JOB smsmcnou: ws RELATIONSHIP TO OCCUPATIONAL, STRATIFICATION AND COMMUNITY VARIABLES Thesis for 9h. Dogma of M. A. MICHIGAN STATE UNIVERSITY Joanne Bubolz Eicher 1956 'L H" .518 IIIIIIIIIIIIIZIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 31046 1 3306 JOB SATISFACTION: ITS RELATIONSHIP TO OCCUPATIONAL, STRATIFICATION, AND COMMUNITY VARIABLES by Joanne Bubolz Eicher AN ABSTRACT Submitted in partial fulfillment of the requirements for the degree of Master of Arts in the Department of Sociology and Anthropology in Michigan State University East Lansing, Michigan 1956 JHESIS‘ "Ia—IO- 54’. 4 . (7‘ l JOANNE BUBOLZ EICHER ABSTRACT In an attempt to understand job satisfaction from a sociological perspective, this study investigated 108 male workers and the relationship of their social backgrounds to job satisfaction. A job satisfaction index was deve10ped from three questions probing job preference in order to measure whether a respondent indicated satisfaction with his present job. Four focal hypotheses guided the study: 1. High job satisfaction will be directly related to high occupational prestige. 2. High job satisfaction will be directly related to upward occupational mobility. 3. High job satisfaction will be directly related to high social position. 4. Job satisfaction will be related to community involvement. For the first hypothesis, Warner's technique for rating occupations, Edwards' occupational groups, and Hatt's situs system of rating occupations were employed as indicators of occupational prestige. All three measures of occupational prestige were found to be related to job sat— isfaction in the direction expected. In addition, upward occupational mobility was relat- ed to high job satisfaction scores to substantiate the sec- ond hypothesis. Mobility was considered from five perspec- tives: 1) number of Jobs held, 2) vertical occupational 2 JOANNE BUBOLZ EICHER ABSTRACT mobility, 5) generational occupational mobility, 4) situs mobility and 5) generational situs mobility. The first three variables upheld the hypothesis in the direction ex— pected, but the last two only demonstrated a trend toward supporting the hypothesis. In the third hypothesis, high job satisfaction scores and high social position were related. General status posi- tion was measured by Warner's Index of Status Characteris- tics and Edwards' social-economic groups. The three com— ponents (house type, occupational rating, and source of in- come) of the Index of Status Characteristics were tested to detect their sensitivity to job satisfaction; source of in- come exhibited the highest sensitivity. Class identifica- tion, education, and amount of income were also included as social position variables. Education was the only one found not to exhibit statistical significance in relation- ship to the job satisfaction index. The final hypothesis, job satisfaction will be re- lated to community involvement, was not borne out. Six variables were included to investigate the hypothesis: age, size of community of socialization, newspaper reader- ship, number of associational memberships, church attend- ance, and party membership. 5 JOANNE BUBOLZ EICHER ABSTRACT In sum, three of the four hypotheses were generally substantiated, and it was concluded for this study that the rewards from life, both on and off the job, are related in a positive manner. JOB SATISFACTION: ITS RELATIONSHIP TO OCCUPATIONAL, STRATIFICATION, AND COXMUNITY VARIABLES by Joanne Bubolz Eicher A THESIS Submitted in partial fulfillment of the requirements for the degree of Master of Arts in the Department of Sociology and AnthrOpology in Michigan State University East Lansing Michigan 195é ACKNOWLEDGMENTS I wish to thank the following members of the Depart- ment of Sociology and Anthropology for their aid: Dr. Charles P. Loomis, Dr. Charles R. Hoffer, and Dr. Duane Gibson. I also wish to acknowledge my debt to Dr. William H. Form, my major professor, for his generous guidance and constructive criticism. Gratitude is expressed to Mr. Gregory P. Stone for his encouragement and to my husband, Carl K. Eicher, for his continued support. TABLE OF CONTENTS CHAPTER I. INTRODUCTION . . . General Review of Literature . Relevant Findings of Literature Appraisal of Empirical Contributions . Appraisal of this Study Problem Reformulated . . . II. METHODOLOGY . . . . . . . . . . . . . . Research Site . . . . . . Sample Design and Composition Social Class Criteria Job Satisfaction Index . . . . . . . . Statistical Techniques . III. JOB SATISFACTION: ITS RELATIONSHIP TO CCUPATIONAL VARIABLES . . . . . . . Occupational Prestige . . . . . . . Warner's Occupational Rating . Edwards' Occupational Groups . . Hatt's Occupational Situs . . . Occupational Mobility . . . . . . . . Number of Jobs Held . Occupational Mobility . . . . . Generational Occupational Mobility . (I) 12 15 l4 l7 I7 18 23 30 32 55 55 57 58 42 45 45 1+8 CHATTER Situs Nobility . . . . . . . Generational Situs Mobility Summary of Chapter . . . . IV. JOB SATISFACTION: ITS RELATIONSHIP +3 () SOCIAL CLASS . Social Class . . . . . . . . . . . . Occupational Rating . . . . . . . Source of Income . . . . . . . . . House Type . . . . . . . . . . . . Other Social Iosition Variables Social-Economic Groups . . . Class Identification . . . . . . . Education . . Amount of Income . . . . . . . . Summary of Chapter . . . . . . . . . V. JOB SATISFACTION: ITS RELAT ONSHIP TO CONNUNITY INVOLVEMENT . . . . Community Variables . . . . . . . Age . . . . . . . . Size of Community of Socialization Newspaper Reaiership . . . . . Number of Organizational Memberships Church Attendance Party Membership . . . . 58 so so 62 65 65 67 7o 73 79 CHAPTER PAGE Summary of Chapter . . . . . . . . . . . . 90 VI. SUMUARY AND DISCUSSION . . . . . . . . . . . 92 Summary . . . . . . . . . . . . . . . . . 92 Discussion . . . . . . . . . . . . . . . . 96 LITERATURE CITED . . . . . . . . . . . . . . . . . IOO LIST OF TABLES TABLE _ PAGE 1. Percentage Distribution of Goldwater Sample According to Warner's and Edwards' Occu- pational Scales . . . . . . . . . . . . . . 19 2. Social Class Composition of Sample . . . . . . 25 5. Distribution of Scores for Job Satisfaction Index . . . . . . . . . . . . . . . . . . . 26 4. Distribution of Responses to Questions on Job Satisfaction . . . . . . . . . . . . . . 28 5. Association Between Warner's Occupational Rating and Job Satisfaction Index . . . . . 36 6. Association Between Edwards' Occupational Groups and Job Satisfaction Index . . . . . 