiment. As mentioned above, 257 SS participated in the study. Only 15 students were drOpped from the experiment, leaving a total sample of 242. Six students were drOpped from the study because they circled two answers for one of the retention questions. Three SS were drOpped because 1Speech 30—-Public Speaking for students in Iowa State University's Technical Institute Speech 211--Fundamentals of Speech (Public Speak- ing) Speech 305-—General Semantics Speech 312A--Business Speaking Speech 336A--Group Discussion 49 they omitted one or two of the retention questions. Likewise, three 85 were not included in the final analysis because they omitted an entire page of the questionnaire. One student was dropped because of an inability to finish after a twenty—five minute attempt.2 One student was drOpped from the study because of the negative nature of his comments toward the purpose of the research. His answers were little more than an attempt at humor. Fi- nally, one student was drOpped from the study because of his knowledge of the purpose of the experiment. This student had participated in the pre-tests and it was be- lieved best not to include his responses in the analysis. After attrition, the total number of SS was 242, 94%.of the original sample. Procedures of the Experiment A four-page message indicting the United States foreign assistance program for Brazil was devised. This 2Average administration time for the project was eighteen minutes. .M 4‘: n 50 particular topic was chosen in an effort to obtain a topic toward which the SS would have an essentially neutral attitude and little specific knowledge. Although Ss might possess some definite attitude toward the United States foreign assistance program in general, or even toward U.S. \1 aid for Latin America as a whole, it was believed that by D '.""“-.-—_. .._-—— f' ' I concentrating upon one Specific country, Brazil, the ex- perimenter would be relatively safe in assuming a neutral topic. Persons familiar with the events of June, 1969, might be inclined to ask if factors other than the exper— imental treatment might have influenced attitudes toward U.S. aid to Brazil. One such factor was the publicity prior to and during Governor Nelson Rockefeller's trip to South America in the early part of JUne, 1969. How- ever, Governor Rockefeller's visit to Brazil did not occur until after the experiment was completed. At any rate, any negative effect toward U.S. aid to Latin America caused by this bad publicity must be assumed to have been randomly distributed throughout all six treatment groups. From the basic 1,226-word message, which contained no verbal cueing of the organizational pattern of the 51 message, the E devised the accurate cueing version (1,311 words) and the inaccurate cueing version (1,336 words). The accurate cueing version, the no cueing version, and the inaccurate version are shown in Appendix A. A cover page was prepared, attributing each of the three versions of the message to one of two fictitious, but believable-sounding, sources. As previouSly stated, the high-credible source was Thomas A. Wilson, former U.S. Ambassador to Brazil, and the low-credible source was Lin Tai, Red Chinese Ambassador to Brazil. In all, there were six treatments: Size (n) of reatment Ee Grpgpg ‘Treatment Groups 1. High—Credible Source--Accurate ‘Cueing 45 2. High-Credible Source--No Cueing 41 3. High-Credible Source--Inac- curate Cueing 4O 4. Low-Credible Source-~Accurate Cueing 39 5. Low-Credible Source--No Cueing 35 6. Low-Credible Source—~Inaccurate Cueing 42 A sample packet of treatments 1 and 5 can be found in Appendices B and C. 52 Instructors of 14 sections of Speech gave permis- sion for the experimenter to visit their class (or classes) during one of the first three days of the first summer session, June 4-6, 1969. Two persons, the E and an asso- ciate, administered the experiment. The following intro- duction was delivered extemporaneously by both the E and his associate. First, I want to thank (Miss, Mrs., or Mr. Instructor] for allowing us to conduct this bit of research. As is stated on the cover sheet of the packet which you will receive in just a few minutes (and I hOpe you'll read that cover sheet carefully), the Department of Speech is interested in isolating some Of the factors that contribute to successful communication. You can help us by Simply reading this four- page message and then by answering some ques- tions on what you've read. So that everyone is playing by the same