A STUDY OF THE EFFECTS 0F ETEGS AND ONE-SIDED VERSUS TWB-SIBED PRESENTATION OF ARGUMENTS IN PERSUASIVE COMMUNICATIGN Thesis for the Degree of M . A . MICHIGAN STATE UNIVERSITY CHARLES D. ERTLE 1967 IIHIIM.I;[311§MW1IWHHIU um“ ‘ 3 ‘2 745 5472 m. " , h. . ‘ in; ‘ Iv . '0 i; ‘.u 29%; V 46% OCT 2 2 2005 1 U 0 43; D i: A STUDY OF THE EFFECTS OF ETHOS AND ONE-SIDED VERSUS TWO-SIDED PRESENTATION OF ARGUMENTS IN PERSUASIVE COMMUNICATION BY Charles D. Ertle A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Speech and Theatre 1967 Accepted by the faculty of the Department of Speech and Theatre, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for ctor of Th sis /%W/ ' ° Guidance Committee: ,, , , Chairman /MW%W 09!ch Q QQQPL the Master of Arts degree. ii ACKNOWLEDGMENTS The writer would like to express his appreciation to all those who offered both their time and advice during the preparation of the thesis. Special credit and apprecia- tion are due to several people: To Dr. James C. McCroskey for his advice, guidance, and encouragement during the planning, execution, and writ- ing of this study; To Dr. David C. Ralph and Dr. William B. Lashbrook, members of the Thesis Committe from the Department of Speech and Theatre, whose guidance and advice were instru- mental in the completion of this study; To Dr. Murray Hewgill and Dr. Leo V. Deal of the De- partment of Speech and Theatre for their assistance in se- curing experimental subjects; To the Senior and Junior Staff instructors of the De— partment of Speech and theatre for their assistance in se- curing class time and for the administration of the actual experiment; To my wife, Marion, for her invaluable services as iii stenographer and endless patience as a wife; To all these persons and many others who offered ad- vice and encouragement, the writer expresses sincere thanks. iv ABSTRACT A STUDY OF THE EFFECTS OF ETHOS AND ONE-SIDED VERSUS TWO-SIDED PRESENTATION OF ARGUMENTS IN PERSUASIVE COMMUNICATION by Charles D. Ertle Body of Abstract Rhetorical theory suggests that both the ethos of the source and the source's message are potent persuasive forces. Prior research has demonstrated that message va- riables vary in their effect with the level of the source's perceived ethos. Several message variables have been studied, but previously reported experimental research has not investigated the possible interaction of ethos and message sidedness in producing attitude change. In the present study a one-sided message is opera- tionally defined as §_persuasive message that presents only the argpments that are ig_agreement with the position advocated. A two-sided message is operationally defined as a_persuasive message advocating_only one position that presents the arguments‘;p_agreement wi£h_the advocated position first, then mentions and refutes counter-arguments. 0n the basis of prior research the writer hypothe- sized that the importance of sidedness varies with the level of the source's perceived ethos; a high-ethos source should find less value in using a two—sided message than would a source with middle—to-low ethos. A one-sided version and a two-sided version of a speech opposing Capital Punishment were constructed. Introductions designed to establish initially high ethos and low ethos were constructed and combined with the two message stimuli to produce four experimental conditions. -Semantic differential measures, found reliable in prior research, were used for the measurement of attitude and the authoritativeness, character, and dynamism dimen- sions of ethos. A pre-test, immediate post-test, delayed post-test, and ethos tests were administered to all eXperi- mental subjects, students enrolled in Speech courses at .Michigan State university. A control group completed only the attitude measures. Primary data analysis con- sisted of analysis of variance and covariance. The results of this study indicated the following: (1) The “high-ethos“ source was perceived as sig- nificantly higher in authoritativeness and character than vi the “low-ethos source but the “low—ethos" was perceived as moderately high in both authoritativeness and character. (2) The “high-ethos“ source with a one-sided message produced more attitude change in the direction advocated than did the “high-ethos“ source with a two-sided mes- sage. Although this result was not significant (.07 >p. <.05) on the immediate post-test, it was signi- ficant: (p. < .025) over time. (3) No significant differences were observed on attitude change between the one-sided and two-sided mes- sages in the “low-ethos“ condition. (4) No significant differences attributable to mes- sage sidedness were observed on any of the three dimensions of ethos. vii ACCEPTANCE.. . . . ACKNOLWEDGMENTS . THESIS ABSTRACT . LIST OF TABLES . . LIST OF FIGURES . LIST OF APPENDICES Chapter ‘I. INTRODUCTION TABLE OF CONTENTS General Statement of the Problem Review of Ethos Studies . . . . . Review of Sidedness Studies . . . Generation of the Experimental Hypotheses . . Hypotheses to Be Investigated . . II. PROCEDURES OF THE STUDY AND CONSTRUCTION OF THE EXPERIMENTAL INSTRUMENTS Research Design . Operational Definitions . . . . . :Construction of the Introduction Stimuli vConstruction of the Speech Stimuli . Measurements and Statistical Analysis Used . Procedures 0 O O O 4 .9. O .1. OOOOO viii Page ii iii xi xii 23 26 3O 32 33 34 35 37 Effects of Ethos and Sidedness on Effects of Sidedness on Perceived III. REPORTING OF RESULTS Attitude Change . Ethos . . . . . Secondary Measures IV. FOR FUTURE RESEARCH :Summary . Conclusions Implications for Future Research LIST OF REFERENCES ix SUMMARY, CONCLUSIONS, AND IMPLICATIONS Page 42 46 50 53 57 63 65 LIST OF TABLES Table Page 1. Immediate Post-Communication Attitude Results . . .~. . . . . . . . . . . . . 45 2. Delayed Post-Communication Attitude Results . . . . . . . . . . . . . . . . 45 3. Authoritativeness Semantic Differential Results 0 0 O O O O O O O O O O O O O O 48 4. Character Semantic Differential Results . 48 5. Dynamism Semantic Differential Results . . 48 6. Factor Analysis Results of Ethos Scales . 49 7. Results on Secondary Measurements of ‘EthoS' . . . . . . . . . . . . . . . . . 51 8. Results on Secondary Measurements for ‘ Sidedness . . . . . . . . . . . . . . . 51 LIST OF FIGURES Figure 7 Page 1. Type of Treatment Administered to Experimental and Control Cells. . . . . . . . . . . . . . . . . 31 xi APPENDIX.A: APPENDIX B: APPENDIX C: LIST OF APPENDICES “ Measuring'Instruments Introduction Speeches . . . . . . . .‘ Experimental Speeches xii 9 Page '67 70 75 CHAPTER I INTRODUCTION General Statement g£_the Problem Source credibility, or ethos, is acknowledged by most modern communication researchers as an element of pri- mary importance in a persuasive communication. Although re- searchers are not in complete agreement as to the actual dimensions of ethos,1 they do seem to agree with the gen- eral opinion held by Aristotle several centuries ago: It is not true that the probity of the speaker contrib- utes nothing to his persuasiveness; on the contrary, we might almost affirm that his character [ethoé] is the most potent of all the means of persuasion (Cooper 1960, 8-9) 0 Arthur Cohen (1964), a more recent contributor to rhetori- cal theory, equated the importance of ethos with that of the message itself: “Who says something is as important as what is said in understanding the effect of a communi- cation on an attitude.“ (p. 23) 1A discussion of the dimensions of ethos appears in Chapter II under the heading Ethos and Attitude Measures. 2 Since the early part of the 1900's there have been many empirical studies undertaken using ethos as an inde- pendent variable. In order to discuss ethos, a clarifica- tion of what is meant by the term is needed. Anderson and Clevenger (1963) refer to ethos as the following: ". . . the image held of a communicator at a given time by a re- ceiver--either one person or a group.“ (p. 60) One of the important elements of this definition is “at a given time" which suggests to us that ethos is not the “same“ at all times. McCroskey (1966a) provides us with an understand- ing of the different types of ethos: Initial ethos is the ethos of a source prior to the beginning of a given communicative act. It is the speaker's ethos just before he begins to speak. Derived ethos is the ethos of a source produced by the act of communicating. It includes the impact of the message, the effects of the circumstances in which the communication takes place, and, if oral, the delivery of the message. Terminal ethos is the ethos of a source at the completion of a communicative act. It is the product of the interaction of the Initial and Derived ethos. (p. 11) Taking the other half of Cohen's statement, we find that what is said is also important in a persuasive commu— nication. Hewgill and Miller (1965) studied the relation- ship of fear-arousing communications and source credibil— ity. The authors hypothesized the following: (1) If a source has high credibility with a listener, appeals that 3 elicit strong fear for persons highly valued by the lis- tener will effect greater attitude change than appeals that elicit mild fear, (2) A low-ethos source with a mild fear appeal will effect greater attitude change than strong fear appeals. The results obtained supported the first hypothesis but not the second. Bettinghaus (1961) looked at five variables he be- lieved responsible for attitude change, one of which was the speaker's delivery. The researcher stated that: The experiment tends to confirm what rhetorical theo- rists have said for centuries: that effectiveness in delivery contributes not only to credibility of the speaker but also to the persuasiveness of the speaker in achieving acceptance of his message. (p. 141) McCroskey (1966a) in his study on ethos and evidence concluded the following: . . .that good use of evidence can be an important as- set to a Speaker who wishes to produce favorable im- mediate post-communication audience attitudes toward his propositions. This will likely be the case for speeches on some topics when the speaker is a moderate-nglow-ethos communicator. (p. 124) Sharp and McClung (1966) studied the effects of or- ganization on the speaker's ethos. The results of the study showed that students exposed to a “disorganized“ speech thought less of the speaker after hearing the speech than before the speech. The students who heard an 4 “organized“ speech shifted their perception of the source's ethos very little. The