38 7. Association Between Hatt's Occupational Situs and Job Satisfaction Index . . . . . . 41 8. Association Between Number of Jobs Held and Job Satisfaction Index . . . . . . . . . . . 45 9. Association Between Occupational Mobility and Job Satisfaction Index . . . . . . . . . 47 11). Association Between Generational Occupational Nobility and Job Satisfaction Index . . . . 49 3—1. Association Between Situs Mobility and Job Satisfaction Index . . . . . . . . . . . . . 52 vii TABLE PAGE 12. Association Between Generational Situs Mobility and Job Satisfaction Index . . . 54 15. Summary of Tests of Association Between Occupation and Job Satisfaction . . . . . 56 14. Association Between Social Class and JOb Satisfaction Index . . . . . . . . . . . . 59 15. Association Between House Type and Job Satisfaction Index . . . . . . . . . . . . 61 16. Association Between Source of Income and Job Satisfaction Index . . . . . . . . . . 63 17. Association Between Social-Economic Groups and Job Satisfaction Index . . . . . . . . 67 18. Association Between Class Identification and Job Satisfaction Index . . . . . . . . 69 19. Association Between Education and Job Satisfaction Index . . . . . . . . . . . . 72 20. Association Between Amount of Income and Job Satisfaction Index . . . . . . . . . . 78 21. Summary of Tests of Association Between Social Position and Job Satisfaction . . . 79 CHAPTER I INTRODUCTION During the years 1953-1955, Michigan State Univer- sity made a study of the social aspects of clothing in a Michigan city of 10,000 peOple. One of the aspects studied was the significance of clothing in the work situation of the men in the sample. Since one of the study aims was to evaluate the role of clothing in occupational adjustment and mobility, a great deal of information was gathered on several aspects of the occupational life of the men. While ‘the research reported on here has nothing to do with cloth- :ing, it is concerned with making use of the unexploited c>ccupational data which were gathered. In general, the purpose of this study was to inves- tzigate the relationship of job satisfaction to the social TDackgrounds of the respondents. The concern with job sat— i.sfaction resulted from an investigation of three questions Curiginally formulated to probe job aspiration. It soon be- <3zame apparent, however, that job aspiration appeared to be <3loselyassociated with job satisfaction. For example, £3Ome reapondents who did not indicate aSpiration for any (DIiher job did reveal a preference for their present jobs. jjrle’basic data for this thesis, therefore, presented two possible themes of study-~one of job aspiration and one of job satisfaction. The writer chose the latter to study. Since diverse occupations have varying class and status positions,l there arises a possibility that job sat- isfaction may differ within the occupational structure of a society. Specific studies will be reviewed which are rele- vant to this problem and to the problem of the relationship of other social factors and job satisfaction. The problem for this thesis has two facets: 1) 0p— erationally defining job satisfaction and 2) examining job satisfaction in reference to the respondents' position in the stratification, occupation, and community systems. General Review of the Literature The studies on job satisfaction have been so numer- ous, it is impossible to review them all. Therefore a few typical ones will be examined.2 Job satisfaction has been 1Emile Durkheim, Division 9f Labor, trans.'George Simpson, Glencoe, Illinois: The Free Press, 1949. Max Weber, "Class,Status, Party," From Max Weber: Essays in Sociolo , trans. and eds. H. H. Gerth and C. Wright MIIls, New York: Oxford Press, 1946. Arthur Salz, "Occupations," Encyclopedia 9f Social Sciences, ed. E. Seligman, Vol. XI, I935. 2The reader interested in the general area of job satisfaction is referred to the following issues in the journal Occupations, in which at least twenty current stud— ies involving facets of job satisfaction were reviewed each time: April, 1958; October, 1940; February, 1945; April, 1945; April, 1948; December, 1948; December, 1949; October, 1950; May, 1951; May, 1952; September, 1955; September, 1954; May, 1955- an appealing subject of study for sociology, industrial psychology, labor economics, and vocational guidance. Al- though their interests and perspectives overlap, the numer- ous studies may be grouped into three general areas: 1) the measurement of job satisfaction, 2) job satisfaction as an employee attitude or as a company morale problem, and 5) job satisfaction as a psychological adjustment variable for the individual. The first area of research, the measurement of job satisfaction, may be further divided into "direct" and "in- direct" measures of job satisfaction. The bulk of the studies have been patterned after HOppock's direct method of measuring this variable. Thus, the respondent is direct- ly questioned about how well he likes his job, whether he has ever thought of changing his job, and so on. The in- direct method of assessing job satisfaction consists in asking the respondent what job he would choose if he could start his occupational life over. That is, he is not asked about how satisfied he is with his job, but his answer pre— sumably will indicate present job satisfaction or job as- piration. The pioneer work by HOppOCk was based on his Job 1 Satisfaction Blank. This index consisted of nine ques- 1Robert HOppock, Job Satisfaction, New York: Harper, 19559 P0 245° 4 . tions, probing satisfaction by means of: A) Four Likert- type questions asking the respondent to choose one of seven alternative responses, B) Four dichotomous-type ques- tions, and C) An item which required the respondent to mark on a horizontal line his estimate of his satisfaction or dissatisfaction. 1 to the employed Heppock administered this index adult population of New HOpe, Pennsylvania. Of the 551 employed adults, 88 per cent replied, and Heppock concluded that about one-third of them expressed dissatisfaction. This technique of measuring job satisfaction was re- vised by Bullock who used essentially the same questions, but reworded them and refined the scoring scale.2 He in- creased the number of items to ten, constructed the ques- tions in identical form with five alternative responses for each question. The items probed areas of satisfaction with the work organization, the job itself, and the respondent's own position in the work group. His sample consisted of the employees and ex-employees of an animal breeder's trade association. Although he questioned the respondents in re- 1A previous and longer index had been administered to 500 teachers from fifty-one urban and rural communities of northeastern United States, ibid., pp. 147-212. 