rules, we ask that you not look ahead at the questions before you read the speech. Second, we ask that you not look back at the speech once you have started to answer the questions. After administering the experiment to about six sections, it was noticed that several students had acci- dently skipped an entire page of the questionnaire. Sub- sequently, the following sentence was added to the in- structions. "And we would ask that you be very careful not to skip any pages." The instructions continued: 53 Now, again we ask that you read the message and answer the questions. Are there any ques- tions before you begin? At this point the packets were distributed. The average time for SS to complete the experiment was 18 minutes, with a range of 14 to 22 minutes. In one case, mentioned above, a subject had not completed answering the questions after twenty-five minutes. The S's form was taken up and discarded from the study. After all sections had completed participation in the experiment, the SS were debriefed. A one-page explanation of the study was given to the instructor of each class used in the study. The instructors were asked to read the explanation of the study to their classes within the next week after completion of the project. A Spot check revealed that all had complied. The debrief- ing statement included the following items: 1. It was explained that both the Speakers (Wilson and Tai) and the message were fictitious. 2. The purpose of the study was eXplained. 3. The hypotheses were briefly presented. 54 4. For those interested in additional information, an address was given. 5. The Ss were thanked for co-Operating. Measurement In order to test the hypotheses of this study, four separate measures were necessary: a) a measure of the perceived organization of the message, b) a measure of the retention of the message, c) a measure of the 85' attitude toward the tOpic of the Speech, and d) a measure of the terminal epppg of the source. Perceived organization of the message was measured by asking the SS to check a point on a five-point scale ranging from "strongly agree" to "strongly disagree," in response to two statements: "I would say that the Ambas- sador's speech was very well organized," and "The main points develOped by the speaker stood out very clearly." Retention of the message was measured originally 12y 13 multiple choice items (Questions 3 through 15, shown iJa Appendices B and C). After the submission of the 55 questions to an item analysis, the top 10 questions were used, and a second retention score was obtained for each S. To secure a measure of the 53' attitude toward the topic of the speech, students were asked to "Please rate the following statement (The U.S. has a well admin- istered foreign aid program for Brazil.) by placing a check mark on the following scales:" Right-Wrong, False- True, Yes-No, Incorrect-Correct, I agree-I disagree, and Foolish—Wise. The final dependent variable to be measured was the terminal ethos of the source. Semantic differential type scales to measure terminal ethos of the Speaker were chosen from previous research by McCroskey (1967). They included six scales for each of the three dimensions of source credibility. The authoritativeness or competence scales were: Informed-Uninformed, Unqualified-Qualified, Reliable-Unreliable, Worthless-Valuable, Intelligent- Unintelligent, and Inexpert-EXpert. The six scales used to measure dynamism were: Aggressive-Meek, Hesitant- EMphatic, Forceful—Forceless, Timid-Bold, Active-Passive, sand Tired-Energetic. The character dimension of terminal _-.¢ 56 ethos was measured by Unselfish—Selfish, Awful-Nice, Friendly-Unfriendly, Dishonest-Honest, Pleasant-Unpleasant, and Sinful—Virtuous. Statistical Inference ian/ .4 91 Statistical analysis was performed in three steps. Before the eXperiment three checks were completed: a test of topic neutrality, a pre—test of the credibility scales, and an evaluation of the messages. After the experiment several preliminary analyses were run on the data. Finally, the data were submitted to a major statistical analysis. The procedures used in these three stages of analysis will be discussed in this section. The results of the prelim- inary checks will be revealed here. The results of the (other analyses will be covered in the next chapter. Preliminary Checks The purpose of the first check was to see if the tozpic chosen was a relatively neutral one. Fourteen stu- derrts were asked to rate the statement, "The U.S. has a 57 well