SOME GGRRELATES 0F ATTITUDE CHANGE,“ RETENTION OF THE ATTITUDE MESSAGE , 'ANDEVALUATIGN OFT‘HECOMMUNICATORI-gi-rg . .. _ ., » Thesis forthe Degree oIRII; D. -: ’ ' MICHIGANSTATE UNIVERSITY ,, WALLACE GEORGE BERGER ‘ 19 6 9 {HF—SSS LIBRJ} :T.’ y IIIIIIIIIIIIIIIIIIIIIIIIIILIIIIIIIIIIII MW w . :; University This is to certify that the thesis entitled SOME COBRELATES OF ATTITUDE CHANGE, RETENTION OF THE ATTITUDE MESSAGE, AND EVALUATION OF THE COMMUNICATOR presented by Wallace George Berger I has been accepted towards fulfillment of the requirements for _PhD._ degree in£synhology $Cg2¢fi proffl Date Q 22) [RéO‘ 0-169 Iw MU ABSTRACT SOME CORRELATES OF ATTITUDE CHANGE, RETENTION OF THE ATTITUDE MESSAGE, AND EVALUATION OF THE COMMUNICATOR By Wallace George Berger The present study explored the relation between three variables commonly used as dependent measures in attitude 3 gal: 25. research. These dependent measures were: 1. attitude change 2. the retention of the attitude message and 3. the evaluation of the communicator. In order to specify more precisely the relation between these three variables, the following three independent variables were utilized: 1. §'s initial attitude 2. the amount of previous factual information g had about the attitude object and 3. the extent of previous contact g has had with the attitude objecto For the most part, previous research has concerned itself with the exploration of the relation between pairs of the above six variables. It appeared profitable at this time to conduct a multivariate study employing all of the above variables since multivariate studies present the re— searcher with a unique opportunity to investigate the inter- relations among variables. Wallace George Berger The attitude object selected for investigation was labor unions. The choice of labor unions stemmed from the following considerations: 1. the topic of labor unions is a contemporary issue 2. pilot work had indicated a range of attitudes toward labor unions in the population under study I.» pilot work had indicated the existence of indi- vidual differences in the amount of information gs had about labor unions 4. gs had an opportunity to expose themselves to labor unions and labor union related material and 5. it was possible to construct messagascomprised of factual material which was unfamiliar to §s. In a series of pilot studies, a set of instruments were developed to measure the variables of interest. In the main study each § participated in two one-hour sessions (1 week.apart). The information collected during the first session was used to individually match pairs of §s on the basis of sex, initial attitude, amount of previous informa- tion, and extent of contact with labor unions. During the second session, half of the matched gs were randomly assigned to receive a pro-labor union video-taped message, and the other half an anti—labor union video-taped message. The results of this study verified the following hypotheses: l. The more favorable cator, the greater 2. The more favorable cator, the greater message. the the, the the evaluation of the communi- attitude change. evaluation of the communi— retention of the attitude Wallace George Berger 3. The greater the amount of previous factual infor— mation, the more polarized the evaluation of the communicator. 4. Post-message attitude is a significantly better predictor of evaluation of the communicator than pre-message attitude. U1 Previous factual information, previous contact, pre-message attitude, post—message attitude and attitude change were not significantly related to the extent to which the attitude message was retained. Although relations between the following variables were predicted, the results indicated that: l. Pre-message attitude, previous factual informa- tion, and previous contact were not significantly related to attitude change. 2. Pre—message attitude.and previous contact were not significantly related to the evaluation of the communicator. The implications of these findings were discussed in light of current research and theory. Approved :I/X—kk W( 96L . g wigg/ adage}. 3h 1W1 V \ \ Thesis Committee: H. C. Smith, Chairman R. H. Davis J. E. Hunter J. S. Karslake SOME”CORRELATES OF ATTITUDE CHANGE, RETENTION OF THE ATTITUDE MESSAGE, AND EVALUATION OF THE COMMUNICATOR BY Wallace George Berger A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1969 09w 7/0 4,2300 To my wife and my parents Ii ACKNOWLEDGMENTS I wish to express my sincere appreciation to Dr. H. C. Smith, my committee chairman, whose suggestions, encouragement and cooperation contributed significantly to this dissertation. Although he spent many hours in consultation with me on various parts of the study, his belief in my capacity for creative self-expression allowed me the freedom to conduct the research as my own. As a result, the writing of this dissertation was a truly en- riching experience. I am also deeply indebted to Drs. R. H. Davis and J. E. Hunter for their aid in the initial phases of the research. Thanks are also due to Dr. J. S. Karslake, the author's academic adviser, for his encouragement and con- structive criticisms. Sincere appreciation is also expressed to Mr. K. Marlow, who helped in the development and pre—testing of the instruments used in this study. Finally, I would like to express appreciation to my wife, Maureen, for her support in those many instances when it seemed that the next hurdle was unattainable. TABLE OF CONTENTS Page DEDICATION . . . . . . . . . . . . . . . . . . . . . . . ii ACKNOWLEDGMENTS. . . . . . . . . . . . . . . . . . . . . iii LIST OF TABLES I o q o o o o I I o c I o c o o o a o I 0 Vi LIST OF APPENDICES O O O I C O C C U 0 O D O O O O O O 0 vii CHAPTER I INTRODUCTION. . . . . . . . . . . . . . . . . . 1 Brief Overview. . . . . . . . . . . . . . . . 1 Section 1: Attitude Change . . . . . . . . . 5 Section 2: Retention . . . . . . . . . . . . 10 Section 3: Evaluation of the Communicator. . 15 Summary of Hypotheses . . . . . . . . . . . . 18 II EXPERIMENTAL DESIGN . . . . . . . . . . . . . . 20 Methodological Difficulties . . . . . . . . . 20 Sequential Outline of the Experimental Design . . . . . . . . . . . . 23 III INSTRUMENT DEVELOPMENT. . . o . . . . . . . . . 25 Attitude Questionnaire. . . . . . . . . . . . 25 Video—Tape Attitude Message . . . . . . . . . 28 Information Test. . . . . . . . . . . . . . . 29 Evaluation of the Communicator Scale. . . . . 30 Contact Scale . . . . . . . . . . . . . . . . 32 Retention Questionnaire . . . . . . . . . . . 33 Evaluation of the Message Position on the Attitude Continuum . . . . . . . . . 34 IV PROCEDURE . . . . . . . . . . . . . . . . . . . 36 SUbjeCtS. a I c o o o o n o o c o o o n o o o 36 MethOd. o I o c o o o o o o o a a o a o o o o 36 iv TABLE OF CONTENTS (Continued) CHAPTER Page V RESULTS I o I I I I I I I I I I I I I o I I o I 39 Some Characteristics.of.the Instruments . . . 39 Attitude Change . . . . . . . . . . . . . . . 44 Retention . . . . . . . . . . . . . . . . . . 47 Evaluation of the Communicator. . . . . . . . 49 VI DISCUSSION. I I C I C O I O O O O D O I O O I I 53 Attitude Change . . . . . . . . . . . . . . . 53 Retention 0 O O O O O O O I I O O O O O 57 Evaluation of the Communicator. . . . . . . . 59 REFERENCES I I I I I I I I I I a I I I I I o I I I I O I 63 APPENDICES I o I I I I I I I I I I I I I I I I I I I I o 67 LIST OF TABLES Table Page 1 Correlations, reliabilities, means and standard deviations for Attitude Questionnaire II . . . . . . . . . . . . . . 28 2 Internal consistencies, means and standard deviations for Information Test I. . . . . . 30 3 Internal consistencies, means and standard deviations for Information Test II . . . . . 3l 4 The reliability, mean and standard deviation of the Contact Scale . . . . . . . . . . . . 33 5 Intercorrelations, reliabilities, means and standard deviations of the scales for the PRO and CON groups . . . . . . . . . . . . . 4O 6 Beta weights, F values and significance levels for multiple regression analysis on final attitude. . . . . . . . . . . . . . 46 7 Beta weights, F values and significance levels for multiple regression analysis on retention . . . . . . . . . . . . . . . . 48 8 Beta weights, F values and significance levels for multiple regression analysis on evaluation of the communicator. . . . . . 50 Vi LIST OF APPENDICES Appendix Page A Audio transcript of the messages . . . . . . 67 B Questionnaire booklet: first session . . . . 71 C Questionnaire booklet: second session - PRO group. . . . . . . . . . . . . . . . . 86 D Questionnaire booklet: second session - CON group. I O I O I I O I I O I O I I I I 98 CHAPTER I INTRODUCTION The.present.study explores some of the variables en- countered in attitude change research. The core of the traditional attitude change paradigm is the exposure of §s to an attitude.related message. The evaluation of the im- pact of the message is generally accomplished by employing a pre and.post attitude test, or a post test only design (where differing messages are presented to the same gs or more commonly where independent random samples of gs receive different treatments). Many of the traditional attitude studies used average attitude change as the dependent measure. Despite its practical implications, the average change in attitude associated with different messages is no longer the central concern of attitude research and theory. Recent investiga- tions, in an attempt to better understand the psychological aspect of attitude change, have focused on the individual differences in attitude change among §s exposed to the same message. The emphasis on the prediction of individual differ- ences in attitude change has provided the impetus to specify the attributes associated with attitudes. Some of the variables which have emerged as potential correlates of attitude change are: l. the retention of the attitude message 2. the evaluation of the communicator 3. g's initial attitude 4. the amount of factual information g had about the attitude object and 5. the extent of contact g has had with the attitude object. Retention of the message and evaluation of the com- municator are of particular interest because these variables illustrate a problem inherent in attitude research. Both variables can be conceived of, and in fact have been used as, independent and dependent measures. As independent variables both have been used as predictors of attitude change and as dependent variables both have used attitude change (or its components—~initial.and final attitude) as predictors. The.difficulty in determining the relation among attitude change, retention, and evaluation is in part attributable to the supposed temporal proximity of the pro- cess involved. Since the prediction of retention and evalua— tion have psychological and practical importance in their own right, the present study will also utilize both these measures as dependent variables. It should be apparent at this point that the phenome- non of interest (i.e., attitude change, retention, and evaluation) can be viewed as multiply determined. For the most part, previous attitude research has.been concerned with the exploration.of the relation between pairs of variables. It appeared profitable at this time to conduct a multivariate study employing all of the variables enumer- ated. Multivariate studies present the researcher with the unique opportunity to investigate the interrelations among variables. The present study employs all of the remaining variables in order to predict a particular dependent vari- able. Certain multivariate methods also permit the investi- gator to side-step a persistent problem encountered in the use of change measures (e.g., attitude change and pre—post differences on a retention measure). The major problems in measuring change are: l. the spurious negative element in the correlation of an initial score with gain on the same scale, due to the sharing (with the opposite sign) of the same errors of measurement and 2. the higher the correlation between pre and post test, other things being equal, the lower the reliability of the difference scores; on the other hand, the lower the correlation between the two measures, the less they can be said to measure the same thing. The above difficulties are commonly subsumed under the label "regression effects.“ Regression effects, by changing the slope and intercepts of the regression line, present serious problems in the interpretation of results.* *For a more comprehensive discussion of the measure- ment of change see Harris (1963), especially Chaptersl & 2. Several multivariate techniques employ the initial score (or pre—test) on the change variable in the prediction of the final score (or post-test) on the change variable. This procedure avoids the above problems while still permit— ting the investigator to explore the correlates of change. The present study employs a multivariate technique of analy— sis of this type and therefore uses initial and final attitude and not attitude change in the prediction of retention and evaluation of the communicator. The remainder of this chapter focuses on the relevant literature and the hypothesized relations among variables. The presentation is organized in three sections with each section devoted to one of the “dependent" variables (i.e., attitude change or retention or evaluation of the communi— cator). Within a section, each of the three independent variables (i.e., initial attitude, information and contact) and the remaining “dependent" variable are discussed in light of their relation to the section variable. For atti— tude change the “dependent" variables used as predictors were retention and evaluation of the communicator. For retention the “dependent“ variables used were evaluation of the communicator and final attitude. Finally for evaluation of the communicator the “dependent" variables used were retention and final attitude. Section 1: Attitude Change a) Initial Attitude Sears and Abeles (1969) in reviewing the literature on the discrepancy between S's attitude position and the attitude position of the message stated: “We could not find any case where discrepancy was negatively related to opinion change“ (p. 257). Sears and Abeles report that the most general finding has been that attitude change is an inverted—U “function“ of the discrepancy (e.g., Bochner and Insko, 1966). McGuire (1969) states that attitude change is likely to be a negatively accelerated increasing function of discrepancy (see Hovland and Pritzker, 1957; Zimbardo, 1960) unless the discrepancy is extreme; in that case, McGuire agrees with Sears and Abeles. g contended that in the present study, the relation between discrepancy and attitude change will be best charac- terized as a linear positive function. Several reasons that led to the above conclusion were: 1) the detection of significant non-linear trends generally entail data from large numbers of §S and 2) the curvilinearities re— ported in the above studies were not extreme. Therefore, it was hypothesized that the more S's initial attitude differs from the attitude expressed in the message, the greater will be the attitude change (Hypothesis l.a.). b) Factual Information In this study, information is measured by the amount of factual material SS possessed at the onset of the study. Although there appears to be little empirical evi- dence relating information about an attitude object and attitude change, there is some theoretical rationale for assuming the existence of such a relation. Knowledge about an attitude object can be viewed as a result of in- terest and/or involvement with the attitude object. In this manner, knowledge can be incorporated into the concept of “salience." Hovland, Janis and Kelley (1953) define "salience" as: ". . . the degree to which, in a given situa- tion, a specific group (of individuals or attitudes) is present and prominent in a person's awareness. . ." (p. 155). They further state that: "It would seem reasonable to assume that the salience of a group determines its availability as a possible source of resistance to change. . ." (p. 157). Lewan and Stotland (1961) experimentally attacked the problem of prior information and attitude change. Their experimental treatment consisted of exposing one half of their S5 to neutral information and then exposing all of their S5 to an "emotional" appeal. The experimental SS changed significantly less than the controls. Unfor— tunately, several problems existed in this study. The test was concerned only with resistance to negative change. The use of "emotional language" in the appeal may have irritated SS previously exposed to the neutral infor- mation. The experimenter and the regular class teacher were present during.both the presentation of the information and the message and finally, the experimental manipulation of information resulted in treating information as a di— cotomous variable (i.e., those gs that were presented with information and those that were not). The present E contended that the amount of previous factual material and final attitude will be related. Specifically it is hypothesized that there will be a nega- tive relation between the amount of previous factual in- formation and attitude change. Thus §S with more previous information will be less likely to modify their attitudes (Hypothesis 1. b.). c) Contact In this study, contact with the attitude object was measured by the extent of exposure to the attitude object and/or attitude related materials. .Contact, like information, might result from an interest or involvement with the attitude object. Sherif gt. El. (1965) have used the concept of previous contact in terms of group membership as an indicator of “ego—involvement.