ACCEPTANCE OF PERSUASIVE INFLUENCE AS RELATED TO THREE DLMENSIONS OF SOURCE EVALUATION Thai: for the Dogma of Ph. D. MECHIGAN STATE UNIVERSITY Roberf .I. Men‘z 1966 x 1 -,,--.. .N.———.. ‘ ”Wm!“ ., ‘ “£7 LIBRARY Michigan State University lflyllllfllwfljfllllllfllljflfllw"1!"le L This is to certify that the thesis entitled Acceptance of Persuasive Influence as Related to Three Dimensions of Source Evaluation presented by Robert J. Mertz has been accepted towards fulfillment ' of the requirements for Ph.D. degree in Communication Major professor \ Date November 28, 1966 0-169 PLACE IN RETURN Box to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE DE‘CM‘SW on w 1-18) F 68;?! 6/01 cJCIRClDateDuepss-pJS J W __ m". '7‘" " ‘7 (“LU , ‘ “'— m ABSTRACT ACCEPTANCE OF PERSUASIVE INFLUENCE AS RELATED TO THREE DIMENSIONS OF SOURCE EVALUATION by Robert J. Hertz Source credibility (the extent to which a message source is favorably evaluated) has been shown to be a useful predictor of reSponse to persuasive communication. This study sought to estimate previous conceptualizations of the credibility variable to take into account (1) the multi-dimensionality of source evaluation, and (2) the source- receiver relationships which is implied by credibility judgments. A multi-dimensional model of source evaluation was developed from the results of recent factor analytic research. The model stipulates receiver judgments of the message source on three independent evaluative .* cfimensions -- safety, qualification, and dynamism, as the principal deter- minant of influence acceptance. A relational conceptualization of credibility was advanced, based on previous research in interpersonal perception. It was purposed that an individual receiver's judgment of his own credibility in a given situation should serve as the principal anchorage for evaluation of the message source. Seven hypotheses were purposed to test the same model. H1: In a situation involving persuasive communication, influence acceptance will be directly related to the perceived Safety, Qualification, and Dynamism of the message source. OJ Robert J. Mertz In a situation involving persuasive communication, acceptance of source influence will be directly related to the direction and magnitude of perceived self-source discrepancy on the Safety, Qualification, and Dynamism dimensions. Knowledge of the direction and magnitude of perceived self-source discrepancy on the Safety, Qualification, and Dynamism dimensions will afford more accurate predictions of influence acceptance than will knowledge of source evaluation alone. There will be a positive relationship between influence acceptance and perceptions of relative source Safety. There will be a positive relationship between influence acceptance and perceptions of relative source Qualification. Among subjects who perceive the message source as higher than self on the Safety dimension, there will be a positive relationship between influence acceptance and perceptions of relative source Dynamism. Among subjects who perceive the message source as lower than self on the Safety dimension, there will be a negative relationship between influence acceptance and perceptions of relative source Dynamism. Robert J. Mertz A list of six topics containing recommendations on current social issues, together with descriptions of potential sources of information on those topics, was submitted to a sample of undergraduate students (N=155). Subjects indicated their opinions on the topics by a seven-point evaluative loaded semantic differential scale, and rated each of the sources on seven- point scale measuring perceived safety qualification. In addition, each subject rated himself as a possible source of information on each topic, using the same scales. Two weeks later, subjects were exposed to a written message arguing _against their Opinion stand on the experimental topic. The message, which was constant across subjects, was attributed to one of eight sources. The sources represent varying levels of perceived safety, qualification and dynamism as established in the pretest source ratings. Following exposure to the persuasive message, subjects re-rated the experimental topic. Multiple correlations were computed to determine the relationship of pre-post topic attitude change scores to (l) the measures of perceived source safety, qualification and dynamism, and (2) measures of perceived self-source discrepancy on the three credibility dimensions. In addition, zero-order correlations between topic attitude change and the measures of absolute source evaluation and self-source discrepancy were computed for each of the dimensions separately. Results obtained in this study, while in accord with the same findings of previous research in the credibility area, did not support the major theoretic hypotheses. Significant positive relationship was found between influence acceptance and the perceived safety, qualification and dynamism of the message source. However, this relationship was significant only Robert J. Mertz for those subjects who perceived the message source as uniformly high or low on all three of the evaluative dimensions. Among subjects who saw the sources as high on one dimension and relatively low on another, no clear results were obtainable. The attempt to predict influence acceptability from multiple regression on the three dimensions of source evaluation failed to produce significant results. The hypotheses relating influence acceptance to perceive self— source discrepancy on the three evaluative dimensions were not supported. The predicted positive relationship between influence acceptance and perceptions of relative source safety was not found. The predicted positive relationship between influence acceptability and perceptions of relative source qualification was supported. This variable was the best single predictor of influence acceptance. The hypothesis relating influence acceptability to perceptions of relative source dynamisms among subjects who saw the source as relatively safe or unsafe were not supported. These results indicate the need for further research to explore the utility of multi-dimensional conceptualization of source evaluation in persuasive communication. The data also seem to suggest that further credibility measures should be designed to better define the source- receiver relationship which is expressed by credibility judgments. Finally, examination of the data suggests that more research is needed to determine the impact of source evaluation in influence situations involving particular topic areas and with audiences of differing demo- graphic and personality characteristics. ACCEPTANCE OF PERSUASIVE INFLUENCE AS RELATED TO THREE DIMENSIONS or SOURCE EVALUATION By ' (n6 .2 Robert JJ’Mertz A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1966 ACKNOWLEDGMENTS The dissertation is rarely the product of solitary effort on the part of the candidate. Moreoften it is the cooperative exercise including the energies of teachers, colleagues, friends, family and a fair amount of total strangers who are dragooned into the project. I cannot hope to discharge my debt to all those who helped in various stages of this study. I can only acknowledge that the dissertation would not have been completed without them. In particular I wish to eXpress my appreciation to my adviser, Dr. David K. Berlo, and to Dr. Bradley Greenberg and Dr. Gerald R. Miller. Their considerations and criticisms were always appreciated if not always followed in time. Al Talbott, Verling Troldahl and Vince Farace gave me moral support and expert advice throughout the study, and Shirley Sherman proved a tower of secretarial strength under arduous conditions. To all these people -- thanks! ii Table of Contents CHAPTER I INTRODUCTION . . . . . . . Statement of Hypotheses II METHOD . . . . . . . . . . Post-test Procedures . . III RESULTS . . . . . . . . . IV SUMMARY AND DISCUSSION . . BIBLIOGRAPHY I O O O O O O O O O O 0 iii 14 18 23 29 49 59 TABLE 10 ll 13 in List of Tables Mean Pre-Post Attitude Change Toward the Experimental TOpic, by Groups . . . . . . . . . . . . . . Observed Mean Differences Between High and Low Sources on Safety Dimension . . . . . . . . . . . . . . Observed Mean Differences Between High and Low Sources on Qualification Dimension . . . . . . . . . . . Observed Mean Differences Between High and Low Sources on Dynamism Dimension . . . . . . . . . . . . . . Regression of Attitude Change Scores in Absolute Source Evaluation Scores . . . . . . . . . . . . . . . . Regression of Attitude Change on Perceived Self- Source Discrepancy Scores . . . . . . . . . . . . . . . Correlations between Attitude Change Scores and Perceived Dynamism of the Message Source (AnalYSis Group 1) O O O O O O O I C O O O O O O O O O 0 Correlations between Attitude Change Scores and Perceived Absolute Dynamism of the Message Source (Analysis Group 2) . . . . . . . . . . . . . . . . Mean Attitude Change for High, Low and Mixed Source Evaluation Groups . . . . . . . . . . . . . . . . Mean Attitude Change Scores for Eight Source Evaluation Categories . . . . . . . . . . . . . . . . . . Analysis of Variance Summary Table of Mean Attitude Change Scores for Subject in the Eight Absolute Source Evaluation Conditions (N = 155) . . . . . . . . . Mean Attitude Change Scores for the Eight Categories of Perceived Self-Source Discrepancy (N = 155) . . . . . Analysis of Variance Summary Table of Attitude Change Scores for Subjects in the Eight Self-Source Discrepancy Conditions . . . . . . . ... . . . . . . . . Mean Attitude Change Scores for Eight Source Induction Categories - Successful Induction Subjects Only (N = 63). iv 29 31 32 33 37 38 41 M2 #3 an an #5 us TABLE 15 List of Tables Page Analysis of Variance Summary Table of Mean Attitude Change Scores for Subjects in the Eight Source Induction Categories -- Successful Induction Subjects only (N=63) O O O C O O O O O O O O O O O O O O LE8 Figure 1. Figure 2. List of Figures Page Eight Source Descriptions and Mean Evaluative Ratings on Three DimenSionS (N=20) o o o o o o o o o o o o o o o 21 Scales of Self and Source Evaluation . . . . . 22 vi List of Appendices Page APPENDIX A TEXT OF THE PERSUASIVE MESSAGE . . . . . . . . . . . . 63 APPENDIX B THE PRE-TEST QUESTIONNAIRE . . . . . . . . . . . . . 65 APPENDIX C THE POST-TEST QUESTIONNAIRE . . . . . . . . . . . . . 87 APPENDIX D INTERCORRELATION MATRIX, ATTITUDE CHANGE, SELF- EVALUATION, SOURCE EVALUATION, AND SELF-SOURCE DISCREPANCY SCORES O O O C O O C C O O C O O O O O O O 93 vii CHAPTER I INTRODUCTION Students of human behavior have long reCOgnized that an individual's response to persuasive communication depends, at least in part, on "who said it". Research, theory, and the overwhelming evidence of common experience affirm the notion that the identity of the communicator plays an important role in determining audience acceptance of new ideas and information. Some sources are clearly more effective than others in getting across difficult ideas and in winning support for controversial points of view. Professional communicators typically go to great lengths to ensure that their messages are presented by appropriate Spokesmen. Advertisers pay nationally known athletes and movie stars to di3play their products, on the assumption that their endorsement will add to the impact of the sales message. Trial lawyers make liberal use of "expert" testimony and call on reSpected members of the community to attest to the good character of their clients. Governnent agencies attempt to soften opposition to controversial new prOgrams by having them announced through the most resPectable channels. All these examples reflect the common belief that Acertain sources of information and Opinion are somewhat more acceptable than others. The movie star's endorsement is presumed to carry more weight than the assertions of an unknown pitchman; the testimony of the expert witness is regarded as more authoritative than the same arguments presented by a layman; controversial government policies become more acceptable when presented by men of recognized integrity and competence. The professional communicator's concern with "who" presents his message is based on intuition and experience; however, the findings of field research support this concern. The differential effectiveness of various kinds Of message sources has also been repeatedly demonstrated in the laboratory findings Of the past four decades. Mausner (1953) and Paulson (195A) for example, have demonstrated that agreement with Opinion state- ments is enhanced when the statements are attributed to prestigeful or authoritative sources. Similar findings have been reported by Haiman (19u9), by Hovland a Weiss (1951), by Kelman (1953), by Kerrick (1958), by Hollander (1961) and many others. Some investigators (Lewis, 19u1; Asch, 19u8) have presented evidence that not only the acceptability Of an Opinion statement, but even its meaning, can be altered by attributing it to different sources. These writers suggest that the supposed authorship of an Opinion statement serves not only to modify the receiver's evaluation Of the statement in the direction Of the source's prestige, but also functions as a part Of the necessary context for determining the meaning Of the statement. One cannot fully understand a statement, it is argued, without knowing who made it or the social context in which it was delivered. Other researchers have demonstrated that the effects Of source image Operate not only in the evaluation Of social, economic, and religious issues, but also affect aesthetic judgment (Bernberg, 1953), personal taste (Cole, 195%), and even food preferences (Ducker, 1938). Still other investigators have eXplored the influence Of source image in interpersonal influence situations (Croner 8 Wills, 1961) and in the context Of the mass media (Highlander, 1953; Adams, 1962). These studies have been conducted under a variety Of descriptive labels which have been used to denote the influence Of source identification on communicative effectiveness -- ethos, prestige, status, image, reputation, and most frequently in recent years, source credibility. Whichever label is employed, researCh consistently supports the conclusion that a highly regarded message source is a more effective transmitter Of ideas and information, whether the basis for this high regard is his apparent social position, knowledge, physical attractiveness, or sincerity. / This single, rather Obvious conclusion - that "highly credible" sources are more effective than "less credible" ones - summarizes much Of what is presently known about the influence of the source's image on ‘ persuasive communication. As Clevenger and Andersen (1963) note in their review Of research in the area: "Deepite the great number Of experimental studies relevant to eghgs, the scope Of this concept is sUch that the findings are not yet sufficiently numerous and SOphisticated tO permit definitive conclusions about the Operation of ethical proof." This somewhat pessimistic evaluation Of the present state Of understanding Of the role Of credibility in the persuasion process reflects the considerable uncertainty which exists in the literature about the answers to such questions as: What are the Specific Characteristics Of a message source which make him an effective