administered foreign aid program for Brazil," on the six semantic differential type scales noted above. The results of this pre-test, administered two weeks before the actual study, revealed that the tOpic was indeed one toward which students held a neutral attitude. As shown in Appendix D, Table 11, the mean attitude score was 24.2, on a scale ranging from 6 to 42, with 24 as the neutral point. Students seemed to be neutral toward the topic chosen for the experimental messages. It was hOped that this neutrality would allow for maximum attitude change. Second, a check was made on the credibility of the two sources used in the study. The fourteen students were asked to rate both Thomas A. Wilson, said to be the former U.S. Ambassador to Brazil, and Lin Tai, supposedly the Red Chinese Ambassador to Brazil, on the eighteen semantic differentail type scales mentioned above. The results of this pre-test are shown in Appendix D, Table 12, .Although no statistical tests for Significant differences were nun on this data, the direction seemed clear. Once again scales with a total range from 6 to 42 were used. The mean rating, across dimensions, was 29.0 for Wilson and 26.5 for Tai. On authoritativeness and character 58 Wilson was judged 2.8 and 5.8 points higher, respectively, than Tai. On dynamism Tai was rated 29.7 to Wilson's rating of 28.5. These results on dynamism are not the first indication this E has received that high dynamism is not necessarily correlated with high credibility. The results of this pre-test were not as clear- cut as hOped in terms of the credibility of Wilson and Tai. However, these sources, with titles, had been used in a previous study3 in which the data indicated that the Ss did perceive a significant credibility difference would occur. Finally, the E asked his colleagues to evaluate the three messages. This evaluation included two steps. First, the three messages were randomly distributed to eighteen English and Speech instructors who were asked to rate the message they received on the following seven- point scales: Clear-Unclear, Not interesting-Interesting, Poorly organized—Well organized, Highly persuasive-Not 'persuasive, Poor reasoning-Good reasonsing, and Good Evidence-Poor Evidence. It was expected that the accurate —— 3McCrockey—-results submitted for publication at thiS‘writing. ‘h. A -« a 1 59 cueing message would be judged more organized, and would be judged to possess more clarity than the no cueing or the inaccurate cueing message. The other scales, besides organization and clarity, were included to cover the pur- pose of the evaluation. Again statistical tests were not used to determine significance, but trends can be noted. In Appendix D, Tables 13 through 16, across all the scales, the messages were ranked as hypothesized: accurate cueing, then no cueing, and finally, inaccurate cueing. However, the pply_single scale reflecting a similar ranking is Clear-Unclear. On three of the six scales (Highly persuasive-Not persuasive, Poor reasoning-Good reasoning, and Good evidence-Poor evidence) the no cueing message was rated highest. The clarity of the message seems to have little "halo" or carry-over effect for this "expert" audience. The other important finding from this faculty evaluation was that although the accurate cueing version 'was rated as being most organized, the inaccurate cueing version was judged to be better organized than was the no cueing version. This result was counter to the pro- posed hypothesis. A second stage of this evaluation was to explain t0 the instructors that the purpose of the study was to 6O examine the effects of verbal cueing which had been de— fined as highlighting or emphasizing the organizational pattern of the message. The "experts" were asked to read the three messages and identify each version. Of the eight instructors who participated in this phase of the 1, evaluation, seven of them accurately identified all three a “-«a- “mf- -! of the speeches. As shown in Appendix D, Table 17, one instructor reversed the inaccurate and the no cueing ver- sions. In summary, what did the preliminary checks ShOW? First, a neutral tOpic had been chosen for the study. Second, students rated Ambassador Wilson higher on author- itativeness and character than they rated Ambassador Tai. Third, when asked to read all three messages, a majority (7 out of 8) of the faculty were able to recognize accur- ate cueing, no cueing, and inaccurate cueing. When the faculty were asked to read and evaluate only one of the three messages, the combined ratings showed that the accurate version was judged more clear and more organized. The inaccurate version was judged more organized, but less clear than the no cueing version. It would seem that while cueing gives some semblance of organization, bad cueing seems to detract from clarity of the message. 