THE RELATIVE IMPORTANCE OF VARIABLES RELATEDZTO ' PARTICIPATION IN'UNI‘ON ACTIVITIES ‘ Dissertation for the Degree of. Ph. D.- ’ - ‘ MICHIGAN STATE UNIVERSITY ' GREGORY EDWARD 'HUSZCZO 1975 m-m.a~m'-m“ 5‘" J ' El! I‘ 3";- ‘ ' ' . IIIIIIIIIIIIIIIIIIIIII _ . . a"; This is to certify that the thesis entitled The Relative Importance of Variables Related To Participation in Union Activities presented by Gregory Edward Huszczo has been accepted towards fulfillment of the requirements for Ph.D. degree in Psychology / 7 Major professor Date WW?" 0-7639 JAN 9 8.12501 AP a I I} ABSTRACT THE RELATIVE IMPORTANCE OF VARIABLES RELATED TO PARTICIPATION IN UNION ACTIVITIES BY Gregory Edward Huszczo Research literature on unions appearing in the psychological journals is scarce, especially over the last 15 years. This dissertation overviews the literature of five subtopics concerning unions: union leadership, treatment of minority groups within unions, labor's involvement with politics, union—management relationships and attitudes of union members. A thorough review of a sixth subtopic, factors related to levels of participation in union activities is presented in table form. A questionnaire containing a broad range of 166 items was administered to 500 union members. A fourteen-item index of parti- cipation of high internal consistency was established. Two separate factor analyses of the remaining sections of the questionnaire produced 14 factors--six measuring attitudes toward various aspects of unions and eight measuring job attitudes, outlooks on life in general, and socio-political attitudes and behaviors. Along with these 14 factors, nine demographic items and two measures of perceived control within local unions were entered into a multiple regression equation as predictor variables and the index of participation was used as the criterion variable. The small shrinkage of the multiple R after Gregory Edward Huszczo double cross-validation stands as evidence of the validity of the equation generated. Two of the indices' of importance described by Darlington [1968], the relationship (zero-order correlation with participation) and relative importance (usefulness index) of each of the predictor variables established the relative importance of variables related to participation in union activities. The results indicate that the active union member has a liberal political orientation, is active in community and political affairs, is very pro—union, believes the rank-and-file has a strong voice in the local union, and is satisfied withlrksor her job. The usefulness index indicates that community- political activities, liberal political beliefs, pro-union philosophy, high standards of involvement with unions and high job satisfaction are the five best predictors of union participation. The results suggest a possible trend. Active union members may view unions as vehicles in a socio-political movement above and beyond their economic and protective functions. Future research is needed for substantiation. This dissertation advances the area of union participation research by establishing a broadly defined, reliable measure of participation. It demonstrates the appropriateness of statistical procedures such as multiple regression and "indices of relative importance" in an area that typically has used less encompassing analyses and the usefulness of attitude factor scores in predicting union participation. The need for further scale development indicates that an important role in union research exists for psychologists. THE RELATIVE IMPORTANCE OF VARIABLES RELATED TO PARTICIPATION IN UNION ACTIVITIES BY Gregory Edward Huszczo A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1975 ACKNOWLEDGMENTS Without the help of some very important people, completion of this dissertation would have been impossible. “I want to take this opportunity to publicly thank some of those people. First and foremost, of all the people who have helped me, no one has helped me more than my parents. They have stuck by me in good times and bad; always giving me their love and support. I am very proud of you, Mom and Dad. I love you both. I wish to also thank the U.A.W. for its support, permission and assistance in this project. I want to note, in particular, the help of Mr. Donald Ephlin, administrative assistant to President Woodcock, and Ms Margaret Neldrett, assistant director of the Walter and May Reuther U.A.W. Family Education Center. The cooperation and openness I experienced with this union impressed me as well as convinced me that an interface of psychology and unions should exist. I wish also to thank my committee members who aided and guided me throughout the dissertation process. Dr. Jack Wakeley, my chairman, was extremely helpful. His editing skills, understanding of the area, encouragement, ability to motivate, openness, and almost immediate feedback along the way contributed greatly to the completion of this task and has re-opened my eyes to consideration of future research. Dr. Neal Schmitt aided me especially in the analysis and inter- pretation of this data set. His knowledge, concern, and ability to ii communicate helped me through some rough moments and guided me in many ways throughout the past year. Dr. Carl Frost offered me support and broader perspectives. I certainly benefitted personally and professionally from our associa- tion over the last few years. I thank you for your time, your openness, your ability to listen, and especially your caring. Dr. Cecil Williams encouraged and supported me throughout my college career. He opened my eyes to many new ideas and values. I thank you for years of help. I would like to give a special thanks to Bryan Coyle whose assistance in computer programming and analysis in general was crucial in completing this dissertation on time. Several undergraduates have worked on this and other research projects with me over the last year. Their help in coding, cleaning-up the data, and general interest and assistance was greatly appreciated. I thank you Elizabeth Bostick, John Comerford, Deborah Bliss, Diana Long, Jerry Schafer, Linda Piasecki, Chalie Mickens, and Ernie Betts. Through many drafts, the secretarial talents of several people rescued me in the face of deadlines and frustrations. I especially thank Rebecca Mulholland, Sharon White, Sue Weesner, Kathy Kohl, and Anna Toth for their assistance. Finally, but certainly not least, I would like to thank a special person--Sue Skarsgard. Thanks especially for helping me believe in myself throughout feelings of despair and thanks for sharing my moments of joy too. iii Chapter TABLE OF CONTENTS I INDUSTRIAL/ORGANIZATIONAL PSYCHOLOGY AND UNIONS . . II A LOOK AT THE LITERATURE . . . . . . . . . . . . . Union Leadership Studies . . . . . . . . . . . . Treatment of Minority Groups Within Unions . . . Union Involvement in Politics . . . . . . . . . Union-Management Relationships . . . . . . . . . Union Attitude Surveys . . . . . . . . . . . . . Research on Union Participation . . . . . . . . III METHODS AND PROCEDURES . . . . . . . . . . . . . . Subjects . . . . . . . . . . . . . . . . . . . . Questionnaire . . . . . . . . . . . . . . . . . Analyses . . . . . . . . . . . . . . . . . . . . IV RE S ULTS O O O O O O O O O O O O O O O C O O O . C C Development Of the Index of Participation . . . Factor Analysis of the Union Attitude Questions Factor Analysis of the Remaining Items . . . . . The Development and Results of the Multiple Regression Equation . . . . . . . . . . . . . . The Relative Importance of the Factors Involved in Union Participation . . . . . . . . . . . . . V DISCUSSION . . . . . . . . . . . . . . . . . . . . BIBLIOGRAPHY . An Index of Participation . . . . . . . . . . . Comparison with Previous Methodologies . . . . . Zero-Order Correlations with Participation . . . The Multiple Regression Equation . . . . . . . . The Relative Importance of the Independent Variables . . . . . . . . . . . . . . . . . . . The Relative Importance of Blocks of the Independent Variables . . . . . . . . . . . . . Summary and Conclusions . . . . . . . . . . . . iv Page 12 13 14 16 18 38 38 38 41 46 46 47 48 49 56 62 63 65 67 71 72 79 80 82 APPENDICES CD RESEARCH QUESTIONNAIRE . . . . . . . . . . . DESCRIPTION OF SAMPLE . . . . . . . . . . . RESULTS OF FACTOR ANALYSIS ON SECTION A OF QUESTIONNAIRE RESULTS OF FACTOR ANALYSIS ON SECTIONS C, D, OF QUESTIONNAIRE . . . . . . . . . . . . . . Page 89 97 99 102 Table 10 11 12 13 LIST OF TABLES Articles in Psychological Abstracts Dealing with Unions . Demographic Variables and Union Participation . . . . . . Personal Associations and Union Participation . . . . . . Personal Attitudes and Union Participation . . . . . . . Characteristics of Job and Union Participation . . . . . Attitudes Toward the Job and Union Participation . . . . Characteristics of Union and Union Participation . . . . Attitudes Toward Unions and Union Participation . . . . . Rationale for Including Variables in Regression Equation. Correlation Matrix of All Variables Used in Regression Equat ion C O O O O I O O O O O O O O O O O C O O O O O O Darlington's Three Measures of Importance on all 25 Independent Variables . . . . . . . . . . . . . . . . . . Darlington's Usefulness Index of Importance Applied to Blocks of Independent Variables . . . . . . . . . . . Number of Significant Correlations Between Individual Items and the Index of Participation . . . . . . . . . . vi Page 23 24 26 28 3O 32 34 51 S3 57 61 66 CHAPTER I INDUSTRIAL/ORGANIZATIONAL PSYCHOLOGY AND UNIONS Unions are organizations representing almost 20 million workers out of a work force of approximately 80 million American workers. Unions are not a "thing of the past"--union membership increased 4,300,000 during the years 1953-1970. Yet of all the graduate studies programs of Industrial/Organizational Psychology, virtually no program includes the study of unions as part of its curriculum. Furthermore, because of the literature, training, and operational practices of industrial/ organizational psychologists, some critics have denounced the whole field as being an adjunct to management. Baritz [1960], for example, states the industrial psychologist has sold out his birthright as a member of the science of psychology in favor of becoming a technical assistant to the practice of personnel management. A systemtatic review of Psychological Abstracts from 1927 through 1974 establishes just what psychologists have published concerning unions. Table 1 displays a frequency count of all articles listed under any topics dealing with Unions or Labor (e.g., Labor-Management Relations, Union Membership, etc.). Considering the thousands of articles psychologists have written concerning aspects of industrial life, it is clear that psychologists have, relatively speaking, ignored the study of unions. Furthermore, even though the total number of articles involving industrial psychology reported by Psychological Abstracts has increased dramatically in the last 20 years, the number 1 of articles concerning unions has significantly decreased during this same time period. Note should also be taken that the literature on unions from recent years focuses on unions' relationships with manage— ment rather than studying aspects of unions as entities in and of themselves. Tfiflel Articles in Psychological Abstracts Dealing with Unions Labor-Management Union-Management Years Union Labor Relations Relations 1927-29 9 1930-32 1 2 1933-35 1 1936-38 1 1939-41 12 1942-44 5 1945-47 2 39 1948-50 6 121 1951-53 1 82 1954-56 69 1957-59 21 30 1960-62 18 2 11 12 1963-65 8 11 13 1966-68 2 37 1969-71 1972-74 14 5 Is there a role for psychologists in labor unions? Peter Dubno thought so in an article he wrote for American Psychologist in 1957. He pointed out that psychologists could be helpful to labor unions in studying industrial unrest, union-management relations, members' attitudes, working conditions, the internal union organization, communication patterns within the union, members' participation (or lack thereof), and union political and educational activity. The arguments presented in this paper indicate that psychology has very much to offer trade unions in terms of specific skilled services. Unions, in turn, inasmuch as they con- tain a population of individuals upon whom relatively meagre research has been inaugurated, can contribute important empirical findings to the existing body of psychological knowledge. Each in its own constituted uniqueness has a great deal to offer the other. It is expected that, for psychology at least, new theoretical concepts in social and industrial psychology may be the happy outcome of the suggested "labor psychology" [p. 215]. This "happy outcome" never happened. The explanation of this is not a simple lack of interest in unions by psychologists; on their side unions have built up a distrust of psychologists. Barkin [1961] points out the roots to some of these problems in his article "Psychology as Seen by a Trade-Unionist." Traditionally, union representatives have been confronted by psychologists "primarily as a member of the manage- ment personnel and engineering team and so considerable suspicion lurks in the minds of the trade unionist about this discipline. Is it neutral or are its premises and techniques primarily designed to serve employers?" [p. 260]. Either as a direct or an "unwitting" agent of the employer, psychologists have provided much information about human characteristics and performance that has directly affected the workers. Even when psychologists' methods were objectively used and intentions pure, the interpretations of the results by management have in the past not always been objective. The results of research by behavioral scientists have often been used to combat unions, weaken support over particular issues, and to promote programs that later became obviously exploitive. Barkin points out that "Psychology has been used primarily in industry as a tool for differentiation among employees and jobs" [p. 260]. Thus, many instruments are not used to measure ability but rather are used to see who has the most ability. This he sees as an inappropriate criterion in most selection, placement, and promotion situations. Unions usually feel that in a promotion situation, for instance, whether an employee can perform the job rather than whether he is slightly better or worse than another person is a more appropriate criterion. Such a criterion would then be more in line with the seniority principle that unions have adhered to for the protection of workers. Furthermore, some of the instruments psychologists use have questionable validity thus economics more than anything else seem to dictate their use. This is especially true in some selection situations but also, as recently pointed out by Irving Bluestone, U.A.W. Vice- President, job satisfaction measures may also be questioned. He states that few workers have had the opportunity to compare work as it might be contrasted with their present job. Unless workers have the chance to experience alternative conditions of work in the performance of the same job, they cannot fairly compare, analyze or react. It seems to me that conclusions which are the result of an analysis of such comparative experiences would be more valid because they would be based on empirical data reactive to actual experience. Otherwise the measurement of 'job satisfaction' is really just a determination of how well the workers, in order to make a living, have adjusted themselves to the nature of the jobs to which they are assigned rather than a measure- ment of how satisfying the jobs are [B1uestone, 1974, p. 14]. Doubts as to the measurement tools psychologists use can be a stumbling block for the mutual acceptance of unions and psychologists, especially when psychologists dogmatically defend such instruments and dogmatically declare that "real reasons" for such problems as job dissatisfaction. Nat Weinberg (former Director of Special Projects for the U.A.W.) says, "I have spent my life in the labor movement but I don't have the certainties about what's on workers' minds that I see reflected in so much of the literature written by people who never saw the inside of a factory or a union hall" [Weinberg, 1974, p. 14]. Distrust of industrial psychologists also results from tieing industrial psychology with Taylor's "Scientific Management," time study, and the exploitive aspects of the Human Relations Movement in many union officials' minds. Harold L. Sheppard [1974] illustrated the distrust by asking, "Is 'Job Enrichment' only another gimmick--1ike the old 'human relations' movement in industry--to keep down wage demands, and to blunt or prevent unionism?"[p. 210]. Feinberg admits, "We may have been affected by the unions lack of clear differentiation between industrial psychology and Taylor's brand of scientific management" [Feinberg, 1961, p. 240]. The exploitive aspects of time study as performed by early human factors-industrial psychologists led the union to train their own staff to perform such functions and protect union members from resulting speed-ups. Thus, past methods and theories of industrial psychologists have contributed to present day distrust by unionists of this profession. Industrial psychologists who espouse Herzberg's theory as a basis of CD efforts (as is frequently the case in many job enrichment programs) provide more basis for doubt to the unionist contemplating working with behavioral scientists. Such doubt centers around Herzberg's discounting of so-called hygiene factors in relation to satisfaction. Union people are interested in the motivator factors (ego and higher order needs) but realize the importance of hygiene factors to their membership. Protecting its members often involves dealing with management on these bread and butter type factors. Distrust of psychologists and management increases when Herzberg's theory is dogmatically espoused. Further doubt arises when research fails to substantiate the validity of this theory. Traditionally industrial psychologists have been tied closely to management. After all, they were paid by management. Although most psychologists can remain objective scientists, suspicion of their relationship with management has been an issue with unionists for years. Industrial psychologists tend to have a social value orientation similar to management. As mentioned previously, unions are a negelected area of study by industrial psychologists. Textbooks on industrial psychology often ignore unions completely. The training of industrial psycholo- gists is management oriented. Writings by psychologists often reflect an attitude that unions are a bothersome interference rather than a deep-rooted social movement. Anti-union articles such as M. Scott Myers' [1971] "Overcoming Union Opposition to Job Enrichment" Show such a bias. Behavioral scientists that have consulted for unions have often oversold surveys, or exploited unions as a research data source without providing meaningful feedback. All of these factors add up to distrust of industrial psychologists by union officials today. Hewever, as Levenstein points out in his article [Levenstein, 1961], lawyers, doctors, economists, and other professionals have been dis- trusted by unionists in the past but are now perceived as helpful resources to the union movement. Although industrial psychologists have obstacles in their paths, they too can provide important inputs to union officials in decisions today in the very ways Dubno pointed out in 1957. The following chapters present a brief overview of five-subareas of research in which psychologists have shown interest in unions, a thorough review of the literature on the correlates of member partici- pation in unions, and an attempt to improve our knowledge of the relative importance of factors related to participation by members in union activities by using improvements in research methodology and by using a large, cross-sectional sample. CHAPTER II A LOOK AT THE LITERATURE A review of the literature helps to define the present interface between industrial