second finding is contrary to what might be expected. The researchers concluded that good organization did not have an effect on attitude because of a classroom situation in that the students expect sound organization. The above studies show that there are many asPects of wh§£_is said. The main point is that they all vary in their effect with the level of the source's perceived ethos. flh§§_is said then does interact with initial ethos and in part is responsible for any possible change between a source's initial ethos and his terminal ethos. Another important aspect of what is said deals with the presentation of the arguments within the communication. Should a communicator present a one-sided speech, i.e., present arguments only on the side he is advocating, or should he present a two-sided speech, i.e., present his arguments and also the opposing arguments to his point? The findings to date do not provide a conclusive answer to this question. Several experimental studies have manipulated ethos as the independent variable in the persuasive communica- tion. Also there have been several experimental studies 5 that have manipulated sidedness as the independent variable within a persuasive communication. As a result of the re- search done in each of these areas, writers have general- ized what effect ethos will have in a persuasive communica- tion and What effect sidedness will have in a persuasive communication. The writer suggests, however, that we may not be able to generalize that these elements will have the same effect when combined as they have when each is a discrete element in an experimental study. For example, will ethos overcome the effects of sidedness when they are combined in a single persuasive effort? As yet, research has not given us the answer to this question. Both the construct ethos and the construct sidedness lend themselves to a dichotomous description.. This is not to negate that both the concepts are continuous variables. The point here is that in discussing ethos, we will con— sider it to be either high-ethos or low-ethos and in dis- cussing sidedness, we consider it to be either a one-sided presentation or a two-sided presentation. Of course, this is strictly a semantic dichotomy created for the ease of discussing each of the constructs. The combining of both of these variables within per- suasive communications and submitting them to empirical 6 study raises two main questions that prior research has yet not answered. Will the audience show immediate atti- tude change or acceptance of a proposition advocated by an initially high-ethos source who presents just a one-sided speech, or will the lack of a two-sided presentation cause a communicator's derived ethos to go down and weaken the effect of high initial ethos, causing the audience's atti- tude to change less? Will the two-sided presentation of arguments cause the derived ethos of an initially low-ethos source to increase and cause a significant incnfise.in at- titude change in the direction advocated by the communica- tor? We need to determine whether the interaction between these variables is sufficiently strong to over-ride a weakness of one with the strength of the other in an effort to bring about a significant change of attitude in a per- suasive communication. In order to determine the extent of the research relevant to the above question, the writer conducted a re- view of the literature. Numerous studies were found deal- ing with either ethos or sidedness of presentation as a separate variable. The more significant of these are re- ported and discussed below. Review 2£_Ethos Studies Studies concerning ethos, also referred to as "Source credibility,“ have confirmed Aristotle's opinion of the importance of ethos in a persuasive communication. Haiman (1951) presented a tape-recorded speech to three groups of students. The communications were identical with the exception that they were attributed to three different sources. One speech was attributed to Thomas Parran, Surgeon General of the united States; the second Speech, to Eugene Dennis, Secretary of the Communist Party in America; and the third speech, to a Northwestern university Sophomore. The subjects voted the Speech at- tributed to Parran significantly more competent than the other two speeches. As measured by the WOodward Shift of Opinion Ballot, the “Parran” speech was significantly more effective in changing attitudes than either of the other.two Speeches. There was no Significant difference found between the “Dennis“ and “Sophomore“ Speeches. Hbvland and Weiss (1961) gave identical written communications to subjects who were college students. The communications consisted of messages on several top- ics and were purported to be excerpts from newspaper and 8 magazine articles. Half of the communications were at- tributed to sources who were considered trustworthy, and the other half of the articles were attributed to sources who were considered untrustworthy, such as Pravda. The subjects were given a pre-test, a post-test, and a delayed post-test one month later. Although the subjects acquired the same amount of information under all conditions, they judged a communication as less fair and the conclusions as less justified when the communication was attributed to a low-ethos source. The change of attitude in the di- rection advocated by the communication was greater when it was attributed to a high-ethos source than when it was attributed to a low-ethos source at the time of the im- mediate post-test. When the delayed post-test was given four weeks later they found that the effects of both high- and low-ethos source had disappeared. There was no Significant difference between the amount of opinion change for a high-ethos source and the amount of opinion change for a low-ethos source. HOVland called this phe- nomenon the “Sleeper Effect.“ The decrease of the atti- tude change of the high-ethos condition was attributed to forgetting, but the increase of the attitude change in the low-ethos condition was explained as a disassociation 9 of the source from the message permitting the “communica- tion“ to have a delayed positive effect. The effect of the source is greatest at the time of the communication and decreases with time faster than the effects of the content. Kelman and Hovland (1953) had different groups of high school students listen to an identical communication in which a speaker advocated leniency in the treatment of juvenile deliquency. Three versions of the credibility of the Speaker were established. In the “positive“ ver- sion the speaker was described as a judge in a juvenile court, a person who was well-trained and honest. In the “neutral“ version the speaker was described as a member of the studio audience who had been picked at random, whose background was unknown. In the “negative“ version the speaker was described as being picked at random from the audience; however, it was established during the in- troduction that the person had been a delinquent during his youth and was at present out on bail on a charge of drug peddling. An attitude questionnaire was given im- mediately after the communication and again three weeks later. The results showed that with identical content in all three conditions the audience perceived the “positive“ lO communicator as being more fair and trustworthy than the “negative“ communicator and the “neutral“ communicator as being in the middle. The group hearing the communication from the “positive“ source showed more attitude change by favoring a more lenient treatment than those hearing the communication from the “negative“ source. Like the study of Hbvland and Weiss the delayed post-test given three weeks later showed that the differences between the experimental _groups were no longer present. The sub- jects who had the “positive“ source showed a decrease in acceptance, and those who had the “negative source showed an increase in acceptance. A study by Hbvland and Mandell (1952) dealt with the “trustworthiness“ of a communicator. A speech on the “Capital Devaluation of Currency“ was presented to col- lege students. Two different intnaductions were used; the first introduction described the speaker as being the head of a large importing firm and as “having some- thing to gain“ if the advocated message were accepted. The second speaker was described as being an economist from a leading American university with “nothing to gain.“ The subjectswere asked to give their opinions on the issue before and after the communication. On the post- 11 test they were also asked to give their reactions to the program and the Speaker. The results Showed that the speaker who “had something to gain“ was perceived by the audience as having done a poorer job and as being less fair and honest than the speaker who was presented as being “impartial.“ Hewever, there was no significant difference in the change of opinion between the two conditions. Thus, the exPSrimental variation produced a difference in per- ceived ethos but failed to do so with regard to attitude change. The results of the last four studies raises the ques- tion of what is ethos or what are the dimensions of ethos. What are the dimensions by which a receiver measures the level of ethos of a source? This is not to be confused with the dichotomy low-ethos/high-ethos but instead Should be thought of as dimensions by Which a person arrives at the notion that a given source has either high or low ethos. The first mention of the dimensions of ethos came from Aristotle when he said: As for the speakers themselves, the sources of our trust in them are three, for apart from the arguments (in a speech) there are three things that gain our be- lief, namely, intelligence, character, and good will. (Cooper 1960, 91-92) These in essence are three dimensions by which we would 12 measure a source's perceived high or low ethos. The study by Haiman seemed to manipulate the dimensions of intelli- gence and character. In the study of Hovland and weiss, on the other hand, they brought up a dimension called trustworthiness. Kelman and Hovland were using the word “honest,“ and Hovland and Mandell seemed to manipulate "trustworthiness.“ In general all of the studies were dealing with ethos; however, the different studies ap- proached it from a different dimension of ethos. Hovland, Janis, and Kelly (1965, pp. 19-25) conceived of two dimen- sions of ethos. They were called “expertness” and “trust- worthiness.