61 Preliminary Analyses Once the data had been gathered, several prelimi- nary analyses were necessary. First, the 13 multiple choice questions were submitted to an item analysis. With emphasis on the difficulty index5 and the discrimina- tion index6 for each question, the top ten items were chosen, and another score for each S was derived. Second, 4Examination of responses of a class to a given question or set of questions is called an item analysis. From an item analysis, easy questions, difficult questions, and non—discriminating questions can be identified. In Short, item analysis is a technique for investigating and improving the correlation or degree of relationship among items. The program used in the present study was devised by John W. Menne and provided by the Student Counseling Service at Iowa State University. The program is explained in full in Student Counseling Service Reports 68-01 and 68—12. 5"The Difficulty Index is simply one minus the per~ centage of those who get the item correct. Thus, for an easy item the percentage will be high and the Difficulty Index will be low; for a difficult item the percentage will be small and the Difficulty Index will be high . . . . The average item difficulty for a good test should be mid-range, from .4 to .6" (Menne, 1968b, p. 2). 6The Discrimination Index is found as follows: Rank the raw scores of an examination and divide them into two equal groups, tOp and bottom. Count the number of those in the tOp group who answered the item correctly. This equals RH. RL equals the number in the bottom group ‘Who answered the item correctly. The Discrimination Index is obtained by dividing RH minus RL by 10 percent of those attempting the item. For a good item this index will be OVer 1.0 (Menne, 1968b, p. 3). 62 the eighteen individual scales for terminal ethos were submitted to factor analysis and varimax rotation. Re- sults of the factor analysis were interpreted as follows: Each item or scale was considered independently. A judgment was made as to which factor, if any, the item "loaded." This decision was based on the relative magni- tude of the factor loadings for the item on each of the factors. The standard established as a criterion was a loading of .60 or above for the factor on which the item received its highest loading, and a loading of .40 or ‘below for the remaining factors. Consider, for example, 'the following hypothetical items, factors, and factor loadings : Item Factor 1 _2_ 2. l .78* .10 .12 2 .05 .79* .02 3 .72 .02 .68 4 .20 .12 .78* The first item above, according to the criterion u£3€3 ‘tihe first factor, since its loading factor under factor °r1€3 .is above .60 and its loadings on all the other factors 63 are below .40. Item two loads on factor two. The third item, however, is split between the first and third factors, and cannot be considered a part of either factor. The fourth item loads on factor three. Once an independent judgment had been made for each item, the scores in each dimension were summed and prepared for the major statistical analysis. Major Statistical Analyses Siegel (1956) wrote that in order to use para- Inetric statistics, the following assumptions must be met: 1. The observations must be independent. 2. The observations must be drawn from nor- mally distributed populations. 3. The populations must have the same vari- ance. 4. The variables involved must have been mea- sured in at least an interval scale. 5. The means of these normal and homosce- dastic populations must be linear combina— tions of effects due to columns and/Or rows The assumption of independence can be met if the Selection of any one case from the population for inclu- SiC>r1 in the sample does not bias the chances of any other i. ~‘vu-‘i . 