transmitter Of ideas and information? How may these characteristics be measured? Is source effectiveness dependent on a single characteristic such as "prestige", or are there multiple bases for credibility? Is credibility Specifiable in terms Of innate source characteristics or is it solely determined by receiver perceptions? Can credibility be defined solely in terms Of perceived attributes of the message source, or do credibility judgments express a particular kind Of influence relationship between source and receiver? How does credibility Operate to influence audience acceptance Of the communicator's message? The confusion which surrounds the answers to these questions reflects the failure Of previous research to place the credibility construct within a coherent theoretic framework or to provide a useful model for the Operation of source credibility in persuasive communication. Such a model should specify in detail: (1) the £2£E§_Of credibility; (2) appropriate measure- ment procedures for establishing credibility differences; (3) the dimensionality Of credibility; and (u) the nature Of the relationship which is expressed by credibility judgments. The present study attempts to supply one model for the Operation Of credibility which meets the above requirements, and provides an Operational test of the preliminary model. The locus Of credibility; Early theorists and researchers in the field commonly assumed the credibility Of a message source to be intuitively determinable through consideration Of such relatively Objective characteristics as the source's age or social status. This assumption was mirrored in the common experimental practice Of employing Senators and college professors as high credible sources and Communist Party Spokesmen and high school freshmen as sources in the low credibility condition. Direct measurement Of audience perceptions of these sources was rarely employed, and then usually only as a check on the experimenter's apriori judgment. Current theory and research, by contrast, place credibility squarely within the domain Of receiver perceptions. In their review Of the literature, Clevenger and Andersen (1963) detail the shift away from the traditional view that ethos or credibility is intuitively determinable, and the developing reCOgnition that the bases Of source effectiveness must be found in the perceptions of the audience. These reviewers note the abandonment Of the Old a priori procedures for assigning experimental sources to "high" or "low" credibility conditions and the growing use of direct audience measures to establish credibility differences. The measurement of credibility: Only a few investigators have made the development of a reliable measure of credibility their main Objective. The bulk Of the research involving credibility or related concepts has been directed toward assessment of the effects Of differing levels of credibility on such dependent variables as learning and attitude change. In these studies the construction Of an index Of credibility has occupied a subordinate position in the design of the research. The methods employed to measure credibility are roughly the same for both types Of investigation. They include: (1) Subjects' rank orderings of source on credibility or some single related dimension (Sherif, 1935; Das, Rath, Das, 1955); (2) Indices Of credibility-related personal characteristics of the source based on sociometric-choice data (Cole, 1951i); (3) "Prestige" indices derived from measures Of attitude change (Kulp, 1931+, Lurie, 1938); (u) subjects' assignment Of sources Of positions on a single linear rating scale tapping credibility or a related dimension of the source's image (Saadi 8 Farnsworth, 193a, Lorge 6 Curtiss, 1936); (5) Credibility measures derived from responses tO several (presumably) related attitude items, utilizing Thurstone or Guttman scaling techniques (Walter, 19ue); (6) Likert-type measuring devices in which the source's credibility is Obtained by summation Of the ratings assigned to him over a number Of related scale items, (Wolfinger, 1955; BerlO 8 Gulley, 1957); and (7) Multidimensional measures Of credibility in which factor analyses or similar techniques are employed to discover those aSpects of a source's projected image which are relevant to his persuasive effectiveness, and separate measures are taken on each of the independent dimensions thus established (Andersen, 1961; Berlo 6 Lemert, 1961; Rarick, 1963). Examination Of these various methods for determining credibility differences among message sources suggests that they vary principally in terms Of: (l) the lgyel_of source measurement which they afford; (2) the extent to which they reflect an attempt tO empirically_determine the dimensions of source image which are relevant to credibility, and (3) the assumptions about the dimensionality_of credibility which underlie use of the particular measuring device. The dimensionality Of credibility: The early theoretic literature is replete with arguments linking persuasive eEectiveness with one or another particular source characteristic. Many of the earliest studies in the credibility area (Farnsworth 8 Misumi, 1931; Arnett, Davidson 8 Lewis, 1931; and Duncker, 1938) assumed a source's effectiveness to be a direct manifestation of his social status or "prestige". Much Of the early research into the effects of source image, in fact, was subsumed under the heading Of "prestige suggestion". Later researchers advanced a wide variety Of personal characteristics to account for persuasive effectiveness, including such diverse attributes of the message source as his sinceriEy (Hildreth, 1953), his physical attractiveness (Haiman, 19%), and his verbal fluency (Miller 8 Hewgill, 196%). Still other writers have presented evidence to show that a source's communicative effectiveness may depend on his perceived Objectivity_ (Hovland 8 Mandell, 1952) or lack Of manipulative intent (Walster 6 Festinger, 1962; Kiesler 6 Kiesler, 196A), his competence (Hollander, 1960; Croner 8 Willis, 1961), and even his apparent sociability (Haiman, 1999, Barnes, 1960). While early theorists and researchers Often disagreed on the Specific antecedents Of source effectiveness, they were unanimous in treating credibility as an essentially unidimensional attribute Of the communicator. As noted in the earlier discussion Of credibility measurement procedures, the initial efforts to establish direct measures Of credibility or ethos indexed a single dimension of the source's projected image. Such indices typically consisted Of a single linear rating scale to tap receiver judgments Of the communicator's likableness (Saadi 6 Farnsworth, 193a), prestige (Adams, 1960), or trustworthiness (Hovland 8 Weiss, 1951). Later studies used several such scales and summated ratings to establish differences between experimental sources on some single dimension of judgment. Wolfinger (1955) and Berlo 6 Kumata (1956), for example, used batteries Of evaluatively loaded semantic differential scales to determine audience attitudes toward message sources. In these studies, the sum Of the audience ratings over the several scales was used as a unidimensional measure Of one aSpect of the source's image - the "evaluative dimension" proposed by Osgood, Suci 8 Tannenbaum (1957). This traditional view Of credibility or ethos as a unidimensional source attribute has been questioned by a number Of writers. As early as 1953, Hovland, Janis 6 Kelley suggested the utility Of a multidimensional model Of source evaluation, and proposed two dimensions Of the communi- cator's image - perceived expertness and greeived trustworthiness, as especially relevant to persuasive effectiveness. It should be noted, however, that these theorists provided little in the way Of direct empirical support for these twO dimensions of source evaluation as the principal components Of credibility. Within the past three years, a number Of researchers have attempted to empirically determine the specific dimensions Of communicator image which are relevant to his acceptance as a source Of information and influence, and to provide reliable instruments to tap the receivers' perceptions Of the source on these dimensions. These investigations, while independently conducted, have generally proceeded along similar lines -- a list Of source-descriptive terms is collected from the theoretic and experimental literature or through solicitation of subjects' descriptions Of public and private sources whom they consider acceptable or not acceptable over a wide range of topics. These descriptions are then translated into semantic differential or Likert-type scales, and a number Of sources Of established acceptability are rated on the scales by a large _group Of subjects. The ratings are then correlated and subjected to factor analysis. A set Of specific scale items is then selected, on the basis Of strength and purity Of loadings, to represent each Of the principal factors. The basic studies in this area were conducted by Andersen at Wisconsin (1961) and Berlo 6 Lemert at Michigan State (1961). More recent investigations by Schweitzer (1966) and by McCroskey (1966) and recent replications Of Berlo's work (Berlo, Lemert 8 Mertz, 1965) have followed essentially the same pattern. When allowances are made for slight differences in the methods Of Obtaining the original scale arrays, types Of sources and subjects employed, and factor solution method, all these studies have yielded remarkably similar results. The findings suggest that there are three relatively independent judgmental dimensions along which an individual evaluates a message source, and which influence the source's acceptance and influence potential. Specifically, the findings indicate that receivers evaluate a message source according to his perceived "Safety" (his manipulative intent and predictability), his "Qualification" (his topic- bound expertness as well as his perceived general ability or intelligence) and his "Dynamism" or "Energy" (something like a combination Of the potency and activity dimensions Of general connotative meaning as discussed by Osgood, Suci 8 Tannenbaum). 10 These studies provide an Operational base for defining the source evaluation process. Their results emphasize the multi-dimensionality of the variable, and they fUrther support the argument that credibility must be defined in terms of the perceptions Of the receiver rather than any innately determined characteristics of the source. Perhaps their greatest immediate significance, however, lies in the fact that they provide the first empirically derived measures Of source judgments. As such, they furnish a stable base for further exploration Of the Operation Of the variable in persuasive communication situations. Credibility as a relational variable: Most discussions of credibility suggest a monadic approach to the conceptualization and operationalization Of the variable. In both traditional and current usage of the concept, "credibility" typically refers to and is defined by, a single element in the persuasive situation — evaluation Of the source Of persuasive communi- cation. The earliest treatments Of ethos or credibility stress inherent source characteristics as the basic determinants Of his influence potential. In these early studies, the persuasive effects of differing levels of source "prestige" or "authority" are assumed to be relatively independent Of other variables in the message situation. Later approaches have similarly obscured the relational nature of the variable. While current conceptualizations of credibility emphasize direct measurement Of the receiver's perceptions of the message source, these measures are generally not interpreted as expressing a source- receiver relationship. As used in most recent studies, indices Of 11 perceived source "competence" or "trustworthiness" are intended only to provide more empirical assessments of source characteristics. Implicit in both these approaches are twO fundamental and related assumptions about the nature Of credibility. First, a source's influence potential in any situation is presumed to depend directly and solely on his evaluative rating on one or more dimensions of audience judgment. Given adequate measures of source evaluation, predictions of persuasive effect may be made without reference to other variables in the situation. Second, credibility judgments are commonly assumed to eXpress relatively absolute source values. A highly evaluated source is expected to be maximally effective in all relevant influence situations. Conversely, a low evaluative rating leads to predictions of minimal persuasive effect in all message situations. In contrast to the singular focus of monadic approaches to the variable, one can conceptualize credibility by stressing the source- receiver relationship which is implied by credibility judgments. This conceptualization of credibility asserts a source's influence potential to be a joint function of (1) receiver evaluations of the message source, and (2) receiver self-evaluations. Under this approach, a source's acceptability as a source of information and Opinion in a given message situation is determined by a series of self-source comparisons on Specific judgmental dimensions. To the extent that the receiver rates the source as 1135119? than self on these dimensions - whatever his absolute rating - he will tend to accept the proferred influence. On the other hand, when the receiver judges the message source lower than self on these 12 dimensions, he will reject the source's influence attempts. Thus, a source's "credibility" and subsequent influence potential are seen to depend, not on his absolute rating, but on his evaluative standing relative to receiver self-evaluation. Self-achieving as a basis fOr credibility has not been advanced in previous theoretic discussions of the variable. TO a considerable degree this omission is attributable to the kinds of message sources employed in most early research. Early studies typically pitted such sources as T. 8. Eliot y§ an unemployed dishwasher on the subject Of poetry, or the Surgeon General Of the United States y§_a convicted narcotics peddler on the need for stronger drug control laws. Use of such extreme sources, while ensuring significant differences between high and low credibility conditions, effectively blocks any consideration Of credibility as a relational variable. Few receivers would consider themselves better qualified than T. S. Eliot on the subject of poetry. By the same token, few receivers would evaluate themselves as lower in credibility than the narcotics pusher. It can be argued, however, that the extreme source conditions imposed by most early research are seldom encountered in real life. The individual receiver is rarely called upon to evaluate himself in comparison to either T. S. Eliot or the convicted criminal. Most instances Of persuasive interaction involve receiver judgments about message sources who are more directly comparable to himself. In these instances, a source- self conceptualization of credibility would suggest that the individual's selfhperceptions furnish the principal anchorage for evaluation Of the source Of persuasive communication. In this study, the concept Of sglff credibility is proposed to describe the receiver's self-evaluations on 13 those judgmental