psychology and unions. Although the psychological literature scarcely deals with unions (especially in the last 15 years), there is material about some important basic topics. If one includes the literature from labor journals, sociologists, political scientists, and especially industrial relations journals, one achieves a more complete picture of the state of research on unions. Six basic themes emerge from the literature: union leadership, treatment of minority groups within unions, labor's involvement with politics, union-management relationships, attitudes of union members, and the factors related to level of participation in union activities. This chapter presents an overview of the first five areas and a thorough review of the literature on the correlates of union participation. Union Leadership Studies Several books have been published about union leadership including A. A. Liverright's Union Leadership Training [1951], Harold L. Wilensky's Intellectuals in Labor Union§_[1956], and Lois MacDonald's Leadership Dynamics and the Trade Union Leader [l959].i Only MacDonald's book is empirically based and is the most thorough treatment of the subject. Strauss and Sayles [1953] using a projective technique (Sayles, 1954] concluded that election to union office is positively related to high 8 in-plant status (pay, skill, and seniority) and opportunities to "move around" the plant, although marked exceptions occur when the plant has been recently unionized or if there has been a high turnover of officers. Miller [1966b] and Rosen and Rosen [1957] investigated personality variableijzrelation to the union leadership role. Miller concluded that inner and outer directed behaviors are important for success as a union official. He also found in an earlier study [Miller, 1966a] that need fulfillment was greater at the higher levels of union leadership positions. Rosen and Rosen's 1957 study used such a small sample, 21, and so many variables, the MMPI, one hesitates to generalize from their results. Bogard [1960] also used personality variables to compare union leader trainees with management trainees. Differences were found in social class identification, social and economic values, femininity, relationships with authority figures, peer group relation- ships, social nonconformity, and early valuation of independence training in childhood. However, the within group variability (individual differences) was so high that the practical operation of the results is lessened. Also, no clear differences between union and management trainees were found on the self-control, psychological-mindedness, dominance, and self-acceptance variables. Another major effort in the area of union leadership appearing in the psychological literature is the Stogdill, Goode, and Day [1964] study. They used a questionnaire to obtain 44 descriptions of the leader behavior of labor union presidents. Factor analysis produced 10 factors: controlling tolerance for uncertainty, motivation of followers, control of the leadership position, consideration of member welfare and 10 expectations, representation of the follower group, considerate tolerance of member freedom of action, anticipatory coping behaviors, reconcilia- tion of conflicting demands, persuasive role enactment, and active role definition. They conclude that these findings are consistent with the demands made upon the role of labor union presidents. A common complaint of management is that union leaders do not truly represent their constituency. Corruption of union leaders in some unions and its subsequent publicity adds further impetus to studying union leadershippespecially the perception of the leaders by the union members. Sayles and Strauss [1953] found that union rank and filers are often suspicious of their own leaders. Members felt that no one would seek the extra work of union office unless there was some personal gain involved. However, as Gorer [1948] points out, Americans in general are suspicious of people who seek public office and often discredit them or use them as scapegoats. Sayles and Strauss use Gorer's theory to conclude that antagonism that could be directed against the union (for various subconscious reasons such as shame for accepting help, ambivalence caused by feelings of resentment for being dominated by the boss vs. feelings of gratitutde for having the job, and fear of reprisal by management for union activity) is projected against the officers of the union since it is psychologically safer to express these feelings towards a culturally approved scapegoat and since most members are clearly convinced of the union's value for security against arbitrary management and economic advancement. Other attempts to measure the per- ceptions of rank and filers' satisfaction with their leadership have fairly consistently shown positive feelings. Major surveys such as 11 Uphoff and Dunnette [1956] and Rosen and Rosen [1955] for instance show most members are satisfied with the job their leaders are doing. The other side of this coin involves the leaders' ability to perceive the attitudes of its membership. Rosen and Rosen [1955] report that union officials consistently overestimate the membership's satis- faction with union policies and actions. More recent studies by Gluskinos and Kestleman [1970] and Howells and Brosnan [1972] also found consistent misperceptions. Union leaders overestimated the importance workers attached to material rewards and underestimated the importance workers attach to good relations. However, it is important to note in these studies that managers also erred in the same manner. Union leadership literature also includes studies dealing with leaders' effects on participation. Miles and Ritchie [1968] studied 110 union officials of all levels and found they advocated that low levels of their organizations participate in decision making and the setting of bargaining goals. They reported that this might not improve the quality of the decisions but it would raise morale and create a greater acceptance of the decisions made by their representatives. Another important study in this area was published by Kahn and Tannenbaum [1957]. They found that participation in union activities was related to the perceived leadership skills of the steward in a number of areas. More will be said about this study later in this chapter. In summary, union leadership research has investigated personality and behavioral factors of the position; has found a general satisfaction with leaders by the rank and file (though not without complaints); has found errors in leaders' perceptions of rank and file members similar to the errors of management personnel; and quite importantly, research 12 on unions has found that a leader's skills effect the level of parti- cipation within union locals. Treatment of Minority Groups Within Unions Many labor unions were (and still are) in the forefront during the civil rights struggles. Other unions, notably the construction unions, have been accused of putting up racial barriers to membership. union treatment of minority groups is thus another area where literature may be found. Sometimes the union's official stand on civil rights issues and nondiscriminatory practices does not influence the union members' own feelings in community issues. Such were the findings of Deitzes [1953]. However, articles by Becker [1953] and Rose [1953] indicate that the union solidarity necessary for a successful union effort has a signifi- cant, although partial, reduction effect on members' racial prejudice and facilitates subsequent inter-ethnic relations even in nonunion spheres of activity. Also a 1965 survey of U.A.W. members and officials [Hero, 1965], found that rank-and-file members were more favorable to racial desegration than the national adult population. U.A.W. leaders were even more liberal than the rank-and-filers in this area. Treatment of minority groups within the union ranks continues to get publicity and since issues like the seniority rule's effect on affirmative action programs are important, the literature in this area will probably continue to grow. If the reader is interested in obtaining a historical per- spective of this issue, an article by Bailer [1943] is highly recommended. Another minority group's treatment by unions has generated litera- ture. Barkin [1950] emphasized the efforts unions have made to help out 13 the older employee. Steadfast support of the seniority rule, special arrangements for job adjustments to relieve older workers of some of the more strenuous aspects of the job and even attempts to reduce production standards in some cases are union efforts reported by Barkin. Kirchner and Dunnette [1954] however, surveyed 18 local unions around the Minneapolis-St. Paul area and concluded that unions did not emphasize procedures designed to foster more efficient utilization of older employees on the job. Not only is the treatment of the older union members an important issue but also the treatment of the newcomers has generated literature. Virtually, all surveys of union membership Show significant attitude and behavioral differences between the older and younger union members. This is not a new problem. Herberg [1953] concluded that a generation gap within union ranks had formed. He states that, "for the oldtimers the union represented the structuring of a socio-ideological ideal to which they brought a willingness for extreme personal sacrifice. For the newcomers the union is, basically, a highly valued but bureaucratized and routinized service organization" [p. 19]. Because of the important implications for unionism as a social movement, the treatment Of particular groups within unions will be an important area of union research in years to come. Union Involvement in Politics Another area that has generated considerable union research is involvement of unions in politics. Rosen and Rosen [1955] found 54 percent satisfied and 17 percent dissatisfied with their union's job on political action. Sheppard and Masters [1957] report similar 14 results (approximately 65 percent favorable and 35 percent against) though they also found that 21 percent of those that approve of political action disapprove of the use of union dues for such purposes. They also found that union members trust voting recommendations of labor groups much more than recommendations of newspapers and business groups. In a more recent survey, Barkan [1967] reports members consistently supported the issues and candidates of their union's (AFL-CIO) Choice except on the open-housing issue. Rosen and Rosen [1955] emphasized that although favorable attitudes in this area prevail, many union members object to being told for whom to vote (40 percent were satisfied with this role of the union, 37 percent were dissatisfied and 23 percent undecided). Kornhauser, Sheppard, and Mayer's book When Labor Votes [1956] provides important background material for the interested reader. Generally, although there is considerable resistance, most union members approve of union involvement in politics. An explanation for these results may lie in the union slogan, "Let's not lose what we gained at the bargaining table through legislation." Union-Management Relationships As unions have become an increasingly integral part of the industrial life of this country, union-management relations have become an important area of research. Probably this area more than any other union research area has involved psychologists. Several reports of successful interventions can be found in the literature. Stagner [1963] points out that management can improve its relations with unions with the aid of a psychologist. Blake, Mouton, and Sloma [1965] report successful improvement of such a relationship through the use of their 15 inter-group conflict technique. A clinical psychologist, Muench [1960], reports his successful intervention in union-mangement relations. Stagner, Chalmers, and Derber [1958] report the development of Guttman- type scales to assess union-management attitudes toward each other. Perhaps the best "primer" on union-management relations from the psychological perspective is Stagner and Rosen's, Psychology of Union- Management Relations [1965]. The purpose of the book is to analyze some of the cognitive and dynamic processes within the human personality which contribute to industrial disputes. It emphasizes the need to study the effects of the union-management relationship as an entity rather than separately studying worker and manager in isolation. During the 1960 American Psychological Association convention, a symposium on psychology's role in labor-management relations was presented and subsequently published in Personnel Psychology in 1961. The speakers emphasized the lack of understanding psychologists have of unions, the need for such understanding, the potential assistance the field of psychology offers in the resolution of union-management conflicts, and the need for more and better research to understand these phenomena. In a similar manner, the potential role of the psychologist in the collective bargaining process is discussed in Personnel Psychology, [1964, pp. 361-383]. Furthermore, the efforts and pitfalls of joint union-management organizational development projects are discussed in Huszczo [1975]. The above three articles will give the interested reader a broader perspective as to the interface of psychology and union-management relations. A related area of union research involves what has been termed the "Dual Allegiance Theory." The classic work in this area is 16 Purcell's book The Worker Speaks His Mind on Company and Union [1953]. Purcell thoroughly documents his thesis that there is a dual allegiance of the worker to company and union and their respective objectives. A symposium reported in Personnel Psychology [1954, pp. 41-80] gives the reader an overview of this area. Rose [1952] found that loyalty to the union surely did not imply disloyalty to the employer. Virtually every attitude survey since then has come to the same conclusion. In general, the literature on union-management relations presents some successful endeavors by psychologists in this area, emphasizes the need for psychologists to obtain a greater understanding of unions, stresses the potential usefulness of psychology in this area, and notes that, at least in the worker's mind, resolution of conflicts between the groups should not be based on the assumption that a worker is only loyal to the union or to management. Union Attitude Surveys A large majority of the worker attitude surveys that appear in the literature were sponsored by management. However, a number of union- sponsored attitude surveys have been undertaken among which are books such as Purcell's The Worker Speaks His Mind on Company and Union [1953], Rose's Union Solidarity [1952], Sayles and Strauss' The Local Union [1953], Rosen and Rosen's The Union Member Speaks [1955], Lipset, Trow, and Coleman's Union Democracy: The Internal Politics of the Inter- national Typographical Union [1956], and Seidman, London, Karsh, and Tagliacozzo's The Worker Views His Union [1958]. Also Kornhauser's The Mental Health of Industrial Workers used a sample primarily composed of union members and provides many valuable insights into their attitudes. 17 Most of the research on attitudes of union members has used opinion surveys and interviews to measure attitudes. Perhaps, the most methodologically sound research in this area was produced by Aylward et al., [1955] in their development and validation of the IRC Union Attitude Questionnaire. Results from the use of this instrument indicate union members' attitudes towards unions are quite favorable and nonmembers show neutral attitudes. Among union members, age and union seniority are positively related to favorable union attitudes (although a slight curvilinear trend exists at the high end of the range of these variables). Of the Six subscales, scores on items dealing with the national union were lowest for both officers and rank and file. There also were indications that "although the great majority of union members are favorable to unionism in a general way, many do not like some of the policies and practices of their local union" [Uphoff and Dunnette, 1956, p. 34]. Norms of each item and for each subscale for officers and for rank and filers are presented. More typically, union attitude survey researchers have measured "attitudes" by responses to individual items. Results are presented as the percent in favor versus percent not in favor of various issues. While the validity and reliability of these results on any given question cannot really be assessed (or at least is not typically assessed), information valuable to union leaders has been obtained through such surveys. One of the few such surveys appearing in the psychological journals, was Davis and St. Germain's 1952 study. Their results were quite typical indicating that union members were reasonably satisfied with their union and its practices (70 percent or greater on all questions of this area). They also found that attitudes toward the 18 employer were, overall, about as favorable though less consistent. However, they reported that a significant minority (25 to 28 percent) were not satisfied with wages, equipment, and work procedures. It should also be pointed out that a significant minority (21 to 27 percent) were displeased with union practices such as provisions of the present contract and the local's communication on union matters. Thus, the main dissatisfactions uncovered by the survey dealt with "areas of communica- tions and practical day-to—day problems, rather than in areas producing a conflict of principle" [p. 290]. These results typify other similar research efforts. Research on Union Participation Although attitudes of union members towards their unions are typically favorable, the relative lack of participation by the member- ship in union activities has been of great interest to union leaders and researchers over the years. While 70 to 80 percent of the member- ship are typically favorable towards the union, figures of 2 to 8 percent are typical for attendance figures at regular meetings of large industrial unions [Sayles and Strauss, 1953]. Perhaps the highest regular attendance figures reported in the literature appear in Kraft's polling of the AFL-CIO where 20 percent of members reported attending almost every local meeting and an additional 14 percent reported attending "quite frequently" [Barkan, 1967]. Why is there such a discrepancy between behaviors and attitudes in this area? What distinguishes the active union member from the inactive? A considerable number of researchers have addressed these issues. The focus of this dissertation is to investigate the relative importance of factors involved in union participation. 