2Robert P. Bullock, Social Factors Related 39 Job §atisfaction, Research Monograph Number 70, Bureau of Busi— ness Research, Ohio State University, 1952, pp. 7-12. gard to their social background} he did not relate their satisfaction scores to their social backgrounds. Bullock‘s technique of measurement was the basis of four studies sponsored by the Psychology Department of Michigan State University. Trier used the job satisfaction scale to measure job satisfaction and related it to the oc- cupational status of 240 employees from six different in- dustrial plants.2 Schell tested the validity of Bullock's scale and the Science Research Associates inventory.5 He obtained a high correlation (r=.740) between the two indi— ces.4 Cheek compared Bullock's study with the Science Re- search Associates questionnaire and discovered that the two job satisfaction inventories correlated highly with each other, and that the shorter technique (Bullock's) could be substituted for the longer one.5 Khan was concerned with 1Ibid., Appendix E, pp. 87-95. 2Howard E. Trier, Job Satisfaction and Occupational Status, unpublished M. A. thesis, Department ofIPsychology, Michigan State College, 1954. 3The Science Research Associates inventory (known as the SBA inventory) is a questionnaire composed of seventy- eight items probing fifteen categories of job satisfaction including factors of job demands, working conditions, and P33- 4 . . . . William A. Schell A Stud of Em athic Ability and of the Validity of Some IndIces 0% 33b Satisfaction, unpub- IIshEd MI A. theEIs, Department 3? Psychology, Michigan State College, 1954. 5Gloria Lee Cheek, A Ps chometric Stud 2; Two Indi- ces 2; Job Satisfaction, unpuEIishéd M. A. t eSis,_Depart- ment of Psychology, 1955. job satisfaction as one variable among eight tests admin- istered to her sample. She reported some occupational dif- ferences.1 The indirect method of measuring satisfaction was used by Fortune magazine. They asked whether they would select a different job if they could choose an occupation 2 Understandably this survey yielded a higher over again. overall percentage of dissatisfaction, because the range of job choice offered the respondents was greater than in the direct technique. An earlier survey by Fortune which used the same technique revealed extreme differences in job sat- isfaction by occupations. In the sample, 29 per cent of the professionals as contrasted to 61.5 per cent of factory laborers signified that they would choose a different career, while 55.5 per cent of the professionals and 21.5 per cent of the factory laborers stated that they would stay in their own career.3 The difference is a striking one. 'Typical of the studies on job satisfaction in refer— ence to morale and company policy are the researches by 1Lilian Khan, Ag Exploratory Study 9: the Relation- shi s Among Personality Characteristics, Work’SItuation, 305 Satisfaction and Abilit £9 Empathize lg 3p Ipdustrial .EEEmework, UnpublIEhed N. A. thesis, Department of PsychoI- Ogy, Michigan State University, 1955. 5 2"The Fortune Survey," 3232223, 35: June, 1947’ pp. ‘6, lo. . 5"The Fortune Quarterly Survey: XI," Fortune, 17: LTan.uary,.l958. Po 86- Reynolds and Shister, Kerr, and Woods. Woods defined morale as: the liking or disliking in some degree the various sit— uations that occur in the process of working with others. It was presumed that degree of morale was represented by attitudes expressed in terms of like or dislike of the various factors of a man's job.1 This definition represents a broader view of factors relat- ing to job satisfaction. Woods was trying to find out whether or not persons liked their jobs, and be emphasized the hours of work, attitude toward company policies, super— vision, and colleague work—relationships. Woods equated liking it with satisfaction and disliking it with dissatis— faction. The results of his study concluded that there is more than one "morale," and that the "causes or origins of morale are to be found somewhere else than in specific work factors. Persons with similar work environment showed varied scores and wide ranges."2 This strongly suggests that the investigation of the social backgrounds of workers may be meaningful for job satisfaction, since scrutiny of the work environment has shown that it alone does not ac- count for contentment with one's work. _1Walter A. Woods, "Employee Attitudes and Their Re- lation to Morale,"_Journal 2f Applied Psychology, 28, 1944, 21bid. Research by Reynolds and Shister viewed job satis- faction as a problem of labor turnover. Those satisfied remained in their jobs--those dissatisfied left, causing problems in turnover, low morale, or lowered productivity. They also inquired about occupational aspiration. In their interpretation of findings, they stated that if a respond- ent signified satisfaction in remaining with his present work, he "had an unusual degree of inertia." These studies were also interested in approval of company policies and management practices in order to bring about job satisfac- tion and better production.1 The third area of study involves the psychological factors of the respondent as exhibited by his job satisfac- tion. Typical of the research in this area are studies similar to Schaffer's which are interested in discovering whether the overall satisfaction would vary directly with the extent to which individual needs are satisfied byajob.2 Relevant Findings of Literature for This Study Since this study is concerned with the occupational, stratification,and community positions related to job sat- 1Lloyd G. Reynolds and Joseph Shister, Job Horizons New York: Harper & Bros., 1949 , p. 77. 2Robert H. Schaffer, "Job Satisfaction as Related to .Need Satisfaction in Work," Psychological Mono ra hs, 67, -NO. 14, American Psychological Association, nc., 55. 