64 case for inclusion, and the score which is assigned to any case does not bias the score which is assigned to any other case. There is no reason to believe that both of the first two assumptions have not been met in the present study. Siegel added that with the possible exception of the assumption of homoscedasticity (equal variances), "these five conditions are ordinarily not tested in the course of the performance of a statistical analysis. lRather, they are presumptions which are accepted" (pp. ;19-20). When the eXperimental groups are relatively equal 111 size and the size of each group is 15 or more, it has been shown (Boneau, 1960) that even the homogeneity of vailriance can safely be violated in most cases. Both of tllee conditions are met in the present study. The fourth assumption listed by Siegel concerns Kerlinger (1964) wrote tllea use of an interval scale. scales "it is probable that most psychological . triait: He suggested approximate interval equality fairly well." thiit: "the best procedure would seem to be to treat ordinal measurements as though they were interval measurements, TFflMI’ :5 65 but to be constantly alert to the possibility of gross inequality of intervals" (p. 427). It was decided that the semantic—differential type scales used in measuring attitude and credibility, the Likert type scales used to measure perceived organization, and the retention scores contained no gross inequality of interval. The fourth assumption, therefore, was not violated- The final assumption listed by Siegel simply states that the effects must be additive. This, too, ‘was assumed for the present study. To test the hypotheses of this study, two para- Inetric statistical tests were employed: analysis of xnariance and petests. AS stated earlier in the chapter, the analysis of variance test indicates Significant dif- ifexrences among combinations of experimental conditions, as well as significant interactions. The p—test (Winer, 1962) was apprOpriate to estimate between-group prob— abilities. The factor analysis7 and analysis of variance ‘Vealree calculated by means of library computer programs \ 7A. Williams, Factor Analysis, Technical Report Nf>‘- 34, Computer Institute for Social Science Research. Michigan State University, 1967. 8W. L. Ruble, Analysis of Covariance and Analysis 01? ‘feiriance With Unequal Frequencies Permitted in Cells—- JTV"_-___.F-( ' ’ 66 on a Control Data Corporation 3600 computer, in operation at Michigan State University. (LS Routine), STAT Series Description No. 18, Michigan State University Agricultural Experimental Station, 1968. ‘ v 3.1-...“ . CHAPTER III RESULTS This chapter will include three major divisions. First, the results of the preliminary analyses will be cited. Section two will present the results of the major statistical analyses, and the final section will summarize the findings of the study. Preliminary Analyses AS stated in Chapter II, two preliminary analyses were completed: an item analysis of the retention ques- tions and a factor analysis of the credibility scales. Item Analysis Retention of the message was measured by a l3—item multiple choice examination, questions 3 through 15 on the questionnaire. In an effort to secure the most 67 68 discriminating measure of retention, an item analysis of the retention questions was performed. Two criteria were employed in order to choose the tOp questions: difficulty and discrimination. It was decided that because questions 10, 6, and 5 ranked lowest among the 13 on both difficulty and discrimination (See Table 18, Appendix E), the exam- ination should be rescored using only 10 items. This score should represent a better measure of retention of the message. Factor Analysis As mentioned in Chapter II, the scales chosen to nneesure terminal ethos for this study had been employed £111 previous research by McCroskey (1967) and were believed t:c> be reliable measures of three dimensions of source Credibility: authoritativeness or competence, character, Eiraci dynamism. In that previous research, however, an oral nReassage was used. Therefore, it seemed necessary to con- firm the ability of these scales to measure and distinguish IDEFtKMeen each dimension of terminal ethos as affected by a was i tten message . 69 The 18 semantic-differential type scales used to measure terminal ethos were submitted to a factor anal- ysis. Results of the factor analyzed credibility data suggest three dimensions of ethos. Table 19, Appendix E, lists the scales and their factor loadings. It will be noted that only two scales failed to meet the .60 criteria, intelligent—unintelligent (.58), and unselfish-selfish (.57). Since both loadings round off to .6, it was not thought necessary to omit these scales and adjust the scores for authoritativeness and character. AS planned, six scales were used to measure each dimension of terminal ethos. AUTHORITATIVENESS: Informed-Uninformed, Unqualified— Qualified, Reliable-Unrediable, Worthless-Valuable, Intelligent- Unintelligent, Inexpert-EXpert. DYNAMISM: Aggressive-Meek, Hesitant-Emphatic, Forceful-Forceless, Timid-Bold, Active—PassiVe, Tired-Energetic. CHARACTER: Unselfish-Selfish, Awful-Nice, Friendly-Unfriendly, Dishonest- Honest, Pleasant-Unpleasant, Sinful-Virtuous. 