“ The con- finement of the concept of contact to group membership is unduly restricting. The present definition permits the substitution of a continuous variable (contact) for a ( dicotomous variable (membership). As used in the context of the present research, contact has not been related to final attitude. Sherif, 33. 21. (1965) have found ego-involvement in terms of group membership to be related to attitude change, with group members more resistant to attitude change on group relevant topics. Sherif and Hovland (1961) suggest that with high involvement the likelihood of rejecting the source increases and results in less attitude change and even boomerang ef- fects. Zimbardo (1960), however, hypothesized that atti- tude change should increase with the level of involvement. Zimbardo further predicted on the basis of dissonance theory, that the relationship.wou1d be particularly evident under conditions of high subject-message discrepancies. E, in agreement with Sherif and Hovland, predicted that there will be a negative relation between the amount of previous contact with attitude related material and the amount of attitude change (Hypothesis 1. c.). Thus §S with more contact will be less likely to modify their attitude. d) Retention McGuire (1957) and Miller and Campbell (1959) after the presentation of a message consisting of a favorable and unfavorable position found a significant correlation between recognition of the specific arguments presented and opinion change. In an extension of the above studies, Watts and McGuire (1964) found a significant positive relation between opinion change and recognition of the .specific arguments presented after presentation of a favor- able message-. These studies did.not.employ a pre-test on the retention material and scored retention as a dicotomous variable. Their.use.of opinion change as.the dependent variable may have also confounded retention and initial attitude. The present E contended that with the control of pre-message familiarity no relation will be found between retention and attitude.change (Hypothesis 1. d.). e) Evaluation.of the Communicator In general, Hovland gt. Ei'~(1953) found greater acceptance of a communicationwhen the communicator had high credibility. Studies such as these employed communi— cators differing in mean credibility ratings, and related these mean ratings to.the.amount of.attitude change ob- tained with each communicator. The present.investigator could not find any studies relating individual differences in evaluation to attitude change (a within group, rather .than a betweenegroup design). E contended that.§s who evaluate the communicator .more favorably will change their attitudes the most in the direction of the message (Hypothesis 1. e.). This Hypothe- sis can be justified in terms of “dissonance reduction" or more parsimoniously by assuming that Ss.who View the com- municator less favorably are less likely to be influenced by the message. 10 Section 2: Retention The process of retention can be defined as the ability to recognize or reproduce sensory stimuli after termination of physical stimulation. a) Initial Attitude There has been much allusion in Social Psychology to the potential influence of attitudes and values on re— tention. Early studies by Bartlett (1932), Levine and Murphy (1943) and others (e.g., Jones and Kohler, 1958) have found that congruence between S's attitude and the attitude message is related to more accurate retention. More recent studies (Greenberg, 1964; Waly and Cook, 1966) found that prior attitude positions were not significantly associated with retention of attitude related material. However, the selective retention issue remains un- settled. Work in the area designated as perceptual defense suggests the operation of personal values on stimulus se- 1ection.* Pavio and Steeves (1963) have approached the question of retention and values directly. Their SS were *Stimulus selection is most consistently used to denote the filtering of the stimulus array before it is stored, whereas retention is the filtering of the stimulus array after it has been stored in memory. Most of the work in this area must be construed to be concerned with the process of retention and not selection, since the dependent measures employed are recall or recognition and not the determination of a physiological response as is implied in the measure of stimulus selection. 11 presented with two tape recordings simultaneously (one to each ear). The contents of the tapes pertained to different Allport—Linzey values. The investigators found that SS tended to recall more of the material pertaining to the value on which they scored more highly. However, their study and others employing techniques such as diadic listen- ing (e.g., Broadbent, 1958) and binocular rivalry (Engel, 1961, and Ittelson and Slack, 1958) are so sophisticated in their methodology and limited in scope that it is extremely difficult to ascertain the relevance of their results to the process of retention given the presentation of stimuli as usually encountered. S's retention score is a function of his pre- vious familiarity with the material plus the increment associated with exposure to the message. The studies cited in this section have not, for the most part, controlled for the effects of pre—message familiarity on retention scores. Furthermore, this influence, if any, would be expected to yield results similar to those supporting selective reten- tion. E contended that when pre—message familiarity is controlled no relation will be found between initial atti- tude and retention of the attitude message (Hypothesis 2.a.). b) Factual Information No explicit empirical evidence relating the amount of previous information with retention was found. If pre- message familiarity is controlled and the selective retention 12 hypothesis is not accepted, then a relation between previous information and retention need not be assumed (Hypothesis 2. b.). c) Contact Once again empirical evidence on the relation between contact and retention was lacking. Since, within the frame— work of this research, information and contact are Viewed as allied concepts, no relation between contact and retention will be expected (Hypothesis 2. c.). d) Evaluation of the Communicator Hovland and Weiss (1951) found that retention tests revealed no difference in learning as a function of the credibility of the communicator. Their procedure involved the assignment of a credibility score to each of several sources on the basis of average "trustworthiness" ratings. This procedure disregards individual differences in credi- bility ratings and restricts the evaluation of the communi— cator to one dimension ("trustworthiness"). The present E felt that if individual differences in evaluation and a more comprehensive evaluation scale, which incorporated salient variables was employed, an association between evaluation of the communicator and re— tention will be found. It appeared that of all the variables which have been related to retention, evaluation of the com- municator possessed the most promise. Evaluation of the communicator is the variable which should, on a priori 13 grounds, be most closely associated with attention and therefore retention. It was the contention of the present E that the failure to find a relation between evaluation of the communicator and retention was due to the use of inade- quate evaluation scales. If individual differences in evaluation and a more comprehensive evaluation scale in- corporating S salient variables are employed, a positive association between evaluation of the communicator and re- tention will be found. The more favorable the evaluation of the communicator the better will be the retention (Hy- pothesis 2. d.). e) Final attitude Weiss (1953) found that final attitude was more con- sistently related with retention than initial attitude. Greenberg (1964) found that although initial attitude was not significantly related to the mean retention, final attitude position was significantly related to mean reten- tion. Greenberg's study, which appears to be the most thorough pre-post attitude and retention study, suffers from the difficulties of a pre—post attitude analysis which concentrated on mean information levels, thus ignoring the familiarity problem. It is possible that Greenberg's results could be attributed to an intervening variable such as evaluation ‘ of the communicator (See Section 2. d.). g contended that W i with the control of pre-message familiarity and evaluation 14 of the communicator no relation will be found between attitude and retention (Hypothesis 2. e.). 15 Section 3: Evaluation of the Communicator a) Initial Attitude Researchers have generally found that initial atti— tude and evaluation of the communicator co—vary. Byrne and Griffitt (1966) found that a communicator is evaluated more favorably by Es with congruent attitudes. In a re— lated study, Waly and Cook (1965) found a negative correla— tion (-.84) between discrepancy and ratings of the plausi— bility of the message. E hypothesized that initial attitude and evaluation of the communicator are related and that the communicator will be evaluated more favorably by §§whose initial atti- tude is more consonant with the message (Hypothesis 3. a.). b) Factual Information No empirical evidence appears to be available relat— ing the amount of previous factual material to the evaluation of the communicator. However, Eagly and Manis (1966) simu- lated ego involvement through creating a message by sex interaction and, as predicted, found that a negative message was less favorably evaluated by the ego involved Es. The Eagly and Manis treatment confounded ego involvement and initial attitude. Their ego involved Es also held favorable attitudes on the attitude issue. Dillehay (1965) found that a positive message was more favorably evaluated by Es with a positive attitude on the issue and high familiarity (as 16 judged by.a one item scale asking Es how familiar they were with the issue). Because of the inconclusiveness of the above studies, E proposed the following association between factual infor— mation and evaluation of the communicator: the relation between factual information and evaluation of the communi— cator will be.determined by the general reaction of Es to the message. Es with more information will evaluate the communicator in a more “extreme" fashion than those Es with less information- Evaluation of the communicator and in- formation will be positively related if the communicator receives a relatively favorable evaluation. Whereas the relation will be negative if the communicator receives a relatively unfavorable evaluation (Hypothesis 3. b.). This effect would probably be accentuated due to the predominantly factual content of the message in the present research. c) Contact There appears to be no research directly concerned with contact and evaluation of the communicator. E contended that the association between contact and evaluation of the communicator was similar to the association between infor- mation and evaluation of the communicator. That is, Es with more contact will evaluate the communicator in a more "extreme" fashion than those Es with less information. Evaluation of the communicator and contact will be positively related if the communicator receives a relatively favorable l7 evaluation. Whereas the relation will be negative if the communicator receives a relatively unfavorable evaluation (Hypothesis 3. c.). d) Final Attitude Although many studies have related evaluation of the communicator and initial attitude (see Section 3. a.), no studies were found relating final attitude and evaluation of the communicator. Since Es' final attitude incor— porate possible changes of attitude associated with the message, E contended that final attitude will be more highly related to evaluation of the communicator than initial atti- tude, with the communicator being more favorably evaluated by those Es whose attitude is most consonant with the message (Hypothesis 1. d.). Since the design employed in this study involves the presentation of a pro and con message, the results generated by one experimental treatment can.be validated by the re- sults of the other treatment. It is especially important to point out at this time that although the hypotheses are generally worded, E will confine his conclusions to the specific attitude and population employed in this study. The attitude object selected for investigation in the present research was labor unions. E selected labor unions as the focus of research for the following reasons: l8 1) the topic of labor unions is a contemporary issue 2 3 4 v in the population under study, pilot work had indicated a range of attitudes toward labor unions v individual differences in the amount of infor— mation that Es had about labor unions was found in pilot work ) Es had an opportunity to expose themselves to labor unions and labor union related materials (news media, classes, etc.) and 5) it was possible to construct messages comprised primarily of factual material which was un- familiar to Es. In summary, the following relations were hypothesized among the variables: Attitude Change a. The more E's initial attitude differs from the attitude expressed in the message, the greater will be the amount of attitude change. The greater the amount of previous factual informa— tion, the smaller will be the amount of attitude change. The greater the extent of previous contact with the attitude object, the smaller will be the amount of attitude change. The extent to which the attitude message is retained and the amount of attitude change will not be sig— nificantly related. The more favorable the evaluation of the communica— tor, the greater will be the attitude change. Retention a. The magnitude of initial attitude and the extent to which the attitude message is retained will not be significantly related. 19 The amount of previous factual information and the extent to which the attitude message is retained will not be significantly related. The extent of previous contact with the attitude object and the extent to which the attitude message is retained will not be significantly related. The more favorable the evaluation of the communi- cator, the greater will be the extent to which the attitude message is retained. The magnitude of final attitude and the extent to which the attitude message is retained will not be significantly related. 