dimensions which are directly relevant to his acceptance or rejection of source influence. Self-Credibility as Anchorage for Source Evaluation: The concept of self-credibility has received little attention in the literature dealing with persuasive communication. As developed here, self-credibility is not meant to be equated with the more general notions of "self-concept" or "self-esteem" as treated in theories Of personality. These concepts have been commonly used to denote an individual's generalized perception of self across all interaction situations (Wylie, 1961; Combs, 1962). The notion of self-credibility, by contrast, is developed directly from the research and theory in source credibility, and refers to a Specific and limited portion Of the individual's total "self-concept". As used here, self- credibility exPresses the individual's evaluation of himself on those dimensions Of judgment which are relevant to his performance in a particular kind Of interaction situation; namely, one in which he acts as a receiver of persuasive communication. It is further proposed that the judgmental dimensions which are most relevant for the individual as a receiver are those which he uses to evaluate the source of the persuasive communication; i.e., Safety, Qualification, and Dynamism. While the notion Of self-credibility has not been explicitly developed in previous research, the literature in the area of Interpersonal Perception Offers some support for this treatment Of the concept. Hastorf, Richardson 6 Dornbusch (1958), for example, report that; "...there is a strong positive relationship between categories which people use in describing other people andii describing themselves." These writers further 1“ indicate that the Specific judgmental dimensions which are applied at any _given time are determined by the activity in which the individual is engaged. They state that: "...a person has a core of generally consistent categories used in describing all people, and a set of more particular categories which depend more on situational factors." On this point, Jones 6 Thibaut (1958) suggest that reCOgnition that persuasive interaction is about to take place arouses a "perceptual set" of relevant judgmental dimnsions . Receiver self-credibility is advanced in this study as a major determinant of communication effect in persuasive situations. Specifically,. it is proposed that an individual's judgments of his own credibility serve as "anchors" or "comparison points" against which he evaluates the message source . Statement of Hypotheses From the preceding discussion the following hypotheses have been developed as to the functioning of credibility in a persuasive communication context. H1: In a situation involving persuasive communication, influence acceptance will be directly related to the perceived Safety, Qualification, and Eypamism of the message source. This hypothesis provides a test of the multi-dimensional model Of source evaluation proposed by Berlo 8 Lemert, Andersen, and others. NO direct test of this model exists in the previous research literature. 15 H2: In a situation involving persuasive communication, acceptance Of source influence will be directly related to the direction and magnitude Of perceived self-source discrepancy on the Safety, Qualification, and Dypamism dimensions. Hypothesis 2 stems directly from the source-self model of credibility judgments advanced earlier. It extends the concept of "credibility" beyond the previous focus on source-evaluation to include also the receiver's judgments of his own credibility within the message situation. H : Knowledge of the direction and magnitude Of perceived 3 self-source discrepancy on the Safety, Qualification, and Dypamism dimensions will afford more accurate predictions of influence acceptance than will knowledge of source evaluation alone. This hypothesis provides a direct test Of the relative adequacy Of source and source-self conceptualizations Of credibility in terms of their ability to predict attitude change in persuasive communication situations. Hypotheses 1 through 3 reflect the principal theoretic interests of this study. The following hypotheses specify in greater detail the predicted relationship between influence acceptance and each Of the component dimensions of credibility. Ho: There will be a positive relationship between influence acceptance and perceptions Of relative source SafeEy. 16 H5: There will be a positive relationship between influence acceptance and perceptions of relative source Qualification. These two hypotheses assert a simple linear correlation between acceptance of source influence and self-source discrepancy judgments on the SafeEy and Qualification dimensions. TO the degree that the receiver judges the message source as more trustworthy or better qualified than himself, he will reSpond favorably to the influence attempt. To the extent that he sees the source as less trustworthy or less qualified than himself, the receiver will maintain his original Opinion on the issue. The relationship between influence acceptance and perceptions of relative source Dypamism is more complicated. Berlo 8 Lemert suggest that judgments on this factor function primarily to intensify the source's perceived standing on the SafeEy factor. Given an initial evaluation of the message source as relatively safe or dangerous, the intensity Of this evaluation will be heightened by perceptions of high source dynamism; i.e., low-safety, high dynamism sources will be seen as potentially more "dangerous" than low safety, low dynamism sources, and their influence attempts will more likely be rejected. Hypotheses 6 and 7 summarize the predicted effect of this inter- action on the influence acceptance variable: H6: Among subjects who perceive the message source as higher than self on the Safesy dimension, there will be a positive relationship between influence acceptance and perceptions of relative source gypamism. 17 Among subjects who perceive the message source as lower than self on the SafeEy dimension, there will be a negative relationship between influence acceptance and perceptions of relative source Dynamism. CHAPTER II METHOD Pretest Procedures: Approximately 200 students enrolled in summer courses in Communication or Sociology volunteered to participate in "a survey of student reaction to controversial public issues". This survey, conducted under the auspices Of an unidentified national research organization, was administered during a class period regularly scheduled for research and discussion. The pretest questionnaire indexed three kinds Of variables; (1) subjects' Opinions on six controversial public issues (Tl attitude), (2) subjects' evaluation of various possible sources Of information and Opinion on these issues (source credibility), and (3) the subject's evaluation Of himself as a potential source of information and Opinion on each of the issues (self-credibility). Pretest Attitude Measurement: In the first section of the questionnaire, subjects were asked to reSpond to a series of Opinion statements dealing with six public issues currently in the news. The topics were: (1) "Extension of medicare to US citizens Of all ages through a national health insurance plan financed by tax funds..." (2) "Abolishment Of the Federal Office Of Economic Opportunity ('war on poverty') administration..." (3) "Revision Of present Selective Service regulations to eliminate student deferments except in a few cases in the national interest..." 18 19 (u) "Restricted use Of nuclear weapons in Viet Nam in non- populated areas..." (5) "Strict University control and supervision Of student political organizations, eSpecially those which Show Signs Of communist domination..." (6) "A mandato jail sentence and one-year suspension Of driver 3 license for anyone convicted Of drunken or reckless driving..." Subjects indicated their Opinions on each of these issues through ratings on five 7-point semantic differential scales. These scales, which are suggested by Osgood, Suci 8 Tannenbaum (1957) to measure response on the evaluative dimension, were: good-bad, wise-foolish, valuable-worthless, fair-unfair, and honest-dishonest. The evaluative scales were presented after each topic statement, and the scale ends were randomly reversed to minimize response set. The subjects' responses were coded l (unfavorable) through 7 (favorable) and summed over the five scales to provide the pretest measure Of attitude on the issue. Source Credibility Measures: Following their rating Of the ex- perimental issues, subjects we presented with brief descriptions of several persons who might serve as sources of information or Opinion on these issues. The source descriptions were selected from a set of 30 descriptions which had previously been rated by a small sample Of students and student wives. On the basis of these pretest ratings, eight sources were selected to represent the eight possible combinations Of high and low on the three dimensions of source evaluation proposed by Berlo and Lemert. The eight combinations Of high-low Safety, Qualification and Dynamism, together with the appropriate source descriptions, are presented in Figure 1. Subjects rated each of the sources separately on 2O twelve semantic differential scales measuring the source's perceived Safety, Qualification and Dynamism. The four scales used for each dimension were selected from a list provided by Berlo, Lemert 8 Mertz (1965). The scales, with their loadings on each Of the three factors, are presented in Figure 2. The source scales were presented in mixed order, and the scales were randomly reversed to avoid response set. Summary ratings for each source were computed within each of the three dimensions. Self-Credibility Measures: The third section of the pretest questionnaire asked the subject to review his own qualifications as a potential source Of information on the issues presented. The topic statements were repeated separately, and the subject was instructed to rate himself on each topic using the same 12 scales as in the measure of source evaluation. The order of presentation Of the selfeevaluation measures and the source-evaluation measures was systematically varied within the sample to control possible order effects. Half the subjects rated the sources first; the other half rated themselves first. Following administration of the topic attitude, source-evaluation, and self-evaluation measures, subjects were asked to comment on the study and its purpose. The experimenter then infOrmed the subjects that the "survey" was completed, thanked them and left. An explicit attempt was made to leave subjects with the impression that the pretest questionnaire 21 Saf. Qual. Dyp. Source Description Hi Hi Hi "A leading national clergyman who has 7 (i;25,6)(§?23,5)(§}25.u) been appointed by the President to a Special advisory commission on social and economic affairs..." Hi Hi Lo "A 76-year Old former Chief Justice 1 (RE27,0)(§}27,0)(§}12,0) Of the Supreme Court of the State of “ Rhode Island. . ." *Hi LO Hi "A housewife and mother Of four who 8 (2}23,5)(§}g.2)(§}24.0) has been asked to write a report on public affairs for her PTA group..." Hi LO LO "A middle-aged male English teacher 2 (RE25.5)(R}8.M)(R}11.8) in a suburban girls' school, on campus for summer courses..." LO Hi Hi "A professional lobbyist who has been a (RE7.2)(7}25.0)(R}26.5) described by members of Congress as 'The most persistent and determined pleader of Special interests ever seen on Capitol Hill'"... LO Hi Lo "A ranking government Official in the 3 (RT7.6)(R}26.0)(R39.0) Truman administration who was involved in a scandal which resulted in his removal from Office. He now leads an inactive life near Washington, D.C. ..." LO LO Hi "The former editor Of a campus magazine 6 (RTB.4)(R}10.5)(R}24.6) of extreme left-wing political Opinion, currently on academic probation for low grades..." Lo LO Lo "A once-controversial Baptist minister, 5 (§}g,o)(7T5.o)(§}7,5) now living in a home for aged clergy. He was censured by his church in 1938 for anti-semitic and racist preaching, and retired from active service shortly afterward..." Figure 1. Eight Source Descriptions Figure 1. Eight Source Descriptions and Mean Evaluative Ratings on Three Dimensions (N:20) * This Source Description was selected on the basis of Evaluative Ratings obtained in an earlier study (Nz60) 22 Factor Loadings Safety Scales Safety; Qual. Dypamism Safe---Dangerous .81 .26 .09 Honest---Dishonest .79 .17 -.08* Qualification Scales Skilled---Unskilled .33 .77 .18 Informed---Uninformed .3“ .74 .18 Qualified---Unqualified .37 .76 .09 Experienced---Inexperienced .25 .80 .19 Dynamism Scales Bold---Timid -.31 -.08 .6“ Energetic---Tired .29 .2“ .6“ Active---Passive .17 .25 .61 Emphatic---Hesitant .Ol .19 .70 *Loadings for this scale Obtained from an earlier u-factor solution. Figure 2. Scales for self and source evaluation 23 constituted the whole study, and that no follow-up measures were anticipated. Post-test Procedures Selection of the Experimental TOpic: 0n the basis of the T1 responses, one of the Six public issues evaluated in the pretest questionnaire had to be selected as the experimental topic. Two criteria guided this selection. First, to allow room for evaluative change, avoid possible "ceiling" effects, and to simplify construction of the persuasive message, the issue had to be one on which there was homogeneity of evaluative response to the recommendation contained in the topic state- ment - either neutral or slightly Opposed. Second, to insure some degree of perceived difference between self-credibility and source-credibility for later analysis, the topic had to be one on which there was a wide range of self-evaluative ratings over each of the three dimensions Of self-credibility. Issue 2 -- "Abolishment of the Federal Office Of Economic Opportunity ('war on poverty') administration." -- came closest to meeting these criteria and.was selected as the experimental topic. Subjects were generally Opposed or neutral to the recommendation to abolish O.E.O. The mean evaluative rating of the proposal was 16.78 on the summary evaluative scale ranging from 5 (unfavorable) through 35 (favorable). The distribution of self-evaluative ratings on this issue also came closest to the ideal. Self-evaluations on the Qualification and Dynamism dimensions were normally distributed around the neutral point (ihlfi.5 on Qualification, §317.9 on Dynamism). Distribution Of self-evaluative ratings on the 24 Safety dimension was skewed toward the high end Of the scale (35 20.91) on this issue; however, since the mean self-safety ratings on the other issues were even higher, it was decided to use this topic. The Post-test Questionnaire: Two weeks after administration of the prdest measures, a different experimenter entered the classroom and enlisted the subjects' cooperation in a different study. The experimenter explained to the students that this study was designed to measure their reactions to various Opinion statements and arguments on the subject Of the Federal "war on poverty" program. The post-test questionnaire consisted Of four sections. In the introductory pages subjects were given a limited amount Of neutral back- ground information On the issue. The introduction noted that there was some degree of public controversy over the operation of the