70 Major Statistical Analyses A two-way analysis of variance revealed no signif- icant interaction effects between verbal cueing and ini- tial ethos. In this section the main effects of cueing and credibility upon each of the four dependent variables will be reported. Effects Upon Perceived Organization H a: Accurate cueing (AC) produces higher perceived organization of the message than does no cue- ing (NC). H b: Accurate cueing produces higher perceived organ— ization of the message than does inaccurate cueing (IC). H c: No cueing produces higher perceived organization of the message than does inaccurate cueing. Two separate measures of "perceived organization" ‘Mere employed. Ss were asked to evaluate the message in 'ternm of both organization and clarity. The perceived organization analysis of variance (summarized in Table 20, Appendix F) revealed a significant cueing effect (F = 4.26, p <: .05). The results from the p—testsl for significant 1Results of all petests are shown in Table 29, Appendix F . lwfl‘“. 4." ’. O 71 differences among the means for the three cueing versions showed a significant difference between the AC version and the NC version (p_= 2.93, p < .05). NO significant differences were noted between the NC version and the IC version (p.= -1.50, p < .20)2 or the IC version and the AC version (p_= 1.46, p < .20). Mean scores for perceived organization and clarity are shown in Table 1. TABLE 1 MEAN SCORES FOR PERCEIVED ORGANIZATION AND CLARITY FOR CUEING CONDITIONS Accurate Cueing No Cueing Inaccurate Cueing (n = 84) (n = 76) (n = 82) Perceived Organization 1.88a 2.23a 2.05 Means Clarity Means 1.94b 2.25b 2.10 a, . . . . . . bMeans Wlth same subscript are Significantly different from each other, p < .05. In the analysis of variance for clarity (sum- marized in Table 21, Appendix F), similar cueing effects 2Negative pyscores indicate that the difference was in the opposite direction than predicted. In this case the IC version was perceived as being more organized that: the NC version. Differences were not significant. P .‘ -‘.-AA‘_ 72 were noted (F = 2.78, p < .06). Again petests revealed a significant difference between AC and NC (t_= 2.41, p < .05), but no significant differences existed between NC and IC (p_= —l.16, p < .30) nor between the AC message and the IC message (5 = 1.27, p < .30). Therefore, there )4?— is a significant cueing effect upon perceived organiza- tion and clarity of the message. The AC message was per- * ceived to possess more clarity and to be better organized than the NC message. There were no significant differ- ences between the other messages. Support for the first hypothesis (Hla) was found. The other two hypotheses were not confirmed (H b and H c). 1 l H : High credibility produces higher perceived or- ganization of the message than does low credi- bility. AS Shown in Table 2 (and Tables 20 and 21, Appen- Forceless Timid : : : : : : Bold Active : : : : : : Passive Tired _:__:_:___:___:_:__ Energetic Unselfish _:__:___:___:__:___:__ Selfish Awful : ‘fi: : : : : Nice Friendly __:___:__:___:____:__:___ Unfriendly Dishonest : : : : : : Honest Pleasant ___:___:__:__:___:_:___ Unpleasant Sinful : : : : : : Virtuous HAVE YOU SKIPPED ANY SCALES? Please turn to the next page. Appendix C (Cont.) 152 2. Please rate the following statement by placing a check mark on each of the following scales: THE U.S. HAS A WELL ADMINISTERED FOREIGN AID PROGRAM FOR BRAZIL . ‘ Right : : : : : : Wrong g“ False : : : : : : True E Yes : : : : : : No E'* Incorrect : :v : : : fiéz Correct I agree : : : : : : I disagree Foolish : : : ° : ° Wise Please answer the following questions by circling the letter beside the best answer. 