19 Three major benchmarks stand out in the literature on union participation: two previous reviews of the literature [Spinrad, 1960 and Perline and Lorenz, 1970] and a thorough research study by Tannenbaum and Kahn [Participation in Union Locals, 1958]. The Spinrad article reviews 35 studies that used a number of different approaches such as case histories, comparisons of the national and local union, and comparisions of individuals and groups within a given local to study participation in union activities. Methodology ranged from impressionistic observation to statistical analysis of survey information. Typical indices of participation reported in this type of research include some combination of the following: holding a union office, serving on a committee, attending meetings, voting in union elections, reading union literature, and using the grievance procedure. Spinrad feels that the literature leaves out some important factors in the study of correlates of participation, especially: unique characteristics of individual unions, the formal structure of the union, the degree of internal democracy, the history of the organization, and the position of the international union within the organized labor movement. Spinrad further concludes that a belief in "unionism" and in the general policies of the national and local union are necessary but not sufficient factors in producing participation. He feels that "participation also requires a belief that activity is functional, that it can achieve observable results" [p. 238]. Beyond these issues, Spinrad groups the correlates of participation into three categories: 1) objective features (the job, pay, residence), 2) personal associations (on and off job contacts, family, friends, belonging to other organizations) and 3) personal 20 orientations (job satisfaction, mobility aspirations, sociability, nonwork interests). His summary emphasizes the interaction of these variables, "Participation in trade unions is enhanced by those factors which makes for greater identification with one's occupational situation and occupational community, and diminished by those influences which foster contrary orientations" [p. 244]. Although published in 1958, Tannenbaum and Kahn's classical work in this area, Participation in Union Locals, was not included in the Spinrad review (a fact that Spinrad apologizes for in a footnote on the first page of his article). This book reports a thorough compara- tive study of four union locals--two with high activity, two with low activity levels. Drawing on 119 references, Tannenbaum and Kahn developed an 87 item questionnaire and used an Index of Participation consisting of a linear combination of: 1) number of regular union meetings attended, 2) number of special union meetings attended, 3) items dealing with behavior at meetings (e.g., asking questions, making motions, etc), 4) holding union office, 5) membership in union committees, and 6) voting in union officers elections. Although methodologically more sound than most of the research in this area, the major statistical analyses performed investigated whether there was a significant difference in percentage of active versus inactive members in responding to the items. Tannenbaum and Kahn found that active members differ from inactives in personality characteristics, group memberships, attitudes, and aspirations to name a few. They also found that participation is influenced by organizational constraints such as control, freedom on the job, leadership, and power. Thus, they saw participation in union activities as a result of many factors both internal and external 21 to the membership of that organization. They pointed out that to the degree that factors increase a stake in onewsjob, that one is more likely to be active. They suggest that this may help explain the "dual loyalty" phenonmenon. The most recent review of the literature of participation in unions is the Perline and Lorenz [1970] article. Perline and Lorenz cite the Spinrad review article frequently. They do however add some literature especially in the area of group variables that effect participation (most notably, the Hagburg [1966] article). They divide the literature in three major parts: 1) individual variables that correlate with participation (e.g., demographic variables, psychological attitudes, and perceptions of job environment), 2) group participation variables (e.g., homogeneity of membership and influence of primary groups), and 3) union structure variables (e.g., size, leadership and control). They conclude that all three areas do indeed influence participation and they speculate that they are interrelated. They state: "Indeed, one possible area of future research would be an investigation which was designed to analyze jointly the relative importance attached to each of the factors affecting participation discussed above, i.e., the characteristics of the individual, the nature of the group, and the union structure" [p. 438]. The focus of this dissertation is to investigate the relative importance of factors influencing participation. In line with this purpose and primarily drawing from the above three references, a thorough review of the findings of research studies on union participa- tion is offered in Tables 2 through 8. These tables provide the 22 reader with a complete, yet concise, up-to-date picture of the state of research on the factors influencing union participation. The tables are organized around three general classifications of factors: 1) individual variables (demographics, Table 2; personal attitudes and associations, Tables 3 and 4), 2) job related variables (job character- istics, Table 5; attitudes toward the job, Table 6), and 3) union related variables (union characteristics, Table 7; attitudes toward unions, Table 8). Active union members have been designated as Afs and inactives as Ifs throughout the tables. Table 2 focuses on the demographic variables as they correlate with union participation. Briefly, it indicates that active union members tend to be older, married, male, have a union family background, and live near the plant where they work. However, there are conflicting results as to whether community where raised and present residence effects level of union activity. No correlation between educational level and activity was found. Tables 3 and 4 deal with active (versus inactive) union members' personal attitudes and associations. Table 3 indicates that active union members tend to be more outgoing, have more off-job contact with fellow workers, and have slightly wider "variety" of friends than inactives. They also tend to join more organizations, spend less time with their families, have more friendships with union leaders and fewer friendships with supervisors and owners than their inactive counterparts. All in all, they seem to be "sociable joiners." Table 4 indicates that union actives tend to desire more power 23 Table 2 Demographic Variables and Union Participation Variable Findings References Age A's older than 1'5 Tannenbaum & Kahn, p. 116 I Dean, pp. 51-52 Univ. of California, p. 175. Marital A's more likely to be married Tannenbaum & Kahn, p. 74 Status than I's (though not true for women) Sex A's tend to be male Strauss & Sayles, p. 40 Community A's tend more to have been Form & Dansereau, p. 11 Where brought up in an urban Tagliacozzo & Seidman, Raised environment than 1'5 p. 548; Whyte, p. 221-233 (Tannenbaum & Kahn, p. 115 shows this tendency but not statistically significant) Present A's tend more to live in town Form & Dansereau, p. 10 Residence in which the plant is located Kyllonen, p. 532 Distance from home to union Miller & Young, p. 41; hall effects participation Purcell, pp. 203-204; level Seidman, London, Karsch & Tagliacozzo, p. 187 WOrkers that live close Sayles & Strauss, p. 202 together tend to b0 more active No overall relationships between Tannenbaum & Kahn, p. 115 present residence and level of union activity Education No statistically significant Tannenbaum & Kahn, p. 117 correlation between education and level of union activity Family Workersvfiifliunion family back- Purcell, p. 214; Seidman, Background grounds tend to be more active et al., pp. 23, 174-175 in unions A's more likely to remain at sane occupational level as Father. (I;§_more likely to be either upwardly or down- wardly mobile) Tannenbaum & Kahn: pp- 143-146 24 Table 3 Personal Associations and Union Participation Variable Findings References Sociability in General Associations with Fellow Workers Friendships with Union Leaders Friendships with Super- visors and Owners Associations with Out-of- Plant Friends A;§_described as "outgoing, liking to deal with people, possessing high activity levels, and nervous tension . . ." . . .more ascendant, outgoing and social. . .a joiner A;§_have more on as well as off job contact with fellow workers and this is signi- ficantly correlated with attendance at union meet- ings and other indexes of participation A;§_more likely to associate with fellow union members in off job recreational activities Workers friendly with union leaders tend to be more active A;§_are more likely than I:§_ to report friendships with local officers (except in friendships with stewards-- this showed no AfI_difference) Workers related or friendly with supervisors tend to be I's A's tend to have a slight but significantly wider variety (wider array of social groups) of friends than 1'5 No significant A-I_difference of having "working class friends" Significant tendency for A's to report friendships with professionals than do I's Seidman, et al., p. 191; Purcell, p. 210; Sayles & Strauss, pp. 100-103; Ginzberg, pp. 92-93; Tannenbaum & Kahn, p. 90, 229; Spinrad, p. 242; Hagburg, p. 17 Lipset, Trow, & Coleman, pp. 127—139; Dean, p. 51; Tannenbaum & Kahn, pp. 78-79 Lipset, Trow, & Coleman, p. 416 Tagliacozzo & Seidman, p. 549 Tannenbaum & Kahn, pp. 85-86 Seidman, et al., p. 131 Tannenbaum & Kahn, p. 85 Tannenbaum & Kahn, p. 85 Tannenbaum & Kahn, p. 84 25 Table 3 (Continued) Variable Findings References Associations No significant A-I_difference Tannenbaum & Kahn, p. 82 with Out-of- on degree of involvement with Plant out-of-plant friendships Friends No support for the hypothesis Tannenbaum & Kahn, pp. 82—83 that attitudes of friends toward the union X degree of involvement with those friends would predict union activity better than either variable by itself Associations ALE report joining other Tannenbaum & Kahn, p. 88 with Other organizations significantly Organiza- more than 1'5 tions '__— A's more than I;§_report Tannenbaum & Kahn p. 89 belonging to sports teams, fraternal groups, & veteran organizations No significant A-I_difference Tannenbaum & Kahn, p. 89 in membership in churches, neighborhood clubs, political organizations Conflicting results as to Spinrad, p. 243 whether A's are more oriented toward leisure pursuits Afi§_tend to spend less time Sayles & Strauss pp. 110-111; with their families and Show Seidman, et al., p. 187 less interest in fixing up their homes I's claim "to much work to do Chinoy, p. 159 on my house." 26 Tfifle4 Personal Attitudes and Union Participation Variable Findings References Power, Recog- nition Needs Aspirations of Upward Mobility "WOrking Class" Identifi- cation Attitude Toward Neighborhood A's conscious of desire to obtain personal power, recog- nition, or financial gain A's aspire upward occupational mobility less than 1'5 Significant negative relation- ship between upward mobility aspirations and union parti- cipation (though the member may aspire to "move up" within the union, not society in general) Potentiality for upward occupational mobility is unrelated to union participation No significant A-I_difference on attitudes concerning a steward becoming a foreman 51E tend to see work group or "working class" or both as a significant reference group and emphasize collective rather than individual efforts for improvement §g§_more than I;§_tend to regard union local as a source of "primary group" satisfaction A;§_tend to see industrial disputes from "workers'" vantage point rather than from management's viewpoint No agreement as to whether A's are more class conscious A's tend to have a lower estimate of their neighbor- hoods and communities as a place to live Seidman, et al., p. 232 Form & Dansereau, p. 12 Tagliacozzo & Seidman, pp. 551-552 Tannenbaum & Kahn, 146- 148 PP- Seidman, et al., p. 183 Tannenbaum & Kahn, 139-141 PP- Seidman, et al., pp. 256- 258; Blum, pp. 43-47 Hagburg, p. 20 Tagliacozzo & Seidman, p. 552; Miller & Young, pp. 41-45 Perline & Lorenz p. 426 Form & Dansereau, p. 9 27 Table 4 (Continued) Variable Findings References Homogeneity A's are more homogeneous than Tannenbaum & Kahn, pp. as a Group 1's in their behaviors and 206-211 attitudes Educational A's enjoy educational Hagburg, p. 20 Activities activities more than 1'5 Political A's are more politically Hagburg, p. 20 Orientations oriented than 1'5 and recognition. They do not however desire traditional upward mobility but regard the "working class" and their local union as significant reference groups. They enjoy educational activities more, are more politically oriented, and are more homogeneous in their behaviors and attitudes than inactives. Tables 5 and 6 display research concerning the relationship of job related variables and union participation. Active members tend to have higher paying, higher status, and less physically exhausting jobs. While no clear cut relationship between job or plant seniority has been found; union seniority, homogeneity of work group, plant size, and department and shift have been found to be related to union participation. There seems to be some question as to whether amount of movement allowed on the job effects union activity though the findings tend more towards a positive relationship between these variables. Active union members tend to be more satisfied and identify with their jobs and there seems to be no relationship between feelings of hostility toward management and union participation. There is evidence that active union members are more interested in intrinsic rewards whereas inactives tend to be more interested in the extrinsic rewards. 28 T&fle5 Characteristics of Job and Union Participation Variables Findings References Pay & Status Seniority Plant Size Department Shift Physicalness of Job Supervisory (Quasi) Nature of Job Accessibility A;§_tend to have higher pay, higher status jobs, more "crucial jobs", and higher skilled jobs. No clear-cut relationship between activity and job or plant seniority However, more union seniority the more union activity Small plant size facilitates union participation Departments within a plant found to differ significantly in terms of participation in union activities Departments successful in winning grievances tend to have a higher proportion of active unionists Amount of union activity found to differ significantly from one shift to the next but no single shift is consistently more active across locals A's have jobs that are less physically exhausting No significant difference between Als and I;§_on jobs requiring quasi-supervisory duties Jobs that permit mingling with many other workers facilitates union activity Sayles & Strauss, pp. 143- 153, 202-208; Seidman, et al., p. 171; Purcell, p. 205; Kyllonen, pp. 528-530; Tannenbaum & Kahn, p. 96, 229 Tannenbaum & Kahn, pp. 98-99 Tannenbaum & Kahn, pp. 97-98 Lipset, Trow, & Coleman, pp. 21-25; Seidman, et al., p. 190; Seidman, p. 223 Tannenbaum & Kahn, pp. 102-103 Sayles & Strauss, pp. 192- 195 Tannenbaum & Kahn, p. 103 Sayles & Strauss, pp. 207- 208 Tannenbaum & Kahn, p. 100 Sayles & Strauss, pp. 148-149 29 Table 5 (Continued) Variables Findings References Accessibility Homogeneity of Work Group No significant difference found between 5:5. and £1.53. on amount of movement and contact with others on the job (as rated by company and union officials) Workers of jobs that isolate workers tend to be apathetic Groups that work close together, have similar pay, skill level and ethnic background, tend to participate more Tannenbaum & Kahn, p. 102 Seidman, et al., pp. 132- 133 Sayles & Strauss, pp. 197-202 30 Tafle6 Attitudes Toward the Job and Union Participation Variables Findings References Job Satisfaction, Interest, Identification Specific Aspects of Their Jobs Attitudes Toward Company and Management A's tend to be significantly more satisfied with their jobs than 1'5 A's tend to be more interested in their work than I's A's tend to have more occupational identification than I's A;§_more interested in intrinsic rewards; I;§_ more interested in extrinsic rewards Union leaders tend to be more critical of specific aspects of the job situation No A-I_difference at Rank- and File level on criti- cism of supervision AL§_more than I;§_tend more to feel the union helped improve working conditions and are responsible for job security and better standard of living No clearcut difference between A's and 1's in Hostility toward management (Though from time to time this may vary as well as the participation level of some members varies from time to time) No A-I difference in perception of the company- union relationship Tannenbaum & Kahn, p. 229 Dean, pp. 51-52; Form & Dansereau, p. 8; Seidman, et al., p. Hagburg, p. 20 178; Lipset, Trow, & Coleman, p. 416 Lipset, Trow, & Coleman, p. 416; Sayles & Strauss, pp. 192—195 Hagburg, p. 20 Rose, 163-164 PP; Strauss & Sayles, pp. 35-36 Tagliacozzo and Seidman, p 548; Univ. of California p. 176 Tannenbaum & Kahn, pp. 229-231 Perline & Lorenz, p. 428 Tannenbaum & Kahn, p 134 31 Table 6 (Continued) Variables Findings References Attitudes No A-I_difference on view Tannenbaum & Kahn, p. 136 TOward of company's profit taking Company and as opposed to sharing more Management profits with employees A's possess favorable Dean, p. 53 attitudes toward union leaders and management Tables 7 and 8 present studies investigating aspects of union structure and attitudes toward unions in relationship with participation. While size of the local union is negatively related to participation, there is some indication that perceived amount of control by rank-and- file and "total control" are positively correlated to union activity. Furthermore, leadership skills of union stewards have also been found to effect participation. Active union members tend to have more favorable attitudes and greater identification with aspects and goals of unionism though the difference between them and the inactives tends to be more a matter of degree than of kind. Active union members Show more support for solidarity, more interest in union education, perceive the local union as providing freedom and spontaneity of expression, and feel unions should have more power. Actives also tend to show more support for "liberal" goals such as equal rights for women, equal rights for Negroes, and general improvement of the welfare of all the people in the community. The overview of studies presented in this review, indicates that many factors are related to union participation. Behaviors related to 32 Table 7 Characteristics of Union and Union Participation Variables Findings References Size of Local Union Amount of Administrative Involvement Control by Membership (Rank-and-File) Control by Officers Total Control (Sum of Control Levels for each of the 4 Hier- archical Levels) Sanctions Within the Union External Power Size of local union is negatively correlated with level of activity Small locals tend to be more democratic internally than large locals but contribute less to democratic processes at the National Union Level As a union becomes more administratively involved. participation level drops Membership control is positively correlated to participation (The above is not true for all locals) Did not find a relationship between officer groups (3 hierarchical levels-- President, Executive Board, Bargaining Committee) control level and member participation Some support that Total Control level is related to member participation A's more likely to give and receive sanctions than 1'3 "External union power appears to be a central variable underlying parti- cipation and numberous other manifestations within local unions." However it is iterrelated with several other variables Stein, p. 47 Faunce, p. 291 Lipset, Trow & Coleman p. 11 Tannenbaum & Kahn, 172-173 PP- Tannenbaum & Kahn, 171, 233 PP- Tannenbaum & Kahn, p. Tannenbaum & Kahn, p. Tannenbaum & Kahn, p. Tannenbaum & Kahn, (not a strictly empirical statement) p. 173 173 197 186 33 Table 7 (Continued) Variables Findings References Union Leadership A;§_more likely to report Tannenbaum & Kahn, p. 219 Skills: Keeping that their steward keeps Members Informed them informed about what is going on than I;§_ Extent to Which More A:§_than IL§_report Tannenbaum & Kahn, p. 221 Stewards Involve that stewards ask their Members in help in deciding what Decisions should be done about union matters Extent to Which A's more than 1'5 report Tannenbaum & Kahn, pp. Stewards Help that they get help from 222-223 Members in Need their stewards when they really need it Extent of Significant relationship Tannenbaum & Kahn, p. 224 Steward's found between participation Interest in and membership perception Members of steward interest 34 TfifleB Attitudes Toward Unions and Union Participation Variables Findings References General