Federal Office Of Economic Opportunity; and that a research team from a "midwestern university" had undertaken a review Of the administration of the anti- poverty program. Subjects were told that the major part of this work had consisted of random interviews with people who lived in communities where anti-poverty programs were in Operation and who held various Opinions on the topic. The subject was then told that he would be able to read the interviewer's report Of one of these interviews and that he was to respond to the arguments on a set of evaluative rating scales. The Source Description: After these preliminary instructions, subjects read a brief description Of the person being interviewed - one of the eight sources rated in the pretest questionnaire. Subject—source pairings were made 25 randomly. The source descriptions were identical with those presented in the pretest questionnaire, with the additional information that this person lived in a community where anti-poverty programs were in Operation. The description also noted that the person being interviewed was in favor of abolishing O.E.O. The Persuasive Message: The subject then read what was purported to be part of the inter- viewer's report Of the source's views on the experimental issue. This "report" constituted the persuasive message used in the study, and was the same for all subjects and source attributions. The message presented a series Of arguments for abolishing the Federal Office Of Economic Opportunity; i.e., it argued in favor of the recommendation contained in the topic statement and against the receiver's own previously stated Opinion on the issue. The message was 533 words and contained 6 grammatical errors and u non-fluencies to provide a degree of plausibility. A specific attempt was made to avoid a too-eloquent presentation Of arguments. In general, the tone and language of the message were consistent with what it purported to be - an interviewer's report Of the Opinions and arguments of one of the eight sources described above. A copy of the message is in Appendix A. Post-test attitude measure: After reading the source's arguments favoring the abolishment of O.E.0., the subject was instructed to give his own Opinion on the issue, using the same five evaluatively loaded semantic differential scales incorporated in the pretest attitude measurement. The measure of post- 26 test attitude completed the questionnaire. Throughout the experiment explicit attempts were made to avoid giving the subjects any inkling that the pretest and post-test questionnaires were related; a different experimenter administered the questionnaires each time; the experimenter Specifically stated each time that his questionnaire was an independent study; and an attempt was made to vary the physical appearance of the two questionnaires. There also was an attempt to change the subjects' context for the study. In the first phase the study was supposedly sponsored by a national research organization. In the second, it was attributed to the Department of Communication at Michigan State University. These efforts, plus the separating Of administration Of the twO questionnaires by a two-week interval, were intended to lull any SUSpicions which the subjects might have had. The subjects were asked at the end Of each questionnaire to comment freely on the study and to make guesses about the purposes behind it. The comments of two subjects suggested that they suspected that the second questionnaire was in some way related to the first. These subjects were discarded. Analysis Procedures: Following administration of the post-test measures, pre- and post-test questionnaires for each subject were matched on the basis of name and student number. A total of 155 subjects returned usable questionnaires both times. From the questionnaires, the following measures were coded for each subject: 27 Pre-test attitude on the experimental topic. Each subject was assigned a pretest attitude score ranging from 5 (unfavorable) through 35 (favorable). This score was Obtained by summing his Tl reSponses on the five evaluative semantic differential scales described earlier. Post-test attitude on the experimental topic. This was Obtained by summing the subject's T2 re3ponses on the five evaluative scales. Pre-post attitude chapge on the topic. The subject's score on the post-test attitude measure was subtracted from his pretest attitude score to provide an index Of the amount of attitude change. Since subjects could change their attitudes in either direction, a constant Of 30 was added to the raw difference between pre- and post-test attitude scores. Thus, a maximum attitude change in the direction advocated by the persuasive message would be scored 60; a "boomerang" maximum negative change would be scored 0 on this index. Source Evaluation: The subject's ratings Of the message source on each of the three dimensions of source evaluation were summed to provide separate measures of perceived source Safety, Qualification, and Dynamism. Summary scores range from A (low) through 28 (high) on each dimension. Self—Evaluation: Summary scores of the subject's self-ratings on the three credibility dimensions were obtained as above. Self-Source Discrepancy: The direction and magnitude Of the perceived difference between the subject and the message source on Safety, Qualification and Dynamism were expressed as a single score, ranging from 6 (source low, self high) through 51+ (source high, self low) for each dimension. Hypothesis 3 calls for a comparison Of the relative adequacy of absolute source evaluative ratings y§_self-source discrepancy measures in predicting attitude change. A searCh Of the statistical literature failed to reveal an adequate test for the significance Of differences between multiple correlations computed.within the Same sample. Accordingly, following computation of the evaluative measures detailed above, subjects were randomly assigned to one Of twO analysis groups. Parallel analyses were performed for the two groups with the following differences: (1) Within Group 1 (n = 77) the dependent variable Of attitude change was related to 28 the measures of perceived self-source discrepancy on the three credibility dimensions. (2) Within Group 2 (n = 78) analysis involved the determination Of the relationship between attitude change and absolute measures of perceived Safety, Qualification, and Dynamism. A detailed description Of the analyses used to evaluate the seven theoretic hypotheses, together with the results of these analyses, is found in Chapter III. CHAPTER III RESULTS Effect Of the Persuasive Message: The design Of the study did not include a control group (which would have received no experimental stimulus) because the hypotheses did not concern themselves with whether there was or was not significant attitude change. Because Of this, no adequate test of the persuasive effect Of the message is possible. It is possible, however, to compare the pre-test and post-test scores for each analysis grOUp (see Table l), and to test for the significance Of difference between those two sets Of scores. Table 1. Mean Pre-Post Attitude Change Toward the Experimental Topic, by Groups Pretest Post-test Mean t p-value Mean Mean Shift Analysis Group 1 (N:77) 17.19 21.19 +9.00 3.96 (.01 Analysis Group 2 (N:78) 16.65 19.38 +2.73 2.8% (:01 Difference in mean shift between analysis groups = 1.27, N.S., P = .43 Both groups were significantly more favorable toward the proposal to abolish the federal Office Of Economic Opportunity after exposure to the message. The mean differences were rather sizable, and we can assume with reasonable confidence that the messages were persuasive. 29 30 Effectiveness of the Source Inductions: Separate analyses were conducted to determine the efficacy of the source descriptions in inducing perceptions Of high and low source Safety, Qualification, and Dynamism. °Comparisons among the eight source inductions on the Safety dimension are summarized in Table 2. The data indicate successful manipulation of source perceptions on this dimension. High-Safety sources were consistently rated significantly higher than low-Safety sources, and there were no significant differences within either the high or low groups. Table 3 presents the comparisons between the eight sources on the Qualification dimension. Again, the data indicate that the source descriptions were generally effective. Differences in the mean Qualification ratings for high and low source inductions were in the right direction in every case, and statistically Significant in all but two. Mean perceived Dynamism ratings for the eight sources are compared in Table A. The comparisons Show that the source descriptions were only partially successful in inducing appropriate source perceptions on this dimension. For twelve out Of the eixteen comparisons, the high Dynamism inductions did produce significantly higher mean ratings. In the remaining four cases, however, low inductions produced higher ratings, although the differences were not significant. The failure to achieve significant differences in these four instances reflects: (1) the general tendency for the subjects to rate all_the sources as relatively high in Dynamism, and (2) the Specific failure of the HiLOHi source induction. This source: 31 umounfl 13 mo. Va .coflpogwo “£an aw moocgmmmwo Hun o A3H.mfiwxv ii: on on on Aom.mammv ea. . --- H: on on AOH.:HHxV mm. 1 om. I Iii 04 H: 0A Amm.:ammv Nb.ar om.Hi or. I iii H: «m on Amm.mfimmv smb.mi «mo.mi «mm.:i «no.2: iii 04 OJ Hm Afim.mfimmv «um.wi «Hm.wr «H:.mi «no.1: mm. 1 run w: 04 H: Amb.ammmv «mm.wl em:.mi «No.5: ewm.or m>.mi am.mi iii 04 Hz Hm m>.mmmmv «mm.mn em:.mi «mm.mi eom.bi mm.mi mm.mi :o.ar run at an an use .aneo .mmm A3H.mfiwmv Aom.mammv AOH.:HmmV Amm.sfimmv Amm.mfimmv AHm.mfimmV Amb.fimmmv Aob.mmmmv coflpOSOCH oqoqoq Hmoqoq oqflzoq amazon oqoqwm Hmoqwz oqwzwm wzfizwm oopfimm son one gene :OMmSOEHQ hpommm so moonoom coozyom mooconommwm coo: oo>nowno .N canoe 32 umou.M zn mo._V.a .Sowuoopwo pzwws cw moocomommwo u e Amm.:finxv --- on on on Amm.sanxv om.m- --- em on on Aem.oamwv m:.m- as. + --- on on em Aeo.mflmmv mm.a+ am.:+ mm.m+ --- em on a: Amm.ommmv «mm.m- sm.m- me.m- «mm.s- --- on em on Amm.mmmmv «mm.m- «om.m- «mm.o- «Hm.OH- mm.m- --- we we on Ame.mmmmv «mm.m- «m:.m- «Hm.m- «ss.oa- e:.m- mm. . --- on a: we . Asm.mm we «ma.m- «mm.m- «os.m- «mm.m- mm.~- me. + Hm.a+ --- we we we and .190 ...mm Amm.efi mo Amm.SHmmv Asm.mfimmv Aso.mfi we Amm.om we Amw.mmmmv Ams.mmmmv Aem.mmmmv composeeH cocoon mmoooq oooowm ozone: omamoo Henson commas enema: monsom .oowwcoEHo cowpmowmwaoso so moonsom sou one nmwm soozuom moooopommwa coo: oo>nomno .m canoe 33 .m.z .SOHpooewo moon: SH moosopommwa u + pmopnw >n mo. a .oowpooewo woman SH moocoeommwo u e :55an ii: on on on 18.1.3.3 as. + in: on w: on Amm.ofimmv 8.: ms. + ii 3 3 2 Status rm. + SH. r we. r iii on H: at Amm.mmmwv «m:.mi «om.wn «mm.mi «mo.mi in: am on on 3981.5 «mm.>i «wm.mi «mo.mr «mm.mi wm.mr an: H: H: on Sesame +H©.H+ +om. + +He. + +bo.H+ oa.e+ w:.m+ in: H: on a: 2.0.2an «Ho.:i «ms.:i «Hm.mi emm.:r m:.a+ :m.m+ mo.mu iii Hm am am .cso .Hmso .mnm Amo.mammv Amm.sammv Amm.mfimmv Amm.bammv Amm.mmmmv Aam.mmmmv Am:.ofimmv aho.mmmmv cowuonooH oqoqoq common oqoqflz oqmmw: «moqoq .mmmzoq aroma: «mmmfim condom commooEHa Emwemmwm so moonsom zoq one swam coozuom mooOOSOMMflQ new: oo>pomno .: canoe 34 "A housewife and mother of four who has been asked to write a report on public affairs fOr her PTA group..." was evaluated as relatively how (.45 above the theoretic neutral point) in Dynamism. In fact, this source was surpassed in perceived Dynamism by all the low induction sources, including the Baptist minister who has been in retirement since 1938. The hypotheses for this study will not deal with the induced groups as such. Rather, subjectS' perceptions of credibility will be used, regardless of the "correct" perception that was attempted by the inductions. 35 A final comment on the overall effectiveness of the source inductions is in order. As mentioned in the preceding chapter, the source descriptions used in this study had been pretested earlier in a small ‘ group of students and wives in married housing. Safety, Qualification, and Dynamism ratings were Obtained on thirty possible sources of information and Opinion on public affairs topics. Six Of the thirty sources appeared to meet the pretest criteria; i.e., they were unanimously judged to be extremely high or low on the three evaluative dimensions. The remaining twO source descriptions were finally selected on the basis of pretest data from another study. Thus, deSpite the limited nature of the pretest, there was a reasonable expectation that the descriptions would produce appropriate source evaluations in the larger sample. This expectation was only partially realized. The source descriptions did induce credibility ratings which were generally consistent with their intent; i.e., "high" source inductions produced consistently higher ratings on the appropriate dimensions, and the mean differences were usually significant. Yet, the results fell short Of expectations. Ideally, the low source inductions should have produced mean ratings of around 8 or 9 on each dimension. None Of the source descriptions produced ratings this low. Mean Safety and Qualification ratings for the low induction sources approached the theoretic neutral point in most cases. On the Dynamism dimension, low source ratings actually exceeded the neutral point. 36 While the high source inductions produced correSpondingly higher ratings on each dimension, and these differences were statistically significant in an ouf of #8 comparisons, the differences between high and low sources were nowhere SO extreme as we would have wished. Some possible reasons for these equivocal results, and their implications for interpretation Of the later findings, are discussed in the following chapter. Test of the Theoretic Hypotheses: H1: In a situation involving persuasive communication, influence acceptance will be directly related to the perceived Safety, Qualification, and Dypamism Of the message source. The first hypothesis predicts that attitude change scores will be a direct function of the combined absolute source ratings on the three dimensions of source evaluation proposed by Berlo and Lemert. In testing this and the other theoretic hypotheses, actual raitngs by the subject were used as the basic data---whether or not those ratings "fit" the ratings eXpected by the inductions. The test of Hypothesis 1 required computation of the multiple regression Of attitude change on the three predictor variables: perceived Safety, Qualification, and Dynamism. If the hypothesis is to be confirmed, the multiple correlation Should be satistically significant. The results Of these computations for subjects in analysis Group 2 (N=78), are summarized in Table 5. 