“ Hovland, Janis, and Kelly in their discussion of perceived eXpertness and trustworthiness made a distinc— tion between credibility and other types of source-related dimensions, such as affection, fear, and power. They do suggest, however, the relevance to credibility of variables such as intelligence and sincerity. In limiting credibil- ity to the dimensions of “trustworthiness“ and “expertness” they state that there is a positive correlation with per- suasion. They state: From the results, it is not possible to disentangle the effects of the two main components of credibility-- trustworthiness and eXpertise--but it appears that both are important variables. (p. 35) 13 Modern researchers in an effort to construct measur- ing instruments for ethos have submitted Aristotle's theo- retical dimensions of ethos to empirical research. Berlo, Lemert, and Mertz (1961) identified three factors of ethos in a factor analytic study using semantic differential scales. The three dimensions of ethos were originally called “competence,“ “trustworthiness,“ and “dynamism.“ As a product of later analysis the labels were changed to “safety," “qualification,“ and “dynamism.“ McCroskey (1966a) using both a Likert-type scale and semantic differential scales found that factor analy- sis produced only two significant factors. The first fac- tor, which was labeled “authoritativeness,” accounted for 47%.of the variance. The second factor, which was labeled “character,“ accounted for 29%.of the variance. It is in- teresting to note that neither Aristotle's theoretical factor “good will“ nor Berlo e3, a1,'s factor “dynamism“ appeared. Theoretically, it seems logical to consider “good will“ and “dynamism“ as dimensions of a source's ethos, but apparently when empirically tested these two dimensions are subordinate to the dimensions of "authori- tativeness“ and “character.“ Later research by McCroskey, as yet unpublished, obtained a clear “dynamism“ dimension, 14 but only when special care was taken to manipulate the "dynamism“ dimension. McCroskey (1966b) used both his Likert-type scale and semantic differential scales in seven experiments and stated the following: On the basis of the above eXperiments it can be con- cluded that these scales are capable of reliably meas- uring either initial or terminal ethos on the two di- mensions of authoritativeness and character. (p. 32) On the basis of the above findings the writer chose to use the semantic differential scales developed by McCroskey and the dynamism scales developed by Berlo, g5. 31, From these studies and numerous others (Anderson and Clevenger 1963) we can reasonably assume that ethos as measured by its dimensions is an important factor in in- fluencing audience attitude. In general an audience will tend to accept a communication from a source perceived to be of high ethos more readily than when they perceive the source to be of low ethos. Review 9; Sidedness Studies Some of the studies that have dealt with the organi- zation of arguments within a Speech have tried to determine the reLative effects of presenting a one-sided message as 15 opposed to presenting a two-sided message. The writer finds that a definition of both “one-sided“ and “two-sided" messages is needed at this time to clarify possible ambig- uity of terms. For the purpose of this study a one-sided message means §_message that advocates §_definite stand §t_ the outset and then proceeds tg_support and prove that stead, A two-sided message means §_message that advocates .33 the start and throughout 9p1y_gge.§1g§_gf_the issue but that mentions and refutes arguments favor;gg_the other side, gglpg.§_climactic order 9§_presentation. Clearly, both speeches favor one side of the issue and both present their stand at the outset of the speech. The essential difference between them, therefore, lies in the fact that the two-sided message after presenting the advocated posi- tion states the arguments on the opposite side and then refutes these arguments. This is the so-called “strawman“ technique which is described in many textbooks on public speaking. The writer has found from personal experience that this technique is often used in business and almost exclusively used in the sales field. Because the “straw- man“ technique is so much a part of our real world the writer prefers to limit his two-sided presentation to this style. 16 One of the earliest experiments on sidedness was con- ducted by HDVland, Lumsdaine, and Sheffield (1949). The experiment presented a communication concerning the end of the war with Japan after the surrender of Germany in 1945 to an experimental group of 214 soldiers and to a control group of 197 soldiers. One eXperimental group was given a communication that presented only the arguments supporting the position that the war with Japan would be long. The other experimental group was given a communication with the same information plus additional information stress- ing the united States' advantage and Japan's weakness. Pre-test and post-test measurements of attitude were given to the experimental group as well as the control group who did not hear the communication. The results showed that neither presentation had any advantage over the other for the audience as a whole. However, the presentation in which they gave pgth_sides was more effective for those men initially opposed to the advocated position, while the one-Sided presentation was more effective for those who initially favored the advocated position. The educa- tion level was also a determining factor. The two-sided presentation was more effective with the "better educated“ [ébove twelfth grade] men, and the one—sided presentation 17 was more effective with the “less educated" [twelfth grade and lower] men. When the initial attitude and level of ed- ucation were considered together, the two-sided communica- tion was more effective with “better educated“ men, regard- less of their initial attiuide; and the one-Sided presenta- tion was more effective with the “less educated" men. Lumsdaine and Janis (1953) extended this study of Lumsdaine and Sheffield. Lumsdaine and Janis were inter- ested in the question of the effectiveness of a one-sided or a two-sided presentation in resisting counter-influences. In this experiment half of the subjects received a communi- cation one week after the initial Speech stimuli contain- ing counter-arguments to the position advocated. The sub- jects had initially received either a one-sided or a two- sided communication on Russia's inability to produce atomic bombs for many years to come. After the counter-arguments were given to half of the subjects in each eXperimental condition, all of the subjects were asked to state their opinions again. The results showed that there were no differences in attitude change for those who did not re- ceive the counter-propaganda. For the subjects who had been exposed to the counter-propaganda and had initially received the two-sided communication, the results showed 18 a greater resistance to the counter-proPaganda than those who had initially received a one-sided communication. Thus it seems that the two-sided presentation prepares the listener to meet and resist later counter-arguments. It seems reasonable to assume that a.two-sided presenta- tion would generally build up a resistance to pressures that could change attitude at a later time. Thistlethwaite and Kamenetsky (1955) conducted an experiment involving the effectiveness of two variables within a persuasive communication. The first variable was refutation versus no refutation. The other variable was a two-Sided presentation versus a one—sided presentation of arguments in a persuasive communication. Two independ- ent samples were obtained; one consisted of recruits in basic training at a military base and the second consisted of high school students. Tape recordings and Slides were used to present communications designed to change the at- titudes of the subjects. A control group listened to an irrelevant communication which had no information concern- ing the Kbrean War. In this experiment ethos was held constant and sidedness and refutation were manipulated. Among the high school subjects refutation evoked signifi- cantly more discounting of the counter-arguments than no l9 refutation. There was no statistically Significant dif— ference in the discounting reactions of the recruits to the refutation and no refutation treatments. Refutation tended to produce greater comprehension of the intended conclusion than no refutation for both samples, although only the difference for the high school subjects was sta- tistically significant. A very significant finding, al- though contrary to prior research, was that a one-sided presentation tended to be more effective in changing at- titudes than a two-sided presentation. Thistlethwaite and Kamenetsky report that attitude change Eepg§_to be greater for subjects whose understanding of the communicator's con- clusion is greater. Introducing unfamiliar facts in sup- port of the other side leads to less Change of attitude; but not including well-known facts on the other side will also weaken the appeal. The degree of attitude change is in direct relationship with the degree that the conclu— sions are understood and the facts familiar. Paulson (1954) conducted a study in which ethos and sidedness were manipulated. A sample of 579 men and 399 women were drawn. The result showed that the high-ethos was significantly better than low-ethos but only with the men. There was no significant difference for either men 20 or women with respect to one-sided versus two-Sided pre- sentation of arguments. It is interesting to note here that Paulson's two—sided presentation of arguments did not contain any refutation. The review of the literature disclosed only one other study that made an attempt to manipulate both ethos and Sidedness within a single communication. WOlfinger (1954) measured attitude change toward both the communica- tion [or concept] and the perceived source of the message. College students rated the concept “Fifth Amendment“ and an imaginary “George Hastings“ against a semantic differ- ential. One group of subjects was then given a two-sided version by “George Hastings“ advocating favorably the “Fifth Amendment." A second group received a one-sided version of the identical Speech in which all opposing ar- guments had been deleted from the tape. A control group heard a speech unrelated to the experimental topic. The results showed that both the one-sided and the two-sided version produced significantly more attitude change in favor of the position advocated than the change in the control group. Contrary to the findings previously cited in the studies above, subjects who initially favored the concept were affected more by the two-sided presentation. 