3. Evaluation of the Communicator a. The more consonant E's initial attitude is with the message, the more favorably the communicator will be evaluated. The greater the amount of previous factual infor- mation, the more polarized will be the evaluation of the communicator. The greater the extent of previous contact with the attitude object, the more polarized will be the evaluation of the communicator. The more consonant E's final attitude is with the message, the more favorably the communicator will be evaluated. This relation is hypothesized to be stronger than the relation between initial atti- tude and evaluation of the communicator (Hypothesis 3. a. above). The experimental design employed in attacking these hypotheses is outlined in the next chapter on Experimental Design. CHAPTER II EXPERIMENTAL DESIGN In an attempt to overcome some of the methodological difficulties in previous research, this study incorporated several refinements. The initial emphasis in this chapter will be directed to these difficulties and the remedial ac- tion taken. The latter part of the chapter is devoted to an overview of the experimental design. The problem of replication, or lack thereof, has plagued attitude research and the social sciences in general. The failure to replicate can often be attributed to the dis— similarity of E populations. Since this study incorporated a "replication" (a pro message and a con message) this pro— blem is of central importance. One procedure used for assuring the similarity of Es on particular variables of interest is matching. The present study matched each E with another E of the same sex on the basis of the inde- pendent variables of interest (initial attitude, information, and contact). One E from each pair was randomly assigned to one of the two groups, and the remaining E was assigned to the other group. 20 21 Another problem encountered in attitude research, and pre-post measurement in general, is the influence of memory factors from one testing situation to another. The effects of memory on post tests is particularly prevalent when test-retest intervals are short and/or few intervening stimuli are introduced (such as,.distracters or other ex- perimental treatments). However, the other horn of this dilemma is that long pre-post test periods and intervening stimuli allow bias to be introduced into the measurement. In an attempt to reconcile the above difficulties, the present study employed both parallel form attitude measures and test-retest measures. The influence of memory factors was attenuated by administering the pre-test for attitude change one week before the attitude message and interposing other experimental stimuli between testings. In order to ascertain whether Es had modified their initial attitude during the one week interval, a second (parallel form) attitude questionnaire was administered during the first session, and during the second session immediately preced- ing the presentation of the attitude message. The above design permitted the use of a test—retest procedure to assess attitude change (thus avoiding calibration problems) and also permitted the measurement of pre-message attitude (through the use of the parallel form attitude question- naire) while avoiding extensive memory effects. The measurement of retention of an attitude message also presents some special problems. Many studies ignore 22 the influence of pre-measure familiarity on retention scores. Pre-message familiarity can be controlled by pre— senting totally unfamiliar material in the message or by pre-testing Es on the retention material. Both pro- cedures entail difficulties. The presentation of totally unknown material can lead to quite artificial or trivial messages. On the other hand, pre-testing for familiarity can draw attention to those items which are to be used to measure retention and possibly contaminate the measurement of retention. The present study used a combination of both of the above methods while attempting to avoid the problems associated with each of these procedures. E selected little known, but relevant factual material in constructing a retention measure and pre-tested the retention measure. The selection of factual material allowed E to construct factual multiple choice objective items which were included in a factual information scale (given one week before the message), thus hopefully disguising its purpose. The question of: "What position on the attitude continuum does the message occupy?", presents another pro— blem of measurement. 'Many studies have not assessed whether, and to what extent, messages do differ on the attitude continuum. Of those studies which did assess the position of the attitude message, few utilized a measure which in— corporated the complexities considered in inferring atti- tude. In the present study Es assessed the position of the message by completing one of the attitude questionnaires, 23 which incorporated the multidimensional aspects considered in measuring a E's attitude. Es were requested to respond to this attitude questionnaire in a manner which was con— sistent with the "content or tone" of the attitude message. Thus, one of the attitude instruments which Es used to indicate their attitude was also used to ascertain Es perception of the position of the attitude message on the attitude continuum. At this point, a sequential outline of the experi- mental design of the main study is appropriate. The de- sign incorporates the solutions to the methodological problems discussed in this chapter. I. Session 1 A. Information test and pre-message retention scales (intermixed but scored separately) B. Contact scale C. First parallel form attitude questionnaire D. Second parallel form attitude questionnaire E. Several biographical items II. One week of intervening time (during which Es were matched on sex, initial attitude, information and contact and then assigned to the PRO or CON group) III. Session 2 A. First parallel form attitude questionnaire B. Video-tape attitude message (pro or con labor union message) C. Retention questionnaire (pro for group receiving pro labor union message, con for group receiving con labor union message) 24 D. Second parallel form attitude questionnaire E. Evaluation of the communicator scale F.. Evaluation of the message position on the atti- tude continuum G. Contact scale A more detailed presentation of the procedure employed in the main study is presented in Chapter IV. Although several problems have been discussed in this chapter, one crucial problem has not yet been explored. This is the necessity of developing relevant, reliable and valid instruments for the measurement of the variable under investigation. The development of the instruments to mea— sure the variables of interest is discussed in the follow— ing chapter on Instrument Development. CHAPTER III INSTRUMENT DEVELOPMENT The procedure used in this study entails the con- struction of several scales and attitude messages (video— tapes). The construction of each of these instruments will be discussed in turn.* Attitude Questionnaire I Subjects 27 males and 28 females (volunteers) enrolled in a sophomore level Psychology course were administered Attitude Questionnaire I. Procedure Uphoff and Dunnette (1956), and Newcomb (1939) developed questionnaires to measure attitudes toward unions. Of approximately 100 items in these questionnaires, 42 were selected on an E priori estimate of their relevance for the E population and their timeliness (wording changes were made in the majority of the 42 items). Twenty items ex— pressed favorable attitudes toward unions and 22 expressed *The raw data for the instruments (in the form of punched cards) can be obtained from John E. Hunter, Dept. of Psychology, Michigan State University, East Lansing, Michigan, 48823. 25 26 unfavorable attitudes toward unions. The items were random- ized and presented.to the Es in a questionnaire format. Es were instructed to.use a 5 point Likert-type response scale (1, Strongly Disagree to 5, Strongly Agree). Analysis The analysis consisted of determining the Operat- ing characteristics of the items and the structure of the item space. The analysis was undertaken in order to be in a better position to rewrite "weak" items, supplement exist- ing clusters, and create new clusters. Attitude Questionnaire II Subjects 53 males and 47 females (volunteers) enrolled in a s0phomore level Psychology course took Attitude Question- naire II. Procedure Five items from Attituderuestionnaire I were rewritten and an additional 35 items were constructed. The 35 new items were for the most part based on information found in Lester (1958), Litwack (1962), Rees.(l962), and Sturmthal (1966). Of the 77 items, 33 expressed favorable attitudes toward.unions and 44 expressed unfavorable atti- tudes toward unions. The items were presented to §S in a format similar to Attitude Questionnaire I. Analysis The sample was divided into two random halves with the restriction that the number of male and female Es were approximately the same in each random half. The 27 item analysis was carried out on one random half of the sample and the validation on the other random half of the sample. The 78.variable correlation matrix (R-matrix) was computed (the.77 items plus the sex of the E). There were no significant correlations between E's sex and item re- sponses, thus further analyses were performed on the com— bined male-female-data (N = 50). The R-matrix was then ordered by.PACKAGE* such that; the variable with the highest commonality is found, then the variable most highly correlated with this first variable is found (Variable 2). Then the variable most highly correlated with the second variable is found, etc. This procedure resulted in 7 "clusters" exhibiting average inter-item correlations above .30 (p §.05). The triplets of items exhibiting the highest inter—correlations were chosen from each of the clusters. This resulted in 10 triplets of items, one of which was assigned to Form A, .the next to Form B and the last to Form C. These three scales each contained 4 favorable and 6 unfavorable items toward unions. The inter-form correlations, reliabilities, means and standard deviations were computed for the item analysis group and for the validation group. These statistics, dis- played in Table 1, demonstrated sufficient reliabilities and stability upon validation for use. *A series of computer manipulations on R-matrices designed and programmed by Hunter and Cohen (1969). 28 TABLE 1. Correlations, reliabilitiesl, means and standard deviations for Attitude Questionnaire II Item Analysis Group , Item.Va1idation Group FORM - FORM A B C A B C A .79 .80 .79 A .81 .82 .81 F F O B .80 .80 O B .77 .72 R R M C .79 M C .76 Mean 30.54 30.32 30.52 Mean 30.36 30.32 28.68 S.D. 5.22 5.04 5.44 S.D. 5.08 4.54 4.77 NOTES: All correlations significant beyond the .01 level. lReliabilities computed from inter-form correlations. For the purpose of this study, Forms B and C were com- bined in order to produce a more reliable 20-item question- naire which would be a more sensitive measure of attitude toward unions. Form A will be referred to as A from this point on. (See Appendix B for a c0py of these questionnaires.) Video-Tape Attitude Message Procedure Two video-tapes were made by the same undergrad- uate male (communicator). One of these tapes expressed a favorable View toward unions; the other an unfavorable View toward unions. Each tape was approximately 5 minutes in duration, and contained comparable factual material. Lester (1958), Litwack (1962), Rees (1962) and Sturmthal (1966) supplied most of the information used in the messages. 29 A transcript of the audio portion of both video tapes is presented in Appendix A. Information Test I Subjects 77 males and 23 females enrolled in a sophomore level Psychology course served as Es in this phase of the study. Procedure A scale consisting of 45 factual items was con- structed. The items were selected and modified from material found in Government Documents, almanacks, periodi- cals, and several objective examinations given in an intro— ductory course on the history of labor unionism in America.* The items were written in a 5 choice format with the correct response appearing with similar frequency in each position. Analysis A random sample of 39 male and 11 female Es were drawn from the original sample of 100 Es. An item analysis was performed on the random sample and validated on the remaining 50 Es. The criterion for selecting the items was internal consistency. The internal consistency, means, and standard deviations for the item analysis group and for the validation group are displayed in Table 2. This data is based on the 16 items selected through item analysis from the original 45 items. Upon validation, the reliability dropped to a level unsatisfactory to E. *The objective examinations were graciously supplied by Dr. Charles Larrowe. 30 TABLE 2. Internal consistencies, means and standard devia- tions for Information Test I Item Analysis Group Item Validation Group K-R 20 .73 K-R.20 .63 Mean 8.66 Mean 7.70 S.D. 3.14 _ S.D. 2.87 Information Test II (INFO) Subjects 28 males and 38 females enrolled in a sophomore level Psychology course served as Es. Procedure Another instrument consisting of the 16 items retained from the original scale plus 14 additional items was developed. This 30 item instrument was similar in for— mat to Information Test I. Analysis A random sample of 14 male and 19 female Es was drawn from this sample of 66 Es. An item analysis was per— formed on this sample and validated on the remaining Es. This data is displayed in Table 3 and is based on the 25 items selected through the item analysis. (See Appendix B for a copy of this test.) Evaluation of the Communicator Scale (EVAL) Subjects 22 males and 12 females (volunteers) enrolled in a sophomore level Psychology course served as Es in this phase of the study. 31 TABLE 3. Internal consistencies, means and standard deviations for Information Test II Item Analysis Group Item Validation Group K—R 20 .76 K-R 20 .73 Mean 13.82 Mean 13.24 S.D. 4.41 S.D. 4.01 Procedure Anderson (1968) had 100 Es rate_555 personality- trait words on "likableness." The present E using the standard deviation of the "likableness" ratings as an index of concensus chose those words having the lowest standard deviation (5.70 on a 7 point scale). This procedure resulted in the selection of 67 words. The communicator and content of the message (video- tape) were described to the Es. The Es were then asked to write 2 descriptions of the communicator incorporating as many of the personality-trait words as they thought would be relevant given the situation. One description was written assuming that the Es felt "favorable“ toward the communicator, the other description was written assuming that the Es felt "unfavorable” toward the communicator. Analysis The frequency of use for each of the personality- trait words was determined for each.condition (favorable affect, unfavorable affect). Those personality-trait words mentioned by 10 or more of the Es under a condition were retained for inclusion. This procedure resulted in the 32 retention of 21 personality—trait words. Three-personality- trait words from this list were deleted since the list con- tained an antonym for each of these 3 words. The final form was composed of the remaining 18 personality—trait words (9 generated under the "favorable" condition and 9 generated under the "unfavorable“ condition. In the main study, the rating procedure was accom- plished.by requesting E5 to indicate on a five point scale (1, Not at All, to 5, Extremely), the extent to which the trait-word characterized the communicator. The evalua- tion score was obtained by subtracting the ratings the communicator received on the 9 "unfavorable" trait-words from his ratings on the 9 "favorable" trait-words. The range of possible scores on this scale was -36 to +36. (See Appendix C for a copy of this scale.) Contact Scale (CONT) Subjects 19 males and 13 females (volunteers) enrolled in a sophomore level Psychology course participated in this phase of the study. Procedure A 12—item yes—no scale designed to assess the extent of varied contact Es had with labor unions and union related materials was constructed. (A similar scale allowing for a gradation of responses was constructed, pre— tested and discarded on the basis of the instability of Es responses.) Es responded to the contact scale on two occasions, one week apart. 