3. The Speaker indicated that A. 8. C. D. E. Brazil has reached the point where no more foreign aid is needed. Brazil has made progress, but additional help is needed. Brazil has made little progress since 1961. Brazil has made little progress since 1961 and all is hopeless. All that is needed is more money for Brazil. ‘4. The Speaker stated that since 1953 Latin America has received A. $60 billion B. $35 billion C. $20 billion D. $ 3 billion E. $ 1 billion in in in in in aid from the U.S. aid from the U.S. aid from the U.S. aid from the U.S. aid from the U.S. 153 Appendix C (Cont.) 5. The largest republic in Latin America is A. B. C. D. E. Brazil. Colombia. Venezuela. Argentina. Chile. Brazil has received of United States aid to Latin America. A. 1/5 B. 1/4 C. 1/3 D. 1/2 E. all The Speaker actually developed indictments of U.S. aid to Brazil. A. B. C. D. E. O‘UI-wa Which of the following points was not actually devel- oped by the speaker? A. B. The A. B. C. D. E. U.S. aid has been wasteful. U.S. aid too often supports the leaders rather than the people. U.S. aid is directed to American companies rather than to Brazilian workers. U.S. aid causes resentment. All of the above points were actually develOped by the Speaker. ambassador suggested that All U.S. aid goes toward capital improvements. Too much U.S. aid goes to capital improvements to the detriment of the needs of the people. More money should go to capital improvements. A11 U.S. aid should go to capital improvements. None of the above. 154 Appendix C (Cont.) 10. According to the Speaker, the least populated area of Brazil is near A. Manaus. B. Jurué. C. Campinas. D. Salvador. E. Recife. 11. The ambassador stated that the primary objective of U.S. policy in Latin America is A. reduction of the population explosion. B. to increase the number of democratic governments. C. to increase the number of military rulers. D. to create stability. E. to create "little America's." 12. Which of the following points was not made by the Speaker? A. A large amount of U.S. aid is diverted to mili- tary interests. B. The U.S. is encouraging competition among the South American republics. C. A wave of militarism has been sweeping Latin America. D. U.S. aid has caused economic harm to its recip- ients. E. All of these points were presented. 13. The Speaker said that our economic aid encourages unrest in Latin America because A. All of U.S. aid is in the form of military assis— tance. B. U.S. economic aid enables the countries to meet current expenses of an economy strained by mili- tary demands. C. The U.S. supports only military rulers. D. The military establishments get nervous about ' each other. E. Both B and D above are correct. “‘r" 3.3,??? 7&4; 155 Appendix C (Cont.) 14. 15. The speaker indicated that A. the USS. is tOo concerned with getting credit for its aid. the U.S. should be more concerned with getting credit for its aid. flood victims in Oregon rejected Russian aid. the over-concern with visibility by the U.S. causes resentment. at least two of the above statements are correct. The speaker suggested that resentment among recipient nations is present because A. the U.S. has given too little aid to Latin Amer- ica. the U.S. has put too much emphasis on loans. the countries feel they must maintain their self- reSpect and demonstrate their independence. the countries_are made to feel like beggars. both B and C are correct. Indicate your reaction to the following statements by placing a check mark in one of the five spaces. 16. 17. 18. 19h Strongly Unde- Dis- Strongly Agree Agree cided agree Disagree I would say that the am- bassador's speech was, very well organized. In general, the speech, was very convincing.’ The-main points developed by the speaker stood out very clearly. In general, the ambas- sador's speech was very interesting. - *wafifiv 156 Appendix C (Cont.) 20. I suSpect that the purpose of this research project is: Other comments: Thank you for your co-operation. -‘V . ions “.05 E ‘3 3......“- ....fi 7 APPENDIX D PRELIMINARY CHECKS w-mn A... “1;”. _agfi*' L APPENDIX D: PRELIMINARY CHECKS TABLE 11 SUMMARY OF PRELIMINARY CHECK OF ATTITUDE TOWARD TOPIC 'i. Class Mean Speech 334A 25.2 b Speech 3343 22.4 Total 24.2* *On a scale ranging from 6 to 42, 24.0 is the neutral point. TABLE 12 SUMMARY OF PRELIMINARY CHECKJOF INITIAL ETHOS Source Combined Authoritativeness Dynamism Character Rating Wilson 29.0 29.9 28.5 28.5 Tai. 26.5 27.1 29.7 22.7 158 ‘. E£3 I‘I“ 159 Appendix D (Cont.) TABLE 13 SUMMARY OF FACULTY RATINGS FOR ACCURATE CUEING MESSAGE fl s..- Instructor Clear Interest Organi- Persua- Reason- EVl' zation sive ing dence A 7 6 7 6 6 6 B 7 4 7 5 5 6 C 7 5 6 5 5 6 D 7 5 5 3 5 4 E 6 5 6 2 5 2 F 7 3 7 3 4 6 Total 41 29 38 24 28 28 TABLE 14 SUMMARY OF FACULTY RATINGS FOR NO CUEING MESSAGE —.