21 and subjects who initially Opposed the concept were more affected by the one-sided presentation; however, these re- sults did not reach satisfactory statistical significance. With respect to the source, the results Showed that the subjects originally favoring the position advocated by the hypothetical source were affected more by the one-sided message. This result also did not reach satisfactory sta- tistical significance. Considering change in attitude to- ward both source and concept taken together, the one—sided presentation produced significantly more favorable changes than did the two-sided version. Although WOlfinger tried to measure the effect of the interaction of ethos with the presentation, the writer fails to see where this was actually accomplished. There was no introduction of the imaginary “George Hastings” so the pre-test showed the source, "George Hastings,“ as neutral. In essence there was no initial ethos condition. The post-test measured ethos but did not reach the necessary level of signifi- cance. If WOlfinger measured ethos at all, then he meas- ured only derived ethos. It would stand to reason that a subject favoring a given communication would favor the source since there was no information concerning the source presented. 22 A study conducted by Jaksa (1964) is of importance because the arrangement of the arguments within the two- sided presentation was manipulated and compared with the effects of a one-sided presentation. Jaksa developed four Speeches: (1) a one-sided speech, (2) a two-sided Speech with a climactic order, (3) a two—sided speech with an anti-climactic order, (4) a two-sided speech with an in- terwoven order. The results showed that both the two- sided climactic and the two-Sided interwoven order of counter-arguments were significantly superior in changing attitude than were a one-sided speech or a two-sided speech with an anti-climactic order of counter—arguments. Based on the above eXperiment the writer used a climactic order of counter-arguments within the two-sided experi- mental speech in the present study. The above studies Show that a one-Sided presentation of arguments is effective in changing attitude in the di- rection advocated by the source when: (l) the audience is not “better educated," (2) the audience initially agrees with the position advocated, (3) the audience is not given any later counter-arguments. A two-sided presentation of arguments is effective in changing attitudes in the direc- tion advocated by the source when: (1) the audience is 23 “better educated,“ (2) the audience initially opposes the position advocated, (3) the audience received later counter-arguments. If a two-sided presentation is used, the counter—arguments should be placed in a “climactic“ or “interwoven“ order of presentation. Generation 9f_the Experimental Hypotheses we can reasonably conclude from the studies dicussed above that the ethos of a source and the type of arguments are both important factors in persuasive communication. The effects of each of the above variables have been singu- larly examined in several studies. However, the interace' tion between these variables has not been adequately in- vestigated. Since both variables are present in all per- suasive communication, they must be studied together if we are to produce useful generalizations. Given a parti- cular condition of ethos coupled with a given condition of the sidedness, the problem is to ascertain the role each variable plays as it interacts with the other. When we are able to understand the interaction between the va- riables, if any, then we will be in a position to know within a given probability of error» which coupling of variables will yield the greatest amount of attitude 24 change in a given situation. The above research has shown that when sidedness is hekiconstant, a high-ethos source produces greater attitude change toward his position than does a low-ethos source. However, when ethos is held constant, we cannot conclu- sively say that either a one-sided or a two-sided presen- tation will produce greater attitude change toward the di- rection advocated. The experimental studies on Sidedness do not agree. The writer suggests that the disagreement may be due to ethos. Ethos seems to be a confounding ele- ment in the sidedness studies because it was not adequately controlled. Paulson (1954) in his study assumed that the high-ethos and low-ethos introductions would be perceived by the receivers as being significantly different from each other, but he failed to obtain an ethos measurement to find out if his assumption was correct. The results of Paulson's study tend to suggest that for the group as a whole there was not a statistically significant difference between the high-ethos condition and the low-ethos condi- tion, which indicates that Paulson's assumption was not correct. The writer suggests that Paulson had a high-ethos source that might be accounted for by the “sponsorship 25 effect.“2 WOlfinger (1954) also did not measure the dimensions of ethos to ascertain whether the hypothetical “George Hastings“ was perceived as a high-, middle-, or low-ethos source. Because there was no prior information establish- ing the ethos level of “George Hastings“ the audience, ac- cording to the pre-test, did not have an attitude toward the source. The writer suggests that during the experi- ment the “sponsorship effect“ may have caused the ethos of “George Hastings“ to be perceived as high. Thus it seems probable that wolfinger actually had a high-ethos source, and when he combined the source with the concept 2The “Sponsorship effect“ was noted in a study by James C. McCroskey and Robert E. Dunham, “Ethos: A Con- founding Element in Communication Research," Speech Mono- graphs, (1966), 456-463. In this study the researchers had to reject the null hypothesis that the unseen, unknown, tape—recorded speaker was a neutral source. The authors concluded that there are theoretical grounds for supposing an unseen, unknown, tape-recorded speaker in an eXperi- mental setting can be perceived as a higher-than-neutral source because of the “sponsorship effect“ of the exPeri- menter who is perceived as a high-ethos source in the ex- periment. A follow-up study was conducted by Paul H. Holtzman, “Confirmation of Ethos as a Confounding Element in Communication Research,“ Speech Monogpaphs, 33, (1966), 464-466. Confounding by "sponsorship effects“ was con- firmed in this study. The author concluded that “. . . all exPerimental designs should account for ethos effects including effects of perceived sponsorship.“ 26 he found that a high-ethos source produced greater attitude change with a one-sided presentation of arguments. The above finding suggests to the writer that ethos is the dominant factor when combined with sidedness. Contrary to the findings that suggest a two-sided presentation is bet- ter than a one—Sided presentation, when the dimensions of ethos are taken into account a high-ethos source has noth- ing to gain from using a two-sided presentation; in fact, the high-ethos source might conceivably be harmed by a two-sided presentation. The writer ventures the following hypothesis: The importance of sidedness varies with the level of ethos of the communicator. A high-ethos communicator should find less value in using a two-sided presentation of arguments than would a communicator with middle- to low-ethos. If this is true, it would tend to clarify the previous re- sults of sidedness research. The present study then is concerned with controlling ppph the variable of ethos and the variable of sidedness in experimental design and analysis. Hypothesesto pp_1nvestigated Based on the above discussion, the following hypo- theses were formulated: l. 3. 5. 27 A message attributed to an initially high-ethos source will produce a significantly greater change of attitude in the direction advocated when measured on an immediate post-test than will an initially low-ethos source. A message attributed to an initially high—ethos source will produce a significantly greater at- titude change in the direction advocated when measured on a delayed post-test than will an initially low-ethos source. A two-sided message will produce a Significantly greater shift of attitude change in the direction advocated when measured on an immediate post- test than will a one-sided message. A two-sided message will produce a significantly greater change of attitude in the direction ad- vocated when measured on a delayed post-test than will a one-sided message. An initially high-ethos source with a one—Sided message will produce a significantly greater change of attitude in the direction advocated when measured on an immediate post-test than will an initially high-ethos source with a two— 6. 10. 28 sided message. An initially high-ethos source with a one-sided message will produce a significantly greater change of attitude in the direction advocated when measured on a delayed post—test than will an initially high-ethos source with a two-sided message. An initially low-ethos source with a two-sided message will produce a significantly greater change of attitude in the direction advocated when measured on an immediate post-test than will an initially low-ethos source with a one-Sided message. An initially low-ethos source with a two-sided message will produce a Significantly greater change of attitude in the direction advocated when measured on a delayed post-test than will an initially low-ethos source with a one-sided message. An initially high-ethos source will have signi- ficantly higher terminal ethos than will an ini- tially low-ethos source. An initially high-ethos source with a one-sided 29 message will produce significantly higher termi- nal ethos than will an initially high-ethos source with a two-Sided message. 