33 Analysis The 9 items which demonstrated the best test— retest reliability were retained. The reliability, mean and standard deviation of the 9 items is presented in Table 4. TABLE 4. The reliability, mean and standard deviation of the Contact Scale Test—Retest Reliability .96 Mean 3.28 Standard Deviation 2.21 A validation of the item analysis was not carried out at this point, but rather in the main study due to the diffi- culty of obtaining Es on two occasions, and the rather high reliability indicated by the item analysis. (See Appendix B for a copy of this scale.) Retention Questionnaire (RET) Subjects 10 males and 10 females (volunteers) enrolled in a sophomore level Psychology course were employed in this phase of the study. Procedure Fourteen 5 choice questions were constructed from a transcript of each of the video tapes (PRO, CON). Due to the design of the study, the items constructed for each video tape could be construed as objective or factual items. Each set of the items (PRO, CON) were written to exhaust the testable material presented in the video tapes. Five males 34 and 5 females read a transcript of each.of the video tapes and then completed the retention questionnaire. . Analysis No classical item analysis was undertaken for this instrument. Es responses were utilized in order to refine the wording of the stems and the choice of foils for the items. Each of the retention tests consisted of 14 recognition items with both retention tests sharing 8 items in common. .(See Appendix C for PRO questionnaire and Appendix D for CON questionnaire.) Evaluation of Message Position on the Attitude Continuum (AC) In order to ascertain the position of the attitude messages (PRO and CON), Es in the main study were requested to complete Attitude Questionnaire A in a manner consistent with the "content or tone" of the attitude message. As a further index of the position of the message a one item scale was constructed. The scale was written in the form of a question: “If you were to compose a message expressing your attitudes toward labor unions, how would its tone compare with the message presented on the video tape?" The response choices for the PRO group were: 1. negative 2. less positive 3. about the same 4. more positive. 35 The response choices for the CON group were: 1. more negative 2. about the same 3. less negative 4. positive (See Appendix C for a copy of these instruments.) CHAPTER IV PROCEDURE Subjects 62 male and 71 female (volunteers) enrolled in a sophomore level Psychology course served as Es. Method Each E participated in two experimental sessions approximately one hour each in duration. Es were adminis- tered the first questionnaire booklet (see Appendix B) in groups consisting of approximately 30 Es each over a period of 5 days. This questionnaire booklet consisted of the Information Test (INFO), the pre-test on the Retention Questionnaire (P-RET), the Contact Scale (CONTl), the Atti- tude Questionnaires (Al and Bl*), and several biographical items. When all Es had been administered the first questionnaire booklet, each E was paired with another E of the same sex whose responses were similar on the following variables: information, extent of their previous contact with and previous attitude toward labor unions. *The subscripts indicate the session. 36 37 This procedure resulted in the formation of 26 male and 31 female pairs. One member of each pair was randomly assigned to receive the pro-union message (PRO) and the other member of the pair was assigned to receive the anti- union message (CON). The time and date for the second session for each E was determined by the date they had taken the first session and by the group to which they were assigned (PRO or CON). Each E was requested by phone to return for the second session one week after their first session at 7:30 p.m. or 8:30 p.m. depending upon the group to which they were assigned. Each video—tape was shown once a night with the order of presentation reversed each night. Es were administered the second questionnaire booklet (see Appendix C and D) in groups of approximately 15 Es each over a period of 5 days. This second session was held in a small (seating capacity approximately 25) classroom facility. Es were instructed to close their booklets when they had completed the Attitude Questionnaire (A2). When all Es completed A2, the video—tape was shown on closed circuit T.V. After viewing the Video-tape, Es were instructed to complete the remainder of the question- naire booklet which consisted of the Retention Questionnaire (RET), the Attitude Questionnaire (B2), the Evaluation of the Communicator Scale (EVAL), the Evaluation of Message Position on the Attitude Continuum (AC), and the Contact Scale (CONTZ). 38 Because 5 Es failed to appear for the second session, the analyzable data consisted of 24 male pairs and 28 female pairs. Thus 52 Es were in each of the two conditions. CHAPTER V RESULTS The results from the PRO and CON groups will be presented together in order to facilitate comparisons. The inter—correlations, reliabilities, means and standard deviations of the scales for both groups are presented in Table 5. The reliabilities of the instruments appear satis— factory with the exception of the Retention Questionnaires. It should be noted that the internal consistencies reported in Table 5 for the Retention Questionnaires are based on the difference score (post-message score on the retention measure minus the pre—message score on the retention mea- sure.) The employment of the difference score was one of the methods by which post—message retention could be mea— sured while controlling for pre—message familiarity. Another method would be the inclusion of the pre-message retention score as one of the predictors of post—message retention. E chose the difference score procedure since its use allowed for a more parsimonious presentation of the data. (In the present study the results obtained by employment of these two procedures were the same.) The 39 ‘+U madam m.nomncouu ma mmaufiaflnmflaou amspfl>flpsfl mcflusmfioo .muoom mosmumwuflp m mo mpHHHQMHHmm .m .imm mama mam u u a m anon 6H0 manage muaaagmaamu was: unamm .v xmmB H ummumulumwe .m mama< m.£omncouo .N Hmcommaa wan sacs omumna mum moaunaanmaamm .H .Hm>oa H0. or» um Hm. cars Hmumwum .Hm>ma mo. mnu pm ucmoamficmam mum mm. swap Hmummnm maoaumHmuuoo "mmsoz ma.m om.HH om.m mH.oH H5.m mo.m ~m.m H¢.m mm.m .Q.m Hm.m mm.m Hm.m ma.m vm.m mm.v mv.v «H.N mw.¢ .Q.m hm.mnm~.m vo.o mm.mm om.om No.am mm.Hm «v.m oo.mH zmmz vm.mv mm.m mh.m Hm.oo oo.om hm.om Nv.am mm.m mm.ma zm AH. om. A¢>m wN.I ma. mvm. .53 mm. mm. mom. Ema mo.u sm.u mo.u Nam. . mm AH. me. «N. we. mm oo. «0.- mm. He. «mm. Hm NH. em. mm. he. «mm. Hm Ho.- om.- ma. so. me. mmm. «a Hm. om. ma. mm. as. mm. ~<, oa. wH.u ma.| vm. mm. mm. Wmm. 4 as ea. mm. mm. Ho. vm. mm. mmm. Hm _ “a J no.1 mH.| mm. om. om. am. am. Mmm. m 9200 mH.I ma. No. ma. mm. vm. mm. mam. Bzou mm.| ov.| ma. mm. mm. mm. ma. am. Wow. a omzH ma.u mm. vH.I Hm. ma. mo. vH. mm. «on. omzH O< A<>m 8mm Nm Hm N< H4 8200 omzH 04 HIE/m 9mm Nm Hm N4 Hfi H.200 omzH zoo 0mm mmsoum zoo tam 0mm map How mmamom on» mo mGOHuma>wG pumpcmum can msme mmHuHHHQMAHmH .mGOHpmamuHoouwuaH .m mqmda H 41 difference score, however, proved to have some undesirable qualities. The pre-test was, in effect, a very specific information test whose difficulty (an average score of 3.56 PRO and 3.67 CON correct out of 14) led to an internal consistency of .00 PRO and .08 CON. The internal consis— tencies of the post measure was .45 PRO and .57 CON. The reliability of the difference score is, in effect, a weighted average of the internal consistencies of the pre and post measures. In a recent study, Bruni (1963), an internal consistency of .73 was found for a 60 item post test con— cerned with the recognition of verbal statements presented on a film. The internal consistency of the post test in the present study corrected to 60 items is .78 PRO and .85 CON. In both groups there was significant attitude change. In the PRO group a mean change of 6.31 (B2 - Bl) yielded a E of 5.95 (p 3.01), and in the CON group a mean change of 4.25 (Bl — B2) yielded a E of 3.24 (p i°01)° ‘ The position of the messages on the attitude con— tinuum were in the expected direction and approximately 7 points from the extremes of the scale. The 3 between AC PRO and Ac CON equaled 29.55 (p $.01). The difference in both groups between the mean initial attitude (A2) and the position of the message (Ac) was approximately 3 standard deviation units. Furthermore, the one item index of dis— crepancy received a mean rating of 2.08 and a standard deviation of .58 in the PRO group (where a 2 on the item 42 indicated that E would construct a message that was “less positive“ than the message presented), and a mean rating of 2.90 and a standard deviation of .79 in the CON group (where a 3 on the item indicated that E would construct a message that was "less negative" than the message presented). The above results indicate that the messages were perceived as being at the extreme ends of the attitude continuum. Sex was not significantly related to any of the de— pendent variables in either the PRO or CON group. Sex was, however, significantly related to INFO and CONT in both ' groups. In the PRO group r = —.34 (p 1.05) and SEX, INFO rSEX, CONT = —.35 (p :.01). In the CON group rSEX, INFO = -.31 (p :.05) and r = -.38 (p :.01). These SEX, CONT correlations indicate that females have significantly less information than males about labor unions and less contact with labor unions. The hypotheses were tested through the use of multiple regression analysis. Multiple regression yields the best linear combination of variables in predicting a criterion. The QEEE weights represent the relative contributions of each of the variables for the prediction of a criterion. Multiple regression also permits the assessment of the relations between several variables and change while avoid- ing the difficulties encountered when dealing with change scores directly (see Introduction p. 3). The use of change scores can be avoided if we note the following: Given true scores, the determination of final attitude can be partitioned 43 into the same components as attitude change (i.e., that partition attributable to initial attitude and that portion attributable to other sources). That portion of final atti- tude not predictable from initial attitude is identical to that portion of attitude change not attributable to initial attitude. Thus, for true scores a significant contribution to R A"A, etc. 18 exactly equivalent to a Significant con— tribution to R For fallible data, there is A'-A'A, etc. usually a considerable increase in statistical accuracy in using final attitude rather than change as the criterion. However, it is crucial that the differences in change attributable to initial attitude is a linear function of initial attitude for the above reasoning to be true. Furthermore, in multiple regression, the assumption of linearity is implied for all the relations among the vari- ables. Therefore, the relations were tested for departures from linearity and no significantly non—linear relations were found. The difference between the correlation of Al and INFO, and A2 and INFO in the CON group was significantly different (3 = 3.34 p :.01, using a test for differences of correlations from the same population). This difference indicates that the relation between attitude and informa— tion shifted from Session 1 to Session 2 without the intro- duction of an experimental treatment (i.e., history effects). Therefore, the relation between B1 and INFO, and B and 2 INFO would be expected to display a difference which was 44 not attributable to the experimental treatment (the message). Due to this postulated difference, and in lieu of the.fact that correlations among A1’ B1' and A2 were rela- tively uniform in both groups, E chose A2 as the most appropriate pre—measure of attitude. The remainder of this chapter is organized around each of the three "dependent" variables (attitude change, retention and evaluation of the communicator). The specific hypothesis will be restated, relevant data will be presented, ‘ and the status of the hypothesis in light of the data will be indicated. Attitude Change Hypothesis l.a., which states that the greater the discrepancy between §'S initial attitude and the mes- sage, the greater the attitude change, was not supported by the data. Initial attitude was a linear function of discrepancy since the ratings (AC) of the positions of the messages on the attitude continuum were more extreme than the attitude of any E. Therefore, the correlations between initial attitude and attitude change are equivalent to the correlations between discrepancy and attitude change. The correlation between initial attitude (A2) and final attitude (B2) in both the PRO and the CON groups (see Table 5) was sufficiently large to make it unlikely that any significant relation existed between initial attitude and attitude 45 change. The existence of a relation between initial atti- tude and attitude change was, however, explored by obtaining rB2_Bl’ A2’ A2 (instead of B1) was chosen as the initial attitude measure since r results in a spurious B2 ‘ Bl' Bl correlation due to the sharing of error varience by both B2,--..Bl (attitude change) and B1 (initial attitude)*. For the PRO group r = .14 and -.03 for the CON group, B2 - Bl,A2 neither of which was significant. Thus, in this study the discrepancy between Es initial attitude and the message position.was not significantly related to attitude change. The remainder of the hypotheses in this section are concerned with attitude change BEE EE' Table 6 presents a test of these hypotheses through a multiple regression analy- sis on final attitude with initial attitude as one of the predictors.** The 2233 weights for INFO, CONT, RET and EVAL represent their contribution to the prediction of attitude change. Hypothesis l.b., which postulated that the greater the amount of information, the smaller the amount of atti- tude change, received no support from the data. The EEEE weight for INFO in the PRO group was only -.04 (p 3.75) and *It is of interest to note that for the PRO group -.43(p<. 01)for'the CON group r r =-o BZ-Bl 1B1: BZ—Bl’Bl (p<. 18). The combined results for the PRO and CON groups yielded a p<. 01 (using Z= 2 normal deviates /V# deviates summed) and were in the predicted direction. **The overall regression represents the contribution of all of the predictors (including initial attitude - A2 ) in accounting for the variance in final attitude. 46 TABLE 6. Beta weights, F values and significance levels for multiple regression analysis on final attitude PRO Overall Regression A2 INFO CONT RET EVAL R .75 B .64 -.04 -.04 .04 .30 F 12.00 F 38.66 .11 .13 .13 7.40 p < .01 p i .01 .75 .72 .72 .Ol CON Overall Regression; A2 INFO CONT RET EVAL R .72 B .60 .02 .08 .00 —.23 F 9.62 F 28.58 .01 .40 .00 4.07 p i .01 p i .01 .91 .53 .99 .05 .02 (p:.9l) in the CON group. change were not significantly Hypothesis l.c., which extent of contact the smaller also received no support from for CONT in the PRO group was (p:.53) in the CON group. Thus information and attitude related in this study. stated that the greater the the amount of attitude change the data. The EEEE weight only —.04 (p:.72) and .08 Thus contact and attitude change were not significantly related in this study. Hypothesis l.d., which attitude change would fail to was not refuted by the data. PRO group was .04 (p:.72) and postulated that retention and show a significant relation The beta weight for RET in the .00 (p:.99) in the CON group. Thus retention and attitude change were not significantly related in this study. 