37 Table 5. Regression of Attitude Change Scores in Absolute Source Evaluation Scores. Predictor Variable Source Safety Source Qualification Source Dynamism TOTAL % Total Variance Accounted for by Variable .025 .027 .001 .053 Multiple Correlation Coefficient: R = .2269; P (R = .2269) = .27 The data clearly indicate that, at least in this situation, knowledge of source evaluation scores would not allow better than chance pre- dictions Of influence acceptance. The low value of the multiple correlation coefficient indicates that only five percent of the total variafility in attitude change scores can be explained by linear regression on combined source Safety, Dynamism, and Qualification scores. Thus, the hypothesis was not supported. H 2: In a Situation involving persuasive communication, acceptance of source influence will be directly related to the direction and magnitude of perceived self-source discrepancy on the Safety, Qualification, and Dynamism dimensions. 38 This hypothesis suggests that acceptance of persuasive communi- cation should be maximized when the subject perceives the message source as "more credible" than self, and impeded when the source is seen as "less credible" than self. Specifically, the hypothesis predicts attitude change as a joint function of the subject's self—evaluation and his evaluation of the message source over the three dimensions of credibility. Accordingly, the regression of attitude change on the measures of perceived self—source discrepancy was computed. Table 6 summarizes the results Of these analyses for Group 1 (N:77). Table 6. Regression of Attitude Change on Perceived Self-Source Discrepancy Scores Self-Source % Total Variance Discrepancy: Accounted fOr by this Variable Safety .003 Qualification .066 Dynamism .013 Total .082 Multiple Correlation Coefficient: R = .2754; P (R = .2754) = .12 The hypothesis is not supported by the data. The value of the multiple correlation coefficient falls considerably short of that required for significance at the .05 level. Only eight percent of the variafility in attitude change can be eXplained by linear regression on the combined self—source discrepancy measures. 39 H3: Knowledge of the direction and magnitude of perceived self-source discrepancy on the Safety, Qualification and Dynamism dimensions will afford more accurate predictions of influence acceptance than knowledge of source evaluation alone. The lack of support for H1 and H2 make inappropriate a statistical comparison of the accuracy of absolute source ratings vs self-source discrepancy measures as predictors of attitude change. Thus, H3 also failed to support. Given the failure of attitude change predictions from combined source Safety, Qualification, and Dynamism ratings, the zero order relationships between influence acceptance and each of the individual dimensions of source evaluation are of particular interest. These relationships are the special province of hypotheses 9 through 7. H4: There will be a positive relationship between influence acceptance and perceptions of relative source Safety. The zero order correlation between attitude change and perceived self-source discrepancy on the Safety dimension was computed for the 77 subjects in Analysis Group 1. The obtained value of r, .065, was not significant. Thus, the hypothesis was not supported. As matter of interest, the zero order correlation between attitude change and absolute source Safety ratings was computed for the 78 subjects in Analysis GrOUp 2. The r-value of .178 was not significant, although it approached significance at the .05 level. There will be a positive relationship between influence acceptance and perceptions of relative source Qualification. H53 ‘40 This hypothesis was supported. A significant zero order correlation between attitude change and perceived self~source discrepancy on the Qualification dimension was found for subjects in Group 1. (r = .266, p .05) The partial correlation between attitude change and selfesource discrepancy on Qualification, minus their common dependence on Safety and Dynamism ratings, was also significant at the .05 level (R = .202) The correlation between attitude change and absolute source Qualification ratings was computed on Group 2 subjects. The obtained correlation (r = .178) was not significant, although it approached significance at the .05 level. The partial correlation value, con- trolling On Safety and Dynamism, was R .112 - considerably Short Of significance. Hypotheses 6 and 7 Specify an interaction between perceptions Of relative source Safety and Dypamism, and predict the effect Of this interaction on the dependent variable of attitude change. H6: Among subjects who perceive the message source as equivalent or higher than self in SafeEy, there will be a positive relationship between influence acceptance and perceptions of relative source Dynamism. 7. Among subjects who perceive the message source as lower than self in SafeEy, there will be a negative relationship between influence acceptance and perceptions Of relative source Dynamism. The relevant data for these twO hypotheses from Analysis Group 1 are presented in Table 7. The results do not support the hypotheses. There was a positive correlation between attitude change and perceptions 41 Table 7. Correlations between Attitude Change Scores and Perceived Dynamism Of the Message Source (Analysis Group 1) Simple r N Source Higher than * Self in Safety.... .26 30 Source Lower than ** Self in Safety.... .14 1+7 * p (r = .301, df = 29) = .05 **p (r = .2u3, df = 95) = .05 Of relative source Dynamism among subjects who saw the source as relatively high in Safety, but the correlation did not differ from chance expectations. Among subjects who saw the source as relatively low in Safety, the correlation was smaller, but still positive. Since neither correlation differed significantly from chance eXpectations, evaluation of the differences befieen the Obtained r's is not possible. A parallel analysis was performed within Group 2, using absolute source ratings on the Safety and Dynamism dimensions. The results Of this analysis are presented in Table 3. Again, the data fail to support the predicted attitude change results Of interaction between perceived source Safety and Dynamism. 42 Table 3. Correlations between Attitude Change Scores and Perceived Absolute Dynamism of the Message Source (Analysis Group 2) Simple r N Source Higher than Self in Safety.... .11 * 35 Source Lower than Self in Safety.... .06 ** A3 *p(r = 296, df 30) = .05 **p (r = 257 df = 90) = .05 Additional Analyses: Given the failure Of six Of the seven theoretic hypotheses tested in this study, it is appropriate to inquire whether there was any overall effect in the attitude change variable which could be attributed to differences in evaluation of the message source. One approach to this question, and one which parallels the design Of most previous research in credibility, involves comparison Of attitude change scores fOr subjects who perceived the source as relatively high on all three dimensions of evaluation against change scores of subjects who rated the source as low on all dimensions. Accordingly, within Analysis Group 2, the median source rating was determined for each dimension of source evaluation. Subjects who rated the source at the median or below on all three dimensions were assigned to the "low source evaluation" group. Subjects who rated the source above the median on all three dimensions were assigned to the "high evaluation" group. The remaining subjects (i.e., those who rated the source as high on one dimension and low on another) were assigned #3 to a "mixed evaluation" group for this preliminary analysis. Mean attitude change Scores were computed for each group, and the differences evaluated by E:tests. Table ‘3 summarizes the results of these analyses. Table $3. Mean Attitude change for High, Low and Mixed Source Evaluation Groups. Source Evaluation Mean Attitude Change SNl_ High/High/High 36.28 18 "Mixed" 32.36 39 LOW/Low/Low 29 .95 19 The data indicate a significant difference in attitude change between the two "pure" source evaluation groups. Subjects who evaluated the message source as high on all three dimensions of evaluation showed significantly more attitude change than subjects who saw the source as low on all three dimensions. ( t = 2.225, p<:.05) These data suggested that differences in evaluation of the message source did have an Overall effect on attitude change, at least among subjects whose perceptions of the source were consistent across the three credibility conditions. Neither group differed significantly, however, from the "mixed" groups. Further analysis was in— dicated to explore the effects of mixed source evaluations. Table 10 shows the mean attitude change scores for subjects in each Of the eight absolute source evaluation categories. The means further support the conclusion that sources who are evaluated highly over all three dimensions are more persuasive than sources who are seen an as low on all characteristics. Mean scores within the "mixed" evaluation categories suggest the possibility of complex interactions between source- evaluative judgments over the three dimensions. To evaluate these Table 10. Mean Attitude Change Scores for Eight Source Evaluation Categories. High Safety Low Safety Hi Qual. Low Qual. Hi Qual. ’TLow Qual. Hi Dyn 36.235 (N=34) 29.000 (N-2) 32.000 (N=16) 34.087 (N=23) Lo Dyn 30.308 (N=13) 32.357 (N=28) 39.125 (N=8) 31.35 (N=31) interactions, an approximate analysis Of variance was computed on the data. The summary of this analysis is presented in Table 11. Table 11. Analysis of Variance Summary Table of Mean Attitude Change Scores for Subjects in the Eight Absolute Source Evaluation Conditions. (N=155) Source df_ HE. _E;_ _p_. Safety 1 9.378 1.126 :>.25 Qualification 1 14.780 1.775 <:.25 Dynamism l .162 - - Safety x Qualification 1 .032 - - Safety x Dynamism 1 6.295 - - Qualification x Dynamism l .293 - - Safety x Qualification x Dynamism 1 45.578 5.474 <:.025 Error 147 8.325 *the approximation employed is described in Walker and Lev, Statistical Inference, Holt, Rinehart and Winston, New York: 1953, p. 381. 45 The analysis produced a significant triple interaction between absolute source evaluations on the three dimensions. Finally, all subjects in the sample were grouped according to perceived self-source discrepancy scores over the three dimensions. Table 12 presents the mean attitude change scores for each of the eight categories of relative source Safety, Qualification 8 Dynamism. An approximate Analysis Of Variance was computed to determine possible main effects and interactions on the discrepancy measures. These data are Table 12. Mean Attitude Change Scores for the Eight Categories Of Perceived Self-Source Discrepancy (N=155) High Safety Low Safety Hi Qual Low Qual High Qual Low Qual Hi Dyn 36.286 (N=28) 30.000 (N=l) 33.395 (N=38) 34.944 (N=l8) Lo Dyn 36.000 (N=8) 30.500 (N=8) 35.071 (N=14) 30.100 (N=40) summarized in Table 13. The analysis produced no Significant F's, though the main effect Of perceived self-source discrepancy on the Qualification dimension approached significance. NO other interpretable effects were Obtained. 46 Table 13. Analysis of Variance Summary Table Of Attitude Change Scores for Subjects in the Eight Self-Source Discrepancy Conditions . _S___°urce 9f. 1‘13. .1. .2. Safety 1 .065 - - Qualification 1 2 8 .910 2 . 179 (.25 Dynamism 1 1.091 - - Safety X Qualification 1 8.746 - - Safety X Dynamism 1 1.430 - - Qualification X Dynamism 1 4.110 - - Safety X Qualification X Dynamism 1 6.671 - - Error 147 13.268 The sizable differences in the number Of subjects in each Of the cells in the above tables, however, suggested considerable cause for concern over the adequacy of the source inductions. Approximately equal numbers Of subjects had been placed by random assignment into each Of the eight source induction conditions. The notable discrepancy bemeen the number assigned to a particular source condition - for example, the High Safety - Low Qualification - Low Dynamism condition - and the number of subjects who actually perceived the source that way, was clear and sufficient evidence that the inductions had failed for a considerable portion Of the sample. This, deSpite the earlier evidence presented in Tables 2, 3 and 4 that overall perceptions of the sources were in line with the inductions. 47 The full implications of this failure of the source inductions to produce unifOrm effects on perceptions Of source Safety, Qualification, and Dynamism will be discussed in the fOllowing chapter. To check the immediate effects on attitude change, a final analysis was undertaken for subjects whose perceptions Of the message source were compatible with the source induction. The results of this analysis are presented in Tables 14 and 15 below. The pattern Of mean attitude change scores pre- sented in Table 15 is consistent with those Obtained in preceding analyses. Again, the only clear and interpretable difference which emerged was between change scores in the High-High-High and Low-Low-Low source conditions. Table 14. Mean Attitude Change Scores for Eight Source Induction Categories - Successful Induction Subjects Only (N=63) High Safety Low Safelzy High Qual. Low Qual. High Qual. Low Qual. Hi Dyn 38.416 (N=12) 31.000 (N=1) 31.727 (N=ll) 36.143 (N=7) Lo Dyn 34.285 (N=7) 33.636 (N=11) 36.857 (N=7) 28.000 (N=7) An approximate Analysis of Variance was computed over the attitude change scores, and is summarized in Table 15. Again, the analysis revealed no significant effects, although the three-way interaction approached significance. Table 15. Analysis of Variance Summary Table of Mean Attitude Change Scores for Subjects in the Eight Source Induction Categories - Successful Induction Subjects Only (N=63) Source Safety Qualification Dynamism Safety X Qualification Safety X Dynamism Qualification X Dynamism Safety X Qualification X Dynamism Error df 1 55 FE 2.656 19.550 2.540 1.640 .289 5.291 50.201 15.009 3.345 (.10 The implications of these results for future research, as well as some possible explanations for the failure of the predictions are outlined in Chapter 4. CHAPTER IV SUMMARY AND DISCUSSION Summary of Results: This study sought to test two extensions from previous theory and research in source credibility. Specifically, the rationale of the study proposed that current definitions of credibility should be extended to take into account (1) the multi- dimensionality Of source evaluations in persuasive communication situations, and (2) the source-receiver relationship which is implied by credibility judgments. A multi-dimensional model of source evaluation was developed from the results of recent factor-analytic studies. The model stipulated receiver evaluations of the message source on three independent dimensions - Safety, Qualification, and Dynamism - as principal determinants Of attitude change through persuasive communication. The study further proposed that predictions of influence acceptance from measures of source evaluation should also take account of the receiver's self-evaluations in the message situation. The rationale argued that influence acceptance should be facilitated when the message source is evaluated more highly than self, and impeded when the source is evaluated negatively in relation to self. A conceptualization of source credibility which defined the variable in terms of source evaluation relative to receiver self-evaluation was suggested as a useful extension of previous approaches. 