Union Attitude Attitude Scale Scores Norm Identification Immediate and Specific Goals Belief in "unionism" and in the general policies of the national and local union is necessary but not sufficient for active participation The favorableness of the attitude of union members toward unionism in general, their local union, and their national unions is in direct proportion to the percentage of meetings attended A's identify more than 1'3 with union norms and attitudes A's are more likely than I's to identify with the nature of the functions the union performs for its members AL§_ more than IL§_showed support for more soli- darity and getting members more interested and education in the union A's more than I's see local union as providing freedom and spontaneity of expression A's significantly more than I;§_reported wanting their union to pursue goals of better wages, working conditions, pensions, etc. (Though a high percentage of both A:§_and I;§_wanted this) Spinrad, p. 238 Uphoff and Dunnette, p. 15 Purcell, p. 205; Tannenbaum & Kahn, p. 231 Lipset, Trow, & Coleman, p. 416 Tannenbaum & Kahn, pp. 125-126 Hagburg, p. 17 Tannenbaum & Kahn, pp. 121-122 No significant A-I difference Tannenbaum & Kahn, p. 128 o — T 0 found concerning increaSIng the union's say within the plant (and only a small per- centage of either A's or I's want this) 35 Table 8 (Continued) Variables Findings References Immediate and A's significantly more than Tannenbaum & Kahn, p. 138 Specific Goals I's desire pushing grievances more aggressively A's feel unions should have Univ. of California, p. more power 177 Broad and General A's more than I's are in Tannenbaum & Kahn, Goals favor of the union pushing pp. 122-123, 126-128 for "righting social Univ. of California, wrongs" such as: equal p. 177 wages for women, equal rights for Negroes and other minority groups, improving welfare of all people in the community, obtaining a guaranteed annual wage Al§_more than I;§_are in Tannenbaum & Kahn, p. 130 favor of the International Union spending money to organize nonunion places (But both A's and 1's emphasized the greater importance of putting the money of into things for the people already in the union) AL§_more than Eli tend more Hagburg, p. 21 to receive satisfaction from union participation above and beyond satisfaction from achievement of higher wages and better working conditions Attitudes of Spouses of A's are more Tannenbaum & Kahn, p. 76 Family favorable toward unions Hagburg, p. 19 than spouses of 1's No support for hypothesis Tannenbaum & Kahn, p. 77 that family's attitudes toward unions would affect women's participation more than men's 36 Table 8 (Continued) Variables Findings References Attitudes of Mixed results concerning Tannenbaum & Kahn, Friends attitudes of in-plant and pp. 78-79, 80-83 out-of-plant friends on participation. (However attitudes toward unions is higher for in-plant friends than spouses or out-of-plant friends) Attitudes of No relationship found between Tannenbaum & Kahn, p. 92 Other the attitude toward unions Organizations in other organizations that A;§_belong to in comparison with attitudes toward unions in organiza- tions I;§_belong to. greater union participation are not isolated features for the active individuals but more of a life-style. They seem to be more active in other organizations, more involved in their jobs and generally more sociable. One might speculate based on the results of previous research that active union members tend to View unions more as a "social movement" whereas inactives view unions more in terms of their protection and economic functions. Individual studies about participation have been of limited scope. Few studies incorporated a broad range of the factors related to parti- cipation and the level of methodological sophistication of these studies has been categorically low. Rarely have indices of participation been statistically justified, individual items rather than scales have predominated as measures, and frequently analyses merely compared the percent of actives agreeing to statements to the percent of inactives 37 agreeing to the same statements. This dissertation advances the research on participation in union activities by: 1. providing an up-to-date consolidation of previous research findings in the descriptive and focused format of Tables 2 through 8, 2. investigating a broad range of factors as they relate to union participation, 3. investigating the relative importance of such factors, 4. using scales (actually factor scores) rather than individual items to measure factors related to participation, 5. using a broad measure of participation and establishing the reliability of such a measure, 6. using a large, cross-sectional, updated sample of union members, and, 7. using a cross-validation design. Through these improvements a more comprehensive, descriptive view of the correlates of union participation is achieved. CHAPTER III METHODS AND PROCEDURES The work reported is a part of a larger study investigating attitudinal and behavioral changes of U.A.W. members and spouses involved in a union sponsored educational program. Subjects Subjects in this study were 500 union members voluntarily involved in two-week educational programs (approximately 100 per two week period) during the summer of 1974 at the U.A.W. education center in Northern Michigan. On the first day of the program, each group of participants completed the research questionnaire. The subjects represent a geograph- ically stratified se'ple of the U.A.W. membership since each geographical region across the 0.8. and Canada sent a quota of voluntary participants for these programs based on membership population in that region. A detailed description of the sample used in this study may be found in Appendix B. Questionnaire The questionnaire employed in this study used items from three major sources: 1) scales used in previous research, 2) questions elicited from top international union officers and representatives, and 3) questions added by this researcher. This researcher, in consultation with an industrial psychologist familiar with the area, shortened an original set of 330 items to a 166 item questionnaire that could be 38 39 completed with a one hour time limit. Guidelines for such judgments came in the form of written and verbal comments of subjects involved in a small pilot study undertaken in August, 1973, comments of union officials involved with the educational programs, experiences of this researcher in previous union educational programs, and the goals of the larger evaluation study. This process resulted in the final 166 item form (see Appendix A). The questionnaire has five sections labelled A through E and one section (the front page of the questionnaire) of demographic variables. All the questions in Section A deal with attitudes towards various aspects of unionism and the U.A.W. in particular. Items Al-A14, A25, A26, A28, A31-A35, A37-A39, A45-A50, A54, and A59 were taken from the IRC Union Attitude Questionnaire developed at the Industrial Relations Center of the University of Minnesota in 1955 [Aylward, Uphoff, Dunnette, and Kirchner, 1955]. The actual IRC Union Attitude Questionnaire contains 77 items in 7 subscales. Evidence of high reliability and validity were established in the original research by Aylward et a1. [1955]. The items of the Unionism in General, Local Union in General, International Union and General Diagnostic subscales (48 items in all) were included in the pilot study to obtain participant reaction and were also sent to the top international union officers and representatives involved with the educational programs. Based on feedback from these groups, 15 items were eliminated from the questionnaire leaving none of the original scales intact. The results of the pilot study showed such skewed distributions on most of the individual items that 7 point Likert-type response formats (rather than 5 point as in the original) were used in the present study with the hOpe that finer discriminations would be possible. 40 The remainder of the items in Section A attempt to survey the attitudes of a variety of other aspects of unions, e.g., cooperation with management, concern for the rank-and-file, power, loyalty, involve- ment of spouses, and union business meetings. Section B was designed to be the major behavioral measures of union activities and includes various measures of control and partici- pation in unions. All except for one of the items used by Tannenbaum and Kahn [1958] in their Index of Participation are included in this section. Items B4 and 85 represent adaptations of items used in Woodward and Roper's Political Activity Index [WOodward and Roper, 1950]. Item B3 comes from the research of Tannenbaum [1956] on control structure in unions. It attempts to measure the perceived amount of control of various hierarchical levels within local unions. The sum of the four parts of B3 is said to measure the perceived amount of total control available within the local union. Section C attempts to gather general life satisfaction measures. This section was designed primarily for use in the evaluation study but was used in this study as it relates to participation. Section D is a conglomeration of individual items and some revised scales aimed at assessing attitudes toward one's job (Dl-Dl7), trust in others (Dl8-D21), self perceptions (D22-D26), attitudes toward education (D27-D30, D44), racial attitudes (D31, D32), political attitudes (D33-DB7), and conservatism in life-style (D38-D43). Items Dl-D7 came from Lodahl and Kejner's Job Involvement Scale [Lodahl and Kejner, 1965]. Items D8-Dl7 deal with various other opinions towards one's 'ob. Items D18—D21 are taken from Rosenber '5 Faith in People 3 9 41 Scale [1957]. Items D38-D43 were used by Campbell et a1. [1960] as a Conservatism Scale. They are a subset of items developed by McClosky [1958] to assess the strength of general (not political) conservative belief in individuals and groups. The remainder of the items in Section D were generated by this author primarily for the evaluation study. However, responses to these items were utilized in relation to parti- cipation in accordance with the purpose of this dissertation. Items El, E4, and E11 deal with attitudes while the remainder of Section E deals with behaviors. Item El uses the five scale titles (work itself, supervisor, co-workers, pay, promotional opportunities) of the JDI [Locke, Smith, Kendall, Hulin, and Miller, 1964] with the addition of "security" and "overall job" to assess various levels of satisfaction with different aspects of one's job. Item E4 is a global measure of one's racial attitudes, and item Ell is a global measure of political attitudes. The rest of Section E deals with such behaviors as work attendance (E2), political-community activity (E3, ES-ElO), and educational activities (E12, E13). Many demographic variables are included in this questionnaire including: Sex, Age, Community where Raised, Present Community, Education, Family Union Background, Company Seniority, Job Seniority, Union Seniority, Size of Plant, and Size of Home Local Union. The relationship between these variables and the union participation measures were investigated. Analyses The major focus of this dissertation is to determine the relation- ship and relative importance of a number of variables to participation in union activities. The following analyses sought to sharpen the focus. 42 Factor and cluster analyses were performed on items B4 through 814 (14 total items, B9 has 4 parts) to determine the advisability of using a single Index of Participation scale. The reliability of such a scale can be determined using Cronbach's Coefficient Alpha [Cronbach, 1951], a measure of internal consistency. Unlike most other reliability formulas, "Cronbach's (alpha) is not restricted to dichotomous scoring, thus it is especially useful for Likert-type scales" [Gilmore, 1968, p. 1]. As noted above, the predominate proportion of previous research investigating factors related to union participation has failed to utilize scale or factor scores as measures. The reliability and validity of the individual items used has been assumed. Debate over the validity of this assumption has taken place for years, e.g. McNemar, 1946; Crepsi, 1946; and Conrad, 1946. In order to reduce linear restraint in the multiple regression equations that are the major analyses of this dissertation, some com- bination of item scores (scale or factor scores) is recommended [Cureton, 1951]. Thus, in order to advance the methodology in this area of research and to reduce the linear restraint in later multiple regression equations, it was decided to factor analyze the attitudinal and behavioral measures. In order to achieve a more content pure solution, two separate varimax rotation factor analyses were performed: First, the 59 items dealing with attitudes towards various aspects of unions (Section A) and the four measures of perceived control within local unions were factor analyzed. Second, all nonunion attitudinal and behavioral measures (the 73 items found in Sections C, D, and B) were similarly analyzed. Standardized factor scores based on the varimax solutions 43 are used rather than scale scores utilizing uni-variate weighting with items loading highest on that factor. Factor scores include all the effects of the items on that factor while allowing the items that load highest to have the greatest effect. Scale scores, although acceptable, arbitrarily assign equal weight to each item and ignore the problems of items that have high loadings on more than one factor. Furthermore, factor scores have a mathematical elegance not found in scale scores and were easier to obtain due to the particular packaged computer programs available. Previous literature shows there are many factors related to participation in union activities. Multiple regression equations are used in the present research to look at the relationship of a number of variables including some factor scores with participation. A measure of participation in union activities is the criterion variable and the predictor variables are individual demographic variables, factor scores, and some individual items. In maximizing multiple correlations, multiple regression procedures take advantage of any correlated errors or specific variations, giving an overoptimistic picture of predictive value of the weighted composite score [Guilford, 1954, p. 405]. Mosier [1951], Cureton [1950], and others have pointed out the importance of using cross-validation pro- cedures to avoid errors of this kind. Double cross-validation procedures yield much more information and require no larger total sample than single cross validation (though more statistical work). The present study uses double-cross validation to avoid the above-mentioned pitfalls of multiple regression. The sample is divided into two equal parts; a 44 set of weights is derived in each of the two subsamples independently and then applied to the other subsample for checking purposes. Shrink- ages in multiple R's are determined in order to provide a more realistic index of validity and to evaluate the appropriateness of the weights [Guilford, 1954, p. 406]. After establishing an estimate of the multiple R for the regres- sion equation on the total sample, examination of the "importance" of individual predictor variables is possible. Darlington [1968] discusses the utility of five measures of importance and concludes that three (zero-order correlations, beta weights, and the usefulness index) are of interest and favors the usefulness measure. Only when all predictors are uncorrelated will these three measures yield the same rank ordering of the variables. The zero-order correlations are presented to enable a discussion of the relationship of each variable to the criterion variable. Zero-order correlations are the only index of importance unaffected by the other variables. The usefulness index is presented in order to discuss the relative importance of the variables in the prediction of the criterion variable. The usefulness index may have little or no relation to the zero-order correlations and is heavily influenced by the nature of the other variables in the regression equation. This statement is also true of beta weights but it can be shown that beta weights are not measures of usefulness when predictor variables are intercorrelated [Darlington, p. 166]. Correlation certainly does not imply causation, but if enough causal direction in the relationships can be logically established, "then there are certain situations in which a multiple regression equation can be used to estimate the importance" [p. 166]. The method assumes the following: 45 1. All variables which might affect the dependent variable are either included in the regression equation or are uncorrelated with the variables included. 2. Terms are included in the regression equation to handle any curvilinear or interactive effects. 3. The dependent variable has no effect on the independent variable [p. 167]. Since the design of this study does not meet these assumptions, beta weights as a measure of importance will not be discussed. They are, however, presented for the interested reader and to be used in support of appropriate alternative interpretations. In summary, 500 union members representing a geographically stratified sample of the U.A.W., completed a 166 item questionnaire. Fourteen of the items (B4 through B14) were analyzed separately to determine the advisability of using them as a single Index of Partici- pation. Factor analysis was performed on the remaining 63 items concerning attitudes towards various aspects of unions and a separate factor analysis was performed on the 73 nonunion attitudinal and behavioral items. Based on the results of the factor analyses, standardized factor scores were calculated for each subject and included with demographic items as independent variables in a multiple regression equation. The Index of Participation was the dependent variable in the multiple regression equation. Double cross validation procedures were used to establish the validity of the equation. The relative importance of the individual predictor variables was then investigated using indices of importance as discussed in Darlington [1968]. CHAPTER IV RESULTS A reliable Index of Participation was developed. The results of the two factor analyses indicated the presence of six union attitude factors and eight factors covering other attitudes and behaviors. A multiple regression equation was developed and cross validated. The relative importance of the factors involved in union participation was then established using Darlington's "usefulness" index. The details of these results follow. Development of the Index of Participati9n_ The results of the principle components factor analysis of the 14 union participation items (see items B4 through B14 of Appendix A) showed all 14 items loading on the first factor. This solution created two principle components--the first accounting for 40 percent of the variance and the second, 8 percent. A varimax rotation solution resulted in two factors: first, a factor with items B4, BS, B8, and B10 through B14 loading highest on it, and then a factor of the remaining six items.