47 Finally, Hypothesis l.e., which stated that the more favorable the evaluation of the communicator, the greater the attitude change was supported by the data in this study. The EEEE weight for EVAL in the PRO group was .30 (p:.01) and.-.23 (p:.05) in the CON group. In the PRO group, a more favorable evaluation was associated with positive atti- tude change, (i.e., in the direction of the message) and in the CON group, a more favorable evaluation was associated with negative attitude change, (i.e., in the direction of the message). Retention The hypotheses incorporating retention as a dependent variable were tested through the use of multiple regression analysis. The results of this analysis is displayed in Table 7. Retention was considered the criterion and A2 (initial attitude), INFO, CONT, EVAL, and B2 (final attitude) were the predictors. Hypothesis 2.a., which proposed that no significant relation will be found between initial attitude and reten- tion, was not refuted by the data. The 2232 weight for A2 in the PRO group was only .04 (p:.83) and .10 (p:.59) in the CON group. Thus in this study no significant relation was found between Es' initial attitude and the extent to which the message was retained. Hypothesis 2.b., which stated that no significant relation will be found between information and retention, 48 TABLE 7. Beta weights, F values and significance levels for multiple regression analysis on retention PRO Overall Regression A2 INFO CONT EVAL B2 R .41 B .04 —.24 .03 .34 .07 F 1.84 F .05 2.64 .04 4.70 .12 p < .12 p i .83 .ll .84 .04 .73 CON Overall Regression A2 INFO CONT EVAL B2 R .34 B .10 .15 .18 .25 .00 F 1.20 F .30 .69 1.20 2.47 .00 p < .31 p i .59 .41 .28 .12 .99 was not refuted by the data. The REES weight for INFO in the PRO group was -.24 (p:.ll) and .15(p:.41) in the CON group. Although the relation in the PRO group approaches significance, the relation in the CON group does not. Therefore,we must conclude that no significant relation was found between the amount of previous information and reten- tion in this study. Hypothesis 2.0., which proposed that no significant relation will be found between contact and retention, was not refuted by the data. The beta weight for CONT in the PRO group was only .03 (p:.84) and .18 (p:.28) in the CON group. Thus in this study, no significant relation was found between the extent of previous contact with the atti- tude object and the degree to which the message was retained. 49 Hypothesis 2.d., which states that the more favorable the evaluation of the communicator the greater the retention, was supported by the data.in this study. The 9223 weight for EVAL in the PRO group was .34 (p:.04) and .25 (pi.12) for_the CON group. The probability of obtaining two 2 values.this small or smaller was :.02.* Therefore, the joint results of both groups.indicated the existence of a positive relation between.evaluation of the communicator and retention. Hypothesis 2.e., which states that a significant re- lation.between final attitude.and retention will not be found was not refuted by the data. The EEEE weight for B2 in the PRO group was only .07 (p:.73) and .00 (p:.99) in the CON group. Thus in this study, no significant relation was found between E's post—message attitude and the degree to which the message was retained. It is interesting to note that the overall regression on retention did not reach a generally accepted level of statistical significance in either group. Evaluation of the Communicator Most of the hypotheses that incorporate evaluation of.the communicator as a dependent variable were tested through the use of multiple regression analysis. The 2 normal deviates *Using Z =/# of deviates summed 50 results of this analysis is displayed in Table 8. Evalua- tion of the communicator was considered the criterion and A2 (initial attitude), INFO, CONT, RET, and B2 (final atti- tude) were the predictors. TABLE 8. Beta weights, F values and significance levels for multiple regression analysis on evaluation of the communicator PRO Overall Regression A2 INFO CONT RET B2 R .57 B -.19 .26 .05 .27 .47 F 4.38 F 1.18 3.83 .16 4.70 7.40 p < .01 p i .28 .06 .69 .04 .01 CON Overall Regression A2 INFO CONT RET B2 R .52 B .13 —o35 .00 .20 ’035 F 3.42 F .57 5.21 ..00 2.47 4.07 P < .01 p i .46 .03 .99 .12 .05 Hypothesis 3.a., which states that the more consonant E's initial attitude is with the message, the more favor- ably the communicator will be evaluated was not supported by the data.. The E233 weight of A2 in the PRO group was only .04 (p:.83) and .10 (p:.59) in the CON group. These results indicate that initial attitude did not significantly con- tribute to the multiple prediction of the evaluation of the communicator in this study. It can also be seen from Table 51 5 that the zero-order correlations between initial attitude and evaluation of the communicator were not significant, but were consistently in the expected direction. Hypothesis 3.b., which proposes that the greater the amount of information, the.more polarized the evaluation of the communicator was borne out by the data in this study. The 9222 weight for INFO in the PRO group was .26 (p:.06) and -.35 (p:.03) in the CON group. The direction of the re- lations, as indicated by the sign of the REES weights, was predictable from the average evaluation of the communicator ‘ in the PRO and CON groups. The average evaluation of the communicator for the PRO group was 8.96, while the mean evaluation for the CON group was 5.25 (E = 1.76, pi.10). Thus, as was hypothesized, in the group which evaluated the communicator more favorably (PRO) the relation between INFO and evaluation was positive. On the other hand, in» the group that evaluated the communicator less favorably (CON), the relation between INFO and evaluation of the com- municator was negative. Hypothesis 3.c., which stated the greater the extent of contact, the more polarized.the evaluation of the communi- cator will be, was not supported by the data. The REES weight for CONT in the PRO group was .05 (p:.69) and .00 (p:.99) in the CON group. Thus no significant relation was found in this study between the extent of previous contact with the attitude object and evaluation of the communicator. 52 Hypothesis 3.d., which stated that the more consonant E's.fina1 attitude is with the message, the more favorably the communicator will be evaluated was supported by the data.in this study. The 2233 weight for B2 in the PRO group was .47 (p:.01) and -.35 (p:.05) in the CON group. Thus in the PRO group, Es with higher final attitudes evaluated the communicator more favorably while in the CON group Es with higher final attitudes evaluated the communicator less favor- ably (as was expected). Furthermore, the relation between final attitude and evaluation of the communicator was hypothe— sized to be stronger than the relation between initial atti- tude and evaluation of the communicator. The relative strength of A and EVAL, and B and EVAL were compared in 2 2 each group through the use of a E test for correlations from the same population (see Table 5 for these correlations). In the PRO group, the 5 between r a was A2,EVAL nd rB2,EVAL 2.17 (p:.05). In the CON group the 3 between rB2,EVAL and rB2,EVAL was 1.68 (p:.10). The probability of obtaining two 2 values this small or smaller was 1.02.* Therefore, the joint results of both groups indicated that the final atti- tude—evaluation of the communicator relation was signifi- cantly-greater than the initial attitude—evaluation of the communicator relation as was predicted. 2 normal deviates * I = USlng Z /# deviates summed CHAPTER VI DISCUSSION The discussion has, for the most part, been organ- ized around each of the "dependent" variables. The find- ings of the present study are summarized, then related to previous theory and research. The implications for future research, generated by the results of the present study, are included in the discussion. Attitude Change As predicted, evaluation of the communicator was significantly related to final attitude. It was found that a more favorable evaluation of the communicator was associated with greater changes in attitude in the direc- tion of the message. This result, although not surprising, emphasizes the importance of individual differences in the perception of the communicator. .It is therefore suggested that future studies incorporate individual differences in the.perception of the communicator. It was hypothesized that the discrepancy between the initial attitude and the attitude expressed in the message would be related to attitude change. The greater the 53 54 discrepancy (which in this study was identical to initial attitude), the greater the attitude change. It was found that discrepancy was not related to attitude change. The failure to find a relation between discrepancy and attitude change is of considerable interest in light of previous research. Several explanations are possible. First, although the attitude change was statistically significant, it may not have been large enough. iIf the treatment effect was not sufficiently strong then this may have precluded the possibility of obtaining a significant relation between discrepancy and attitude change. Secondly, it was found that the messages were per— ceived as extreme by a large majority of Es. Therefore, it is not unreasonable to assume that the range of dis— crepancies between the initial position of Es and the position of the message was restricted (i.e., low discrep- ancies were non-existent). Since the treatment effect was sufficient to produce a relation between evaluation of the communicator and atti- tude change, the second explanation appears to be the most parsimonious. It should also be noted that most previous studies have used the initial attitude score as a component of attitude change, thus introducing a spurious relation between initial attitude and attitude change. The spurious relation would tend to produce a positive association be- tween discrepancy and attitude change. The present study avoided this spurious relation by employing a parallel form 55 attitude scale on the initial attitude measure and a pre- post measure on another attitude scale to assess change. It was hypothesized that the greater the amount of previous factual information or the greater the extent of previous contact with attitude object, the less will be the attitude change. No such significant relations were found in the present study. The lack of an association between information and attitude change and also contact and atti- tude.change is of some interest. Both dissonance theory (e.g., Zimbardo, 1960) and ego-involvement theory (e.g., Sherif and Hovland, 1961) would predict significant, but differing in direction, relations between information, contact and final attitude. The reliabilities of the in- formation and contact scales used in the present study make it unlikely that the results are due to errors of measurement. The degree to which information and contact can be incorporated into theories of attitude change has not been specifically enunciated by attitude theorists. Therefore it is also possible that these two variables may not be the appropriate operational definitions of concepts " as was intended. such as "centrality" or "involvement, It is appropriate at this time to examine the operat- ing characteristics of the contact scale. This scale pre- sents 9 situations through which an individual may expose himself to union-related material. The total score on the scale represents the variety of situations in which the E has had contact with union-related.material and not the 56 amount of contact the E has had with union related materials. Although variety and amount are, in all likelihood, related empirically, theoretically the concepts are distinct. E, in pilot work, found the operationalization of amount of contact particularly difficult and consequently chose variety as a measure. It is suggested, however, that the hypothe— sized relation between contact and attitude change might be uncovered by using a measure of amount of contact. Previous research has suggested the existence of a relation between attitude change and retention which was not found in the present study. (The present study hypothesized that no significant relation would be found.) Unfortunately, several explanations are possible. First, the internal con- sistency of the retention measure used in the present research was unquestionably low and may have masked these relations. The low internal consistency raises a fundamental issue: Is internal.consistency, as used in this study, the appropriate index of reliability for retention (e.g., if retention is viewed as a random attention process with the intervals of attention varying as a function of the perceiver and the situation then Cronbach's Alpha would not be an appropriate index of reliability)? Unfortunately, this study is in no position to attack this question. Second, the majority of the previous retention studies employed no pre-measures on the retention material. If, as is quite possible, a positive relation pre-existed between familiarity of the retention material and attitude, 57 the continued existence and/or augmentation of such a relationship after the presentation of the message would not be surprising. Third,.it is feasible that the association of atti- tude change with retention may be mediated by evaluation of.the communicator. Evaluation of the communicator may be of particular importance in understanding the relation between final attitude and retention. As was noted in the Introduction, recent research has tended to find a stronger association between final attitude and retention than initial attitude and retention. The present study found that final attitude was more closely associated with evalua- tion of the communicator than was initial attitude. It is therefore.quite possible that evaluation of the communicator is the mediating variable in the attitude-retention associa— tion. E could not find any research in which evaluation of the communicator was held constant or partialed out of the initial attitude-retention or attitude change—retention relation. Retention Evaluation of the communicator, as predicted, was positively related to retention. Thus a more favorable evaluation of the communicator was associated with the re- tention of a greater amount of the attitude message. Due to the supposed temporal proximity of evaluation of the communicator and retention, previous research has not 58 discerned the directionality of the relation (i.e., cause and effect). However, if one was to speculate about direc— tionality, it would be reasonable to assume that positive affect toward the communicator would result in greater attention to the message, and consequently better retention. As hypothesized, no significant relations were found between initial attitude or final attitude and retention. It is unfortunate that several explanations for the lack of significant associations between initial or final atti- tude.and retention are possible: 1) the internal consist- ency of the present retention measures were quite low, 2) the failure of previous research to ascertain pre-message familiarity may have resulted in the reported association between initial attitude or final attitude and retention, 3) the association between initial attitude or final atti- tude and retention may have been mediated in previous re- search by another variable, namely evaluation of the communicator. (For a more extensive discussion of the preceding three explanations, refer to the section on Attitude Change.) As hypothesized, no relation was found between pre- vious factual information about the attitude object or extent of previous contact with the attitude object and retention. E found no rationale for predicting any such relations. .However, an allied area of considerable practi- cal importance, which was not explored in the present