9 iAsfaction, the literature will be reviewed in reference to these areas. Occupational Position. Hoppock's results reported tihat the mean satisfaction index of each group increased crver the group below it when the occupations were scaled according to Beckman's scheme: Group V Professional, managerial, and execu- tive Group IV Sub-professional, business and minor supervisory Group III Skilled manual and white collar Group II Semi—skilled Group I Unskilled manual 31:1 Trier's study the workers' occupations and those of tstieir fathers were ranked according to a five class occupa- 13i.onal status ladder, ranging from professional at the tOp t3C> unskilled at the bottom. Trier discovered that the high- ‘31? the occupational status, the greater the job satisfac- tion. He also discovered that workers at a higher occupa- t3jl<3nal status than their fathers' were more satisfied with their jobs.2 Super investigated the relationship of occu- PE113ional level to job satisfaction and found that in the Wiliite collar level, 86 per cent of the professionals, 74 P533? cent of the managers, 42 per cent of.the commercial 513<>ups were satisfied. In the manual category, 56 per cent --_n 1Hoppock, _p. 213., pp. 56-57. 2Trier, pp. gi§., pp. 55-54. 10 of the skilled workers were satisfied and for semi—skilled workers,48 percent were satisfied.1 The Fortune survey al- ready cited2 reported the same trend. Stratification. Hoppock equated Beckman's occupa- tional scale with social status;5 thus the findings re- ported above would hold true in this category. The studies surveyed did not relate such status ratings as Warner's or Edwards' to job satisfaction, but some of the factors of stratification such as income and education were often in- cluded in the research. Trier stated that workers reCeiv- ing higher wages were significantly more satisfied, while workers with higher education were shown to be only slight- ly more satisfied than those with lower education. (This was a contrary finding in his research, as he expected. workers with less education to be more satisfied.“) Thom- son investigated a sample of 181 men from the 1926 graduat- ing class of a midwestern university and found 67 per cent to be satisfied. Fourteen per cent expressed disSatisfac- tion and 18 per cent qualified their answers. These quali- 1Donald E. Super, "Occupational level and job satis- faction," Journal pg Applied Psychology, 25, 1959, pp. 547- 564. . . 2"The Fortune Quarterly Survey: XI." 92- cit., p- as. 3H0ppock. loc. cit. 4Trier, pp. cit., p. 54. 11 fications involved financial considerations, and Thomson concluded for this group that dissatisfaction seemed direct- ly related to income.1 Community variables. Not much work has been done relating community variables to job satisfaction. Among the variables tested, age and community background are most frequently used. Trier found that older workers in the same occupation were more satisfied than younger workers.2 Heppock found that the mean age of the teachers satisfied with their jobs was 7.5 years older (57 contrasted to 29.5) than those not satisfied.5 In his New Hope survey which covered all the occupations of the community, the correla— tion of age and job satisfaction was rather low, .21:t.O4.4 Kessler found that the veterans of his sample who were sat- isfied with their jobs were on the average eighteen months younger than the dissatisfied.5 As for childhood residence in relationship to job satisfaction, HOppock's study of 500 teachers revealed that 1William A. Thomson, "Eleven Years After Graduation," Occupations,l7, May, 1959, pp. 709-714. 2Tl‘ier, _O-Ee Cite, p. 540 3Hoppock, 92. cit., p. 40. 4Ibid. 5Milton S. Kessler, "Job Satisfaction of Veterans Rehabilitated Under Public Law 16, " Personnel and Guidance IQEEQEI 35. October. 1954, pp. 78- 81. 12 a higher percentage of those satisfied with their job were brought up in a community the same size as the community they were teaching in.1 Appraisal of Empirical Contributions The disadvantages of many of the above studies lay in the fact that the samples in several instances were too small, making certain statistical manipulations either dif- ficult or impossible to apply. In addition, generaliza- tions drawn about the job satisfaction of a particular oc- cupation could not be applied to a complete occupational structure of a community. The New Hope study was an excep— tion. The majority of the studies measuring satisfaction Were executed by the "direct" method, which contains the drawback of suggesting satisfaction or dissatisfaction to the respondent. This may put the respondent on the defen- sive; he may feel compelled to admit satisfaction. Both these techniques, and those of the morale surveys, were ad- ministered to workers in industrial work plants or offices and were not applied to independent operators or to profes- sional persons. The morale studies, too, seemed dominated by the assumption that if the worker exhibited high morale, this meant satisfaction, which in turn pointed to a low —__ 1Hoppock, _p. cit., pp. 201-202. 15 probability of turnover. Finally, although many of the studies tested the relationship of job satisfaction to some social variables, none of them studied the relationship of job satisfaction to these important sociological factors systematically or extensively. Appraisal of This Study This study has certain drawbacks of its own. The sample does not include the women or Negro workers in the community. In other areas, however, it overcomes some of the weak spots of previous research. The sample contains 108 men who had work experience. The sample was taken to represent the male occupational structure of a community of 10,000 pOpulation. Thus certain generaliza- tions may be drawn from this study which transcend a spe- cific occupational group. The "indirect"method of ascer- taining job satisfaction does not focus the reSpondent's attention on satisfaction, but allows for a spontaneous ad- mission of satisfaction or aspiration. In addition, the questions as asked were generally applicable, not only to manual and office workers, but also to entrepreneurs and professionals. The questions measuring job satisfaction - were not concerned with equating morale and satisfaction, but with the respondent's acknowledgment of satisfaction with his work.' This study will attempt to contribute a l4 sociological perspective to the field of job satisfaction, which can be added to the storehouse of psychological and vocational guidance information. Problem Reformulated Although empirical studies contain some contradic- tory findings, they do suggest that job satisfaction may be understood in the context of a person's social position as described by his position in the occupational, stratifica- tion, and community systems. None of the studies reviewed here have investigated these systems in an orderly fashion. However, they have shown that, since men with the same jobs exhibited large ranges in their job satisfaction, there must be other reasons responsible for job satisfaction than the factory environment and the job itself. This suggests the possibility that the social milieu from which a person de- rives may be a factor of considerable importance. It is to this question that this thesis is addressed. Several studies reviewed have suggested that job sat- isfaction varies directly with the status ranking of occupa- tions. They have also suggested that the general social position (which includes the occupational position, income, education, and style-of-life factors) also is related to a person's job satisfaction. In addition, job satisfaction appears to be related the person's community integration. 