11. An initially low-ethos source with a two-sided message will produce significantly higher termi- nal ethos than will an initially low-ethos source with a one-sided message. A study was designed and conducted to test these hypotheses. Chapter II sets forth the procedures and the experimental instruments the researcher employed in the study. Chapter III presents the results obtained. Chap- ter IV includes the conclusions derived from those results and suggestions for possible future research in the area. CHAPTER II PROCEDURES OF THE STUDY AND CONSTRUCTION OF THE EXPERIMENTAL INSTRUMENTS Research Design The writer conducted an experimental study to test the hypotheses generated in Chapter I. At this point a statement of the research design is appropriate. The design of this study included six cells; four cells received a form of experimental manipulation, and two cells were used as control groups. The experimental subjects for each cell were students enrolled either in Speech 108 or in Speech 116 at Michigan State University.3 Six sections of Speech 116 were used, and twelve sections of Speech 108 were used. Each section was randomly as- signed to an exPerimental treatment or a control group. Each cell of the study contained three sections for a 3Speech 116 is a basic course in group discussion and Speech 108 is a basic course in voice and articulation. Each of these courses contains multiple sections, and each course is offered during every quarter within the Department of Speech and Theatre at Michigan State Uni- versity. 3O 31 combined total of at least 55 subjects for any one experi— mental or control cell. The total N used was 273 subjects for the experimental cells and 113 subjects for the control cells. Six of the experimental sections held class in the afternoon, which enabled the researcher to partially con- trol for a mediating variable “time of day of experiment,“ which might otherwise have been a confounding element within the experiment. Each cell received a different treatment as shown in the following figure: FIGURE 1 TYPE OF TREATMENT ADMINISTERED TO EXPERIMENTAL AND CONTROL CELLS Type of Type and Measurements Administered Experimental or Time of Pre- Post- Delayed Control Ce11f Seppion Test Test ngt-tesp, Ethos (l) High-Ethos 116 Morning X X X X One-Sided 116 Afternoon X X X X 108 Afternoon X X X X (2) High-Ethos 116 Morning X X X X Two-Sided 108 Afternoon X X X X 108 Afternoon X X X X (3) Low—Ethos 116 Morning X X X X One-Sided 116 Morning X X X X 108 Afternoon X X X X (4) Low—Ethos 108 Morning X X X x Two-Sided 108 Morning X X X X 116 Afternoon X X X X (5) Control I 108 Morning X X 108 Morning X X 108 Afternoon X X (6) Control 11 108 Morning X X 108 Morning X X 108 Afternoon X X 32 Operational Definitions The experimental stimuli consisted of a high-ethos introduction, a low—ethos introduction, a one-sided speech, and a two-sided Speech. The primary dependent variables were immediate post-communication attitude, delayed post- communication attitude, and terminal ethos of the communi- cator. The secondary dependent variables were the audience members' evaluations of the speech stimuli. Ethos is operationally defined as a score on a sem- antic differential for each of the three dimensions of ethos: (l) Authoritativeness, (2) Character, (3) Dynamism. Attitude is operationally defined as a score on a semantic differential. A one-sided Speech is operationally defined asxa per- suasive communication advocating a position against “Capi- tal Punishment“ that presents only the arguments in agree- ment with the position advocated. A two-sided speech is operationally defined as a persuasive communication advocating a position against “Capital Punishment" that presents the arguments in agree- ment with the position advocated first and then presents arguments against the position advocated with refutation of the counter-arguments. 33 A high-ethos introduction is operationally defined as a speech of introduction depicting “William T. Anderson“ as a man high in Authoritativeness and Character as meas- ured by the ethos semantic differentials. A low-ethos introduction is operationally defined as a speech of introduction depicting “Anthony L.Capelli“ as a man low in Authoritativeness and Character as meas- ured by the ethos semantic differentials. Construction pf_the 1ntroduction Stimuli One of the primary variables manipulated in this study was the initial ethos of the speaker. The manipula- tion of ethos was accomplished through the induction of a high-ethos introductory speech describing “William T. Anderson“ and a low-ethos introductory Speech describing “Anthony L. Capelli.“ The introductions used are reported in Appendix B. In 1965-1966 McCroskey (1966a) used the topic “Capi- tal Punishment.“ For this study McCroskey developed two introductory Speeches for use in establishing an initially high-ethos source and an initially low-ethos source. Both speeches were pre-tested by McCroskey and found to be successful in producing a statistically significant 34 difference between the perceived high-ethos source and low-ethos source as measured by a semantic differential for the ethos dimensions of character. The high-ethos in- troduction was perceived as being Significantly high on authoritativeness, but the low-ethos source failed to pro- duce perceived authoritativeness Significantly lower than neutral. The writer decided to use the high-ethos intro- duction of McCroskey's intact with the exception of chang- ing dates to bring the period of time up to the present. The writer developed a new low-ethos introduction using much of the same information from McCroskey's low-ethos introduction in an attempt to reduce the perceived authori- tativeness level. A discussion of the outcome of this change is included in Chapter IV. Construction pf_the Speech Stimuli The topic "Capital Punishment" has already been used in a study by Cathcart (1955) and in a study by McCroskey (1966a). Both researchers found that the audiencestended to be evenly disposed on their attitude toward the t0pic. Both studies also found the t0pic was of interest to the audiences. Because the topic has already been successfully employed this writer decided to use “Capital Punishment" 35 in the present study. The speeches were prepared by the writer. These speeches are included in Appendix C. Both the one-sided speech and the two-sided speech were pre- sented to members of the Senior Staff in the Department of Speech and Theatre at Michigan State University. Both speeches were judged by them and found to meet the opera- tional definitions set forth in the beginning of this chapter. Measurements and Statistical Analysis Used As mentioned in Chapter I a considerable amount of work has been done in the development of instruments to measure the dimensions of ethos. Knowing from prior re- search that both the Likert-type scales and the semantic differential scales are very reliable instruments for measuring the dimensions of ethos, this writer chose to use the semantic differential scales. The decision was made from a pragmatic viewpoint. The semantic differen— tial is a much easier measurement scale to administer be- cause of its design and clarity than is a Likert-type scale. Based on the work done by both Berlo pp, 1. (1961) and McCroskey (1966b), all three dimensions of ethos as 36 measures were included in this study. As a secondary part of this study all three of the dimensions (authoritative- ness, character, and dynamism) were submitted to factor analysis, and the results are reported in Chapter III. Again from a pragmatic point of view this writer chose to use the semantic differential for “Capital Punishment“ de- veloped by McCroskey because it was already in existence, had been used, and had been found to be reliable. (McCroskey 1966a, pp. 33-45) The secondary measurements used to measure the audi- ence's perception of the Speech were chosen by the writer on the basis that each set of adjectives is representative of those generally used in the field of Speech for this purpose. Copies of the measuring instruments are included in Appendix A. Statistical procedures for the treatment of the collected data include the following: (1) analysis of co- variance to test hypotheses relating to main effects, (2) .prtests for hypotheses concerning the effect of one vari- able at a given level on the other variable, (3) factor analysis to test the stability of the ethos dimensions. The data collected were punched on cards, and the 37 information was submitted to the computer at Michigan State University for computation of the necessary statis- tics using programs that have been developed by the Com- puter Laboratory for this purpose.4 Procedures The study was conducted during the Spring Term of the school year, 1966-1967, at Michigan State university. Prior to the start of classes, all necessary measuring instruments, instructions, taped speeches, and introduc- tions were prepared. The Speeches were recorded on tape by a member of the Speech Department faculty. The intro- ductions of the high-ethos and low-ethos sources were read from a manuscript by the experimenter in all cases. The procedures for administering the experiment were handled by the experimenter, and each section was handled 4The program used for the control group's data was flBastat Routine," Statt Series Description No. 5, devel- 0ped by the Agricultural ExPeriment Station, MiChigan State university. The program used for the experimental group's data was “Analysis of Covariance and Analysis of Variance with Unequal Frequencies Permitted in the Cells: L. S. Routine,“ developed by the Agricultural Experiment Station, Michigan State university, Statt Series Description No. 18. The factor analysis program used for the ethos scales was “Factor Analysis: Factor A; Principal Components and Orth- ogonal Rotations,“ developed through the Computer Institute for Social Science Research, Michigan State university. 