study is the issue of selective exposure. That is: "To what 59 extent‘will previous contact with, or previous factual in- formation about, an object 1ead.to purposeful exposure to additional information about that object?" In all likeli- hood, the answer to this question will depend upon the type of contact and information Es have previously had, the affect associated with this information and contact, and the characteristics of the additional information, etc. Although the area of selective exposure entails many com- plexities, it appears worthwhile to explore the relation between selective exposure and previous contact and infor— mation. Evaluationuof the Communicator As hypothesized, final attitude added significantly to the prediction of the evaluation of the communicator. In the group which received.the pro-union message, more favorable attitudes toward unions were associated with more favorable evaluations of the communicator; while in the group that received the anti-union message, favorable atti- tudes.toward unions were associated with less favorable evaluations of the communicator. However, a similar relation between initial atti- tude and evaluation of the communicator, although hypothe- sized, was not found. This lack of,a relation is explainable if it is assumed that Es evaluate the communicator in terms of their post-message attitudes. If the message affects a change in the highly polarized Es, then these Es 60 may evaluate the communicator quite differently than would be expected on the basis of their pre-message attitude. The above considerations would lead one to conclude that final attitude and evaluation of the communicator would be more highly related than initial attitude and evaluation of the communicator (which was in fact the case). The rela- tion between initial attitude and evaluation of the communi— cator, while not significant, was consistent and in the predicted direction. As predicted, the relation between amount of previous , ( information and evaluation of the communicator appeared to be associated with the level of the evaluation of the com- municator. Thus in the group in which the communicator was more favorably evaluated (pro-labor union message), a positive.relation existed between the amount of previous information and evaluation of the communicator. On the other hand, in the group where the communicator was less favorably evaluated (anti-labor union message), the rela- tion between amount of previous information and evaluation of the communicator was negative. It appeared that Es with more previous factual information were more sensitive to the expertise and credibility of the communicator. These results require additionalistudy. It is pos— sible.that the relation.between information and evaluation of the communicator is peculiar to a factual message such as the present one or to other idiosyncratic aspects of this study. 61 Previous contact with the attitude object, which was hypothesized to be associated with evaluation of the communicator in the same manner as previous factual infor- mation was not significantly.related to evaluation of the communicator. The lack of such a relation may be attribut- able to the Operating characteristic on the contact scale, which was previously discussed in the section on Attitude Change. Although the extent of previous contact was not found to be related to evaluation of the communicator, the amount of previous contact might be related to evaluation of the communicator. Overall, it appears that evaluation of the communi- cator was the variable which was most consistently related to the other variables in this study. The amount of pre- vious factual information, the extent to which the attitude message was retained, and the final post-message attitude were all significantly associated with evaluation of the communicator. It should be noted, however, that the evalua- tion scale employed in this study.was specifically designed for: l) a particular communicator-(an undergraduate Sociology student), 2) a specific content area (i.e., labor unions), 3) a message of a particular length (e.g., 5 minutes-in.duration), and.4) a specific mode of presentation (i.e., videOrtape). Although.the above 4 factors may vary from study to study, the evaluative dimensions which Es select to use may display consistency within a factor (e.g., regardless of the duration of the presentation) or 62 across factors (e.g., regardless of the mode of presenta- tion or topic). Therefore, the specification of the rele- vant evaluative dimensions for specific combinations of the above factors is, in itself, a worthwhile endeavor. REFE.RENCES REFERENCES Anderson, N. H. Likeableness ratings of 555.personality- trait words. Journal gf-Personality_and Social Psychology, 1968, 9, 272-279. Bartlett, C. Remembering: a study in experimental and ~socialpsycholo‘gy. Cambridge: Cambridge University press, 1932. Bochner,.S., and Insko,.C. A. Communicator discrepancy, -source credibility, and opinion change. Journal of Personality and Social Psychology, 1966, 4, 614-21. Broadbent, D. E. Perception and communication. London: Pergamon Press, 1958. Bruni, E- .A film test of accuracy in observing people and .its correlates. Unpublished-Master's thesis, Michigan State University, 1963. Byrne, D. and Griffitt,.W. .Similarity versus liking: A -clarification. -Psychonomic Science, 1966, 6, 295-296. Dillehay, R. C. Judgmental processes in response to a persuasive communication. Journal of Personality and Social Psychology, 1965, 1, 631-641. Eagly, Alice H., and Melvin Manis. "Evaluation of Message and Communicator as.a Function of Involvement," Journal of Personality and Social Psychology, 1966, 3, 483-485. Engel, E. Binocular methods in psychological research. .In.F. P. Kilpatrick (ED-), Explorations in trans- iactional_psychology. .New York: New York University Press, 1961.- Greenberg, B. S. On relating attitude change and informa- tion gain. Journal of Communication, 1964, XIV, 157-171. Harris, C. W. (Ed.) Problems in measuring change. Madison: University of Wisconsin Press, 1963. 64 65 Hovland, C- I. and I-.L-.Janis, and H. H..Kelley. Communi— cation and persuasion. New.Haven, Conn-: Yale University Press, 1953. Hovland, C. I. and Pritzker, H..A. Extent of opinion change as a function.of amount of change advocated. .Journal of Abgprmai-and Seaial Psychology, 1957, 34, 257-261. Hovland, C. I. and Weiss, W. The influence of source credibility on communication effectiveness. Public .Opinion Quarterly, 1951, 15, 635—650. Hunter, J. E., and Cohen, S. H. Package: A system of com- puter routines for the analysis of correlational data- .Educational and.Psychological_Measurement, 1969, 29, 697-700. Ittelson, W. H. and Slack, C. W. The perception of persons as visual objects. In Tagiuri, R., and Petrullo, L. (Eds-) Personal perception and inter-personal be- havior, Stanford: Stanford University Press, 1958, pp. 210—228. Jones, E. E. and Kohler, R. The effects of plausibility on the learning of controversial statements. Journal of Abnormal and Social Psychology, 1958, 57, 315-320. Lester, R. A- As unions mature. Princeton: Princeton University Press, 1958. Levine, J. M. and Murphy, G. The learning and forgetting of controversial material. Journal of Abnormal and Social ngchology, 1943, 38, 507-517. Lewan, P..C. and Stotland, E. The effects of prior informa- tion on susceptibility to an emotional appeal. Journal of Abnormal and Social Psychology, 1961, 450-453. Litwach, L. The American labor movement. New Jersey: .Prentice-Hall, 1962. McGuire, W. J. Order of presentation as a factor in "con— ditioning“ persuasiveness. In C. I. Hovland (Ed.) .The order of presentation inEpgrsuasion. New Haven: Yale University Press, 1957, 98-114. McGuire, W. T. The nature of attitudes and attitude change. In G. Lindzey and E. Aronson (Eds.), The handbook of social p§ychology, 2nd.Ed., Reading, Massachusetts: Addison—Wesley, 1969, 134-314. 66 Miller, N.,-Campbell,.D. J. Recency and primacy in persua- sion as.a function.of the timing of speeches and measurements.. Journal of.Abnormal and Social Psy- chology, 1959, 59, 1-9. Newcomb, T. Mr .Labor unions as seen by their members: an -attempt to.measure attitudes.~ In Hartman, G. W., rand Newcomb, T. M. (Eds.), Industrial conflict, New Yorke.CordonuPress,‘1939,'313-338. Paivio,.Ar and Steeves, R._ Personal values and selective perception of speech, Perceptual and Motor Skills, 1963, 17, 459-464. Rees, A. The economics of trade unions. Chicago: University of Chicago Press, 1962. Sears,.D. O. and.Abeles, R. P. Attitudes and opinions. -Annualeeview Of Psychology, Volume 20, 1969, 253-288. Sherif, Carolyn W., Sherif, M., and Nebergall, R. Attitude wand attitude change:-The.socia1_judgment-involvement .approach. -Philadelphia: Saunders, 1965. Sherif, M- and Hovland, C. I. (Eds.), Social judgment. New Haven, Connecticut: Yale UniverSity Press, 1961. Sturmthal, A. White-collar trade unions. Urbana: University of Illinois Press, 1966. Uphoff, W. H. and Dunnette, M. D. (Understanding the union member-. Minneapolis: University of Minnesota Press, 1956. Waly, P- and Cook, 3. W.. Effect of attitude.on judgments of plausibility. .Journal of Personality and Social Psyghology,.l965, 2, 745-749. Waly, Patricia.and Cook, S. W.. Attitude as a determinant of.learning and.memory: “A failure tO confirm. .Journal of Personality and Social Psychology, 1966, 4, 280-288. Watts,.W. A. and McGuire, W. J...Persistence of induced Opinion change and retention Of inducing message content. -JOurnal of Abnormal and Social Psychology, .1964, 68, 233-41. Weiss, W. A. »A “sleeper" effect in Opinion change. Journal .of Abnormal and Social Psychology, 1953, 48, 173-180. Zimbardo, P- G- Involvement and communication discrepancy as determinants Of Opinion conformity. Journal Of .Abnormal and Social Psychology, 1960, 60, 86-94.' APPENDIX A Audio transcript of the messages PRO PRESENTATION Good day. My name is Ken Marlow and I am an undergradu- ate sociology student. I have 'been asked for the purpose of this experiment to present both my own beliefs and supportive factual information. I Obtained my data from sources such as Government Labor Statistics, several books on 1a- bor unions in relation to con- temporary society and current newspaper and magazine articles. I will say at the onset, that I am very much in favor Of labor unions. Some of the reasons for the view that I hold are these: Of the approximately 80 million people over the age of 16 who are presently employed in the United States about 20 million are union members. I consider this to be a significant number. Unions provide their members with increased security, self-esteem, better working conditions & wages. Unions have been instrumental in the introduction of safety features in industry and most dramatically in the area Of better wages. Unions have not only made work more endurable for their members, but have simultaneously caused non~ union industries to keep pace and thus raised the standard Of living for the nation as a whole. For instance, in manufacturing which accounts for about 45%.Of the labor force, wagesehave through negotiations with unions, increased by approximately 40% in the last 10 years. These increases have been hard won & in fact were necessary to keep up with the cost of living in this period. CON PRESENTATION Good day. My name is Ken Marlow and I am an undergradu- ate sociology student. I have been asked for the purpose of this experiment to present both my own beliefs and supportive factual information. I Obtained my data from sources such as Government Labor Statistics, several books on labor unions in relation to con- temporary society and current newspaper and magazine articles. I will say at the onset, that I am very much opposed to labor unions. Some Of the reasons for the View that I hold are these: Of the approximately 80 miilion people who are over the age Of 16 presently employed in the United States only 20 million of these are union members. Unions in fact determine not only whether their members shall work but also dictate whether others can work. Thus the . unions are interfering with the right to work Of the non-union workers. Their Often unrealistic demand for wages and benefits have dis- placed workers from jobs & plac- ed heavy burdens On particular industries and on the economy in general. For instance, in manufacturing, which accounts for about 45% of the labor force, wages not even including fringe benefits have in the last 10 years increased by an unprecedented 40%. These increases are responsible for a substantial amount of the cur- rent inflation. Of those companies which for the first time last year, cooperated ‘with a union in drawing up a labor contract, 100% of these contracts included an increase in wages, and thus an increase in the standard of living of its members. Negotiations over the conference table promoted by the union have caused increases in the standard of living. Negotiations through the collective bargaining processes have resulted in these increases ‘without increasing the number of strikes. Although in the last 10 years there has been an increase in the total number Of unions and em- ployers, the number of strikes has not increased but has remained constant. Union leaders are necessary in or- der tO represent their members to the college-educated professionals of industry. ‘We can attribute the increased bene- fits tO the dedication of union officials to the betterment Of their Inembers. It is interesting to note that the president of the United Steel WOrkers Union receives for his work approximately fifty thou- sand dollars, which is less than 1/5 of the wages Of the president of U.S. Steel. The misconception concerning union corruption can readily be dispelled *when one considers the fact that only four union Officials have been prosecuted for illegal use of funds over the last ten years. Whereas in industry in the last year alone over 6,000 non-union industry Offi- cials were convicted for embezzle- ment. Of those companies which for the first time last year signed union contracts, 100%.were required to grant their employees wage in- creases, regardless Of the fin- ancial situation of the company. Over the last 10 years there has been a two-fold increase in the number of workers involved in strikes, not to mention the yast number of people affected out- side of the particular company struck. Last year alone, thirty million man-production days were lost to strikes. While, in principle, collective bargaining seems desirable, in practice, through threatening to strike, the union has developed dispro- portionate power. Union leaders generally lack the education necessary to under- stand the complexities Of indus- try and the economics Of our society and are, therefore, realistic in their demands. un- Not only do union leaders indis- criminately seek higher wages without regard to the situation of the industry, but they them- selves receive large sums of money from the workers in ex- change for their services. The president of the United Steel Workers Union receives more than fifty thousand dollars a year. During the last ten years, no less than four union Officials have been prosecuted for ille— gal use Of funds. While the union may be the first to suffer the effects of criminal activia: ties conducted by union offi-ii: cials, in the long run it is the public that pays for this racke- teering. Through the structure Of union ac- tivities, implemented by union leaders, the union is an extension of the democratic process into the industrial community. On their own time, approximately 60%.of the members attend the important meet- ings in order to discuss and develop union policy. In reference to pressing social issues, union policy has been ex— ceptionally progressive and humani- tarian. In the area Of public housing for impoverished families, several labor unions have sponsored about seventy non-profit housing projects. The labor movement is a major force affecting the broad range of welfare and social legis- 1ation in the federal and state legislature. In conclusion, I believe that the facts and conclusions presented here are sufficient to justify a favorable view toward labor unions. Union policy and domineering union leaders tend to discour- age a large portion of the mem- bers from attending the union meetings. Within the 60% Of the union membership that does attend the most important meetings, dissention and discussion are discouraged. Union policy has ignored the pressing social issues although they are in a key position to produce social change. In the area of minority group relations most, if not all, of the actions taken by the union in the direc- tion of lessening discrimination have been due to federal inter- vention. The union's attitudes are reflected in the fact that the AFL—CIO constitution carries no sanctions against unions using discriminatory practices. In conclusion, I believe that the facts and conclusions pre- sented here are sufficient to justify an unfavorable View to- ward labor unions. APPENDIX B Questionnaire booklet: first session SESSION 1 Student Number Phone Number Notes: 0 indicates the correct response PRO indicates an item used in retention test-PRO group CON indicates an item used in retention test-CON group + indicates a favorable item - indicates an unfavorable item INFO and FEE-RET Listed below are 45 factual items about labor unions. Please choose the correct answer for each item and circle the number corresponding to that answer. Please answer all the items. 