15 Such factors as age, organizational membership and attend- ance are gxtgg-job factors which could account for persons in the same job exhibiting differences in satisfaction. With these findings in mind, four guiding hypotheses for this study were formulated. Since each has been developed in some detail throughout the text, they will only be briefly stated here. 1. High job satisfaction will be directly related to high occupational prestige. 2. High job satisfaction will be directly related to upward occupational mobility. 3. High job satisfaction will be directly related to high position in the stratification system. 4. Job satisfaction will be related to community involvement. Procedure Before any of these hypotheses could be tested, a job satisfaction scoring device had to be improvised. Chapter II is devoted to this subject, as well as to a gen- eral review of the methodological background of the re- search. The first two hypotheses will be considered in Chap- ter III. To test whether or not the high job satisfaction scores are related to high occupational prestige, three l6 methods of rating occupations will be used: Warner's, Ed- wards' and Hatt's. To test whether or not high job satis- faction is related to upward mobility, five variables will be investigated: number of jobs held, occupational mobil- ity, generational occupational mobility, situs mobility and generational situs mobility. Chapter IV will analyze the hypothesis that job sat- isfaction will be directly related to high status position. To do this, the variables of social class (including the source of income, occupational rating, and house type which compose social class), social-economic groups, class iden- tification, education, and amount of income will be consid— ered. In Chapter V, the fourth and most generally stated hypothesis, job satisfaction will be related to community involvement, will be develOped. Variables of age, sociali- zation-place, newspaper readership, number of organization- al memberships, church attendance, and party membership will be analyzed as indices of community involvement and will be related to the job satisfaction index. Finally, Chapter VI will summarize the findings of the study and present some of its conclusions. CHAPTER II METHODOLOGY Since the data for this study were taken from a 1 the research site and sample selection for larger study that project are relevant for the present investigation. Therefore, the general background of the larger study will be reviewed. Research Site The research site was selected for several reasons. The general research had as its central postulate that clothing functions in social life as a symbol of social status.2 More specific purposes of the general study were to examine the relevance of clothing in everyday life--the work situation for the males and purchasing situation for the females-—in a community context. Therefore it was necessary to select a community of a small enough size to 1The study, "Consumer Problems in the Purchase of Clothing and Textiles," has been sponsored by the Michigan State University Agricultural Experiment Station and con- ducted under the general supervision of Gregory P. Stone. The Department of Sociology and Anthropology collaborated with the Department of Textiles, Clothing and Related Arts. 2Gregory P. Stone and William H. Form, "Instabilities in Status: The Problem of Hierarchy in the Community Study of Status Arrangements," American Sociological Review, 18, April. 1955. P- 153- 18 study in detail, yet with a wide enough range of occupa- tions and industries to reflect an urban influence. It was also desirable that a rural population be included nearby in order to compare rural-urban clothing variations.l Considerable data were already available for Branch county in south-central Michigan.2 Goldwater, the county seat, had a population of 10,000 and fulfilled the urban requirements. Its principal industries were wholesale and retail trade, metal casting, wood and metal manufacturing, 3 transportation (trucking) and miscellaneous services. Sample Design and Composition For the sample, about one hundred cases were select— ed which represented the complete range of male occupations in the city. In order to draw the sample, all the occupa- lGregory P. Stone and William H. Form. Clothing in- ventories and preferences among rural and urban families. Michigan State College Agricultural Experiment Station, East Lansing. Technical Bulletin 246, March, 1955. 2The social composition of the county had previously been completed for another study by the Department of So— ciology and AnthrOpology at Michigan State University. See John B. Holland, Attitudes toward Minority Grou s in Rela- tion to Rural Social Structure, unpublished . . disserta— tion,-Department of Sociology and Anthropology, Michigan State College, 1950. 3William H. Form and Gregory P. Stone. The social significance of clothing in occupational life. Michigan State College Agricultural Experiment Station, East Lansing, Technical Bulletin 247, June, 1955, p. 8. The writers use "Vansburg" as a pseudonym for Coldwater. IIhlEpill ‘H[ u I 19 tions of the married males in Coldwater were classified and ranked according to the seven-point occupational prestige scale of Warner.1 Within each of these occupational strata, male heads-of-households were drawn at random in prOportion to the percentage of all occupations held by married males in the city. This resulted in a stratified sample of 108 married men who were interviewed. A matched second sample was also drawn for selection of cases to fill in for re- fusals or chronic not-at-homes. Table 1 depicts the occu- pational composition of the men in the sample. This job satisfaction study concerned itself with the male respond- ents only. TABLE 1 PERCENTAGE DISTRIBUTION OF COLDWATER SAMPLE ACCORDING TO WARNER'S AND EDWARDS' OCCUPATIONAL SCALES‘ Warner's Occupational Prestige Scale Edwards' Occupational Scale 8 Professional . . . . . . . . ll Proprietors . . . Stratum one . . 