49 50 Seven hypotheses governing the Operation of source credibility in persuasive communication were proposed to test the implications of these two arguments. The results, while in accord with the general findings of previous research in the credibility area, did not support the major theoretic hypotheses. One direct and statistically significant relationship was found between attitude change and the perceived Safety, Qualification, and Dynamism Of the message source. However, this relationship appeared to be significant only for those subjects who perceived the message source as unifOrmly high or low on all three of the evaluative dimensions. Among subjects who saw the source as high on one dimension and relatively low on another, no clear results were Obtained. The attempt to predict in- :fluence acceptance from multiple regression on the three dimensions of absolute Source evaluation (H1) failed to produce significant results. The hypotheses relating influence acceptance to perceived self- source discrepancy on the three evaluative dimensions together (Hzand H3) were not supported. Knowledge of perceived self-source discrepancy over the three dimensions failed tO provide better than chance prediction Of influence acceptance. Given these findings, no meaningful test was possible for differences between the two conceptions of credibility. The hypothesized relationships between influence acceptance and perceived self-source discrepancy on each of the individual judgmental dimensions were partially supported. The predicted positive relationship between influence acceptance and perceptions of relative source Safety (Hg) was not found. However, the hypothesized positive relationship between influence acceptance and perceptions Of relative Source Qualification (H5) 51 was supported. This variable was the best single predictor Of influence acceptance. The hypothesis relating influence acceptance and perceptions of relative source Dynamism among subjects who saw the source as relatively Safe (H6) was not supported; nor was the reverse prediction (H7) that relative source Dynamism would be negatively related to influence acceptance among subjects who saw the source as relatively Dangerous. Additional analyses Of the data confirmed the conclusion Of a significant difference in influence acceptance between two grOUpS Of subjects. Specifically, those who evaluated the message source as high on all three dimensions Of evaluation Showed significantly greater attitude change than subjects who saw the source as low on all three dimensions. This relationship emerged in three separate analyses. However, this clear overall effect Of differences in source evaluation was only Obtained among subjects whose perceptions of the message source were consistent across all three credibility dimensions. InSpection Of mean attitude change scores fOr subjects in the six conditions representing mixed source evaluation failed to yield any interpretable results. The consistent predictive failure of the theoretic hypotheses requires explanation. Examination of the design of the study and the underlying theoretic rationale suggests three possible explanations. These explanations, and their implications for future research, are dis- cussed below. Discussion. One possible explanation for the failure Of the hypotheses involves the stipulation Of non-linear relationships between attitude change and the predictor variables. For example, one might infer 52 from certain findings in the literature on group leadership that a curvilinear relationship Should exist between perceived self-source discrepancy on the Qualification dimension and acceptance Of source influence. The principal method of analysis employed in this study (linear correlation and regression) only provided a measure Of the significance Of the linear relationship between these variables. To the extent that attitude change is curvilinearly related to source evaluation or perceived self-source discrepancy on any of the three dimensions, we would expect the corresponding r-values to be reduced. A measure Of the extent to which the relationships between variables depart from linearity is provided by computation of Eta for each of the zero-order correlations between attitude change and the six predictor variables used in this study. The Eta values for Attitude Change X Absolute Source Safety, Qualification, and Dynamism were .255, .248, and .177, resPectively. For Attitude Change X Self-Source Discrepancy on Safety, Qualification, and Dynamism, the values were .278, .238, and .200. While these values slightly exceed the correSponding r-values in all cases, they are not significantly different from zero or from the simple correlation value. Thus, the data do not support an explanation of the results of this study in terms Of non-linear relationships between the variables. A second possible explanation of the data is derived from the "Personality and Persuasibility" area Of theory and research. Findings from this area have consistently demonstrated an inverse relationship between the strength of an individual's generalized self-concept or self-esteem and his susceptibility to persuasive influence. Thus, it 53 might be argued that the attitude change results Obtained in this study merely reflect differences among the individual subject's self- perception. Subjects who evaluated themselves as low in Safety, Qualification and Dynamism should be eXpected to accept the recommendations of the persuasive message more than subjects who rated themselves more highly, regardless of the source of the message. Support for this explanation of the data would require the demon- stration of negative correlations between attitude change and the measures Of self-evaluation. The zero-order correlations between attitude change and self-evaluation on the three credibility dimensions were: for Attitude Change X Self-Safety, -.015; for Attitude Change X Self- Qualification, -.051; and for Attitude Change X Self-Dynamism, -.057. These low correlations do not support an eXplanation of the data through consideration of subject self-evaluation scores alone. The most likely explanation for the lack of support for the theoretic hypotheses lies in the failure Of the source inductions. While appropriate and statistically significant differences were found in the overall source ratings on the three dimensions, the inductions failed in a large number Of individual cases. Subjects' perceptions of the message source deviated from the inductions on at least one dimension in over half the cases. Most of the failure of the inductions can be traced to the Dynamism dimension. There were more discrepancies on this dimension (61 "errors") than there were on Safety (21) or Qualification (18) combined. The failure of the source descriptions to induce appropriate perceptions of the source introduced a contaminating factor which was not allowed for by the original design Of the experiment. The analysis design 54 rested on the assumption that the "message" which the subject received should be constant for each level of combined source Safety, Qualification, and Dynamism. This total message consisted of two parts - the persuasive arguments contained in the text Of the message preper (held constant across all conditions), and the infOrmation about the supposed author of those remarks (held constant within each source induction condition). For proper control, it was necessary that both parts of the total message he unifOrm within each Of the eight categories of source evaluation. TO the extent that individual subjects at each level of perceived source Safety, Qualification, and Dynamism based their evaluations on different source descriptions, then, essential control Of part of the message was lost within that source category. This contamination of the eight source categories might plausibly account for the Observed significant and un- interpretable triple interaction between absolute source ratings over the three dimensions. AS Lindquist notes, such interactions are frequently attributable to extrinsic Type G error. It seems unlikely, however, that the above considerations account for all the results obtained in this study. Even when subjects' per- ceptions of the message source were wholly consistent with the inductions, the analyses failed to produce the predicted findings. A second kind of explanation involving the failure Of the inductions relates to the lack of extreme differences between high and low source ratings on the three dimensions. These differences, while statistically significant in nearly every case, were considerably smaller than anticipated or desired. The lack Of Spread between high and low sources 55 could have affected the results in twO ways. First, the differences in evaluative ratings may not have been large enough to produce differential attitude change eEects. Subjects may have interpreted these differences as too small to be taken into account in their decision to accept or reject the proferred source influence. Related to this notion is the possibility that whatever effects were produced by the source inductions were subsequently wiped out by exposure to a lengthy and persuasive message which was constant across all induction groups. Some inferential support for this explanation of the data is Offered by the clear and consistent finding that sources who were evaluated high on all three credibility dimensions were more effective than sources who were perceived as low on all the dimensions. It may be that substantial and meaningful perceived differences between sources were only produced when the subjects received descriptive information about the source which was internally consistent. When the source induction contained information which was contradictory across the three dimensions, the subject may have rejected the description as a basis fOr decision-making and formed his topic Opinion entirely on the message content. Thus, the constant message may have overwhelmed the perceived differences between sources except where the sources were clearly and uniformly separated on all three dimensions. One other explanation, statistical in nature, is possible. It involves the failure of the experimental inductions to produce large differences between high and low sources. To the extent that most of the sources occupied positions Of near neutrality or higher on all the evaluative dimensions, the range of possible scores on the predictor variables was restricted. 56 This restriction of the range inevitably produced smaller correlations between attitude change and source evaluations. These effects are particularly apparent in the data relating to Hypothesis 5, which predicted attitude change from self-source discrepancy scores on the Safety dimension. Here, the restriction of the range Of source evaluation scores was coupled with a skewed distribution of self- evaluation scores on safety. The range of combined self-source dis- crepancy scores on this dimension was thus greatly curtailed. Con- ceivably, this could account for the low zero-order correlation between attitude change and self-source discrepancy on the Safety dimension, and the negligible contribution Of this variable to explained variance in the multiple correlation. Implications for Future Research: Although the results of this study do not permit any definitive conclusions as to the Operation of source credibility, they do suggest some areas for future exploration. First, the failure Of the experimental source inductions points up the need for research to determine the information bases which underlie receiver evaluations Of the source. Specifically, what kinds Of in- formation about a potential influence agent lead to the attribution Of high or low Safety, Qualification, and Dynamism? To what extent do receivers vary in their interpretation of different kinds Of source in- formation in relation to these Specific evaluative dimensions? What kinds of personal and situational factors in the persuasive situation underlie individual differences in interpretation of the same information about a message source? Previous research in the credibility area has not touched on these questions. 57 Within the context of the general failure Of the theoretic hypotheses, the predicted positive relationship between attitude change and perceived self-source discrepancy on the Qualification dimension was supported. This finding suggests the utility Of additional research to further explore the use Of self-source interval measures as predictors of influence acceptance in persuasive communication situations. Research is also indicated to explore the relative weightings of judgments on the separate dimensions of source evaluation, and to determine what kinds Of factors in the message situation influence these weightings. Previous research by Berlo, Lemert 8 Mertz (1965) suggests that judgments cm the Safety dimension Should be heavily weighted in proportion to judg- ments on the other dimensions. In this study, however, judgments on the Qualification dimension appeared to be more heavily weighted. Part of the explanation for the very small apparent effect Of safety judgments may lie in the type of topic which was used in the present study. The experimental message dealt with a topic area which might be presumed to be relatively non-involving and technical for most members Of the audience. Both these topic characteristics would be expected to lead to greater emphasis on judgments Of source Qualification. Other research using more personally involving and less technical topics is needed to explore the effects of this variable on the weighting assigned to each Of the credibility dimensions. The effects Of differing receiver characteristics on the relative weighting of the evaluative dimensions should also be investigated. One might speculate that such factors as age and education should effect the relative weighting which a subject should assign to judgments Of source 58 Safety or Qualification. Older and less educated receivers might be expected to place high reliance on Safety considerations, while younger and more highly educated subjects might be expected to weight the source's Qualification more heavily, as happened in this study. Personality factors may also be involved. Mention has already been made of the possible influence of self-esteem on overall susceptibility to source influence. More particular predictions might be made from Rokeach's theory Of beliefstructure. His theory would seem to predict that high dogmatic subjects would be inclined to lay heavy emphasis on determining the Safety of the potential message source. 59 B IB LI OGRAPHY Adams, J. B. Effects of reference group and status on Opinion change. 