‘ Using these two factors as scales, Hunter's multiple groups cluster analysis [Hunter and Cohen, 1971] showed the two scales to intercorrelate .72. Based on content considerations and the statistical results, it was decided to combine all 14 items into one scale, to be called the Index of Participation. The internal reliability coefficient of this scale is a = .80. Items were standardized and missing data were 46 47 replaced by the mean of that item. The Index equals the sum of the standardized items and represents a measure of union participation of broad content range with high internal consistency. Factor Analysis of the Union Attitude Questions A varimax rotation factor analysis was performed on questions A1 to A59 (questions dealing with many aspects of unionism) plus item B3 (Tannenbaum's control in local unions questions). Items were reflected such that the direction of the response categories on the items indicated a "pro-union" attitude. A 19 factor solution resulted but only six factors had eigenvalues greater than the Kaiser criterion of 1.00. Tannenbaum's control variables loaded separately on two factorSL-being the only variables loading highest on those factors. Based on this result and on the previous manner of usage of these variables by Tannenbaum, these items were dropped from this factor analysis and later used as separate variables. The remaining 59 items were utilized in a forced six factor varimax solution. The factor loadings of all the items on all the factors, and the eigenvalues of the factors are presented in Appendix C. Although the eigenvalue of the sixth factor was less than 1.00 (it equalled .87), it was kept for further analysis due to its purity of content-—i.e., only two items (A15 and A16) had their highest loadings on this factor and these two items are the only items in the questionnaire dealing with attitudes toward union-management cooperation. Based onthe content of the items loading highest on each factor, the factors were labelled: Factor 1 Attitude towards Local Union, Factor 2 Attitude towards Union Power Relationships, 48 Factor 3 = Philosophy of Unionism in General, Factor 4 = Attitude towards the International Union, Factor 5 = Beliefs in Union Participation, and Factor 6 = Attitude towards Union-Management Cooperation. Standardized factor scores (2 scores) were calcualted for each of these factors for each subject. Factor Analysis of the Remaining Items A varimax rotation factor analysis was performed on the remaining 73 attitude and behavioral items (Sections C, D, and E of the question- naire--seeAppendix A). Appropriate variables were reflected such that the direction of the scores on the items indicated a "liberal" attitude. A 22 factor solution resulted but only 10 factors had eigenvalues greater than 1.00. A reduced 10 factor solution was then performed. Because the tenth factor had an eigenvalue of less than 1.00, and the ninth factor had an eigenvalue of 1.00 exactly, and because of the content of the highest loadings produced by this solution, it was decided to attempt an eight factor forced solution. This produced eight factors all of eigenvalues greater than 1.00 and factors of sufficient content clarity. The factor loadings of all the items on all the factors and the eigenvalues of the factors are presented in Appendix D. Based on the content of the items loading highest on each factor, the factors were labelled: Factor 7 = General Job Satisfaction-Involvement, Factor 8 = General Satisfaction with Life, Factor 9 = Socio-Political Liberalism, Factor 10 = Faith in People, 49 Factor 11 Job Satisfaction-Specific Aspects, Factor 12 = Anti-Conservatism in Life Style, Factor 13 = Community-Political Activities, and Factor 14 Voting Behavior. Standardized factor scores (2 scores) were calculated for each of these factors for each subject. The Development and Results of the Multiple Regression Equation A multiple regression equation was developed and tested in a double cross validated design. The total sample pool was divided into two halves of 250 subjects each for this purpose. The Index of Participa- tion was the dependent variable and a total of 31 variables were con- sidered as possible inputs as independent variables. Casewise deletion was used to handle missing data. This left 213 usable subjects in one- half and 201 in the other. The independent variables considered were the 14 standardized factor scores, 12 demographic variables (Item 115 - sex, Item 116 - age, Item 118 - number of children living at home, Item 120 - size of community where raised, Item 121 - size of community of present resi- dence, Item 123 - level of education, Item 124 - whether either parent belonged to a union, Item 125 - company seniority, Item 127 - job seniority, Item 128 - U.A.W. seniority, Item 129 - plant size, and Item 130 - local union size) and Tannenbaum's five variables concerning perceived control in local unions (four items from B3 plus the sum of the four items, "amount of total control"). The final equation con— tained variables that either: (1) had a significant zero order correla- tion between the variable considered and the Index of Participation, or 50 (2) remained in either half-sample when step-forward multiple regression equations were generated in both halves using all 31 variables and an F inclusion level of p < .10, or (3) were considered important by reason of previous research findings. Twenty-five of the 31 independent variables considered were accepted into the final equation. Table 9 summarizes the rationale for including each variable. Table 10 presents the intercorrelations of the 25 independent variables. Zero order correlations between the Index of Participation and Factors 1, 2, 3, 5, 9, 13, and both of the control variables were each significant at the .001 level. The corresponding zero order correlations with Factors 1, 12 and demographic items, level of education and plant size, were significant at the .01 level. While the correla- tion between Participation and sex was significant at the .05 level, the remaining correlations with the Index of Participation were not significant. The multiple R produced by the equation utilizing all 25 indepen- dent variables equals .6936 in the first half-sample and .6899 in the second half-sample. Applying the Beta weights generated in the second half-sample on the first half-sample produced a shrinkage in the multiple R from .6936 to .5518. Applying the Beta weights generated in the first half-sample on the second half-sample produced a shrinkage in the multiple R from .6899 to .5358. When this set of variables was applied to the total (N = 414) sample, the multiple R was .6537; thus accounting for approximately 43 percent of the variance in the criterion variable. 51 Table 9 Rationale for Including Variables in Regression Equation Variable Significant Results of Previous Zero Order Step Forward Research Correlation Regression Findings with Partici- Equations pation Index Sex X X Age Size of Community where raised X X Size of Community of Present Residence X Level of Education X Whether Parents were in union X U.A.W. Seniority X X Plant Size X X X Local Union Size X X Perceived Control of Rank & File X X Perceived Control: Total Amount X X Factor 1: Attitude towards Local Union X X X Factor 2: Attitude towards Union Power Relationships X X Factor 3: Philosophy of Unionism in General X X X Factor 4: Attitude towards the International Union X X Factor 5: Beliefs in Union Participation X X X Factor 6: Attitude towards Union— 1 Management Cooperation X Factor 7: General Job Satisfaction- Involvement X X X Factor 8: General Satisfaction with 2 Life X X Factor 9: Socio-Political Liberalism X X X3 . . 2 Factor 10: Faith in People X 52 Table 9 (Continued) Variable Significant Results of Previous Zero Order Step Forward Research Correlation Regression Findings with Partici- Equations pation Index Factor 11: Job Satisfaction- Specific Aspects X X Factor 12: Anti-Conservatism in Life Style X X Factor 13: Community-Political Activities X X X Factor 14: Voting Behaviors 1Generalizing from literature on dual allegiance theory and on job procedures as related to union activities (Table 6). 2 . Generalizing from literature on general SOClablllty and outgoxng nature of union activities (Table 3). 3See Hero, 1965. 53 ooo.a ed omm.- ooo.H I ma omo.u moo. ooo.H NH mao.- mmo.- mas. ooo.H as moo. moo. mmo.- omo. ooo.H oH oeo. mmo. moo. moo.- moo.- ooo.H m mmo. HHo.u «Ho.- moo. mHo.- moo.n ooo.a m Hoo. mmo. mmo.- mao. omo. moo. moo.- ooo.H s noo.- mma.- moH.- omo.- omo. Hos. omo.- Hmo.- ooo.H o mmo. mmo. Hmo.u moo.- Nmo.- mom. moo. omo. mmo. ooo.H m moo.- omH. oHH. omH. moo. was. mom. mom. moo. NHo.- ooo.H o moo. Hos. Hmo. who. one. mmm. omo. mmo. omo. mmH. moo. ooo.a m oHo. mom. mmm.- ooo. Hoo. Rom. mmH. moo.- moo.u moo. was. Hmo. ooo.H N moo. omm. moo. mom. mam. mmo. mHo. mmH. mmo.u moo.- «NH. Hmo. moo. ooo.H H umosoz SH ma NH so oH m o m o m o m m a manamum> acmumoom COAwmoumom ca cow: Hmoaommum> Had mo xmuumz COwumHOHHOO OH OHQMB 54 ooo.- Hmo.- mmo.- mmo. MHo.u ooo.H omH mmo. mHo. mmo. moo. mmo.: mmo. ooo.H mmH moH. mmo. mmH.- mHo. mmH.- Hmo.: omo.- ooo.H HNH omo. omo. mHH.) mmo. oHo. mo~.- mmo.- mmo. ooo.H omH omo.- ooo. moo. mmo. ooo. mmH.- moH.- mmo.- Hmo. ooo.H mHH moo. moo. moo. HHo. omH. mmo. omo. moH.- mmo.- mmo. ooo.H mHH omo. mmo. mHo.- mHo. omo.- Hmo.- mmH.- omo. mmo.- mmo.- moo. SH mmH. mmH. mmo.- moH. HHo. ooo. mmo. mmH. mmo. mmo. ooo.- MH moo. mmo. mmo.- mmo. mmH.- moo. moH. mmo.- omo.- mmH.- mmo.- NH omo. mmH. mmo. mmH. mmH. Hmo.- mmo.- mmH. mmo. mmo. ooo. HH mHH. oom. mHo.- omo. omo. moo. mHo. mmH. omo. HHH. moo.- oH HoH. mHH. oHH. moo. mmo.- ooo. mmo. ooo.- mmo.- moH.- mmo.- m HHH. moo. mHo.- Hmo. mmo.- mmo.- HHH.- omo. HHo.- mmo.- omo.- m mmH. Hmo. mmo. mHo. mmo. omo. moo. moH. mmo.- moH. mmo. m mmo.- mmo. mHo.- omo. mHo.- omo. mHo.: omo. mmo.- ooo. omo.- o mHH. mmo. mmo.- HHo. mHo. omo. mmo.- omo. mHo. moo. omo.- m mmH. HmH. moo. moo. mHo.- moo. mmH.: omo. mmo. Hmo. mmo.- o moH. moH. mmo.- mmo. mmH. moo. omo.- mmo.- mmo.- moH. ooH. m mmH. HmH. moo.- omo. mmH.- mmo.- mmo. mm-., mHo.- ooo.- mmo.- m omo. oom. Hmo.- mmo. Hmo. moo. omo.- moH. Hmo. HmH. ooo. H mHHm Houuooo loom Ixcmm .8852 Hence uHofiooo oMH mNH omH omH mNH HNH omH oHH mHH mHanum> Homscmuooov oH oHnme 55 .comumoam>o ucoucoo «0 come you oouOOmeu mums OmH one .mmH .mmH .mmH moaoomum> mo mcomumaouuoo .Hoummoo was» om oocmmoc >HmeH>OHm mm anH muouomm u «Hid .onwm Hmocq n oma Honmm pomHm u mmH “mumHOmcom .3.¢.D u mmH “some: cw ucouom n «ma “Hoboa Hmcoflumooom n mNH «Hammond >umcoeeou u HNH moomflmm SHHCOEEOU u omH Home u mHH axom u mHH ..m.o .oocoHucoE mamoOH>oum madness Houcmm com scum on Home“ mumofidc oaoomum> H ooo.H Houuaoc Hmuoe Hmo. ooo.H mHHm-oomuxomm ”Honoooo mmo. moH. ooo.H omH mmo. HHH. mom. ooo.H mmH omo. mHo. mmH. mom. ooo.H mmH mHHm Houuoou loomuxcmm Honfipz Hmuoe "Homecoc omH mmH mmH «NH mmH HNH omH mHH mHH mHanum> IomooHucooo oH mHome 56 The Relative Importance of the Factors Involved in Union Participation Table 11 displays the rank ordering of the relative importance of the factors used to predict participation in this study. This rank ordering is based on Darlington's usefulness index [1968]. Factors 13 (Community-Political Activities), 3 (Philosophy of Unionism in General), 9 (Socio-Political Liberalism), 7 (General Job Satisfaction-Involvement), 5 (Beliefs in Union Participation), and 4 (Attitude Towards International Union) are all significant beyond the .01 level. Plant Size and Factor 12 (Anti-Conservatism in Life Style) are significant at the .05 level. Factor 8 (General Satisfaction with Life), Perceived Control of rank and file, age, and sex Show F's at the .10 level of significance. The remaining 13 variables arermwsignificantrelative to the first 12 variables. The 25 independent variables can be blocked into four content categories: 1) union attitude factor scores, 2) nonunion attitude and behavior factor scores, 3) demographic variables, and 4) perceived control in the local union variables. Table 12 shows the results of using Darlington's indices when dropping the four blocks of variables one at a time. Dropping either set of factor scores shows very signi- ficant effects (an F of p < .01). Dropping the set of demographic variables produces an F right at the borderline of significance (p < .05). DrOpping the two perceived control variables shows no significant effect. Overall, the results Show the relationship of the 25 variables in this study to a reliable Index of Participation. Evidence of the validity of the multiple regression equation was established through the small shrinkage in the multiple R upon double-cross validation. 57 oumm woman m mmo. 0H «emvH.| m Rmamoo. mNH aouH some: Hmcowuo: nuoucH Chases opsumuuc o mMH.I mm moo. o «ammmHo. v Houoom acmummwowuumm coma: cw mwowaom v mva. v «comm. m :«vomao. m uouomm ucoso>ao>qn Icofiuommmmumm non Hmuocoo m HvH. m «equ. v «RMHmHo. m wouomm Emmaononfiu HmomumHOAIOHOOm m Hmm. H Ramon. m «emommo. m uouoom Hmumqoo cw Emmemcs mo anaemonsm m mam. m «emvm. m «emmmmo. m uouoom momum>mu04 Honeymaomnwumcssfioo H oom. N «evmm. H «RHHmmo. ma Houomm Acowummmowuummv Hoouo unemoz Hoouo cowwmuwuo saws Moose and cm money xddm muom xcmm :OHumHmuHoo xcmm mmocdomomo mmHQMmHo> ucoocomoocH mm Ham co mocmuuomEH HO HA UHQMB moudmmoz coupe m.:oumcwauoa 58 COHumuomooo ucoeooocmz IGOHCD cumsoe ooouHuu< oH N«o.| mH mmo.: mH «mHoo. o uouomm HOH>osom moHuo> «H omo. NN mNo.| «H mNNoo. «H uouoom mmHsmcoHuMHom nosed :oHcD cumsoe oosuHuuc NH moo. o RRONN. MH «mmoo. N uouomm xom NH Hmo. mH mmo. NH «o«oo. mHH EouH com oH mmo.: «H ««mo.u HH mm«oo. mHH EouH oHHmuoootxqmm mo Homecoo oo>Hoouom m mOH. m ««««N. 0H «m«oo. «MH souH OMHH :uH3 :OHuommmHumm Hmuoooo HH mmo. NH «Hmo. m o«moo. m uouomm 33m OMHH CH EmHum>uomcoulHuqd m mmo. HH «MHH. m ROHmoo. NH nouomm HooHummHoHuummo Hoouo usmHoz “mono :OHuouHuO nuHs “mono HNm :H monov xdmm ouom scum SOHumHOHHOU Hood mmocHsmomD HowocHuoooo HH oHnme 59 mHaowm cH :uHmm MN MHO.) «N mHo. MN oHooo. 0H uouoom oomHmm when: wuHcoEEOO mo ONHm HN HNo.I mH Mmo.l NN HMooo. ONH EouH MOHHOHcom .3.¢.D mH oNo.u oN o«o.| HN mmooo. wNH EouH COHumsom mo Ho>oH NN 0No.| MH «mmo. 0N oMooo. MNH EouH GOHCD HMOOH GH Houucou Hmuoe om>Hmouom mH Hmo.: m «««mH. mH m«ooo. moH EouH ONHm :oHCD HMOOH 0N 0N0. mH Mmo.| mH omooo. oMH EouH muoommd OHMHoomm IcoHuommeuom non mH H«o.| MN oNo.| mH mHHoo. HH Houoom oOHcD HMOOH Cumsoa mosuHuud MH Hmo. m «eHON. 0H H«Hoo. H Houomm HooHudeoHuummo woouo uanoz moose :oHuouHuO suHs noouo ANm CH money Hoax ouom xdmm :oHuoHouwoo zoom mmochmomD HomsoHuooOo HH «Home 60 .Ho. v a no neMOHmHomHm «« .mo. v o um unmonHomHm * oocochom ucomoum mo muHCQEEOU mo ONHm «N NHo. HN MMo. MN 0Hooo. HNH smuH oOHda :H vacuum MN HHo. 0H moo. «N HHooo. «NH EouH HooHummHOHuuomv nacho uanoz Hoouo :oHuouHHO :qu noose HNm :H money zoom ouom Hana COHUMHOHHOO Hana mmooHamomD lemocHucooo HH mHome 61 Table 12 Darlington's Usefulness Index of Importance Applied to Blocks of Independent Variables Usefulness F (Drop in R2) Dropping Factors 7 - 14 (The "other" attitudes and behaviors factor scores) .12661 10.72** Dropping Factors 1 - 6 (The union attitudes factor scores) .08838 9.98** Dropping the 9 demographic items .02504 1.88* Dropping the 2 items on Perceived Control (Control: rank-and-file, control: total amount) .00559 2.06 F8, m, .99 = 2.51 F6, m, .99 = 2.80 F9, m, .95 = 1.88 F2, m, .90 = 2.30 *Significant at p < .05 **Significant at p < .01 Darlington's indices of importance provide data as to the relative importance of the variables involved. A detailed discussion follows. CHAPTER V DISCUSSION Generalizations of the results of studies on unions must be made with caution. Generalization requires that the sample used in the study be representative of union members and that the measures used be reliable and valid. Robert Hoxie's words on the difficulties of researching unions should be heeded: Unionism is in essence one of the most complex, diffuse and protean of modern social phenomena. There is not one local union, but probably 30,000; there is not one national union, but about 130, each with its own problems to solve and its own aims, policies, attitudes, and methods. These unions do not amalgamate into a single general organization and move- ment but there are many independent unions and several groups and general associations with vitally different viewpoints, fundamental purposes, and ways of attaining them. What is true of one union or group may not be true at all of another. No judgments may be rendered nor generalizations made in regard to unionism as such from the study of any union or any small number of unions or any group. And, moreover, in the realm of unionism everything is in a state of flux, of constant change and development. Positive conclusions, therefore, are almost impossible to secure, and tentative generalizations can be made only as the result of the most broad and painstaking examination of the facts and an ability to get beneath appear- ances to discount deliberately false and prejudiced statements [Hoxie, 1920, pp. 1-2]. Time has not simplified these matters, but rather the situation today is even more complex. Fortunately, methodological sophistication has grown considerably since Hoxie's time and when used properly thus aids generalization and interpretation of results. The sample used in this study was large and geographically 62 63 stratified, but contains subjects from only one international union. The average subject in this sample was male, married, white, and 40 years old. He had slightly less than a 12th grade education and had been in the union 13 years. A detailed description of the sample can be found in Appendix B. The applicability of the results of this study become increasingly less certain to the extent that a given population is dissimilar to this sample. As is true of virtually all union studies reported in the litera- ture, both active and inactive members of this sample showed definite pro-union attitudes. On 72 percent of the union attitude questions asked, 70 percent or more of the subjects responded in a pro-union manner. Behaviorally, however, this sample seems to be more active in terms of meeting attendance than other samples reported in the literature. A full 30 percent reported attending all local union regular meetings and another 20 percent reported attending almost all of them. Barkan [1967] reported 20 percent attended almost all meetings and 14 percent attended quite frequently. Keeping these potential limitations and generalizability in mind, let us now scrutinize the results of this study. An Index of Participation The major focus of this dissertation is to determine the relation- ship and relative importance of a number of variables to participation in union activities. Previous researchers have investigated many corre- lates of union participation as Tables 2 through 8 indicate. Also, these researchers used differing measures of participation. What does participation in union activities include? 64 Participation is a complex phenomenon in many organizations including labor unions, and students of labor organizations have distinguished these categories. (1) Participation may occur in either the formal or the informal activities of the union. (2) Participation may occur either in the union hall (off the job) or within the plant [Tannenbaum and Kahn, 1958, p. 49]. This study accepted the idea that participation is a behavioral concept made up of various forms of behaviors and tried to obtain an index of participation that reflected a variety of union related activities. Most previous studies investigated only one aspect of union participation (usually attendance at meetings, though frequently it was holding or seeking union office). Although meeting attendance has been found to discriminate active from inactive union members better than any other single item [Tannenbaum and Kahn, 1958] and holding union office is by definition a-sign of greater union activity, no single item truly captures the range of behaviors involved in union participation. Tannenbaum and Kahn did make use of an index of parti- cipation involving a number of behaviors but report no measure of reli- ability and fail to standardize the items used in their combined index score. The index of participation used in the present study utilizes a wide range of behaviors (discussing union topics with friends, meeting attendance, behavior at meetings, union voting behaviors, seeking office and filing greivances) that have at least face validity as measures of union activity. The index is of established high internal consistency and is composed of the sum of the 14 standardized items. While the multi-behavior scale used in this study represents an improvement over measures used in previous studies, further research and scale development is needed to establish the validity of such an index. 