12 Managers and officials . . . . . 9 Stratum two . . Stratum three . . 18 Office clerks and sales . . 27 Skilled workers and foremen . . 14 Semi- skilled and unskilled workers . . . . . . . . . . 24 Stratum seven . . 10 Domestics, attendants, and others . . . . . . . . . . . . 8 Stratum four . Stratum five . Stratum six . Totals . . 100 Totals . . . . . . . 100 (N = 108) (N = 108) j —— *Table reproduced from Form and Stone, qp. gt” 11.9. 1W. Lloyd Warner, Marcia Meeker, and Kenneth Eells, Social Class ig America, Chicago: Science Research Associ- ciates, 1939, p. . 20 The sample included married males who were experi— enced workers. Since there were no reapondents younger than twenty years of age, the just-out-of-school workers were excluded. The actual ages ranged from twenty 1K) seventy-nine. The sample was composed largely of native- born whites with native-born parents (only twelve respond— ents stated one or both parents were foreign-born). The sample also represented the social levels of the community. Social Class Criteria An investigation of the social levels was included in the research task. The definition of social class used in this research and in the larger study from which these data came, is based on a modified version of Warner's Index of Status Characteristics.1 The familiar upper-upper, lower-upper, upper-middle, lower-middle, upper-lower and lower-lower status groups were the result of this Index. In parts of the country newer than the East, Warner sug- gested, there may be only five classes with no differentia- tion made between a group of upper-uppers and lower-uppers. He also predicted the possibility that one of the four fac- tors comprising the Index of Status Characteristics--occupa— tional status, source of income, house type and dwelling 1Loc. cit. 21 area--may not be available. The Index is still effective, tiowever, if only three characteristics are used in comput- ‘:ing it, but certain alterations must be made in the weight- :ing of the items.1 In setting up the larger study, two changes were Inade in the Index of Status Characteristics because one of ‘the four characteristics was not obtainable and another had two be modified.2 Dwelling area ratings were dropped from ‘tlie Index because it appeared that the community did not Tbtave a clear-cut ecological structure reflecting clear-cut <1j_fferences. Although the lower status homes tended to be fcrund together, the high and middle status homes were often sfi_tuated next to each other. The modification of the Index also involved the oc— Culoational ratings. The ratings as developed by Warner did nOt: seem to reflect adequately the status system of occupa- . ticxns in the community. For example, the truck drivers lixring in Coldwater made up about five per cent of the ap- Droximately 2700 employed males. From observation it ap— Pealmed that their level in the community was thought to be kligher than in those communities Warner had studied.3 To lIbid., p. 185. 2Stone and Form, "Instabilities in Status;zlp cit,1xl54. 3Loc. cit. 22 correct this, ten local long-time residents of diverse oc- cupational backgrounds were asked to rate the occupations in the sample on a seven point scale.1 The individuals of the sample were then rated and the scores were included in computing the Index of Status Characteristics. a As a result, the Index of Status Characteristics as a computed for Coldwater included a modified rating of occu- pational status, source of income, and house type. The so- *1 cial classes of Coldwater, computed by this system, are shown in Table 2.2 Social class is considered in this study as a crude index of prestige. Two of the principal re searchers in the study found that for Coldwater, Warner's conception of a hierarchical status system was imprecise.3 There was, to be sure, a status system, but it did not con- Sist of a unidimensional ranking; the status groups at the top exhibited a split. Four factors were especially re- SPOnsible for the above findings: 1) status arrangement was not clearly reflected in community ecology, 2) lack of adequate "status reputation" existed for a sizable segment or the community, 5) consensus on status extremes and dis- agre ement existed in the middle range, 4) invasion of the "cosmOpo— 23 lites" threatened the upper class.1 These findings will be taken into consideration when summarizing the present study. TABLE 2 SOCIAL CLASS COMPOSITION OF SAMPLE Social Class Per Cent Upper . . . . . . . . . . 5.8 Upper Middle . . . . . . . . 15.5 Lower Middle . . . . . . . . 26.9 Upper Lower . . . 56.5 Lower Lower . . . 17.5 Total . 100.0 (N = 108) To obtain Job Satisfaction Index a measure of job satisfaction, questions were constructed which probed job preference in three dif- ferent dimensions. The "dimensions" refer to occupational Dreferences in reference to the respondent's present place 0f work, his general knowledge of jobs, and his own job hi Story. If the respondent spontaneously named his own jobs in these three dimensions, this was coded as indicating a high degree of job satisfaction. As posed by the statement 0f the problem in Chapter I, this was conceived of as a way ‘ 1Ibid., pp. 155-155. 24 of testing whether job satisfaction could be indirectly measured. The job satisfaction index Was constructed from three key questions from the work situation schedule: 1. Which job would you most like to have at your present place of work? 2. Of all the jobs you know of, what job would you most like to have? 5. Which of all the jobs you have had, did you like best? To refresh his memory about the work positions he had held, the respondent was asked to outline his job his- tory. This job history was also used to gauge occupational mobility of the worker. ‘ From these questions, an index of satisfaction with Present job was constructed by the following method. When answering "I like my own," "Satisfied with what I have," "I "1 S0"; her, Chum," or naming his own job as "Painter, a scozne of "one" was given for that question. To those ex- Pressing preference for a job different than their own, a SCOIWe of "zero" was recorded. The scores were added and a fou1u~point scale ranging from zero to three resulted. The E , 1The respondent had been asked "what is your present 30b?" Therefore, occupational titles given as answers were comPared to his own occupational label. 