38 the same with the exception of presenting the different types of experimental treatment. The pre-test and the delayed post-test were administered by the individual class instructors. A chronological explanation of the procedures is ap- propriate. During the first week of class each instructor was given a pre-test measurement form for each student en- rolled in his sections. The pre-test form contained the necessary directions for its completion, which enabled each instructor to say as little as possible when adminis- tering the test in class. The pre-test contained four other distractor concepts as well as the experimental con- cept in an effort to control for later test awareness and possible bias. After the completion of the test forms the instructor collected them and returned them to the exPerimenter. The experimenter then prepared an IBM data card for each student, recording the pre-test attitude score on the experimental concept. During the third and fourth weeks of the Spring Term the exPerimenter arranged a time with each instructor for the administration of the experiment in each section used in the study. The total time of the experiment was approximately twenty-five minutes. Upon entering the 3 9 classroom the eXperimenter distributed the necessary ex- perimental measuring instruments as well as a placebo Speech rating form. At this time the subjects were not told the purpose of the study but were told that they were participating in the development of a new speech rating form.5 The subjects were then told that they were to fill out the rating forms after hearing the tape-recorded speech; first, however, the experimenter was going to read some information concerning the speaker they were about to hear. The experimenter then read one of the ethos intro- ductions, turned on the tape recorder, and sat down. After the speech was over the subjects were asked to fill out the forms; when they were finished the experimenter collected the forms and left the classroom. The collected data was then punched on the subjects' data card. During the ninth week of class the experimenter again had each instructor distribute and collect the 5Under the auspices of a Federal grant William B. Lashbrook, a Senior Staff member of the Department of Speech and Theatre at Michigan State University, is in charge of a research project to develop a new rating form for public speaking. The project was initiated at Michigan State university in the Fall Term of 1966 and is still in progress at Michigan State university and several other colleges. With permission from Dr. Lashbrook the experi- menter successfully used the existing project as a placebo for the present study. 40 delayed post-test. After the delayed post-test was com- pleted each instructor was given a prepared statement to read to the subjects reporting the true nature of the pres- ent study. The data collected from the delayed post-test was also enetered on the subjects' data card. The sections that were used as control subjects were divided into two groups. The first group, Control I, re- ceived a pre-test during the first week of classes and an immediate post-test during the fourth week of class. The second group, Control II, received a pre-test during the first week of class and a delayed post-test during the ninth week of class. In both control groups the tests were administered by the instructor in the same manner as the pre-test and delayed post-test were given to the experimental groups. The subjects were not informed as to the true nature of the experiment until the experiment was over for all groups, both experimental and control. The subjects were informed through a prepared statement read to them by their instructor. The data for the control groups were then punched on data cards. All of the data were then submitted to the computer for analysis. Both forms of the speech stimuli were recorded by 41 the same speaker in an effort to control for individual differences of style and delivery between Speakers, which might otherwise have been confounding elements in the pres- ent study. The use of the tape recorder as the channel of communication was Chosen to reduce possible non-verbal confounding elements that could not be controlled in a live, taped television, or motion picture presentation. Only the speaker's voice and the message content could change the ethos level from the initial ethos established through the introductory speech. The results obtained from the above procedures are reported in Chapter III. CHAPTER III REPORTING OF RESULTS Chapter I preposed eleven hypothesesthat were tested in the present study. The present chapter reports the re- sults of the hypothesesunder two headings: (1) Effects of ethos and sidedness on attitude change, (2) Effects of sidedness on perceived ethos. The audiences' perception of the speech is reported under a third heading: Secon- dary measures. Effects pf_Ethos and Sidedness pp_Attitude Change Hypothesis 1; A message attributed to an initially high-ethos source will produce a significantly greater change of attitude in the direction advocated when meas— I ured on an immediate post-test than will an initially low-ethos source. The hypothesis was not confirmed. The high-ethos treatment was not perceived as being more ef- fective than the low-ethos treatment (F—.23, p. < .32) . Hypothesis 3; A message attributed to an initially high-ethos source will produce a significantly greater 42 43 attitude change in the direction advocated When measured on a delayed post—test than will an initially low-ethos source. The hypothesis was not confirmed (F-2.10, p.p.<.05) on the immediate post-test, it was significant (p.<:.025) 61 over time when measured on the delayed post-test. This finding iS consistent with Wolfinger (1954) and Thistle- thwaite and Kamenetsky (1955). Thistlethwaite and Kamenetsky pointed out that introducing unfamiliar facts and not including familiar facts when mentioning the other side tends to lead to 1e§§_attitude change. This finding might account for the weakness of the two-sided argument used in the present study. The writer suggests that a high-ethos source needs only to persuade, not to inform. It seems that a high-ethos condition is perceived as be- ing enough of an authority that a two-sided message is not needed. If a high-ethos source wants to acquire more ethos, he Should use a two-sided message, but if his in- tention is to change attitude in the direction advocated, a one-sided message is best. Sidedness did not affect the low-ethos condition. The main effect of ethos resulted in a non- significant difference because there was a middle-ethos condition instead of a low-ethos condition. Intuitively it seems reasonable to assume that if a high-ethos condi- tion produces a change in one direction and a low-ethos condition produces a change in an opposite direction, then the middle-ethos condition might not have any effect 62 at all. This may be what happened in this study. The same assumption may be made concerning the main effect of sidedness. Apparently it is possible to reach a point of diminishing returns for both sides of an element. The writer suggests that this point is reached in the middle of the scale. In Short, when a measurement reveals an ethos condition or a sidedness condition as being per- ceived in the center of the scale, this level of the con- dition will result in a non-significant difference. If the above assumption is correct, it may well account for the results of this study. There are four reasons or combinations of reasons that could account for the results of the present study: (1) The ethos introductions, although being perceived as different, were not strong enough to adequately test the hypotheses. (2) The two-sided message, although perceived as being different from the one-sided message, was not “two- sided“ enough. (3) Both Speeches, although they could be equal, were perceived as being of too high a quality in delivery, content, and organization. 5 (4) The hypothesesthemselves are incorrect. 63 unfortunately, within the limitations of the design and execution of this study we cannot be certain what pro- duced the results. 1mp1ications for Future Research Obviously, the first consideration of future research Should be the main effects of ethos and sidedness. The introduction and speech stimuli must be pre-tested to ascertain their position on the scale in order to ade- quately test proposed hypotheses. In short, now that we have tested the high-ethos and middle-ethos conditions, we need to test the low-ethos condition. Also, now that we have tested a one-sided and a middle-of-the—road message, we need to have an adequate test of the two-sided condi- tion. This study found that with a high-ethos condition a one-sided message was more effective than a two-sided mes- sage in changing attitude over time. The question then arises--will a one—sided argument attributed to a high-ethos source inoculate against counter-arguments at a later time? This study found that with a high-ethos condition a two-Sided message produced higher terminal ethos than did a one-sided message. Will these conditions inoculate 64 against later counter-arguments? Counter-arguments used in prior research normally have been one—sided messages attributed to a high-ethos source. What happens to the elements of ethos and sided- ness when used in later counter-arguments? To date there are no data available that Show the best combination of ethos and sidedness in counter-arguments. As yet we are unable to generalize about which combination of ethos and sidedness is best to inoculate against counter-arguments nor which combination of ethos and sidedness is best to use in the counter-argument to break down the original message. The writer suggests that since oral messages and counter-messages account for the greatest proportion of all communication the above questions need to be answered. Only then will we be able to establish valid generaliza- tions concerning the interaction of ethos and message variables in oral communication. LIST OF REFERENCES Anderson, K. and Clevenger, T., Jr. A summary of experi- mental research in ethos. Speech Monographs, 1963, 30, 59-78. Berlo, D. K., Lemert, J. B., and Mertz, R. J. Dimensions for evaluating the acceptability of message sources. unpublished paper based on a paper presented at Speech Association of America Convention, New York, 1961. Bettinghaus, E. P. The Operation of congruity in an oral communication Situation. Speech Monographs, 1961, 28, 131-142. Cathcart, R. S. An experimental study of the relative ef- fectiveness of four methods of presenting evidence. Speech Monographs, 1955, 22, 227-233. COOper, L. The Rhetor19_p§_Aristotle. New York: Apple- ton-Century-Crofts, Inc., 1960. Haiman, F. An exPerimental study of the effects of ethos and evidence in persuasive communication. unpub- lished Doctoral dissertation, Northwestern univer- sity, 1951. Hovland, C. L., Janis, I. L., and Kelley, H. H. Communi- cation and Persuasion. New Haven: Yale university Press, 1965. Hovland, C. L., Lumsdaine, A. A., and Sheffield, F. D. Experiments pp_Mass Communication. New Jersey: Princton university Press, 1949. Hovland, C. L. and Weiss, W. The influence of source credibility on communication effectiveness. Public Opinion Quarterly, 1961, 16, 635-650. 