65 Comparison with Previous Methodologies Had this study followed the pattern of the bulk of previous studies in this area, either the scores on the index of participation would have been dichotomized and the differences between "actives" and "inactives" would have been reported or the scores on the index would have been correlated with the remaining 152 individual variables and the signi- ficant correlations would have been reported. Because of a belief that participation is a continuous not a dichotomous concept and because of the Spuriously high number of "significant" results that would have resulted from correlating so many variables in a sample so large, neither of these procedures were seriously utilized. However, out of curiosity, this researcher did correlate all 150 variables with the Index of Participation and Table 13 shows the large number of statis- tically significant findings that would have resulted. One should keep in mind that with the present sample size a correlation of .115 is significant at the .01 level and a correlation of .083 is significant at the .05 level. The proportion of the variance accounted for at those levels would be around 1 percent and thus pragmatically of very little use. Only three items could have accounted for 10 percent or more of the variance. They were: E6 (What activities concerning politics have you participated in the last year?) (r = .366), D30 (I find my union activity is educational for me.) (r = .363), and A43 (I would like to see my spouse become more involved with the union.) (r = .335) . In order to establish more reliable measures, 14 factor scores were established through factor analysis in lieu of using the individual 66 items. Factor scores better represent the content domains resulting from factor analyses than any subset of items. The six factors generated with items Al-A59 accounted for 25 percent of the variance of these items. This indicates "the extent to which the variables tend to have common factors among them rather than only specific factors" [Nunnally, 1967, p. 264]. The 29 percent of the variance accounted for in the 8 Factor solution of the items in Sections C, D, and E (73 items) Should be interpreted in the same manner. Having thus reduced 132 variables into 14 factor scores and standardizing those factor scores, it is appropriate to look at the correlations between these measures and the index of participation and compare these findings with the findings of previous research. Table 13 Number of Significant Correlations Between Individual Items and the Index of Participation Significant Significant Significant Not Statis- at .001 at .01 at .05 tically Level Level Level Significant Items dealing with Attitudes Toward 28 9 ll 11 unions (Section A) Items Dealing with Other Attitudes and 27 9 12 25 Behaviors (Sections C, D, and E) Items Dealing with Perceived Amount of 2 l 0 2 Control (B3 plus "Total Control") Demographic Items 0 2 1 10 Tbtal 57 21 24 48 67 Zero-Order Correlations with Participation As Table 11 indicates, seven of the factor scores and three individual variables correlate significantly at the p < .01 level with the index. The highest correlation is with Factor 9 (Socio-Political Liberalism, r = .366). This is in general agreement with Hagburg [1966] who found union actives more politically oriented than inactives. Previous research support as to the "liberal" orientation is found in Tannenbaum and Kahn [1958] who found actives more in favor of the union pushing for "righting social wrongs" (e.g., equal rights for Negroes, obtaining a guaranteed annual wage) than inactives. Hero and Starr [1970] noted "Adults in union families have been more inclined than other Americans to consider themselves liberals rather than conservatives" [p. 79]. They found that rank-and-file members of the same international union (U.A.W.) used in thi§_study were more liberal than the other unions they studied. Furthermore, they found union leaders and activists of the U.A.W. even more liberal than the rest of the members of that union. Thus, the finding of this study that liberal social and politi- cal attitudes are quite highly related to union activity replicates previous findings and is probably accentuated due to the sample used in this study. In connection with the above discussion, Factor 13 (Community- Political Activities) had nearly as high a correlation with the index of participation (r = .354). Apparently union actives are politically more conscious behaviorally as well as attitudinally. This generally agrees with another Tannenbaum and Kahn [1958] finding; i.e., that actives are more homogeneous than inactives in their behaviors and attitudes. One of the most consistent findings in all research done on 68 active union members indicates that they are "joiners"--outgoing personalities that possess high activity levels (see Table 3). One might speculate that this high activity level of union activists is directed toward political and community activities as well. Attitudes toward various aspects of unions showed strong rela- tionships to union activity. Positive correlations between the index of participation and four of the union attitude factors showed significant (p < .01) correlations (Factor 3 - Philosophy of Unionism in General (r = .345), Factor 5 - Beliefs in Union Participation (r .289), Factor 2 - Attitude towards Union Power Relationships (r .220), and Factor 1 - Attitude towards Local Union (r = .201)). The factors dealing with general, philosophical issues of unionism bore stronger relationships with participation than the factors involving specific aspects. These findings are all in agreement with previous literature reported in Table 8. As mentioned previously,the subjects of this study (both actives and inactives) showed pro-union attitudes. Thus, the findings reported here are to be interpreted as matters of degree not kind. As Spinrad [1960] pointed out, pro-union attitudes are necessary though not sufficient for active participation. The correlation between the index of participation and Factor 4 (Attitude towards the International Union) was curiously very low (r = .005). We shall discuss the particular aspects of this factor later. Two other variables related to unionism showed significantly high correlations with the index of participation. .These were Tannenbaum and Kahn's measure of (l) perceived control of the local union by the rank and file, and (2) perceived total control available within the 69 union local (r = .244 and r = .194, respectively). The results are in direct agreement with Tannenbaum and Kahn's [1958] findings (see Table 7). These variables broaden our understanding of union partici- pation by focusing on union democracy (and power) in the measurement of distribution of control and by looking at the effects of outside forces (albeit indirectly) in the measurement of total control available to the local union. The correlations found in this and Tannenbaum and Kahn's study indicate that union participation is related to factors outside as well as within the local union. One such "outside" influence is the work situation. Factor 7 (General Job Satisfaction-Involvement) correlated significantly (r = .164) with the index of union participation. The items loading highest on Factor 7 emphasize the degree of involvement felt toward the job. Thus, members who are more active in their unions tend also to be more involved and satisfied with their jobs. This is in agreement with a number of previous reseachers (see Table 6). Furthermore, the low correlation between union participation and Factor 11 (Job Satisfaction- Specific Aspects) (r = -.026) agrees with the findings of Rose [1952] and follows the trend we saw in the union attitude measures, i.e., general factors correlated higher with participation than factors tapping specific aspects. One other variable correlates at the p < .01 level with partici- pation. That variable is Item 129 - Plant Size (r = -.142). This replicates the findings of a number of other studies (see Table 5). Basically, the smaller the plant in which a member works, the more likely it is that a member will participate. 70 Four variables significantly correlate with participation at p ‘<.05. Factor 12 (Anti—Conservatism in Life Style) correlates with participation .113. The six items that load highest on Factor 12 (and thus influence the factor the most) came from a scale developed by McClosky [1958] to assess the strength of general, not political, conservative beliefs in individuals and groups. The items reflect resistance to change (e.g., D41 "It's better to stick by what you have than to be trying things you don't really know about.") and thus anti- "active" behaviors. All items in this scale had been reflected so that responses to all items in the questionnaire indicated liberal responses thus the label "Anti-Conservatism in Life Style." The relationship found in this study lends further evidence to the generalization that individuals active in unions are also more active in other aspects of their lives (i.e., "joiners," outgoing personalities). It further indicates that liberal attitudes as a correlate of union activity are not restricted to social and political attitudes. Factor 8 (General Satisfaction with Life) correlates .091 with participation indicating that actives are generally more satisfied than inactives. Perhaps, this is an outcome of actives' more outgoing sociable nature. It is probably also a function of the items that loaded highest on this factor. They refer to aspects of the members' lives (e.g., job, family, union) that, as has been previously pointed out, are viewed positively. The positive correlation between level of education and level of union activity (r = .088) is in disagreement with the Tannenbaum and Kahn study. The results thus indicate that the union member of the 1970's who is more educated would be more likely to participate in union 71 activities. This was not the case in 1954 with the data of the Tannenbaum and Kahn study were gathered. Union members today have a higher educa- tion level in general (as is true of our whole society). There is also the possibility that this result occurred as a function of this particular sample. The members had volunteered for the programs at the education center and perhaps were more likely to have had a higher level of education. Finally, a significant negative relationship exists between union participation and age (r = —.084). The average age of this sample was 40. Thus, it might be suggested that members in this sample that are under 40 are more likely to be active than those over 40 years old. However, a curvilinear relationship (as was found in Uphoff and Dunnette, 1956) could exist. Future research could substantiate this. The remaining 11 variables have nonsignificant relationships with the index of participation. However, because of the criteria developed in Chapter IV (see Table 9, p. 51), they were still included in further analyses. The Multiple Regression Equation Before looking at the relative importance of the factors discussed above, a look at the factors as a group is appropriate. Using a multiple regression design with double cross validation, the results indicate that the group of 25 variables as a whole account for at the very least 29 percent of the variance in the criterion variable (the index of parti- cipation). Since the multiple R's calculated in each half-sample did not "shrink out of sight" upon cross validation, we are confident that at least some of our measures are important correlates of union 72 participation. The multiple R calculated on the entire sample (without considering possible shrinkage) was .6537, indicating that the 25 variables may account for up to 43 percent of the variance. These results also indicate evidence thatthe measures used in this study do have respectable reliability. The Relative Importance of the Independent Variables The research shows that this group of 25 variables contains important correlates of union participation. But which ones are more important? Which variables account for more unique variance in the criterion variable than the others? As mentioned previously, Darlington [1968] discusses five different measures of importance of predictor variables--three of which are of interest. We have just completed a discussion of one of these--zero-order correlations with the criterion variable. This is the only index that is not affected by any of the other variables. Another index, beta weights, is presented in Table 11, but since this study does not meet the necessary requirement for a causal analysis of the data, beta weights as an index of relative importance will not be discussed. Beta weights will only be discussed as they aid interpretation. Darlington's article indicates that use- fulness is clearly the measure of importance of the greatest interest here. The rank ordering of the 25 variables based on the usefulness index (the drop in R2 when a given variable is removed from a multiple regression equation and the beta weights are recalculated) is also presented in Table 11. The results generated by the use of this index will be discussed in conjunction with the zero-order correlations to achieve a more thorough understanding. It should be noted that any 73 interpretation of the results on any individual variable must be made relative to the variables included in this study, not as statements of importance in the absolute sense. Relative to the other variables, Factor 13 (Community—Political Activities), Factor 3 (Philosophy of Unionism in General), and Factor 9 (Socio-Political Liberalism) stand out as far more important predictors of union participation. These three factor scores were also the only variables studied that correlated .35 or greater with participation. These three factor scores are influenced more heavily than any of the other variables by items involving philosophical statements. Although each contributes a large share of unique variance, together they indicate that active union members hold liberal, pro—union philosophies and are active beyond the union hall. Based primarily on this finding, this researcher speculates that active union members espouse a theory of unions and union activities beyond their economic and protective functions. Active union members see their union as a vehicle in a social-political movement. In fact, this View of unionism may distin- guish active from inactive (even though loyal) union members better than any single factor. Future research is of course needed for sub- stantiation. Three other factors showed significant F's beyond the .01 level. Their unique variance would thus contribute very highly to the prediction of union participation. Factor 7 (General Job Satisfaction-Involvement) would have ranked only ninth if only zero-order correlations had been utilized. The use of Darlington's usefulness index highlights the greater importance this variable possesses in union participation 74 research when used in conjunction with variables similar to those used in this study. Active union members like and are involved with their jobs. The myth that union activists are nothing but disgruntled workers has again been repudiated. Purcell's [1953] "dual-allegiance" theory has seemingly been re-substantiated with this up-to-date data. As Hagburg [1966] points out, union actives (more than inactives) tend to receive satisfaction from union participation above and beyond the satisfaction gained from achievement of higher wages and better working conditions. This coincides with the finding that Factor 7 grossly out- ranks Factor 11 (Job Satisfaction-Specific Aspects) in usefulness as a predictor of union participation. Certainly, active union members show involvement with their union and their jobs, as well as the social and political system to which they belong. About as important as job satisfaction-involvement in the predic- tion of union participation is the union member's attitudes concerning how much participation in union activities is "right" (Factor 5). Active union members indicate that unionists "should" be more involved and act on that proclaimed ideal. This is again in basic agreement with Tannenbaum and Kahn's finding that actives are more homogeneous than inactives in their behaviors and attitudes. Attitude towards the International Union (Factor 4) ranks sixth on this index of importance. Thus, it is quite "useful" in predicting participation in union activity. However, the correlation between union participation and Factor 4 is very low (r = .005). Two general explana- tions emerge. First, although the relationship with participation is quite small, what does exist is very unique relative to the other 75 variables. Thus it is very useful. Second, Factor 4 is a suppressor variable. Its beta weight is relatively high and negative. Thus, Factor 4 is very "useful" in predicting participation because it "subtracts" out variance that is unrelated to participation in other predictor variables. Since the relationship between Factor 4 and the index of participation is so small as to be almost nonexistent, the second general explanation seems more probable. Thus, there exists some elements in Factor 4 shared with other variables that when removed aids the prediction of union participation. Although there is no easy content explanation of this, certain facts are available: First, Factor 4 correlates highest with Factor 13 (Community Political Activities) (r = .190) and Item 134 — Perceived Control of Rank-and-File (R = .191). Second, the highest loading item of Factor 4, Item A36, deals with concern of the International for the rank-and-file. Third, there is evidence that actives and inactives have pro-International Union attitudes. Pro-International Union responses averaged 72 percent across the 7 items loading highest on Factor 4. A total of 96 percent of the subjects answered "satisfied" or "very satisfied" to the global question C8 (Rate your feelings toward the U.A.W.) but this item is not part of Factor 4 because it was not part of the same factor analysis. Item C8 loaded highest on Factor 8 (General Satisfaction with Life). Fourth, Factor 4 and Item C8 correlate significantly (r = .256) but Item C8 correlates significantly with the index of participation (r = .302) whereas Factor 4 did not. Two possibilities emerge from the facts given above: First, Factor 4 operates as a suppressor variable in the regression 76 equation "subtracting out" that variance dealing with "individual concern" (as opposed to collective action) that is unrelated to level of union activity but which is a part of Factor 4, Factor 13, and Item 134. If this is true, union activists must stress collective rather than individual action. This is in agreement with Tannenbaum and Kahn [1958, pp. 125-128]. This would also lend further support to this researcher's speculation that active unionists view the union as a collective social- political movement above and beyond its functions of protecting individuals. Second, possibly in action to the above, Factor 4 may be slightly mislabeled or insufficiently developed. Although Factor 4 does correlate significantly with the global rating of the international union (Item CB), it may be concentrating on some specific aspects of attitudes toward the international union and not including enough general measures. As we have seen previously, specific aspects' measures correlate lower with participation than the general, philosophical measures. Future scale development is recommended. Two variables have usefulness indices significant at the .05 level: Item 129 (Plant Size) and Factor 12 (Anti-Conservatism in Life Style). The finding concerning Factor 12 indicates that union activitsts are not only involved with their union, their jobs, their community and politics but maintain a life style that is open to progress--to change. This lends further support to this author's speculation that union activity represents more to union activists than economic gain and protection. Union activists see unions as part-and-parcel of deeply rooted socio-political movements. This activist characteristic apparently carries over into almost all aspects of life. 77 The high index of usefulness found with the plant size variable combined with its significant negative correlation with participation demonstrates the importance of this variable in the understanding of union participation. In fact, the results of this study indicate that plant size is a better predictor of union participation than local union size. The two size variables do correlate highly with each other (r = .403). Apparently, they both tap much of the same variance in the participation index but plant size accounts for a greater amount. Four variables have usefulness indices that are significant at p < .10 and thus "suggest" possible importance in understanding union participation. However, interpretations of the four should be made with extra caution. Item 134 (Perceived Control of Rank—and-File) has a high correlation with participation (r = .244) and a usefulness index in this "suggestive" range (p < .10). Item 167 (Perceived Total Control) also had a high correlation with participation (r = .194) but the two "control" variables correlated so high (r = .69) that very little unique variance is contributed by the Total Control variable. Thus, perceptions of union members as to the distribution of control at the grass roots level of the local union may be important in predicting union participation but perceptions of total control adds nothing significant to our predictive ability. Factor 8 (General Satisfaction with Life) does have a significant though smaller correlation with participation (r = .091). It ranks ninth among our variables in importance using Darlington's usefulness index. Union actives apparently tend to be more satisfied with the various aspects of their lives and the unique variance of this satisfac- tion as measured by Factor 8 aids our ability to predict union activity. 