25 scores indicated the following: 0 - Present job was not given in response to any question. 1 - Present job was given as response to one ques- tion. 2 - Present job was given as response to two ques- tions. 5 - Present job was given as response to all three questions. The index was deveIOped in order to fulfill the ffiirst objective of this thesis: to operationally define j<>b satisfaction by setting up a means of indirectly judg- izig job satisfaction from questions about job aspirations. 111 addition, it was further reasoned that with three dif- ferent job aspiration dimensions, a score could be devel- oped indicating a range of satisfaction from those persons coritented with their present job in all three dimensions to thc>se not expressing satisfaction with their jobs at all. The distribution of the index of satisfaction with Present job1 appears in Table 5. There it is observed that roughly half of the respondents (fifty-one) fell into the 'munre satisfied" categories with scores of three and two. They* designated their own job as a preference to all three qwestions or to two of them. The other half (fifty-eight) 1Hereafter expressed as job satisfaction index. 26 fell into the "less satisfied" category; they preferred their job in only one dimension or in none. TABLE 5 DISTRIBUTION OF SCORES FOR JOB SATISFACTION INDEX — t ’ _—: Score Number Per Cent 5 22. 20. 4 2 29 26.8 1 58 55.2 0 19 17.6 Total 108 100.0 1 J - Most job satisfaction studies have probed only sat— i.sfaction with the present job, and their usual findings Iweport that very few of the respondents express dissatis- faction.1 Since eighty-nine of the one hundred eight re- Spondents of this sample expressed job satisfaction in at lweast one dimension, these results are in line with previ— ous ones. This index of job satisfaction is only a crude ap- IIPoximation of a scale and should not be considered as an absolute yardstick of job satisfaction. To test this, the Wt‘iter attempted to prepare a Guttman—type scale. This at- 1’ 1The percentages for dissatisfaction reported in the l1Wereture have run well below thirty-three per cent, and as 0" as three per cent. as tempt was not successful. Only three questions were in- volved, and the co-efficient of reproducibility for the three was .92. To be meaningful, it wouldhave had tobe per- fect (1.00) for three questions. Even then, the validity of a scale on the strength of three items might be ques- tioned. This study assumed that answering "present job" to any of the three questions carried equal weight. The possi- bility remains, however, that the dimensions probed are not equal in contributing to job satisfaction. The distribution of answers to each dimension is as follows: Seventy men selected their own job in answer to "Which job would you most like to have at your present Place of work?" Thirty-two men selected their own job in answer to "Of all the jobs you know of, what job would you most like to have?" Sixty men selected their own job in answer to "Which of all the jobs you have had, did you like best?" The combination of answers will also be reported here so that the incongruous combinations may be noted. 1‘able 4 depicts the number of responses to each of the eight possible combinations. As previously stated and Shown in Table 5, twenty-two of the reapondents selected their own job in answer to all three questions, while nine- teen indicated no preference for their present job. Of those twenty-nine men who selected their present job 28 TABLE 4 DISTRIBUTION OF RESPONSES TO QUESTIONS* ON JOB SATISFACTION Selection of present job in answer to Number Per Cent None of the questions 19 17.6 Question 1 only 22 20.4 Question 2 only 2 1.8 Question 5 only 14 15.0 Questions 1 and 2 5 4.6 Questions 1 and 5 21 19.4 Questions 2 and 5 ' 5 2.8 Questions 1, 2, and 5 22 20.4 Total I68 100.0 ’For questions, see beginning of this section. 29 twice, twenty-one chose question one and three, five chose one and two, and three, two and three. When these responses are inspected, possible incongruous combinations arise. For instance, why would a man admit satisfaction with his present job to questions one and two and not to three, or why would he omit question one and select two and three? The plausible combination is one and three in which a man indicates preference for his job at his present place of work and in relationship to all the jobs he had held, but perhaps, if choosing his life work over again, he would choose a different occupation. The selection of two and three or one and two will be noted here as seemingly il- logical combinations, but for the purposes of this thesis, the cases will be retained in the study. In other words, the respondents' answers will be taken at face value as in- dicating some degree of satisfaction. 0f the thirty-eight respondents who named their job only once, twenty-two said they liked their present job at their present place of work (question one), two liked their present job best of any they knew (question two), and four- teen liked their present job best of any in their job his— tory (question three). The only incongruous choice in this group seemed to be selection of their present job in answer to question two alone. It is strange for a man to say that his present job is the best of any he knows, yet not also 50 admit satisfaction to either one or both of the other di- mensions. The term "satisfied" will be used to distinguish those workers scoring three and two on the Index, while "dissatisfied" will denote those scoring one and zero. The writer realizes that respondents with zero scores actually may be content with certain aspects of their work world. Statistical Techniques A Chi-square test of significant association was com— jputed to ascertain the probability of independence between ‘the Job Satisfaction Index and the control variables.1 In (describing statistical significance or lack of it, the null krypothesis statement will be used. The following probabi1~ isty ranges will be denoted by the qualifying adjectives of: .05>;p>'.01 -- moderately significant .01) p:>.001 - highly significant .001) p - extremely significant 1The procedure for determining Chi-square followed iihat described by Margaret J. Hagood and Daniel 0. Price, {Statistics for Sociologists (New York: Henry Holt and Co., 319525, p. 5697 The formuIa used was an alternative formula, 2 12 = S-N, where s 2.2%; . 