65 66 Jaksa, J. A. An experimental study of one-sided and two- sided argument, with emphasis on three two—sided Speeches. Speech Monographs, 1964, 31, 234. Lumsdaine, A. A. and Janis, I. L. Resistance to counter- pr0paganda produced by one-sided and two-sided propaganda presentations. gpblic Qpinion Quarterly, 1953, 17, 311-318. McCroskey, J. C. Experimental studies of the effects of ethos and evidence in persuasive communication. Uh- published Doctoral dissertation, Pennsylvania State University, 1966. (a) McCroskey, J. C. Scales for the measurement of ethos. Speech Monographs, 1966, 33, 65-72. (b) Paulson, S. F. The effects of the prestige of the speaker and acknowledgment of Opposing arguments on audience retention and shift of Opinion. Speech Monographs, 1954, 21, 267-271. Sharp, H., Jr. and McClung, T. Effects of organization on the speaker's ethos. Speech Monographs, 1966, 33, 182—183. Thistlethwaite, D. L. and Kamenetsky, J. Attitude change through refutation and elaboration of audience counter-arguments. Journal p§_AbnO§mal and Social Psychology, 1955, 51, 3-12. WOlfinger, R. E. Attitude Change toward source and issue resulting from one-sided and two-sided communication. unpublished Master's thesis, university of Illinois, 1954. APP ENDIX A 67 68 Secondary Measurements pf_the Audience's Objective .Biased Good Content One-Sided Good Delivery Disorganized Capital Harmful Good wrong Fair Negative Wise Perception p£_the Speech Subjective Unbiased Poor Content Two-Sided .Poor Organized Punishment Semantic Differential Beneficial Bad Right Unfair Positive Foolish 69 Authoriativeness Semant;p_Differential Informed : : : : : : : : Uninformed Unqualified : : : : : : : : Qualified Reliable : : : : : : : : unreliable WOrthless : : : : : : : : Valuable Intelligent : : : : : : : : Unintelligent Inexpert : : : : : : : : Expert Dynamism Semantic Differential Agressive : : : : : z : : Meek Hesitant : : : : : : : : Emphatic Forceful : : : : : : : : Forceless Timid : : : : : : : : Bold Active : : : : : : : : Passive Tired : : : : : : : : Energetic ‘ Chagacter §emantig_Differential 1 Unselfish : : : .: : : : : Selfish 1 Awful : : : : : : : : Nice Friendly : : : : : : : : Unfriendly ‘ Dishonest : : : : : : : : HDnest Pleasant : : : : : : : : Unpleasant Sinful : : : : : : : : Virtuous APPENDIX B 70 71 LOW-ETHDS INTRODUCTION We are trying to develop a new rating scale for the rating of speeches. The rating sheets I have passed out were designed to rate both a taped TV speech and a taped radio speech so some of the items, such as eye contact, will have to be ignored. The speaker we will hear is Anthony L. Capelli, who discusses the question of whether society should continue capital punishment. Mr. Capelli is presently an inmate at San Quentin Prison. Two years after dropping out of the university of Chicago Mr. Capelli was arrested on a charge of extortion in connection with the numbers racket. He received a sen- tence of 3 to 5 years and was paroled after serving 3 1/2 years of his sentence. Five years later Mr. Capelli was picked up on a narcotics charge in connection with the smuggling and illegal distribution of dOpe. However, he was not convicted of the charge. In 1963 Mr. Capelli was arrested in San Francisco on a charge of first degree mur- der in connection with a gangland slaying. He was con- victed and sentenced to die in the gas chamber, but on 72 June 15, 1964, Governor Brown, then Governor of Califor- nia, commuted his sentence to life imprisonment without parole. The speech we will hear was originally recorded at San Quentin for broadcast on the NBC radio program, “The Public Speaks." I should like to stress that the views exPressed by Mr. Capelli are his own and are not necessarily those of NBC. The series of programs on which this speech appeared was devoted to talks on highly controversial topics and the speakers themselves were highly controversial in some cases . 73 HIGH-ETHDS INTRODUCTION We are trying to deve10p a new rating scale for the rating of speeches. The rating sheets I have passed out were designed to rate both a taped TV speech and a taped radio speech, so some of the items, such as eye contact, will have to be ignored. The speaker we will hear is war- den William T. Anderson, who discusses the question of whether society should continue capital punishment. Mr. Anderson has spent his life in the field of criminology and is recognized as one of the outstanding criminologists in the united States. Mr. Anderson received his law degree from the university of Pennsylvania. He then joined the F.B.I. and spent eight years in their service. At the end of WOrld war II Mr. Anderson was assigned to the Allied Intelligence Agency to aid in the trials of Nazi war cri- minals as part of his F.B.I. assignment. In 1949 Mr. tinderson accepted a position as Assistant warden of the South Dakota State Penitentiary. In 1951 he moved to war- den, the position he has held for the past 15 years. Dur- ing his life warden Anderson has witnessed over 30 74 executions. The speech we will hear was originally re- corded for broadcast on the NBC radio program, "The Public Speaks." I should like to stress that the views expressed by Mr. Anderson are his own and are not necessarily those of NBC. The series of programs on which this Speech appeared was devoted to talks on highly controversial toPics and the speakers themselves were highly controversial in some cases . APPENDIX C 75 76 CAPITAL PUNISHMENT ONE-SIDED PRESENTATION Good evening, ladies and gentlemen. During 1966 so- ciety killed only one man. Now don't confuse this death with the thousands of deaths that occur on the highway each year or the loss of men in the Viet Nam war. Society as a whole did not kill these people. The only person who was killed by society was James D. French, age 30, whose execution occurred August 10 in the Oklahoma State Peni- tentiary under the existing laws of capital punishment. A jury made up of people just like yourselves found James French guilty. Having no other option under the law of capital punishment, society executed Mr. French. A contro- versy over capital punishment has raged in this country for the past century. Every week there is an article in the paper concerning this form of punishment. Being in the news so much, it is not surprising that nearly everyone has an Opinion on the question. What is surprising is that the American public is nearly evenly divided in their Opinions. It is difficult to understand how a people who 77 pride themselves on judging the desirability of a policy in light of cold hard facts can be so divided. Either the American pe0ple are irrational or the important facts on this question have not been made available. In my opinion the latter is the case. Thus, my purpose in this talk is to discuss the facts concerning capital punishment. Let's look at the five main arguments against capital punishment and the facts which support those argu- ments. The first main argument against capital punishment is that the administration of capital punishment is full of discrimination, discrimination on the basis of both race and sex. Let's look at the facts. First, racial discri- mination. Certain laws, in particular the law making rape a capital crime, are enforced only against Negroes in some states. Whites who commit the same crime receive a lesser penalty. Over the nation as a whole more Negroes are put to death than whites even though the Negro is only a small minority of the people compared to the white p0pulation. As clearly as the facts support the contention that Negroes are discriminated against in the administration of the death penalty, the facts based on sex are even more lop- sided. WOmen are almost never executed, no matter what their crime has been. Of the hundreds of people that 78 have been executed since the end of WOrld War II, only a small handful have been women. Actually, a man has about a hundred times greater chance of being executed than a wo- man for the same crime. Another major argument against capital punishment is that with all of the executions that have taken place some completely innocent men have been put to death. The exact number, of course, is impossible to estimate. In nearly every capital punishment state there have been cases where a real murderer has confessed to the crime on his deathbed, after a totally innocent man has been executed. Once a man has already been executed for a crime, it is impossible, of course, to correct this grave mistake. The only way we as a society can be absolutely certain that an innocent man is not executed is not to exe- cute anyone at all. It is interesting to note that after a recent publication of a case of this type two states have abolished capital punishment, and crime commissions in three others have recommended to their State Legisla- tures that these states should abolish capital punishment also. The third argument against capital punishment points out the fact that the threat of the death penalty has no effect on potential criminals. States that have the death 79 penalty have higher homicide rates, for example, than neigh- boring}. states which do not have capital punishment. We would expect that not having the death penalty as a threat to criminals should cause the crime rate to increase, but this is clearly not the case. The facts tell us that in reality capital punishment has the opposite effect if, in- deed, it has any effect at all. Capital punishment does not deter crime. The fourth major argument against capital punishment is that killing a man is not the best way to protect so- ciety against criminals who have already proven that they are a menace to society. Very few persons are actually executed each year anyway--only one last year. To suggest that executing this small number of criminals protects so- ciety is absurd. Obviously, keeping a man in prison is just as effective a method of keeping him from society as killing him. There is more than adequate space in our prisons to keep all capital offenders securely restrained so that they can't molest the rest of society. When you average out the few hundred persons who are convicted of capital crimes each year, there are only a handful of in- dividuals for each state. Certainly this small group of criminals can be retained in prison without overburdening 80 our vast prison system. But the main problem is that soci- ety cannot protect itself under our present laws. At pres- ent, in many cases, a jury has two options: one is to put the defendant to death, and the other is to acquit him and turn him loose in society. What happens when a jury is 99% sure about a man's guilt, but they don't want to impose the death penalty? The jury under present