78 Although Age (Item 116) does have a significant correlation with partici- pation (r = -.O84), it is correlated with a higher level with some of the other predictors. It does contribute some unique variance and is thus ranked in this middle range of importance. As mentioned earlier, the possibility of curvilinear trends in the relationship of this variable to participation is suggested future research. Sex (Item 115) does not have a significant correlation with parti- cipation (r = .078). It contributes something unique to the prediction of participation; however, only to the p ‘<.10 significance level. Its correlation indicates a possible trend that males participate in union activities more than females. This in agreement with Strauss and Sayles' [1953] finding in two different locals. In relation to other variables in this study, sex is not a very useful measure in predicting participation. Four variables had significant zero-order correlation with parti- cipation but nonsignificant usefulness indices. Items 167 (Perceived Total Control) and 123 (Level of Education), and Factors 2 (Attitude towards Union Power Relationships) and l (Attitude towards Local Union) fall into this category. Their lack of "importance" is based on their high correlations with other predictor variables. They thus do not explain unique variance in the criterion variable. This emphasizes an advantage of considering the "usefulness" index in interpreting the results of multiple regression when the predictor variables are intercorrelated. Nine of the variables have neither a significant correlation with the index of participation nor a significant F using Darlington's 79 usefulness index. These variables are: Factors 6 (Attitude towards Union-Management Relations), 10 (Faith in People), 11 (Job Satisfaction- Specific Aspects), and 14 (Voting Behavior), plus Items 120 (Size of Community Where Raised), 121 (Size of Community of Present Residence), 124 (Whether Parents were in a Union), 128 (U.A.W. Seniority), and 130 (Local Union Size). Although these findings contradict some previous literature, they were found in this 1974 sample of union members to be unrelated and "unimportant" to the understanding of participation in union activities. The Relative Importance of Blocks of the Independent Variables A brief word is in order concerning the relative importance of certain "blocks" of variables. Table 12 shows that the two sets of factor scores (the set of six factors dealing with attitudes towards various aspects of unions and the set of eight factor scores measuring a range of "other"--nonunion attitudes and behaviors) are far more important to the prediction of union participation than either the set of two items concerning control within the local union or the set of nine demographic items. The fact that the eight factor set ranked highest in importance lends further evidence to the speculation that the active union is involved in a socio-political movement above and beyond the economic and protective functions of the union. As stated earlier, active and inactive union members tend to have pro-union attitudes and perhaps the distinguishing quality of active unionists is this interest and activity in broader social movements. Since the group of nine demographic variables taken as a group indicated very little importance, further research on their relationship 80 to union participation should be de-emphasized. Such variables dominate previous literature in terms of frequency. It is now time for this area of research to "come of age" and move on to potentially more important aspects. Factor scores (and possibly scale scores) are highly recommended for future research. Indeed, in this area in particular, there appears to be a role for psychologists and other social scientists for the advancement of union research. Summary and Conclusions Tannenbaum and Kahn [1958] say, One of the major processes of social science is extra- polation from the specific to the general, from a sample population in hand to a universe beyond reach, from vari- ables and measures of limited scope to ideas concerning broader social processes. To stop short of such generaliza- tions is to be less than scientific, but to attempt them from insufficient data is also less than scientific, and perhaps more dangerous. Moreover, the line between appropriate speculation and rank guesswork is often difficult to discern" [pp. 235-236]. This final chapter has attempted to stick to the facts and explain as well as to speculate and generalize beyond the facts. This study draws the picture of the active union member as a person of high energy level, active in community and political behaviors, and with political attitudes leaning toward liberalism. The member is a strong believer in unions; believes that unions should have power; believes that workers should join and get involved in union activities; likes the local and international union; perceives the union as having considerable "grass roots" democracy, in fact, perceives that all levels within his union have a say about how things are decided in his local. He/she is satisfied and involved with the job and disagrees with 81 statements espousing resistance to change type attitudes. If one wishes to predict union participation, community-political activities, liberal philosophical attitudes, general pro-union philosophy, and involvement in one's job would be the most important factors to consider. All of these generalizations are, of course, in contrast to the inactive union member but only in degree not kind. To speculate beyond the confines of statistical significance, it appears that a trend in the results is indicated. That trend centers around the active, progressive, liberal, philosophical nature of the majority of significant results. Perhaps then, the active union member sees union activity as a means of participating in a social movement. As proposed in Chapter II, this dissertation provides a general overview of union research and an in-depth review of the research on union participation in the focused format of Tables 2 through 8. The relationship and relative importance of twenty-five variables to a broadly defined and reliable index of participation are presented. This dissertation shows the improvement factor scores provide in pre- dicting union participation at least over the use of demographic variables. Future research is indeed called for. Substantiation of the speculations presented, inclusion and investigation of more factors, and a greater degree of scale development should have high priority. This dissertation used a large, up-to-date, geographically stratified sample but from only one international union, and the results need to be sub- stantiated in other samples before generalizations to more union members is certain. Progress has been made, but much more needs to be accomplished. BIBLIOGRAPHY BIBLIOGRAPHY Aylward, M. S., Uphoff, W. H., Dunnette, M. D., and Kirchner, W. K. Development and validation of a union attitude questionnaire. Mimeographed Release. No. 7, June, 1955. Bailer, L. H. The negro automobile worker. Journal of Political Economy, 51, 1943, 416-417. Baritz, L. The Servants of Power. Middletown, Conn.: Wesleyan University Press, 1960. Barkan, A. E. The union member: Profile and attitudes. AFL-CIO American Federation. August, 1967, 1—6. Barkin, S. Union policies and the older worker. In Derber, M. The Aged and Society. Champaign, 111.: IRRA, 1950. . Psychology as seen by a trade-unionist. Personnel Psychology, 14, 1961, 259-270. Barre, E. W., Kerr, C. and Andod, C. W. Unions, Management and the Public. Chicago: Harcourt, Brace and World, Inc., 1967. Becker, W. Conflict as a source of solidarity: Some notes on the California farm labor scene. Journal of Social Issues, 2, (l), 1953, 25-28. Blake, R. R., and Mouton, J. S. Some effects of managerial grid seminar training on union and management attitudes toward supervision. Journal of Applied Behavioral Science, 2_(4), 1966, 387-400. , Mouton, J. S. and Sloma, R. L. The union-management inter- group laboratory: Strategy for resolving intergroup conflict. Journal of Applied Behavioral Science, 1, 1965, 25-57. Blueston, I. The union and improving the quality of worklife. Unpublished paper, January, 1974. Blum, A. A., Moore, M. L., and Fairey, B. P. The effect of motivational programs on collective bargaining. Personnel Journal, 52 (7), 1973, 633—641. Blum, F. A. Toward a Democratic Work Process: The Hormel-Packinghouse Workers' Experiment. New York: Harper, 1953, 43-47. 82 83 Bogard, H. M. Union and management trainees: A comparative study of personaltiy and occupational choice. qurnal of Applied Psyphology, ‘44, 1960, 56-63. Campbell, A., Converse, P. E., Warren, W. E., Stokes, D. E. The American Voter. New York: John Wiley and Sons, Inc., 1960, 210-212. Chinoy, E. Local union leadership in Gouldner, A. W. (ed.). Studies in Leadership. New York: Harper, 1950, 157-173. Conrad, H. S. Some principles of attitude measurement: A reply to "opinion-attitude methodology." Psychological Bulletin, 42, 1946, 571-588. Crespi, L. P. "Opinion-attitude methodology" and the polls--a rejoinder Psychological Bulletin, 42, 1946, 562-569. Cronbach, L. J. Coefficient alpha and the internal structure of tests. Psychometrika, 16, 1951, 297-334. Cureton, E. E. Validity, reliability, and baloney. Educational and Psychological Measurement, 19, 1950, 94-96. . Approximate linear restraints and best predictor weights. Educational and Psychological Measurement, 11, 1951, 12-15. Darlington, R. B. Multiple regression in psychological research and practice. Psychological Bulletin, §§, 1968, 161-180. Davis, K. and St. Germain, E. E. An opinion survey of a regional union group. Journal ongpplied Psychology, §§, 1952, 285-290. Dean, L. R. Social integration, attitudes, and union activity. Industrial and Labor Relations Review, §_(l), 1954, 50-58. Dual allegiance to union and management (A symposium). Personnel Psychology, 2, 1954, 41-80 Dubno, P. The role of the psychologist in labor unions. American Psychologist, 13, 1957, 212—215. Faunce, W. A. Size of locals and union democracy. The American Journal of Sociology, 1959, 291-298. Feinberg, M. R. Symposium introduction: Implications of psychology in labor-management relations. Personnel Psyphology, 14, 1961, 239-241. Form, W. H., and Dansereau, H. K. Union member orientations and patterns of social integration. Industrial and Labor Relations Review, 1954, 273-382. Gilmore, B. M. Cronbach's alppa inner-consistency_test reliability. Technical Report 89, Computer Institute for Social Science Research, Michigan State University, East Lansing, Michigan, 1968. 84 Ginzberg, E. The Labor Leader: An Exploratory Study, New York: Macmillan, 1948, 92-93. Gluskinos, U. M. and Kestleman, B. Management and union leaders per- ception of work needs as compared with self-reported needs. Proceedings of the Annual Convention of the A.P.A., §_(Part 2), 1970, 603*604. Gorer, G. The American Character. London: The Crescent Press, 1948. Guilford, J. P. Psychometric Methods. New York: McGraw-Hill, 1954. . Fundamental Statistics in Psychology and Education. New York: McGraw-Hill, 1965. Gullahorn, J. and Strauss, G. The field worker in union research. Human Organization, 1}, (3), 1959, 153-165. Hagburg, E. C. Correlates of organizational participation: An examination of factors affecting union membership activity. Pacific Sociological Review, 2_(l), 1966, 15-21. Herberg, W. The old—timers and the newcomers: Ethnic group relations in a needle trades union. Journal of Social Issues, g_(l), 1953, 12-19. Hero, A. The U.A.W. and world affairs, Report on a survey of members opinions. Unpublished report to U.A.W., July, 1965. , and Starr, E. The Reuther-Meany foreign_pclicy_dispute. Dobbs Ferry, New York: Oceana Publications, Inc., 1970. Howells, J. M., and Brosman, P. The ability to predict workers' preferences: A research exercise. Human Relations, 2§_(3), 1972, 265-281. Hoxie, R. Egade Unionism in the United States. New York: D. Appleton and Co., 1920. Hunter, J. E., and Cohen, S. H. Package Version I. 0. East Lansing: Michigan State University, Computer Institute for Social Science Research, June 1, 1971. Huszczo, G. E. Joint union-management organizational development efforts. Unpublished Paper, January, 1975. Kahn, R. L., and Tannenbaum, A. S. Union Leadership and member partici- pation. Personnel Psychology, 191 1957, 277-292. Karsh, B. Union traditions and membership apathy. Labor Law Journal, 2_(9), 1958, 641-646. 85 Kirchner, W. K., and Dunnette, M. D. Survey of union policy toward older workers. Journal of Personnel and Adminstration Industrial Relations, 1, 1954, 156-158. Knapp: M. L., and McCroskey, J. C. Communication research and the american labor union. The Journal of Communication, 18, 1968, 160-172. Kornhauser, A. The mental health of the industrial worker. New York: John Wiley and Sons, 1965. Kyllonen, T. B. Social characteristics of active unionists. American Journal of Socilogy, 56, 1951, 528-533. Lawler, E. E. III, and Hall, D. T. Relationship of job characteristics to job involvement, satisfaction, and intrinsic motivation. Journal of Applied Psychology, §§_(4), 1970, 305-312. Levenstein, A. The psychologist joins the labor conflict. Personnel Psychology, 14, 1961, 250-259. Lipset, S. M., Trow, M. A., and Coleman, J. S. Union Democracy: The Internal Politics of the International Typographical Union. Glencoe, Ill.: The Free Press, 1956. Liveright, A. A. Union Leadership Training. New York: Harper and Brothers, 1951. Locke, E. A., Smith, P. C., Kendall, L. M., Hulin, C. L., and Miller, A. M. Convergent and discrimant validity for areas and methods of rating job satisfaction. Journal of Applied Psychology, 48, 1964, 313-319. Lodahl, T. M. and Kejner, M. The definition and measurement of job involvement. Journal of Applied Psychology, 42_(1), 1965, 24-33. MacDonald, L. Leadership Dynamics and the Trade Union Leader. New York: New York University Press, 1959. Mahoney, T. Factors in union meeting attendance. Labor and Nation, Jan.-March, 1952, 1-4. McClosky, H. Conservatism and personality. American Political Science Review, 52, 1958, 27-45. McNemar, Q. Opinion-attitude methodology. Psychological Bulletin, ‘43 (4), 1946, 289-294; 301-304. Miles, R. E., and Ritchie, J. B. Leadership attitudes among union officials. Industrial Relations, 8 (1) 1968, 108-117. Miller, E. L. Job satisfaction and national union officials. Personnel Psychology, 19_(3), 1966, 261-274. 86 __, Job attitudes of national union officials: Perceptions of importance of certain personality traits as a function of job level and union organization structure. Personnel Psychology, 55_(4), 1966, 395-410. Miller, G. W. and Young, J. E. Membership participation in the trade union local. American Journal of Economics and Sociology, 55, 1955, 36-45. Mosier, C. I. Symposium: The need and means of cross validation. Educational and Psychological Measurement, 51, 1951, 5-11. Muench, G. A. A clinical psychologist's treatment of labor-management conflicts. Personnel Psychology,_lg (2), 1960, 165-172. Myers, M. S. Overcoming union opposition to job enrichment. Harvard Business Review, May-June, 1971, 37-49. Nunnally, J. C. Psychometric Theory, New York: McGraw-Hill, 1967. Perline, M. M., and Lorenz, V. R. Factors influencing member partici- pation in trade union activities. The American Journal of Economics and Sociology, 25, 1970, 424-438. Perlman, M. Labor Union Theories in America: Background and Development. Evanston, Ill.: Row, Peterson and Company, 1958. Psychologist in collective bargaining process. Personnel Psychology, 12_(4), 1964, 361-383. Purcell, T. U., S. J. The Worker Speaks His Mind on Company_and Union. Cambridge, Mass.: Harvard University Press, 1953. Reitzes, D. C. The role of organizational structures: Union vs. Neighborhood in a tension situation. Journal of Social Issues, 2_(l), 1953, 39-44. Rose, A. Union Solidaripy, Minneapolis: University of Minnesota Press, 1952. . The influence of a border city union on the race attitudes of its members. Journal of Social Issues, 5_(l), 1953, 20-24. Rosen, H., and Rosen, R. A. The Union Member Speaks. Englewood Cliffs, N.J.: Prentice-Hall, 1955. . Personality variables and role in a union business agent group. Journal of Applied Psychology, 4;, 1957, 131-136. Rosenberg, M. Occupations and Values. Glencoe, Ill.: The Free Press, 1957, 25-35. 87 Sayles, L. R. Field use of projective methods: A case example. Sociological Research, 55, 1954, 168-173. and Strauss, G. The Local Union. Chicago: Harcourt, Brace and WOrld, Inc., 1967. . What the worker really thinks of his union. Harvard Business Review, May-June, 1953, 94-102. Seidman, J., London, J., Karsh, B., and Tagliacozzo, D. L. The Worker Views His Union. Chicago: The University of Chicago Press, 1958. Sheppard, H. L. Task enrichment and wage levels as elements in worker attitudes. IRRA 26th Annual Winter Proceedings, 1974, 210-218. , and Herrick, N. Q. Where Have All the Robots Gone? New York: The Free Press, 1972. , and Masters, N. Union political action and opinion polls in a democratic socity. Social Problems, 5, 1957, 14-21. __, Mayer, A. J., and Kornhauser, A. When Labor Votes. New York: University Books, Inc., 1956. —_ Spinrad, W. Correlates of trade union participation: A Summary of the literature. American Sociological Review, 25, 1960, 237-244. Stagner, R. The psychologists function in union-management relations. Personnel Administration, 25_(1), 1963, 24-29. , Chalmers, W. E., and Derber, M. Guttman-type scales for union-management attitudes toward each other. Journal of Applied Psychology, 32, 1958, 293-300. , Rosen, H. Psychology of Union-Management Relations. Belmont, Ca1if.: Wadsworth Publishing Co., Inc., 1965. Stein, E. The dilemma of union democracy. Annals of the American Academy of Political and Social Science, 310, 1963, 46-52. Stogdill, R. M., Goode, O. S., and Day, D. R. The leader behavior of presidents of labor unions. Personnel Psychology, lz_(l), 1964, 49-57. Strauss, G. Control by the membership in building trade unions. American Journal of Sociology, 55, 1956, 527-535. , Sayles, L. R. Occupation and the selection of local union officers. American Journal of Sociology, 55, 1953, 585-591. . Patterns of participation in local unions. Industrial and Labor Relations Review, 1954, 31-43. 