0/ Joum 0 Quart o , 1960 , 37 ’ no 8"”12 O Adams, J. B. The relative credibility of 20 unnamed news sources. Jrn. Quarto’ 1962’ 39, 79-820 Andersen, K. E. "An Experimental Study of the Interaction of Artistic and Non-artistic Ethos in Persuasion," unpub. diss., Wisconsin, 1961. Andersen, Kenneth E. and Clevenger, Theodore, Jr. "A Summary of experimental research in ethos" §M_Vol. XXX (June, 1963) pp. 59-78. Arnett, C. E., Helen H. Davidson and H. N. Lewis, "Prestige as a Factor in Attitude Changes," Sociology and Social Research, Vol. 16, pp. 49-66, (1931). Asch, S. E., "The doctrine Of suggestion, prestige and imitation in social psychology," Psychol. Review, 1948, 55, 250-76. Barnes, Ted J., A Re-examination Of the Relationship Between Speaking Effectiveness and Personality Traits. Paper presented at the Central States Speech Association Conference, April, 1960. Berlo, David K. and Hideya Kumata, "The Investigator: The Impact Of a Satirical Radio Drama," Journalism Quarterly, Vol. 33, pp. 287-298. \ (1956). ‘— ‘ - fibfi ‘tBerlo, D. K., Lemert, J. B. 8 Mertz, R. J. "Dimensions for evaluating the ‘W acceptability Of message Sources", submitted to P00. Bernberg, Raymond E., "Prestige Suggestion in Art as Communication," Jrn. Of Soc. Psychology, Vol. 38, pp. 23-29, (1953). ’Ckue, David. "'Rational Argument' and 'Prestige-Suggestion' as Factors Influencing Judgment," Socio., 17 (1954), 350-54. Combs, A. W. (Chairman) Perceiving, behaving, becoming. Yearbook Association for Supervision and Curriculum Development, 1962, National Education Association, Washington , D. C . Croner, M. D. 6 Willis, R. H. "Perceived Differences of Task Competence and Asymmetry of Dyadic Influence." J. Abnorm. Soc. Psychol., 62 (1961), Das, J. P., Rath, R., and Das, R. S. "Understanding vs. Suggestion in the Judgment Of Literary Passages," J. Abnorm. Soc. Psychol., 51 (1955), 624-28. I 60 Ducker, K. "Experimental Modification Of Children's Food Preferences Through Social Suggestion." J. Abnorm. Soc. Psychol., 33 (1938) 489-507. Farnsworth, Paul R. and Issei Misumi, "Further Data on Suggestion in Pictures," American Journal Of Psychology, Vol. 43, p. 632, (1931). Gulley, H. E., 8 Berlo, D. K. Effect Of intercellular and intracellular speech structure on attitude change and learning. Speech Monog§., 1956, 23, 288-97. Haiman, F. S. "An Experimental Study of the Effects of Ethos in Public Speaking," _peech Monog., 2 (1949), 190-202. Hastorf, A. H., Richardson, S. A., Dornbusch, S. M. "The Problem of relevance in the study of personal perception." in R. Tagiuri and L. Petrullo (eds.) Person Perception and Interpersonal Behavior Stanford: Stanford University Press, 1958. Highlander, John. "Audience analyzer measurements and informational effects Of Speaker variables in radio talks." unpubl. diss. Wisconsin, 1953. Hildreth, R. A. "An Experimental Study of Audiences' Ability to Distinguish Between Sincere and Insincere Speeches," unpubl. diss., U. S. C., 1953. Hollander, E. P. Competence and conformity in the acceptance of influence. J. Abnorm. Soc. Psychol., 1960, 61, 365-69. Hollander, E. P. "Some Effects Of Perceived Status on Responses to Innovative Behavior," J. Abnorm. Soc. Psyghol., 63 (1961), 247-50. ' HOvland, Carl 1., Janis, Irving L. and Kelley, Harold H. Communication and Persuasion, New Haven, Yale University Press, 1953. Hovland, C. I., and Mandell, W., "An Experimental Comparison Of Conclusion- Drawing by the Communicator and by the Audience," Journal Of Abnorm. and Soc. Psychology, Vol. 47, pp. 581-588, (1952). V'Hovland, Carl 1., and Walter Weiss, "The Influence Of Source Credibility on Communication Effectiveness," Public Opinion Quarterly, Vol. 15, pp. 635-650, (1951-1952). Jones, E. E. and Thibaut, J. W. "Interaction Goals as Bases of Inference in Interpersonal Perception" in H. Tagiuri 8 L. Petrullo (eds.) Person Perception and Interpersonal Behavior, Standord: Stanford University Press, 1958. Kelman, H. C. "Reinstatement Of the Communicator in Delayed Measurement Of Opinion Change." J. Abnorm. Soc. Psychol., 48 (1953), 327-335. 61 Kerrick, Jean S. The effect Of relevant and non-relevant sources onl attitude change. J. soc. Psychol., 1958, 47, 15-20. Kiesler, C. A. 6 Kiesler, S. B. Role of forewarning in persuasive communications. J. Abnorm. Soc. Psychol., 68 (1964), 547-49. Kukp, D. H. Prestige, as measured by single-experience changes and their permanency. J. Educ. Res., (1934), 27, 663-72. Lewis, Helen Block. Studies in the principles Of judgments and attitudes: IV. The Operation of prestige suggestion. J. Soc. Psychol., 14, (1991), 229-256. Lorge, Irving, and Curtis, Carl. Prestige, as measured by single-experience changes and their permanency. J. Educ. Res. 27 (1934), 663-72. Lurie, W. A. The measurement of prestige and prestige-suggestibility. J. Soc. Psychol., 9 (1938), 219-25. Mausner, Bernard. Studies in social interaction: III. Effect of variation in one partner's prestige on the interaction of observer pairs. McCroskey, James C. "Scales for the measurement of ethos" SM, Vol. XXXIII, (March 1966), pp. 65-72. Miller, G. R., and Hewgill, M. A. The effect Of variations in nonfluency on audience ratings Of source credibility. QJS, 50 (1964), 36-44. OSgOOd, Charles E., Suci, George J., and Tannenbaum, Percy H. The Measurement of Meaning, Urbana, University of Illinois Press, 1957. Paulson, Stanley. Experimental study of Spoken communications; the effects Of prestige Of the speaker and acknowledgment of Opposing arguments on audience retention and shift of Opinion. Unpubl. diss., Univ. of Minn., 1952. Also briefly reported in SM, 11 (1954), 267-71. Rarick, G. Effects of two components Of communicator prestige. Paper de- livered to the fourth annual conference, Pacific Chapter of AAPOR, reported briefly in POQ, Spring, 1964, 149. Saadi, Mitchell and Farnsworth, P. The degrees Of acceptance Of dogmatic statements and preferences for their supposted makers. J. Abnorm. Soc. Psych01., 24 (1934), 143-50. Schweitzer, Don A. "Factors Of communicator credibility" in Secord 8 Backmanux“ (eds.) Problems in Social Psycholggy (in press). Sherif, M. An Experimental study Of stereotypes. J. Abnorm. Soc. Psychol.,cx’ 29 (1935), 371-75. 62 Walster, E. and Festinger, L. The dfectiveness of overheard persuasive communications. J. Abnorm. Soc. Psychol., 65 (1962), 395-402. Walter, Otis M. The measurement of ethos. Unpubl. diss., Northwestern Univ., 1948. Wolfinger, R. E. Attitude change toward source and issue resulting from one- sided and two-sided communication. Unpubl. M. A. thesis, Univ. Of 111., 1955, reported in Osgood, Suci and Tannenbaum, The Measurement Of Meaning, Univ. Of I11. Press, Urbana, 1957. Wylie, Ruth. The Self Concept, Lincoln: University of Nebraska Press, 1961. APPENDIX A: TEXT OF THE PERSUASIVE MESSAGE 64 "Nobody is against an effective anti-poverty program. But the federal Office of Economic Opportunity isn't capable of administering the program...too much scandal... Since it began Operation a year or so ago, the agency has piled up an amazing record Of waste. Inefficiency all over. ...so far, the people who seem to have benefited most from the millions Of our tax dollars that have been poured into anti-poverty funds have been the agency's staff members. They make fantastic salaries. I think most people are in favor of some kind Of aid program for the underprivileged, but the men in charge have bungled the jOb from the beginning. "Experts who have studied the problem admit that the organization and direction of our war on poverty program can be criticized on four important counts: "First, the federal agency is too isolated from local problems. A lot of the failures in our anti-poverty campaign happen because O.E.O. insists on handling everything from Washington. Most times they ignore the advice of their own men in the field who have the necessary knowledge Of local conditions. Another reason is that O.E.O. has failed to coordinate its own programs with other government and state and local agencies who are engaged in anti- poverty Operations. There has been wasteful duplication Of effort in many areas. In other cases, O.E.O.'s programs have undermined the effectiveness Of other agencies' campaigns. Again, most experts believe that these problems are caused by the federal agency insisting on trying to run its whole program from the Washington office. "The federal Office Of Economic Opportunity is staffed by a large number of inexperienced people. These people believe in eXperimenting for experi- mentation's sake. They even refuse to admit that some of these experimental programs fail, or to modify them, even when they have overwhelming evidence that they are not working. "Finally, worst of all, O.E.O.'s fumbling administration Of our national anti-poverty program is destroying the pe0ple's respect for all anti-poverty programs. As a result, many effective state and local prografig'are in danger Of losing public support. TOO many times, in tOO many communities, the failure of one of the federal agency's pet projects has left the community leaders and social workers holding the bag. "The plain truth is that we are in danger of losing our war against poverty before it even gets started unless some drastic action is taken immediately. The federal Office Of Economic Opportunity has been given more than enough time to straighten out its affairs. The result has been more waste...more bungling...and a black eye for anti-poverty programs all over. It is time to admit that the federal agency cannot do the job it was created fOr. Let's abolish the federal Office of Economic Opportunity, and give the direction Of our anti-poverty campaign to those who can best do the jOb. State and local agencies have proven that they are most capable of administering good, effective programs in their own areas. Abolishment of the federal Office of Economic Opportunity would result in releasing fUnds to properly support these many effective local agencies. Abolishing the federal Office Of Economic Opportunity Ought to be the first step in straightening out our anti-poverty program." APPENDIX B: THE PRE-TEST QUESTIONNAIRE 66 Confidential - For Research Purposes Only DO Not Write Here C 1-3 Project NR C 4 Card NR C 5 Phase NR C 6 Sub Deck NR C 7-9 Respondent NR PUBLIC OPINION STUDY Please fill in the fOllowing: Name: Year in School: Sex: Residence Hall: 67 SECTION ONE This research study is one of a series concerned with the way college students react to controversial public issues. Please note that the information you provide will be held confidential and will be used only for the purpose of this research. In this study we are interested in three kinds of information. We'll deal with them one at a time and give you the necessary instructions as we go along. First, we'd like your Opinion on several public issues which are currently in the news. On the following pages you will find a number of propositions or statements of opinion dealing with these issues. You will be asked to rate each statement on a number of 7—point scales like the following Good : : : : : : Bad VERY QUITE SLIGHTLY DON'T SLIGHTLY QUITE VERY KNOW, NEUTRAL Here is how the scales work. Suppose you were asked to rate the statement: "The voting age in Michigan should be lowered to 18 years." on the above scale. If, in your Opinion, lowering the voting age is a ve good idea, you should put a check in the extreme left-hand p031tion on the scale (next to the adjective "GOOd"). If, in your Opinion, lowering the voting age would be a very bad idea, you should mark the extreme right-hand position of the scale (next to the adjective "Bad"). If you could not decide whether lowering the voting age would be good or bad, or if you held a neutral position with regard to the statement, then you should mark the center position on the scale. Use the same procedure in marking the other scales which follow the statement. Be sure to mark each scale only once; do not skip any scale. If you have any questions about how to mark the scales, please ask them now. Ok, please turn the page and begin. worthless 68 Extension of Medicare to US citizens of all ages through a national health insurance plan financed by tax fUnds... valuable good bad wise foolish honest dishonest fair unfair Abolishment of the Federal Office Of Economic Opportunity ("war on poverty") administration... worthless : valuable good : - bad wise foolish honest ° : dishonest fair unfair 69 Revision of present Selective Service regulations to eliminate student deferments except in a few cases in the national interest.... worthless valuable good : bad wise : foolish honest : dishonest fair : z : unfair Restricted use of nuclear weapons in Viet Nam in non-populated areas... worthless : : : valuable good : ° : : bad wise foolish honest : : : dishonest fair : unfair worthless 7O Strick University control and supervision of student political organizations, eSpecially those which Show Signs of communist domination... : valuable good : : bad wise foolish honest ° : : dishonest fair : unfair A mandatory jail sentence and one-year suSpension of driver's license for anyone convicted of drunken or reckless driving... worthless : valuable 800d bad wise foolish honest dishonest fair : unfair 71 NEW SECTION -- PLEASE READ CAREFULLY From time to time, all of you have probably been asked to state your Opinions or to provide information on various public issues. You may have been assigned to write papers or give in-class presentations on these topics; friends or relatives may have asked your Opinion; or you may have been interviewed in a public opinion survey like this one. In this section of the study, we'd like fOr you to think Of yourself as a possible source of infOrmation and opinion on the topics you have judged. The following Six pages contain a listing Of the topics, with 7-point self—rating scales for each topic. The self-rating scales fellow the same format as the scales you have previously marked; use the same procedure to mark these new scales. Remember, you are to rate yourself as a source of information on these issues... Any questions? 0k, please turn the page and begin 72 How would you rate yourself as a source of information on the issue: ”Extension of Medicare to US citizens of all ages through a national health insurance plan financed by tax funds..."? Just Unjust Unqualified : Qualified Skilled : Unskilled Dangerous : Safe Hesitant Emphatic Informed : Uninformed Energetic Tired Kind : Cruel IneXperienced : EXperienced Bold Timid Active Passive Honest Dishonest How would you rate yourself as a source of information on the issue: "Strick University control and supervision of student political organizations, especially those which show Signs of communist domination..."? Just Unqualified Skilled Dangerous Unjust Qualified Unskilled Safe Hesitant Emphatic Informed Uninformed Energetic Tired Kind Cruel Inexperienced Experienced Bold Timid Active Passive Honest Dishonest 74 How would you rate yourself as a source of information on the question: "Abolishment of the Federal Office of Economic Opportunity ("war on poverty") administration..."? Just Unjust Unqualified Qualified Skilled Unskilled Dangerous Safe Hasitant Emphatic Informed Uninformed Energetic Tired Kind Cruel Inexperienced Experienced Bold Timid Active Passive Honest Dishonest How would you rate yourself as a source of information on the question: "Restricted use of nuclear weapons in Viet Nam in non-populated areas..."? Just Unjust Unqualified Qualified Skilled Unskilled Dangerous Safe Hesitant Emphatic Informed Uninformed Energetic Tired Kind Cruel Inexperienced Experienced Bold Timid Active Passive Honest Dishonest 76 How would you rate yourself as a source Of information on the issue: "Revision of present Selective Service regulations to eliminate student deferments except cases in the national interest..."? Just Unjust Unqualified : : : Qualified Skilled : Unskilled Dangerous Safe Hesitant Emphatic Informed : : : : : Uninformed Energetic : : : Tired Kind : : Cruel IneXperienced Experienced Bold Timid Active : : : : : Passive Honest : : : A : : Dishonest How would you rate yourself as a source of information on the question: "A mandatory jail sentence and one-year suspention Of driver's license for anyone convicted of drunken or reckless driving..."