law must turn the man free in society. Now a possible solution to this problem is a law that will give the jury the Option of sending a man to prison without the possibility of parole. In the event later evidence is found in his favor, the man is still alive to have another trial. The final argument against capital punishment is based on the moral issue. Capital punishment itself is based on the philosophy of an eye for an eye, a tooth for a tooth. A recent case in point is the Richard Speck trial for the mass murder of the nurses in Chicago. The courts are having a very hard time in finding a jury that does not want revenge. The philosophy of revenge is noth- ing but the law of the jungle and such a philosophy is be- neath the dignity of civilized man. Every major religion in the world has denounced the philosophy of revenge as im- moral. As far as I have been able to determine, only the 81 Russian and Chinese Communists and a few primitive tribes in Africa hold the philosophy of revenge as a basic rule of their lives. well, I don't believe I need to carry the issue any further. We have examined the facts and we have seen that discrimination based on race and sex flourishes in the administration of the death penalty. We have noted that innocent men have been executed. we have seen that capital punishment does not deter crime. We have seen that society can be protected against criminals without killing them. And finally, we have seen that the alleged moral basis for capital punishment has been renounced by every major religion in the world. I have decided what I think should be done about the laws that permit capital punishment. I would guess that you have also. -82 CAPITAL PUNISHMENT TWO-SIDED PRESENTATION Good evening, ladies and gentlemen. During 1966 so- ciety killed only one person. Now don't confuse this death with the thousands of deaths that occur on the highways each year or with the loss of men in the Viet Nam War. So- ciety as a whole did not kill these peOple. The only per- son who was killed by society was James D. French, age 30, whose execution occurred on August 10 in the Oklahoma State Penitentiary under the existing laws of capital punishment. A jury made up of peOple just like yourselves found James French guilty. Having no other option under the law of capital punishment, society executed Mr. French. A contro- versy over capital punishment has raged over this country since the 17th century. Every week there is an article in the paper concerning this form of punishment. Being in the news so much, it is not surprising that nearly every- one has an opinion on the question. What is surprising is that the American public is nearly evenly divided in their opinions. It is difficult to understand how a people who 83 pride themselves on judging the desirability of a policy in light of cold, hard facts can be so divided. Either the American peOple are irrational or the important facts on this question have not been made available. In my opin- ion the latter is the case. Thus my purpose in this talk is to discuss the facts concerning capital punishment. Let's look at the arguments for both sides, and the facts which apply to those arguments. The first major argument advanced by the opponents of capital punishment is that the administration of capital punishment is full of discrimination, discrimination on the basis of both race and sex. Let's look at the facts. First, racial discrimination. Certain laws, in par- ticular the law making rape a capital crime, are enforced only against Negroes in some states. Whites who commit the same crime receive a lesser penalty. Over the nation as a whole, more Negroes are put to death than are whites, even though the Negro is only a small minority of the peo— ple compared to the white p0pu1ation. As clearly as the facts support the contention that Negroes are discriminated against in the administration of the death penalty, the facts based on sex are even more lopsided. WOmen are al- most never executed, no matter What their crime has been. .84 Of the hundreds of peOple that have been executed since the end of WOrld War II, only a small handful has been wo- men. Actually, a man has about a hundred times greater chance of being executed than a woman for the same crime. Another major argument advanced by those who oppose capital punishment suggests that with all of the executions that have taken place certainly some completely innocent men have been put to death. The exact number, of course, is impossible to estimate. In nearly every capital punish— ment state there have been cases where the real murderer has confessed to the crime on his deathbed after a totally innocent man has been executed. Now once a man has been executed for a crime, it is impossible, of course, to cor- rect the grave mistake. The only way that we as a society can be absolutely certain that an innocent man is not exe- cuted is not to execute anyone at all. It's interesting to note that after recent publication of a case of this type two states have abolished capital punishment, and crime commissions in three other states have recommended to their State Legislatures that capital punishment Should be abolished in these states also. On the basis of these two arguments of the opponents of capital punishment it certainly does seem clear that 85 this form of punishment Should be abolished. However, be- fore we make any decision On a controversy, we really Should look at both sides of the story. So what are the arguments advanced by those who favor capital punishment? The most frequent argument advanced by people sup- porting capital punishment is that the threat of the death penalty acts as a deterrent to crime. It deters potential criminals from committing crimes. This deterrent influence supposedly helps reduce the crime rate for those offenses against society to which capital punishment applies. That this is astrong argument for retaining the death penalty is attested to by the fact that the majority of the Ameri- can public believes that it's true. Let's examine the the- ory of deterrence very closely. This theory says that by threatening potential criminals with the death penalty, we will reduce the crime rate. Now turning this around, we would expect that not having the death penalty as a threat to criminals, the crime rate would increase. Clearly, the best way to test this theory is to look at the crime rate where the death penalty is legal and where it is not legal to see if there is a difference. The figures for murder, the most common crime for which the death penalty applies, show that the homicide rates are lower for the states that 86 have done away with capital punishment than for states that still retain it on the law books. Well, what do these facts tell us? Well, they suggest that homicide rates are exactly opposite to what they should be if the theory of deterrence were correct. While there are, of course, other reasons for retaining capital punishment on the law books. Many argue that it is essential to elimi— nate those criminals who have already proven that they are a menace to society so that society will not be threatened by them again. Now this certainly makes sense. If we run over a nail and get a flat tire, we want to remove that nail so that it won't give us any more flat tires in the future. Only a fool would leave the nail where it was. Now Opponents of capital punishment readily agree that so- ciety must be protected from criminals, but they suggest that capital punishment is not necessary to accomplish that end. They argue that very few persons are actually executed each year, and, of course, they are right; there was only one executed last year. Thus, to suggest that executing this small number of criminals actually protects society is absurd. Obviously, keeping a man in prison is just as effective a method of keeping him away from soci- ety as killing him is. There is more than adequate space 87 in our prisons to keep all capital offenders securely re- strained from molesting the rest of society. When you av- erage out the small number of persons who are convicted of capital crimes each year, there are only a handful of indi— viduals for each state. Certainly this small group of cri- minals can be retained in prison without overburdening our vast prison system. But the main problem is that society cannot protect itself under our present laws. At present in many cases a jury only has two Options: one is to put the defendant to death, and the other is to acquit him of the charges and turn him loose in society again. Now what happens when the jury is 99% sure about the man's guilt, but they do not want to impose the death penalty? The jury under present law must turn that man free in society. A possible solution to this problem is a law that will give the jury the option of sending the man to prison without the possibility of parole. In the event later evidence then is found in his favor, the man is still alive to have another trial. ' The final argument that is advanced by those in favor of capital punishment is the age-old concept that says so- ciety needs revenge. An eye for an eye, a tooth for a tooth. we've had a recent case in point on this subject 88 with the trial of Richard Speck who committed the mass mur- der in Chicago. The courts are having a difficult time in finding a jury who does not want revenge. Opponents of capital punishment say that such a philosophy is nothing but a law of the jungle and is beneath the dignity of a civilized man. Both sides quote the Bible to support their arguments. Every major religion in the world has denounced the philosophy of revenge as immoral. As far as I've been able to determine, only the Russian and Chinese Communists and a few primitive tribes in Africa hold the philosophy of revenge as a basic rule in their lives. Well, I don't be- lieve I need to carry this issue any further. We've ex- amined both sides of this controversy. We've looked at the facts. we've seen that discrimination based on race and sex flourishes in the administration of the death pen- alty. we have noted that innocent men have been executed. We have seen that capital punishment does not deter crime, that society can be protected from criminals without kill- ing them, and that the alleged moral basis for capital pun- ishment has been denounced by every major religion in the world. I have decided what I think should be done about the laws that permit capital punishment. I would guess that you have also. 111111111111115