88 Symposium on psychologists in the union-management relationship. Personnel Psychology, lg_(4), 1961, 239-284. Tagliacozzo, D. and Seidman, J. A typology of rank and file union members. American Journal of Sociology, 61, 1956, 546-553. Tannenbaum, A. S., and Kahn, R. L. Participation in Union Locals. Evanston, 111.: Row, Peterson and Company, 1958. Tannenbaum, A. S. Control structure and union functions. American Journal of Sociology, 55_(6), 1956, 31-43. . Mechanisms of control in local trade unions. The British Journal of Sociology, Z, 1956, 275-282. University of California's Institute of Industrial Relations. Attitudes toward unionism of active and passive members. Monthly Labor Review, Feb., 1966, 175-177. Uphoff, W. H., and Dunnette, M. D. Overall favorable attitudes: Understanding the union member. Bulletin 18, Minneapolis: The University of Minnesota Press, July, 1956. Weinberg, N. Job Enrichment. An address presented at the 27th Annual Conference on Labor, New York City, June 14, 1974. Whyte, W. F. Who goes union and why. Personnel Journal, 35, 1944, 221-223. Wilensky, H. L. gptellecutals in Labor Unions. New York: The Free Press, 1956. Winer, B. J. gtapistical Principles in Experimental Design. New York: McGraw-Hill, 1962. Woodward, J. L. and Roper, E. Political activity of american citizens. American Political Science Review, 44, 1950, 872-885. Zeller, F. A., and Hagburg, E. C. Participation, the key to union effectiveness. The American Teacher Magazipg, Feb., 1965, 5-7. I 4 . APPEND ICES APPENDIX A RESEARCH QUESTIONNAIRE R ESEARCH QUESTIONNAIRE Background Information (all information will be kept confidential by the researcher). No Union official will see any individual's responses, however, I will need an address to send you another questionnaire six months from now. Address (Do not include your name) Are you a U.A.W. member a spouse of a U.A.W. member a spouse of a U.A.W. member and a member myself other (specify) Sex: M F Birthdate: Month Day Year Marital Status: Never married Married Previously married but no longer married No. of children living at home If married, do both Spouses have jobs? Yes, both full time, Yes, 1 full-l part time, No In what type of community were you raised? a) In the country b) Town of less than 2,000 c) Town of 2,000 or more but less than 10,000 (1) Independent city of 10,000 to 100,000 e) Suburb of a larger city f) Independent city of more than 100,000 or more In what type of community do you presently live? (Use same choices as above question and put your letter choice here ). What racial background do you consider yourself? Black, White, Oriental, Indian, Mexican-American, Other How many years of school have you completed (circle highest grade completed). 6 7 8 9 10 ll 12 some college or courses or less . beyond high school college beyond college degree Were either of your parents in a union? Yes No (Next 6 questions are for U.A.W. members only). How long have you been employed by the company you now work for? years months. What is your job title? How long have you been at your present job years months. How long have you been a member of the UAW? years months. About how many people are employed at the plant where you work (just guess)? About how many people are in your home local union? 89 90 For the following statements, make an X in the column of the answer that best describes the degree to which you agree or disagree with the statement. >‘ >40) m >s r4 5: r4ir4 m m )4 on u m L1H H 00 c m o .c m n .5 no G o m m now u u>m m 0 u u u -a u a -H w w H and Gov-l a) HH '1'" u cn-< 4: U)<: 2: UJCJ Cl ”3 Disa ree A1. If it were not for unions, we'd have little pro- ? tection against favoritism on the job. ”*0. A2. I think the best person should be kept on the ‘ job regardless of seniority. : A3. Unions impose too many restrictions on employers} . ~a 5.-..«np —.. .. A4. Charges of "racketeering" in unions are greatly i exaggerated. AS. Employees of a firm have better wages and working 3 conditions when all of them belong to unions. ; Q--(n—"-' A6. Unions should have something to say about whom the employer hires. i 6 A7. A nonunion shop usually pays lower wages than a t i union shOp: A8. Every worker should be expected to join the~ union where the person works. A9. The high wage demands of unions reduce chances for employment. A10. The growth of unions has made our democracy 1 stronger. i All. The selfishness of employers can be fought only 45y strong unions. IAI2. .Labor unions should be regulated to a greater extent by the federal government. A13. In a factory where there is a union, workers who i are not members should be required to pay the re- gular union fees if they are getting union rates} ofgpay. - i A14. If the majority of workers in a plant vote to l have a union, the others should be required to l gjoin. A15. I would like to see more joint union-management l . efforts concerning work procedures and work issues. 1 A16. I would like to see more cooperation in general 9 between union and management. . * A17. Without conflict with management, unions would not survive. A18. The average worker suffers from the conflicts between union and management. A19. Company management is very concerned about the welfare of the rank and file worker. A20. There is a conflict of opinion in this union between the older union members and the younger ones. - F”"""‘"i"“' ...... A21. Conflicts between groups within the union hinders success. 1 A22. My local union has too much power in our plant. i -0 :p-os “in--- - A23. The company management has too much power in I our plant. ' 91 Strongly Agree Slightly Agree ~Neutral Slightly Disagreg _Disagree Strongly Disagree Agree “I’d A24. In times of crises, the union should abandon democratic procedures in order to act more swiftlyi A25. I feel that too many things are already decided before the union meetings are held. A26. If you don't agree with the officers of our local union, you might as well stay home. A27. The union seems to be too closely tied to the Democratic party, A28. The international union spends too much time and money on political action. A29. Boycotts are a legitimate union weapon. A30. Young workers should be encouraged to seek union careers. A31. The officers of the international union are _paid too much. ' A32. Our international union takes its share of our dues but gives us very little help. A33. Our international union interferes too much in our local affairs. A34. We don't get enough help for our union educa- tional program from the international union. A35. Our international union provides the necessary facts and helps at negotiation time. A36. The international union is very concerned about the welfare of the rank-and-file workers. A37. The paid officers of my local are worth the money we_pay them. A38. I regard my union dues as a good investment. A39. My union got a "good deal" for me when the last contract was sigged. A30. The elimination of unions or drastic reduction of their strength would mean a quick reduction in my living_standard. A41. The international union has the support of most of its members. A42. I feel offended when someone criticizes my union» A43. I would like to see my spouse (wife or husband) become more involved with the union. A44. My spouse spends too much time with his or her involvement with the union. A45. There isn't a better union than the one I f belongito. A46. In case of a strike, I'm sure we'd stick togethet. ‘A47. Every union member should attend at least I two out of three of his/her local union meetiggs. 1A48. My union makes new members feel that it is worth- while for them to belong. A49. My union is quick to defend any member who doesn'tgget a fair deal from his/her boss. >.m o >~0 .3" Z? HHQ div-‘0 92 on u «Supt-dang}; : m '0 .5 m u .5 N) N) m omuoomucommom uuuwuswmmufi u xi no.4 “i o r4~4 ~4 5;c: m<m m o o :2“ 23:22:33.: 3:2 maqiqumgqmp D25. I believe my ideas are worthwhile for others to hear. D26. The answers I have given on this question- ~ naire have been swayed a bit towards what I thopght the union wanted to hear. D27. Generally Speaking, education for peOple who have been'out of school for over 10 _years is not very worthwhile. ifs. I want to obtain more education. D29. Federal aid to education is not high enough. D30. I find my union activity is educational for me. D31. I feel at ease in social situations with people of races different than my own. The more contact I have with people of races other than my own, the more I appreciate_peop1e of all races. D33. The average person has really no say in what goes on in this country. baa . Individual candidates are more impor- tapt than political party_affiliation. is. I believe the energy crisis was rigged by big business. D36. 1' m in favor of natibnal health care ___: legéglation. - D37. I‘m in favor of the Women' 3 Equal Rights _émendment. D38. I perfer the practical person to the person of ideas. D39. If you start trying to change things very_much, you usuallygmake them worse. D40. If something grows up after a long time, there will always be much wisdom to it. D41. It's better to stick by what you have than to be trying things you don't really know about. D42. We must respect the work of our fore- fathers and not think that we know better than they did. D43. A person doesn't really have much wisdom until he is well along in years. D44. tional programs. The union spends too much money on educa- 96 ' line, now the last section... Please answer. BI. 22. B3. B4. 85. : 87. EB. B9. £10. 811 O 312. £13. Haw satisfying to you is each of the following~aspects of your job; Neutral . a) BBB work itself ' b) isor c) e WOI’ d) t . e)lt securi ' f) tional rtunities ‘ g) t overall ob My work attendance in the last 6 months has been: '8) almost perfect b) better than most 5c) average d) below average a) poor ‘ . In the last six months have you written to a newspaper to suggest some- thing or let them know your opinion_ Yes ”No About what? Rate your racial attitude on this scale: , Very liberal “ . ' ' Very conservative l 2' 3 - 4 5 6 7 When you get together with your friends how often would you say you discuss tOpics like politics, taxes, or legislation? 3) very frequently b) often c) occasionally d) seldom.e) never What activities concerning politics (elections or legislation) have you participated in the last year (e. g. campaigning, fund raising, letter writing campaigns, etc.) were you a member of any community organizations or involved with any community services in the past year? _Yes :‘No. If yes, what organizations or services ’ In the last six months have you written or talked to your congressman or senator to suggest something or let them know your Opinion? Yes _No. About what? I vote in national elections Always Frequently Occasionally Seldom.Never I vote in local political elections Always Frequently Occasionally Seldom Never I would describe my political attitudes as: very very liberal . . conservative 1 2 3 4 5 . 6 7 Bow'many books have you read in the last six months? Have you attended any classes (adult education, U.A. W. courses, college courses, etc.) in the last six months? APPENDIX B DESCRIPTION OF SAMPLE APPENDIX B Description of Sample Item 115 Sex: 436 Males; 64 Females 116 Age: Mean 40.0 years (Range 20 to 63) Standard Deviation 8.7 years. 117 Marital Status: 13 Never Married; 448 Married, 36 Previously Married but no longer married 118 Number of Children: Mean 2.7 (Range 0 to 12) Standard Deviation 1.6 119 If married, do both spouses have jobs? 111 Yes, both full time, 69 Yes, 1 full time--1 part time, 192 No. 120 In what type of community were you reared? a) In the country 109 b) Town of less than 2,000 33 c) Town of 2,000 or more but less than 10,000 68 d) Independent city of 10,000 to 100,000 114 e) Suburb of a larger city 34 f) Independent city of more than 100,000 121 121 In what type of community do you presently live? a) In the country 51 b) Town of less than 2,000 33 c) Town of 2,000 or more but less than 10,000 58 d) Independent city of 10,000 to 100,000 136 e) Suburb of a larger city 97 f) Independent city of more than 100,000 103 122 What racial background do you consider yourself? Black, 58 White 420 Oriental 3 Indian 1 Mexican—American 8 Other 4 123 How many years of school have you completed? Mean = 11.7 Standard Deviation = 1.427 Less than High School (6-11) 131 High School (12) 215 More than High School 147 97 Item 124 125 126 127 128 129 130 98 Were either of your parents in a union? Yes No Company Seniority: Mean = 12.87 years Standard Deviation II (I) O) N Job Title: Not on union payroll Full time union job Job Seniority: Mean = 8.25 years Standard Deviation 7.33 UAW Seniority: Mean - 13.026 Standard Deviation = 8.22 Plant Size: Mean = 3830.8 Standard Deviation 4592.6 Local Union Size: Mean = 4771.527 Standard Deviation 6438.8 267 225 452 40 APPENDIX C RESULTS OF FACTOR ANALYSIS ON SECTION A OF QUESTIONNAIRE APPENDIX C Results of Factor Analysis on Section A of Questionnaire1 Factor 1- Factor 2 Factor 3 Factor 4 Factor 5 Factor 6 A58 .675* .163 .101 .188 .108 -.046 A55 .669* .083 .167 .121 .039 -.071 A49 .674* .067 .232 .037 .096 -.003 A37 .559* .062 .160 .116 .074 -.O62 A48 .500* -.062 .041 .173 .102 -.020 A50 .473* .156 -.188 .125 -.086 -.007 A56 .455* .118 .085 -.000 .111 .038 A59 .435* .119 .053 .214 .122 .151 A51 .425* -.092 .087 .011 -.003 .120 A25 .422* .386 .037 .160 -.168 -.068 A39 .325* -.019 .167 .321 -.051 -.118 A23 -.302* .073 -.028 -.128 .218 .081 A54 .294* .036 .244 .031 .055 .055 A22 -.035 .551* .018 -.077 .040 .154 A28 .139 .546* .173 .326 .067 -.008 A27 .156 .486* .074 .132 -.057 .069 A52 .144 .430* -.019 .140 -.047 -.062 A26 .239 .413* .102 .067 -.093 .122 A31 .133 .397* .174 .350 .069 -.075 A33 .089 .392* -.019 .373 .008 .042 A9 .052 .386* .105 .015 .110 .045 A19 .139 -.363* -.l74 .120 -.041 .042 A3 .090 .354* .168 .043 .161 .027 A12 -.063 .352* .216 -.035 .086 .055 99 100 Appendix C (Continued) Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Factor 6 A44 .062 -.329* .064 -.l44 -.203 .084 A2 .068 .308* .246 -.041 .127 -.180 A24 .062 -.274* -.080 -.002 .015 .095 A57 -.095 -.265* -.038 -.010 —.246 -.113 A18 -.184 -.237* -.044 -.lo7 -.045 .072 A17 .023 -.120* -.041 .064 -.004 -.104 A1 .122 .104 .445* .016 -.072 .154 A10 .104 .204 .391* .063 .139 .066 All -.027 .067 .387* -.015 .112 .268 A46 .334 .006 .385* .204 .010 -.023 A38 .177 .131 .375* .171 .216 -.030 A40 .111 .082 .370* .055 -.008 .060 A45 .251 .130 .367* .113 .130 -.l34 A5 .138 .122 .283* .039 .083 .055 A30 -.072 .131 .260* .009 .247 .046 A29 -.010 .085 .222* .006 .105 .003 A7 .097 .035 .213* .047 .039 -.123 A36 .223 -.001 .250 .646* .075 .100 A35 .112 -.031 .136 .558* .091 -.002 A32 .235 .364 .088 .464* .050 -.056 A34 .142 .213 -.075 .329* -.085 .074 A41 .240 -.110 .258 .278* .042 .066 A53 -.l99 -.121 .218 -.269* .154 -.019 A4 .128 .099 .145 .151* .034 -.108 A14 A43 A13 A47 A8 A42 A20 A21 A6 A15 A16 101 Appendix C (Continued) Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Factor 6 .160 .085 .200 .081 .515* .159 .123 .157 .168 .108 .440* -.074 .118 .046 .143 .081 .415* .126 .169 .047 .061 -.010 .396* -.052 .134 .175 .366 .033 .372* .122 .249 .126 .293 .092 .334* -.126 -.057 -.177 .001 -.176 .302* .060 -.O73 -.003 -.011 -.026 .214* .014 -.012 -.058 .104 .036 .112* -.034 -.010 .037 .098 .049 .198 .504* .045 .001 .063 .029 .001 .607* Eigenvalue 7.30 2.47 2.07 1.09 1.01 .87 Proportion of Variance Accounted For .124 .042 .035 .018 .017 .015 1Questionnaire may be found in Appendix A. Item numbers referred to in this Appendix are the same as found in the questionnaire. Asterisks indicate factor on which item loaded highest. APPENDIX D RESULTS OF FACTOR ANALYSIS ON SECTIONS C, D, AND E OF QUESTIONNAIRE APPENDIX D Results of Factor Analysis on Sections C, D, and E of Questionnairel Factor 7 Factor 8 Factor 9 Factor 10 Factor 11 Factor 12 Factor 13 Factor 14 D6 .781* .031 .018 .013 .037 .078 .142 .016 D2 .757* -.005 .163 .035 -.024 -.027 -.053 -.015 Ela .667* .069 .029 .031 .130 .047 -.070 -.096 D7 .637* .116 -.032 .012 .102 .006 .193 .037 D9 .634* -.017 -.007 .099 .295 .060 .011 -.037 E13 .625* -.026 .068 .047 .431 .100 -.060 .003 C3 .602* .221 .109 .057 .247 -.016 -.164 -.053 D4 .592* .000 -.070 -.156 -.080 -.196 .213 .113 D1 .548* .179 .142 -.003 -.078 -.065 .019 .041 D10 .518* -.011 .009 .131 .221 -.O37 .013 -.055 D5 .503* -.016 -.036 -.126 -.010 -.244 .222 .151 D14 .475* .035 .006 .200 .172 -.096 -.005 -.099 D16 .451* -.032 -.077 .149 .249 .082 -.035 -.059 D17 .447* .018 .095 .114 .008 -.059 .041 -.070 D3 .425* .037 .058 .008 -.156 -.071 .183 .136 E2 .215* .012 -.010 .096 -.037 -.033 .019 -.l40 C5 -.002 .619* .045 .039 -.151 .051 -.001 -.040 C11 .050 .616* .021 -.123 .029 -.184 -.137 .067 C4 .161 .614* .136 .187 .041 .040 -.094 -.058 C9 .090 .585* -.062 .104 .128 -.067 .060 .110 C8 .051 .563* .127 .000 .096 -.023 .272 -.047 010 .101 .510* -.033 -.008 .039 -.017 .155 -.059 06 .130 .417* .045 .255 .114 -.097 -.008 .009 C7 .154 .398* .097 .062 .332 .003 .343 -.039 C1 -.044 .350* .092 -.031 .025 .040 -.084 -.067 102 103 Appendix D (Continued) Factor 7 Factor 8 Factor 9 Factor 10 Factor 11 Factor 12 Factor 13 Factor 14 032 .039 .124 .503* .116 .136 .003 .075 .068 031 .090 .113 .493* .100 .078 .100 .076 .052 04 .049 -.011 .452* .126 .015 .175 .027 .009 024 .002 .055 .451* -.129 -.034 -.025 -.041 -.026 013 .043 .021 .444* .034 -.005 .026 —.008 -.004 025 .031 .056 .413* -.114 -.122 .035 -.005 —.092 015 .030 .002 .395* -.043 -.049 -.057 .052 .023 028 -.016 .080 .379* -.O6O — 076 .087 .128 .083 E11 -.039 .003 -.357* -.042 .051 -.127 -.058 .151 030 .075 .063 .354* .077 .127 .048 .300 .134 026 -.128 -.o2’. .314* -.021 -.003 .190 -.040 .066 037 -.013 -.000 .297* .021 .045 .085 .217 .085 036 .047 -.035 .272* -.042 -.094 -.024 .185 -.073 023 .052 -.013 .251* .034 .098 -.011 .065 -.064 044 -.075 .237 .244* .004 .012 .143 .111 .070 029 -.092 .053 .242* .047 -.126 -.017 .196 -.o75 011 .112 -.013 .224* -.003 -.181 -.101 -.054 -.091 035 -.042 .026 .209* -.151 -.098 -.041 .084 -.096 E13 .065 .072 .186* .043 -.009 .087 .161 -.016 018 .072 .069 .031 .696* .092 -.060 .002 -.072 019 .148 .032 .108 .661* .050 -.138 -.001 .022 020 .047 -.070 -.085 .533* -.037 .166 .133 .041 021 .083 .128 .043 .488* .100 .218 .125 .079 Elc .133 .121 -.032 .264* .128 .008 -.078 -.016 033 .111 .070 .047 .248* .207 .211 .224 -.060 104 Appendix D (Continued) Factor 7 Factor 8 Factor 9 Factor 10 Factor 11 Factor 12 Factor 13 Factor 14 Elf .189 .030 .002 .020 .610* -.021 .017 .035 e .004 -.000 -.042 .072 .543* -.029 .051 .025 d .051 .020 -.038 -.018 .443* -.039 -.003 .072 D12 .089 .050 .062 .111 .396* .068 .388 .069 CZ .345 .165 -.093 .053 .392* -.021 —.134 .151 Elb .152 .077 -.054 .103 .340* -.046 -.036 .126 D8 .207 .058 -.000 .118 .297* .032 .021 .104 D43 -.053 -.005 .129 .036 .020 .560* .081 .044 D41 -.118 .043 .179 .087 .011 .550* .118 .046 D42 -.070 -.046 -.019 -.057 .026 .438* .012 .002 D39 -.052 .038 .279 .037 .002 .400* .023 .058 D40 .061 -.082 «.133 .004 .052 .364* -.026 .078 D38 -.011 -.069 .109 -.019 .051 .352* -.033 .067 D27 -.038 .217 .146 -.011 .097 .243* .006 .062 D22 .022 .048 -.110 .079 .102 .212* .116 .103 D34 .006 -.009 -.062 -.047 .083 -.172* -.138 .019 E6 .064 -.030 .163 .019 .050 .140 .407* .020 E7 .027 .061 -.056 -.000 .050 —.022 ~.343* .147 E8 -.066 -.058 -.069 -.056 .023 -.023 -.342* .130 E3 .076 .047 .121 -.119 .044 -.016 .225* .075 E5 .036 .000 .178 .093 .081 .002 .208* .195 E9 -.061 -.031 .007 .062 -.035 -.054 -.183 .735* E10 -.062 4.067 -.050 .053 -.046 -.050 -.337 .728* 105 Appendix D (Continued) Eigenvalue Factor 7 Factor 8 Factor 9 Factor 10 Factor 11 Factor 12 Factor 13 Factor 14 7.04 3.73 2.52 2.32 1.70 1.39 1.29 1.14 Proportion of Variance Accounted For .096 .051 .035 .032 .023 .019 .018 .016 1Questionnaire may be found in Appendix A. Item numbers referred to in this Appendix are the same as found in the questionnaire. Asterisks indicate factor on which item loaded highest. HICHIGRN STQT 312931 111011