—_‘ Y— T Instructor Clear Interest Organi- Persua- Reason- EVl' zation sive ing dence G 6 6 6 6 7 3 H 6 4 l 3 3 6 I 5 4 4 3 3 4 J 6 5 3 6 6 6 K 7 2 2 6 7 6 L 7 7 5 7 7 7 {Total .37 28 21 31 33 32 160 Appendix D (Cont.) TABLE 15 SUMMARY OF FACULTY RATINGS FOR INACCURATE CUEING MESSAGE Instructor Clear Interest Organi— Persua- Reason— EVI- zation sive ing dence M 7 5 6 5 6 6 N 3 4 2 4 3 3 O 5 ‘4 2 2 3 4 P 5 4 6 5 5 5 Q 6 6 6 5 6 4 R 6 6 5 6 6 6 Total 32 29 27 27 28 28 TABLE 16 SUMMARY OF FACULTY RATINGS FOR THE THREELMESSAGES Organi- Persua- Reason- Evi- .Message Clear Interest . . . Total zation Sive Ing dence Acc Cue 41 29 38 24 28 28 188 No Cue 37 28 g 21 31 33 32 182 Inacc Cue 32 29 27 » 27 28 28 171 161 Appendix D (Cont.) TABLE 17 SUMMARY OF FACULTY PLACEMENT OF THE THREE MESSAGES Instructor Accurate Cueing No Cueing Inaccurate Cueing a. * OOHQMUOUJ vwmwvnsnszu ZZZZZHZZ HHHHHZHH 100.0% 87.5% 87.5% APPENDIX E PRELIMINARY ANALYSES APPENDIX E: PRELIMINARY ANALYSES TABLE 18 ITEM ANALYSIS RESULTS Question DiffiCU1tY Discrimination Index Index 12 .51 1.40 13 -46 1.98 7 ~43 1.53 14 .40 2.60 8 -40 1.53 15 ~36 2.02 4 ~27 1.57 9 ~26 1.49 3 ~25 1.69 11 .23 1.45 10 .15 .95 6 ~06 .41 5 ~02 .21 163 164 Appendix E (Cont.) TABLE 19 FACTOR ANALYSIS OF TERMINAL ETHOS RATINGS Dimensions and Factor Loadings: Scales Authorita- Dynamism Character tiveness Informed-Uninformed .78* .10 .12 Qualified-Unqualified .66* .07 .19 Reliable-Unreliable .62* .04 .37 Valuable-Worthless .63* .18 .35 Intelligent-Unintelligent .59* .28 .20 Expert-Inexpert .73* .26 .00 Aggressive-Meek .05 .79* -.02 Emphatic-Hesitant .07 .82* .04 Forceful-Forceless .14 .81* .09 Bold-Timid .15 .84* -.ll Active-Passive .18 .78* .05 Energetic-Tired .25 .70* .08 Unselfish-Selfish .31 -.04 .57* Nice-Awful .09 .02 .83* Friendly-Unfriendly .10 -.12 .78* Honest-Dishonest .34 .ll .68* Pleasant-Unpleasant .15 .01 .76* Virtuous-Sinful .14 .14 .69* PrOportions of variance (59%) .17 .22 .20 APPENDIX F MAJOR STATISTICAL ANALYSES I“ - hi APPENDIX F: MAJOR STATISTICAL ANALYSES TABLE 20 ANALYSIS OF VARIANCE SUMMARY FOR PERCEIVED ORGANIZATION ‘_- -7 ‘—.7 _ ‘_‘— Source of Variance df M.S. F Probability Credibility 1 0.02 0.03 0.86 Cueing 2 2.40 4.26 <0.05 Credibility X Cueing 2 0.34 0.61 0.55 Error 236 0.56 TABLE 21 ANALYSIS OF VARIANCE SUMMARY FOR CLARITY _‘?} wu‘x‘i ' ' 9 —~ ' Source of Variance df M.S. F Probability Credibility 1 0.01 0.01 0.92 Cueing 2 1.84 2.78 0.06 Credibility X Cueing 2 0.25 0.38 0.69 Error 236 0.66 166 Appendix F (Cont.) ANALYSIS OF VARIANCE SUMMARY FOR 10 167 TABLE 22 RETENTION ITEMS J Source of Variance df M.S. F Probability i ‘ Credibility 1 0.01 0.00 0.96 If Cueing 2 25.40 7.84 <0.05 F Credibility X Cueing 2 0.59 0.18 0.83 Error 236 3.24 TABLE 23 ANALYSIS OF VARIANCE SUMMARY FOR ATTITUDE SCORES Source of Variance df M.S. F Probability Credibility 1 323.43 7.62 <0.05 Cueing 2 36.12 0.85 0.43 Credibility X Cueing 2 8.68 0.20 0.82 Error 236 42.46 168 Appendix F (Cont.) TABLE 24 ANALYSIS OF VARIANCE SUMMARY FOR AUTHORITATIVENESS RATINGS Source of Variance df M.S. F Probability Credibility 1 1185.76 49.12 <0.05 Cueing 2 15.87 0.66 0.52 Credibility X Cueing 2 52.32 2.17 0.12 Error 236 24.14 TABLE 25 ANALYSIS OF VARIANCE SUMMARY FOR DYNAMISM RATINGS Source of Variance df M.S. F Probability Credibility 1 0.09 0.00 0.96 Cueing 2 23.01 0.75 0.47 Credibility X Cueing 2 0.87 0.03 0.97 Error 236 30.65 169 Appendix F (Cont.) TABLE 26 ANALYSIS OF VARIANCE SUMMARY FOR CHARACTER.RATINGS Source of Variance df M.S. F Probability Credibility 1 1033.13 47.65 <0.05 Cueing 2 1.91 0.09 0.92 Credibility X Cueing 2 38.71 1.79 0.17 Error 6236 21.68 TABLE 27 ANALYSIS OF VARIANCE SUMMARY FOR PERCEIVED INTERESTINGNESS Source of Variance df M.S. F Probability Credibility 1 0.02 0.03 0.87 Cueing 2 1.15 1.82 0.17 Credibility X Cueing 2 0.07 0.11 0.90 Error 236 0.63 170 Appendix F (Cont.) TABLE 28 ANALYSIS OF VARIANCE SUMMARY FOR PERCEIVED CONVINCINGNESS Source of Variance df M.S. F Probability Credibility 1 0.55 0.79 0.38 g ” Cueing 2 2.83 4.08 <0.05 E u Credibility X Cueing 2 0.42 0.06 0.94 Error 236 0.69 TABLE 29 RESULTS OF EfTESTS ON ORGANIZATION, CLARITY, RETENTION. AND CONVINCINGNESS FOR CUEING CONDITIONS r __ —-_ J r ‘L AC vs. NC NC vs. IC AC vs. IC DV X X X X X X E/P E/b E/b Perceived 1.88 2.23 2.23 2.05 1.88 2.05 Organization E_= 2.93 E_= -1.50 £_= 1.46 p < .05 p < .20 p»< .20 Perceived 1.94 2.25 2.25 2.10 1.94 2.10 Clarity t_= 2.41 3.: —l.l6 t = 1.27 p < .05 p < .30 p < .30 Retention 6.90 6.57 6.57 5.82 6.90 5.82 10 Items E_= 1.15 E_= 2.61 E_= 3.85 p < 30 p < .05 p < .05 Convincingness 1.98 2.33 2.33 2.25 1.98 2.25 3.: 2.65 5.: -.60 E_= 2.09 p < .05 p < .60 p < .05