1. 2. PRO CON 4. PRO CON The most significant factor contributing to the rise of unionism during the 1930's was: 1. 2. (9 4. 5. return of prosperity role Of the Communists sympathetic Federal administration sympathetic management increase in women membership The number of people employed in the U.S. is approximately: 1. 20 million or less 2. 40 million 3. 60 million 80 million 5. 100 million or more The percent of the working force which is unemployed is approximately: 1. 2% or less (9 3-6% 5. 16% or more The number of union members in the U.S. is approximately: 1. 5 million or less @ 20 ‘million 3. 50 million 4. 80 million 5. 110 million or more The "company town": 1. gave basic security to the worker without employer interference 2. provided real savings at the company store 3. made available jobs become dependent on demonstrated membership in the union decreased labor mobility operated on a union, closed shop basis -2- 6. unions have most dramatically influenced: P30 1. 2. 3. @ 5 0 job security self esteem safety conditions wages fringe benefits 7. A group of people generally excluded from membership in industrial unions are: (:3 managers 2. unskilled workers 3. semi-skilled workers 4. custodial workers 5. guards 8. Unions determine whether their members should work and also CON whether: ' 1. production should change 2. supervision should change 3. a person can become a member 4 extra holidays should be introduced 6? others can work 9. If one were tO attempt tO select the most important factor affecting the success or failure of unions in the past 69 years, one would point to: WNH 00 U1 0 economic.setting Socialism the use of lockouts role of Government technological change 10. In the last 10 years the number of workers involved in strikes has: CON Q 22‘ 3. 4. 5. doubled increased slightly remained constant decreased'slightly decreased by half 11. 12. CCUJ 13. 14. PRO CON’ 15. 16. PRO CON -3- A federal law considered anti—union is: H O Uib‘» o o The The The The The Civil Rights Act of 1964 Taft-Hartley Act McClellan Act Fair Labor Standards Act Nixon Labor Bill Unions have: m.b63poka . O O (*3 *3 5 m UlbbJN 0.. Approximately what percent Of the labor force decreased increased increased increased increased overtime the number of strikes the burden on the economy socialiunrest the power of their leaders AFL was based on: exclusive craft jurisdiction radical leadership company unions union shops social revolution in manufacturing? NH .0 “9(9) 0 Which union most recently broke away from the AFL-CIO? f—l O ('3 ‘11th o The salary Of the president of the United Steel Workers 25% 35% 45% 55% 65% AFL UAW or less or more Teamsters United Steel Workers American Federation of Teachers Union is approximately: 1. (9 3. 4. 5 $30. $50. $70. $90. 000 or less 000 000 000 $110,000 or more is employed -4- l7. Unions have generally not supported: 1. social security 2. progressive income tax 3. minimum wage @ open shops 5. unemployment compensation 18. The number of union officials prosecuted for illegal use PR of funds in the last 10 years was approximately: CON 1. 1 (Z) 4 3. 9 4. 15 5. 24 19. The President of the UAW is: 1. George Meany (2) Walter Reuther 3. James Hoffa 4. I. W. Abel 5. John L. Lewis 20. The number of white collar officials convicted for embezzlement in industry last year was approximately: PRO 1. 6 2. 6O 3. 600 6,000 5. 60,000 21. President of the AFL—CIO is: E) E?» (0 George Meany Walter Reuther James Hoffa I. W. Abel John L. Lewis UQWN no. 22. Approximately what percent of the union members attend important union meetings? ' PRO CO“ 1. 10% or less 2. 20% 3. 40% 60% 5. 80% or more 23. 24. PRO 25. 26. CCHJ 27. 28. IPRO COKI -5, The lowest level union official is: H . bargaining agent foreman shop stewart . local president . national negotiator QB“ 01-h Approximately how many housing projects have been sponsored by labor unions? 1. none 2. 20 (3) 70 4. 130 5. 210 Under fifiesent law the use of strikes are illegal for: 1. farm workers 2. transportation workers 6) government employees 4. hospital employees 5. communication workers In reference to discrimination the AFL-CIO has: 1. discriminated against women 2. expell discriminating unions 3. a vested interest in discrimination been involved in legal suits no constitutional sanctions Pensions and seniority have: Q3 discouraged job change increased mobility reduced "job consciousness” increased discontent discouraged collective bargaining U10wa .0. In manufacturing over the last 10 years wages have increased by approximately: 1. 20% or less 2. 30% 3 40% 4. 50% 5 60% or more -6- 29. In recent times unions have been strongly opposed to: Q Communism 2. Socialism 3. democracy 4. a welfare state 5. collectivism 30. Of those companies which for the first time last year PRO signed union contracts approximately what percent raised CON‘ the wages of their employees? 1. 10% or less 2. 25% 3. 50% 4. 75% (ED 90% or more 31. The new type of unemploymentghas struck its most telling blow at workers who are: Q unskilled 2. skilled 3. technicians 4. professional people 5. women 32. In the last 10 years the number of strikes has: P30 1 doubled 2. increased slightly C:) remained constant 4 decreased slightly 5 decreased by half 33. Labor's share of the national income has been in recent years: slowly decreased kept the same . decreased through union action 4. determined by legal regulations CE; gradually increased 0 WNH o 34. Overtime pay attempts to accomplish all but one of the following: 1. force regularization of hours 2. create more jobs 3. increase pay of senior union members 4. increase number of employees employed increase safety standards 35. PBC’ 36. 37. 38. 39. -7- Union leaders are necessary in order to: IthH 0 determine union policy raise funds . negotiate with competing unions determine fair wages and standards for employers represent their members to the college-educated professionals of industry A "wild cat" strike is a strike called by: (”NH 4 63 company managers non-union workers competing industries a national union a local union Unions have in recent years demonstrated themselves to be politically and economically: l. revolutionary . conservative . socialistic 4. overly democratic 5. Communistic Which of the following has been used by unions against companies? 1. lockout 2. strikebreakers (:) featherbedding 4. company unions What 2 major non AFL—C10 unions have recently joined tOgether to form an association? @ U'Itb-WN o UAW - Teamsters Teamsters - Steel workers Steel Workers - UAW Steel Workers - Electrical Workers Electrical Workers — Building Trades 40. 41. CON 43. 44. CON 45. -8— The early labor parties failed primarily because: 1. intellectuals were kept out 2. members didn’t understand what was involved (:) intellectuals splintered the organizations 4. they pushed for too little reform 5. they were un—American Last year the number of man-production days lost to strikes was approximately: 1. 1 million or less 2 10 million (:D 30 million 4. 50 million 5. 100 million or more The President of the United Steel Workers Union makes what fraction of the salary of the President of U.S. Steel? 1. 1/2 2. 1/3 3. 1/4 l/5 5. 1/6 National Labor unions have generally come out in support of the: l. Socialists (:) Democrats 3. Republicans 4. Independents 5. Radicals In reference to industry and society union leaders: Ci) lack the education needed to understand the complexities 2. are out of touch with their members 3. support conservative political candidates 4. distort industries motives 5. deny their responsibility The present Secretary of Labor is: 1. David Howe 2. John James (:) George Shultz Winton Blount Walter Goodman U1 45 o CONT Below are some questions concerning the amount of contact you have had with labor unions and union related material. Please circle the response (under each question) which most nearly expresses your answer to the question. Is or was anyone in your immediate family (a relative you have lived with) been a member of a labor union? 1. no 2. yes .Are any of your friends members of labor unions? 1. no 2. yes Have you ever worked in a unionized company? 1. no 2. yes ‘Nere you (or are you now) a member of a labor union? 1. no 2. yes ZHave you discussed labor unions with a labor union member ‘within the last year? 1. no 2. yes .Have you discussed labor unions with a non-union member within the last year? (roomate, etc.) 1. no 2. yes Have you read any books in the last year which contained sections dealing with labor unions? 1. no 2. yes Have you had any courses in college (or in the last year) that were devoted in part or completely to the study of labor unions? 1. no 2. yes Do you think that you pay more attention to news concerning labor unions than other students? 1. no 2. yes A 1 Listed below are some statements about labor unions. Please circle the response (under each question) which most nearly expresses your answer to the question. Respond by using the following scale: strongly disagree disagree neither agree nor disagree agree strongly agree U'IDWNH o + Unions represent the best ideal in modern life. 1 2 3 4 5 - The high wages demanded by unions reduce the chances for employment. 1 2 3 4 5 + tnions exert a strong influence for good government. 1 2 3 4 5 - thions are decreasing in value to society. 1 2 3 4 5 *‘tkuons should have more power in their negotiations with management. 1 2 3 4 5 - Jnions rules often interfere with the efficient running of the company. 1 2 3 4 5 - Unions thrive on hatred and greed. 1 2 3 4 5 - Farm laborers should not be permitted to organize because of the small amount of money farm owners are making. 1 2‘ 3 4 5 - Strikes or slowdowns by sanitation workers are a violation of the rights of the community. 1 2 3 4 5 + Unions are necessary for a democratic society. 1 2 3 4 5 Listed below are some statements about labor unions. Please circle the response (under each question) which most nearly expresses your answer to the question. Respond by using the following scale: 1. strongly disagree 2. disagree 3. neither agree nor disagree 4. agree 5 strongly agree The federal government should regulate unions more closely. l 2 3 4 5 Unions are detrimental to our society. 1 2 3 4 5 Unions adjust themselves to changing times and conditions. 1 2 3 4 5 The concept of unions appeal to man's highest nature. 1 2 3 4 5 Strikes or slowdowns by firemen are a violation of the rights of the community. 1 2 3 4 5 A white collar worker who joins a union is lowering himself to the level of a factory worker. l 2 3 4 5 Unions are too big to be of service to their employees. 1 2 3 4 5 Unions give meaningfulness to the lives of many. 1 2 3 4 5 Many present union activities should be outlawed. l 2 3 4 5 Union members generally have an important voice in union affairs. 1 2 3 4 5 BI (continued) .Listed below are some statements about labor unions. Please circle the response (under each question) which most nearly expresses your answer to the question. Respond by using the following scale: . strongly disagree . disagree . neither agree nor disagree . agree . strongly agree U'lowaH Unions should try to become powerful enough to help decide how plants should be managed. 1 2 3 4 5 Unions are too radical in their views and actions. 1 2 3 4 5 Unions are necessary to our society as it is organized. l 2 3 4 5 Unions actually hinder the people they are supposed to help. 1 2 3 4 5 There is too much control of unions by state and federal laws. 1 2 3 4 5 Unions are interested in gaining power without responsibility. 1 2 3 4 5 Unions are more concerned with their own welfare than the welfare of society. 1 2 3 4 5 Strikes or slowdowns by police are a violation of the rights of the community. 1 2'34 5 Union leaders reflect the opinions of the rank and file union members. 1 2 3 4 5 Supply and demand does a better job of dictating wages and hours than unions. 1 2 3 4 5 Listed below are 5 questions about you. Please answer each question by circling the appropriate response. 1. What is your sex? 1. male 2. femal 2. In what college is your major area of study? Business Education Socia 1. 2. 3. Natur 4. 5. Other e al Science 1 Science 3. What is your class level? . Fresh man . Sophmore . Senio . Other 1 2 3. Junior 4 5 r 4. What is your G.P.A.? O l 2 O 2 2 l — 3. 2.6 — 4 3 l — 5 3 5 O to other 1. more 2. more 3. more 4. more 5. more r below 2.5 3.0 3.5 r above do you think you know about labor unions relative students? than 90% (much more) than 75% than 50% than 25% than 10% (much less) APPENDIX C Questionnaire booklet: second session - PRO group S E S S I O N 2 F O R M A Student Number Notes: 0 indicates the correct response + indicates a favorable item or trait word - indicates an unfavorable item or trait word A 2 Listed below are some statements about labor unions. Please circle the response (under each question) which most nearly expresses your answer to the question. Respond by using the following scale: strongly disagree disagree neither agree nor disagree agree strongly agree UTv-bWNl-J 0 Unions represent the best ideal in modern life. 1 2 3 4 5 The high wages demanded by unions reduce the chances for employment. 1 2 3 4 5 Unions exert a strong influence for good government. 1 2 3 4 5 Unions are decreasing in value to society. 1 2 3 4 5 Unions should have more power in their negotiations with management. 1 2 3 4 5 Unions rules often interfere with the efficient running of the company. 1 2 3 4 5 Unions thrive on hatred and greed. 1 2 3 4 5 Farm laborers should not be permitted to organize because of the small amount of money farm owners are making. 1 2 3 4 5 Strikes or slowdowns by sanitation workers are a violation of the rights of the community. 1 2 3 4 5 Unions are necessary for a democratic society. 1 2 3 4 5 STOP HERE DO NOT CONTINUE RET Presented below are 14 statements or questions which were discussec on the video tape. Please circle that answer which was presented by the speaker on the video tape. 1. The number of people employed in the U.S. is approximately: 1. 20 million or less 2. 40 million 3. 60 million 80 million 5. 100 million or more 2. The number of union members in the U.S. is approximately: 1. 5 million or less @ 20 million 50 million 80 million 110 million or more U'I-bw 3. Unions have most dramatically influenced: 1. job security 2. self esteem 3. safety conditions wages 5. fringe benefits 4. Approximately what percent of the labor force is employed in manufacturing? 1. 25% or less ‘ 2. 35% ® 45% 55% UI J> o 65% or more 5. In manufacturing over the last 10 years wages have increased by approximately: 1. 20% or less 2. 30% 40% . 50% 5. 