? Just Unqualified Skilled Dangerous Hesitant Informed Energetic Kind Inexperienced Bold Active Unjust Qualified Unskilled Safe Emphatic Uninformed Tired Cruel Experienced Timid Passive Honest Dishonest 78 NEW SECTION -- PLEASE READ CAREFULLY The issues with which we have been concerned in the preceding pages have aroused a great deal of public interest. Government officials, business and professional leaders, and many ordinary citizens have expressed their Opinions on these issues. On the following pages you will find brief descriptions Of some of the people who have presented their views on these topics. Following each description is a set of rating scales like the ones you marked before. On the basis of the information given in the description, we would like you to rate each person as a source of information on these kinds of public affairs topics, using the scales provided. Note: Each of the descriptions contains a limited amount Of information. You may feel that you can't adequately judge the person on the basis Of the description provided. Please do the best you can. We are primarily interested in how you would "Size up" another person as a source Of information on public issues if you had only a little information about him. Remember, in this section you are to rate each person described as a source of information and opinion on public issues. Any questions? Ok - please turn the page and begin 79 A 76 year Old former Chief Justice of the Supreme Court of the State of Rhode Island. . . On the basis of this information, what do you think this person would be like as a source Of information on public issues? Just Unqualified Unjust Qualified Skilled Dangerous Hesitant Informed Energetic Kind Inexperienced Unskilled Safe Emphatic Uninformed Tired Cruel EXperienced Bold Timid Active Passive Honest Dishonest A middle-aged male English teacher in a suburban girls' school, on campus for summer courses... On the basis Of this information, what do you think this person would be like as a source of information on public issues? Just Unjust Unqualified Qualified Skilled Unskilled Dangerous Safe Hesitant Emphatic Informed Uninformed Energetic Tired Kind Cruel Inexperienced Experienced Bold Timid Active Passive ' Honest Dishonest A ranking government Official in the Truman administration who was involved in a scandal which resulted in his removal from Office. He now leads an inactive life near Washington, D.C. ... On the basis of this information, what do you think this person would be like as a source of infOrmation on public issues? Just Unjust Unqualified Skilled Qualified Unskilled Dangerous Hesitant Safe Emphatic Informed Energetic Kind Uninformed Tired Cruel Inexperienced Bold EXperienced Timid Active Honest Passive Dishonest 82 A professional lobbyist who has been described by members of Congress as "The most persistent and determined pleader of Special interests ever seen on Capitol hill"... On the basis of this information, what do you think this person would be like as a source Of information on public issues? Just : : : : Unjust Unqualified : : : : : : Qualified Skilled : : : : : Unskilled Dangerous : : : : : : Safe Hesitant : : : : : Emphatic Informed : : : : : : Uninformed Energetic : : : : : Tired Kind : : ' : ° Cruel Inexperienced : : : : : : Experienced Bold : : : . : Timid Active : ° : : : Passive Honest : : : : : Dishonest A once-controversial Baptist minister, now living in a home for aged clergy. He was censured by his church in 1938 for anti-semitic and racist preaching, and retired from active service shortly afterward... On the basis Of this information, what do you think this person would be like as a source Of information on public issues? Just Unjust Unqualified Qualified Skilled Unskilled Dangerous Safe Hesitant Emphatic Informed UninfOrmed Energetic Kind Tired Cruel Inexperienced Bold Active Experienced Timid Passive Honest Dishonest 84 The former editor Of a campus magazine of extreme left-wing political Opinion, currently on academic probation for low grades... On the basis of this information, what do you think this person would be like as a source of information on public issues? Just : : : : . Unjust Unqualified : : : : : (Qualified Skilled : : : : : Unskilled Dangerous : : : : . : Safe Hesitant : : : : : Emphatic Informed : : : : Uninformed Energetic. : : : : : Tired Kind : : : Cruel Inexperienced : : : : : Experienced Bold : : : : : Timid Active : : : : : : Passive Honest : : : : : : Dishonest 85 A leading national clergyman who has been appointed by the President to a special advisory commission on social and economic affairs... On the basis Of this information, what do you think this person would be like as a source of information on public issues? Just Unjust Unqualified : Qualified Skilled : Unskilled Dangerous : Safe Hesitant Emphatic Informed : : Uninformed Energetic : Tired Kind Cruel Inexperienced Experienced Bold : Timid Active : : Passive Honest : : : Dishonest A housewife and mother Of four who has been asked to write a report on public affairs for her PTA group... On the basis of this information, what do like as a source of information on public Just you think this person would be issues? Unjust Unqualified Qualified Skilled Unskilled Dangerous Safe Hesitant Emphatic Informed Uninformed Energetic Tired Kind Cruel Inexperienced Experienced Bold Timid Active Passive Honest Dishonest _ . . . . a... . _ n I L 1 . J I ~ .7 c. ' - . ' - a ,. . ., ' . 7 IO :,, ‘ ‘ o o . , . I _ A t I 1 .) L .1 . - w . . . v y . . . - . .1 9 .‘ . - . e. .. -‘ :. ' g . 1' ' .u. .\, (1'. _. .. I-w . u . - . ' . ' .- I . c . A ‘ ' - A . - _ . .‘ .. ' L‘ ' o u I I I ‘ o o . . x . - ..__.~. .12 ...-.... . . - -... - - .— _. . - - ...—.— - .. - - - . . . , , ._ . _ 1‘ . I n v . o - ' x -‘ a .. ,. .. 1 ... - - .. ‘. ....-- ...--..... _ - _ . -..... --. . .. - I 9 D ‘e I ' . Q . o 4 n v. i I" -.- . ..- _ ...- .. - -.. . ..- -. ... .-.-.... . o 1 u 0 r - . , . I C Iv . I ..~ . _ , _ _ .—.. ”1..-... . - -_ - .. _ - ... - ‘ . . o . . a Y ! ‘ 1 . u | u ‘9 I ‘ -- ‘ -l _ _ ..- - .1. -— - —...-- ._ r .-A . _.........1 - , l p h. d-‘v.-" ‘ v o . o . r ' ~ - . - .v 1 .. - . . . . .-.. -4” F ...- . . .. ....tl , .. ...— . o I o o . ‘ - ~ ‘ , . A . . . - . _- ..-. .- . ...-. --- ....» -r. -.. -- ..-. - . ‘ -.--..“ . , \ . I u - . r . I A C I - O .' \ - A-.. .. .- --.a .--- . _.... ... -_ — —.. —. m.-- u - , g o o O i) - Jth‘ "{‘ ."'-’ o I - c I . Ax ‘.. - » . - - .... _x --.—1. ._.._ W-..- 1 - .... _- . ...-- . .. c - I I o f I ' U ' ' - . .a. -- . . . .._. ...“... ,...—... . —- .. -- _ .... .. -.1 - 1h- ... .n .1 . o n o . ,‘ {f ‘ ‘ c o - e , ... . . -.. ..- .-- -....- - . .-..1-_1__...-- ——.—— ._.-.. —..-.._ ~.... - o o o o ‘ -' : , . o o o I . , - . . . a . _- . ... ...-.....- - .-—.... v— *h-Qv—tuu -. . r - APPENDIX C: THE POST-TEST QUESTIONNAIRE 88 FOR RESEARCH PURPOSES ONLY COMMUNICATION SURVEY I-B Please fill in the following: Name: Year in School: Residence Sex 89 INSTRUCTIONS This study is one of several research projects now being conducted on the MSU campus. This research is aimed at learning how college students in general reSpond to new ideas and information on controversial public issues. For each topic we have collected a number of arguments which have been presented on these issues by various kinds Of people. You will be asked to evaluate some of these arguments and the people who presented them, and to give us your OWn Opinion on the public issue involved. The next two pages contain some background information on one of the public issues, a brief description Of the person whose arguments you are to evaluate, and a summary of his or her arguments on the question. This information is designed to help you reach a decision on the issue; please read it carefully. Further instructions will be given as needed. If you have any questions, please ask the research assistant who is administering this questionnaire... Please turn the page and begin... 90 Background Information The "War on Poverty" program has aroused considerable controversy in its first year Of Operation. Critics have pointed to examples Of waste and inefficiency and have labelled the program "...a vast boondoggle..." Defenders of the program have argued that some growing pains are inevit- able in any new undertaking, and that the program as a whole is functioning well. Much of the criticism Of the national program has centered around the Office Of Economic Opportunity (O.E.O.) -- the federal agency which is re- sponsible for overall administration of anti-poverty funds. Critics Of O.E.O. charge that the federal Office has proven unable to cope with the problems of directing the national prOgram. A short time ago, a research team from a midwestern university undertook an extensive review of the anti-poverty program and its administration, National and regional directors of the program were interviewed. In addi- tion, the researchers sent interviewers to 45 communities where federal anti-poverty programs had been in Operation for several months. Detailed interviews were obtained in these communities with civic leaders, business and professional people, students, and housewives on a random basis. The people interviewed were asked their Opinions on the prOgress of anti- poverty programs, what impact these programs had had in their communities, and what changes they thought should be made in the program. Among those who were randomly selected by the interviewers was Mr. J. P. Ritchie, a middle-aged English teacher in a suburban girls' school. At the time Of the interview, Mr. Ritchie was attending summer school classes at a midwestern college. The federal anti-poverty program has been in Operation in Mr. Ritchie's home community for six months. Ritchie stated that, in his Opinion, the Federal Office Of Economic Opportunity should be abolished. The fellowing is from the interviewer's report Of Ritchie's views: A. 91 "Nobody is against an effective anti-poverty prOgram. But the federal Office of Economic Opportunity isn't capable Of administering the program ...tOO much scandal... Since it began operation a year or so ago, the agency has piled up an amazing record of waste. Inefficiency all over. ...so far, the people who seem to have benefited most from the millions Of our tax dollars that have been pured into anti-poverty funds have been the agency's staff members. They make fantastic salaries. I think most peOple are in favor Of some kind of aid program for the underprivileged, but the men in charge have bungled the jOb from the beginning. "Experts who have studied the problem admit that the organization and direction of our war on poverty program can be criticized on four important counts: "First, the federal agency is too isolated from local problems. A lot of the failures in our anti-poverty campaign happen because O.E.O. insists on handling everything from Washington. Most times they ignore the advice of their own men in the field who have the necessary knowledge of local conditions. Another reason is that O.E.O. has failed to coordinate its own prOgramS with other government and state and local agencies who are engaged in anti-poverty Operations. There has been wasteful duplication Of effort in many areas. In other cases, O.E.O.'s programs have undermined the effectiveness Of other agencies' campaigns. Again, most experts believe that these problems are caused by the federal agency insisting on trying to run its whole program from the Washington Office. i "The federal Office Of Economic Opportunity is staffed by a large number of inexperienced people. These people believe in experimenting for experimentation's sake. They even rerSe to admit that some Of these ex- perimental programs fail, or to modify them, even when they have overwhelming evidence that they are not working. "Finally, worst Of all, O.E.O.'s fumbling administration of our national anti-poverty program is destroying the people's reSpect for all anti- poverty programs. As a result, many effective state and loc§I_programs are in danger of losing public support. TOO many times, in tOO many communities, the failure Of one of the federal agency's pet projects has left the community leaders and social workers holding the bag. "The plain truth is that we are in danger of losing our war against poverty before it even gets started unless some drastic action is taken immediately. The federal Office Of Economic Opportunity has been given more than enough time tO straighten out its affairs. The result has been more waste...more bungling...and a black eye for anti-poverty programs all over. It is time to admit that the fedeal agency cannot do the jOb it was created for. Let's abolish the federal Office of Economic Opportunity, and give the direction Of our anti-poverty campaign to those who can best do the jOb. State and local agencies have proven that they are most capable of adminis- tering good, effective programs in their Own areas. Abolishment Of the federal Office Of Economic Opportunity would result in releasing fUnds to properly support these many effective local agencies. Abolishing the federal Office Of Economic Opportunity ought to be the first step in straightening out out anti-poverty program." 92 Now that you have read this person's arguments on the issue, we'd like your own Opinion. In general, how do you rate the proposal to do away_ with the Federal Office of Economic Opportunity administration? Please use the rating scales below to indicate your Opinion. "Abolishment Of the Federal Office of Economic Opportunity ('war on poverty') administration." wise : : : : : : foolish unfair : : : : : : fair valuable : : : : : : worthless g good : : : : : : bad dishonest : : : . : : honest APPENDIX D: INTERCORRELATION MATRIX, ATTITUDE CHANGE, SELF-EVALUATION, SOURCE EVALUATION, AND SELF- SOURCE DISCREPANCY SCORES Variable 10 13 Variable I‘ 1.000 .365 .053 -.006 -0089 -0102 .824 .305 .109 .136 2 Source rating on Trustworthiness" 1.000 .577 -0002 -011“ .030 .300 .748 .429 .212 Content 1.000 .171 -0070 .045 -.054 .434 .747 .112 Source rating on 99 1.000 .384 1.000 .453 .593 1.000 -.572 -.290 -.341 -.258 -.745 -.376 -.169 -.449 -.630 -.015 -.051 -.058 5 5 7 "SafeEy/ dimension "Qualification/ Competence" dimension Source rating on "Dypamism" dimension Subject's self-credibility rating on "Safety/Trustworthiness" dimension on the experimental issue Subject's self-credibility rating on "Qualification/Competence dimension on the experimental issue 1.000 .396 1.000 .185 .588 1.000 .120 .177 .125 8 9 10 Range 04 - 28 04 (low), 28 (high) 09 - 28 09 (10W), 28 (high) 04 - 28 09 (16w), 28 (high) 04 - 28 O4 (ION), 28 (high) 09 - 28 09 (low), 28 (high) Variable 10 13 95 Content Subject's self-credibility rating on "Dypamism" dimension on the experimental issue Self-source discrepancy: " afeEy" dimension. Self-source discrepancy: "Qualification" dimension. Self-s ource dis crepancy: "Marxism" dimension. Pre-post attitude change on experimental issue Range 09 - 28 09 (low), 28 (high) 06 - 54 06 (source low, self hi) 54 (source hi, self low) 06 - 54 06 (source low, self hi) 54 (source hi, self low) 06 - 54 06 (source low, self hi) 54 (source hi, self low) 00 - 60 maximum "boomerang" 00 (negative change) 60 (maximum favor- able change) "I71111111111711lilies