60% or more 10. Of those companies which for the first time last year signed union contracts approximately what percent raised the wages of their employees? ' H N be; 5. 10% or less 25% 50% 75% 90% or more In the last 10 years the number of strikes has: 1. 2. <9 4. 5. doubled increased slightly remained constant decreased slightly decreased by half Union leaders are necessary in order to: UIQWNH ”coco determine union policy raise funds negotiate with competing unions determine fair wages and standards for employers represent their members to the college-educated pro— fessionals of industry The salary of the president of the United Steel Workers Union is approximately: H 0 <9 01.5w $30,000 or less $50,000 $70,000 $90,000 $110,000 or more The President of the United Steel Workers Union makes what fraction of the salary of the President of U.S. Steel? 1. 2. 3. @ 5. 1/2 1/3 1/4 1/5 1/6 11. 12. 13. 14. -3- The number-of.uni0n.officials prosecuted for illegal usefi of funds in the last 10 years was approximately: 1. 1 (Z) 4 3. 9 4. 15 5. 24 . w .- The number of white collar Officials convicted for embezZlement in industry last year was approximately: 1. 6 2. 60 3. 600 6,000 5. 60,000 Approximately what percent of the union members attend important union meetings? 1. 10% or less 2. 20% 3. 40% 60% 5. 80% or more Approximately how many housing projects have been sponsored by labor unionS? 1. none 2. 20 Q) 70 4. 130 5. 210 Listed below are some statements about labor unions. Please circle the response (under each question) which most nearly expresses your answer to the question. Respond by using the following scale: 1. strongly disagree 2. disagree 3. neither agree nor disagree 4. agree 5 . strongly agree -The federal government should regulate unions more closely. l 2 3 4 5 -Unions are detrimental to our society. 1 2 3 4 5 i'Unions adjust themselves to changing times and conditions. 1 2 3 4 5 "*The concept of unions appeal to man's highest nature. 1 2 3 4 5 -Strikes or slowdowns by firemen are a violation of the rights of the community. 1 2 3 4 5 -A white collar worker who joins a union is lowering himself to the level of a factory worker. l 2 3 4 5 "Unions are too big to be of service to their employees. 1 2 3 4 5 +Unions give meaningfulness to the lives of many. 1 2 3 4 5 .. Many present union activities should be outlawed. l 2 3 4 5 "*Union members generally have an important voice in union affairs. 1 2 3 4 5 B2 (continued) Listed below are some statements about labor unions. Please circle the response (under each question) which most nearly expresses your answer to the question. Respond by using the following scale: strongly disagree disagree neither agree nor disagree . agree . strongly agree mwal-I 9 Unions should try to become powerful enough to help decide how plants should be managed. l 2 3 4 5 Unions are too radical in their views and actions. 1 2 3 4 5 Unions are necessary to our society as it is organized. l 2 3 4 5 Unions actually hinder the people they are supposed to help. 1 2 3 4 5 There is too much control of unions by state and federal laws. 1 2 3 4 5 Unions are interested in gaining power without responsibility. 1 2 3 4 5 Unions are more concerned with their own welfare than the welfare of society. 1 2 3 4 5 Strikes or slowdowns by police are a violation of the rights of the community. 1 2 3 4 5 Union leaders reflect the opinions of the rank and file union members. 1 2 3 4 5 Supply and demand does a better job of dictating wages and hours than unions. 1 2 3 4 5 EVAL Presented below are some characteristics that can be used in describing people. Indicate the extent to which you believe that each characteristic describes or applies to the speaker on the video tape by using the following scale: not at all slightly somewhat very extremely U'lprNH 0 Please circle yum answer under each characteristic. + 1. interesting - 10. insincere l 2 3 4 5 l 2 3 4 5 - 2. narrow-minded - ll. inexperienced l 2 3 4 5 l 2 3 4 5 + 3. intelligent + 12. bright l 2 3 4 5 l 2 3 4 5 - 4. annoying + 13. capable l 2 3 4 5 l 2 3 4 5 + 5. honest ‘ l4. obnoxious 1 2 3 4 5 1 2 3 4 5 + 6. well-read + 15. entertaining l 2 3 4 5 1 2 3 4 5 - 7. incompetent ' 16. phony l 2 3 4 5 l 2 3 4 5 + 8. alert + 17. level headed l 2 3 4 5 1 2 3 4 5 ‘ 9. irritating ' 18. uninspiring l 2 3 4 5 l 2 3 4 5 Listed below are a series of statements about the Video tape. Please circle the appropriate response. If you were to compose a message expressing your attitudes towards labor unions, how would its tone compare with the message presented on the video tape? 1. negative 3. about the same 2. less positive 4. more positive Respond to the following statements in a manner which would be con- sistent with the content or tone of the video tape message. Do not respond by the way ygu feel about these statements, but respond in a manner consistent with the video tape message by using the following scale: 1. strongly disagree 3. neither agree nor disagree 2. disagree 4. agree 5. strongly agree 'Unions represent the best ideal in modern life. 1 2 3 4 5 The high wages demanded by unions reduce the Chances for employment. 1 2 3 4 5 Unions exert a strong influence for good government 1 2 3 4 5 Unions are decreasing in value to soCiety. l 2 3 4 5 Unions should have more power in their negotiations with management. 1 2 3 4 5 Unions rules often interfere with the effiCient running of the company. 1 2 3 4 5 'Unions thrive on hatred and greed. l 2 3 4 5 .Farm laborers should not be permitted to organize because of the small :amount of money farm owners are making. 1 2 3 4 5 £3trikes or slowdowns by sanitation workers are a violation of the .rights of the community. 1 2 3 4 5 Unions are necessary for a democratic society. 1 2 3 4 5 CONT Below are some questions concerning the amount of contact you have had with labor unions and union related material. Please circle the response (under each question) which most nearly expresses your answer to the question. Is or was anyone in your immediate family (a relative you have lived with) been a member of a labor union? 1. no 2. yes Are any of your friends members of labor unions? 1. no 2. yes Have you ever worked in a unionized company? 1. no 2. yes Were you (or are you now) a member of a labor union? 1. no 2. yes Have you discussed labor unions with a labor union member within the last year? 1. no 2. yes Have you discussed labor unions with a non-union member within the last year? (roomate, etc.) 1. no 2. yes Have you read any books in the last year which contained sections dealing with labor unions? 1. no 2. yes Have you had any courses in college (or in the last year) that were devoted in part or completely to the study of labor unions? 1. no 2. yes Do you think that you pay more attention to news concerning labor unions than other students? 1. no 2. yes APPENDIX D Questionnaire booklet: second session - CON group S E S S I O N 2 F O R M B Student Number Notes: 0 indicates the correct response + indicates a favorable item or trait word - indicates an unfavorable item or trait word A 2 Listed below are some statements about labor unions. Please circle the response (under each question) which most nearly expresses your answer to the question. Respond by using the following scale: strongly disagree disagree neither agree nor disagree agree strongly agree U'II-bUJNl" 0 Unions represent the best ideal in modern life. 1 2 3 4 5 The high wages demanded by unions reduce the chances for employment. 1 2 3 4 5 Unions exert a strong influence for good government. 1 2 3 4 5 Unions are decreasing in value to society. 1 2 3 4 5 Unions should have more power in their negotiations with management. 1 2 3 4 5 Unions rules often interfere with the efficient running of the company. 1 2 3 4 5 Unions thrive on hatred and greed. l 2 3 4 5 Farm laborers should not be permitted to organize because of the small amount of money farm owners are making. 1 2 3 4 5 Strikes or slowdowns by sanitation workers are a violation of the rights of the community. 1 2 3 4 5 Unions are necessary for a democratic society. 1 2 3 4 5 STOP HERE NOT CONTINUE Presented below are 14 statements or questions which were discussed on the v was presented by t l. The number of l. 20 million 2. 40 million 3. 60 million 80 million 5. 100 millio 2. The number of union members in the U.S. is approximately: 1. 5 million (a?) 20 million . 50 million 80 million . 110 millio m b I 3. Unions determine whether their members should work and also whether: 1. production 2. supervisio 3. a person c 4. extra holi others can 4. Unions have: H 0 decreased increased increased increased increased N a m;bg§ '5. Approximately what percent of the labor force is employed in manufacturi l. 25% or les 2. 35% @ 45% 55% m b 65% BET ideo tape. Please circle that answer which he speaker on the video tape. people employed in the U.S. is approximately: or less n or more or less n or more should change n should change an become a member days should be introduced work overtime the number of strikes the burden on the economy social unrest the power of their leaders ng? S -2- 6. In manufacturing over the last 10 years wages have increased by approximately: 1. 20% or less 2. 30% @ 40% 4. 50% 5. 60% or more 7. Of those companies which for the first time last year signed union contracts approximately what percent raised the wages of their employees? 1. 10% or less 2. 25% 3. 50% 4 75% 5 90% or more 1M 8. ever the last 10 years the number of workers involved in strikes has: (i) doubled' 2. increased slightly 3 remained constant 4. decreased slightly 5. decreased by half 9. Last year the number of man-production days lost to strikes was approximately: 1. 1 million or less 2. 10 million C:) 30 million 4. 50 million 5. 100 million or more 10. In reference to industry and society union leaders:. <:> lack the education needed to understand the complexities are out of touch with their members support conservative political candidates distort-industry's motives deny their responsibility WUPWN coo #34 11. The salary of the president of the United Steel Workers Union is approximately: 1. $30,000 or less $50,000 3. $70,000 4. $90,000 5. $110,000 or more 12. The number of union officials prosecuted for illegal use of funds in the last 10 years was approximately: 1. l @ 4 3. 9 4. 15 5. 24 13. Approximately what percent of the union members attend important union meetings? 1. 10% or less 2. 20% 3. 40% 60% 5. 80% or more 14. In reference to discrimination the AFL-CIO has: discriminated against women expell discriminating unions a vested interest in discrimination been involved in legal suits no constitutional sanctions 45.me 0. + Listed below are some statements about labor unions. Please circle the response (under each question) which most nearly expresses your answer to the question. Respond by using the following scale: strongly disagree disagree neither agree nor disagree agree strongly agree UIJ>LUNH O The federal government should regulate unions more closely. l 2 3 4 5 Unions are detrimental to our society. 1 2 3 4 5 Unions adjust themselves to changing times and conditions. 1 2 3 4 5 The concept of unions appeal to man's highest nature. 1 2 3 4 5 Strikes or slowdowns by firemen are a violation of the rights of the community. 1 2 3 4 5 A white collar worker who joins a union is lowering himself to the level of a factory worker. l 2 3 4 5 Unions are too big to be of service to their employees. 1 2 3 4 5 Unions give meaningfulness to the lives of many. 1 2 3 4 5 Many present union activities should be outlawed. l 2 3 4 5 Union members generally have an important voice in union affairs. 1 2 3 4 5 B2 (continued) Listed below are some statements about labor unions. Please circle the response (under each question) which most nearly expresses your answer to the question. Respond by using the following scale: 1. strongly disagree 2. disagree 3. neither agree nor disagree 4. agree 5 . strongly agree Unions should try to become powerful enough to help decide how plants should be managed. 1 2 3 4 5 Unions are too radical in their views and actions. 1 2 3 4 5 Unions are necessary to our society as it is organized. l 2 3 4 5 Unions actually hinder the people they are supposed to help. 1 2 3 4 5 There is too much control of unions by state and federal laws. 1 2 3 4 5 Unions are interested in gaining power without responsibility. 1 2 3 4 5 Unions are more concerned with their own welfare than the welfare of society. 1 2 3 4 5 Strikes or slowdowns by police are a violation of the rights of the community. 1 2 3 4 5 Union leaders reflect the Opinions of the rank and file union members. 1 2 3 4 5 Supply and demand does a better job of dictating wages and hours than unions. 1 2 3 4 5 EVAL Presented below are some characteristics that can be used in describing people. Indicate the extent to which you believe that each characteristic describes or applies to the speaker on the video tape by using the following scale: not at all slightly somewhat very extremely WIhUONH 0 Please circle ymm'answer under each characteristic. + l. interesting - 10. insincere l 2 3 4 5 l 2 3 4 5 7 2. narrow-minded - ll. inexperienced l 2 3 4 5 l 2 3 4 5 + 3. intelligent + 12. bright l 2 3 4 5 l 2 3 4 5 - 4. annoying + 13. capable l 2 3 4 5 l 2 3 4 5 '+ 5. honest - l4. obnoxious l 2 3 4 5 l 2 3 4 5 + 6. well-read + 15. entertaining 1 2 3 4 5 l 2 3 4 5 ' 7. incompetent - l6. phony l 2 3 4 5 l 2 3 4 5 +8. alert +-l7. level headed l 2 3 4 5 l 2 3 4 5 ' 9. irritating ‘ 18. uninspiring l 2 3 4 5 l 2 3 4 5 ll Listed below are a series of statements about the video tape. Please circle the appropriate response. If you were to compose a message expressing your attitudes towards labor unions, how would its tone compare with the message presented on the video tape? 1. more negative 3. less negative 2. about the same 4. positive Respond to the following statements in a manner which would be con— sistent with the content or tone of the video tape message. Do not respond by the way ygg feel about these statements, but respond in a manner consistent with the video tape message by using the following scale: 1. strongly disagree 3. neither agree nor disagree 2. disagree 4. agree 5. strongly agree 'Unions represent the best ideal in modern life. 1 2 3 4 5 -The high wages demanded by unions reduce the chances for employment. 1 2 3 4 5 'Unions exert a strong influence for good government. 1 2 3 4 5 Unions are decreasing in value to society. 1 2 3 4 5 Unions should have more power in their negotiations with management. 1 2 3 4 5 thiions rules often interfere with the efficient running of the company. 1 2 3 4 5 thiions thrive on hatred and greed. l 2 3 4 5 JFarnIlaborers should not be permitted to organize because of the snnall amount of money farm owners are making. 1 2 3 4 5 :itrikes or slowdowns by sanitation workers are a violation of the rights of the community. 1 2 3 4 5 [unions are necessary for a democratic society. 1 2 3 4 5 CONT Below are some questions concerning the amount of contact you have had with labor unions and union related material. Please circle the response (under each question) which most nearly expresses your answer to the question. Is or was anyone in your immediate family (a relative you have lived with) been a member of a labor union? 1. no 2. yes Are any of your friends members of labor unions? 1. no 2. yes Have you ever worked in a unionized company? 1. no 2. yes Were you (or are you now) a member of a labor union? 1. no 2. yes Have you discussed labor unions with a labor union member within the last year? 1. no 2. yes Have you discussed labor unions with a non-union member within the last year? (roomate, etc.) 1. no 2. yes Have you read any books in the last year which contained sections dealing with labor unions? 1. no 2. yes Have you had any courses in college (or in the last year) that were devoted in part or completely to the study of labor unions? 1. no 2. yes Do you think that you pay more attention to news concerning labor unions than other students? 1. no 2. yes WICHIGQN STATE UNIV. LIBRARIES WWWWIWIWIUIUIHHIWWWHI 31293101413478