2Originally used by George W. Snedecor, Statistical Methods: A lied to Experiments _i_;_1 Agriculture and Biology, €Unes: Iowa tate Callege Press, 4th ed., 1946. 'Reproduced 1I1 Margaret J. Hagood and Daniel 0. Price, Statistics for .fhbciologists, New York: Henry Holt and Company, 1952. P- 51 In addition, the term "approaching significance" will be used when the probability range is .10; p> .05.. The index *was dichotomized when running the Chi-squares for every “table (the cases scoring three and two were collapsed, as *were those scoring one and zero). A coefficient of contingency, C, was computed and corrected1 for all the Chi-squares which were significant, to determine the degree of association between the vari- ables investigated. 1The formula used for computing the C was I 2 (2 a a%§———, ibid., p. 570. The correction for C was - N C’ = , as found in Thomas C. McCormick, Elementary S2- C tr tc Cial Statistics, New York: McGraw, Hill, 1941, p. 207. W CHAPTER III JOB SATISFACTION: ITS RELATIONSHIP TO OCCUPATIONAL PRESTIGE AND MOBILITY Two hypotheses relating to occupational position and experience provide the framework for the findings discussed in this chapter. Many of the previous job satisfaction studies have suggested or investigated by the direct method, ‘the first hypothesis, that the higher the prestige of the ‘xjcib, the higher the job satisfaction. To implement and ex- paind these researches, this study introduces a comparison 01? three different techniques of classifying jobs hierarchi- caally and relates them to the indirect measurement of job saitisfaction. The three techniques are: Warner's occupa— tiAonal strata,1 Edwards' occupational groups,2 and Hatt's sifrus system.3 It was felt that comparing three methods of Jot) ranking to job satisfaction would provide a more thor- Ouégh test of the hypothesis. _-_ 1W. Lloyd Warner, Marcia Meeker, and Kenneth L. EGIJls, Social Class in America, Chicago: Science Research ASSOciates, 1949, pT-l . , 2A1ba m. Edwards, Alphabetical Index 3; Occupations, Washington: United States Government Printing Office, 1937, p, 3, 3Paul K. Hatt, "Occupation and Social Stratifica- Eign'" American Journal 2.1.? W. 45. May. 1950. pp. ~539. 53 The second hypothesis, that high job satisfaction will be related to upward occupational mobility, was im- plicit in the literature, but this thesis will investigate the factors of mobility in greater detail and relate them to job satisfaction. Five variables will be included: ruumber of jobs held, occupational mobility, generational occupational mobility, situs mobility, and generational si- tus mobility. Occupational Prestige Results of previous job satisfaction studies showed T3a definite relationship between high .prestige jobs and high satisfaction. HOppock reported that each higher occupa- tional group in New Hope had a higher mean satisfaction in- dex than the one below.1 Trier ranked occupations on a fivepoint scale and substantiated his hypothesis that the higher the occupational status, the greater the job satis- fac.=‘t:ion.2 Super found white collar workers more satisfied than manual and within these two breakdowns, professionals more satisfied than managers, and managers more satisfied than commercial workers; as for the manual workers, skilled \ Harper, 1935 lRobert HOppock, Job Satisfaction, New York: 53 pp' 36-370 St 2Howard E. Trier, Job Satisfaction and Occupational - atus, unpublished M. A. theEis, Department of PsychOIOgy, M . man State College, 1954. P- 35' 54 inorkers were more satisfied than semi-skilled. Khan found tflnat supervisors experience the highest satisfaction, execu- 1Jives somewhat lower and workers the lowest. Walker and hdarriott found in their study of attitudes to factory work tihat skilled jobs carried a sense of status and higher sate j_sfaction than unskilled jobs.3 Centers also reported: There are distinct differences in satisfaction and dis— satisfaction among occupational groups with respect to their jobs. Whereas the peOple in the top occupational stratum are all satisfied with their jobs, large num- 4 bers in the lower occupational groups are dissatisfied. These studies point to the expectation of finding in t:txis study that people in the top occupational groups would exhibit more three and two scores in the job satisfaction figcuflex than the lower occupational groups. The contribution <31? this thesis in investigating prestige and job satisfac- tion is the comparison of three different techniques of ¥ 1Donald E. Super, "Occupational Level and Job Satis— faction," Journal _o__f_ Applied Psychology, 25, 1959, pp. 547- 5(54L., 2Lilian Khan, Ag Exploratory Study pf the Relation- s Among Personality Characteristics, Work Situation, sk1j_ %:§:§1%Satisfaction, and Ability to Empathize lg 32 Industrial 1?£irnework, unpublished M. A. thgsis, Department of Psychol- 0857, Michigan State University, 19559 P- 30- t 3J. Walker and R. Marriott, "A Study of Some Atti— Iidles to Factory Work," Occupational Psychology, 25, July, 1951, pp. 181-191. t‘ ALRichard Centers, "Motivational Aspects of Occupa- ZLCDIlel Stratification," Journal 2; Social Psychology, 28, 1948 , pp. 187-217. 55 ranking jobs and their relationship to job satisfaction. Warner's occupational ranking. Since Warner's occu- pational schema was basic to this whole study, prestige as measured by his ranking method was the first variable in- vestigated. Details of his ranking method will not be pre- sented here, but it is important to note that his occupa— tional strata are not composed of functional occupational groupings. His strata cut across different" occupational families, so that the top level includes doctors, lawyers, businessmen whose businesses are worth “375,000 or more, and t o p management officials . The data in Table 5 confirm the expected results. All of the first stratum have job satisfaction scores of tIIITee and two, while the second stratum is characterized by a modal score of three. Below the two upper strata (from 5 to 5), job satisfaction scores are scattered with most of them scoring one or two. The bottom strata are then heavi- ly concentrated in the dissatisfied scores of one and zero. The Chi-square computed for the table bears out the signif- ic"‘Eillrlce of the association between job satisfaction and co- cupational prestige, for the probability of the two factors being related falls between the .01 and the .001 level. The null hypothesis is rejected, for this is a highly sig- Also, the C of .47 indicates a fairly ni_ - . . fl cant assoczlation. St: . . . . . Pong degree of correlation. Warner's class1flcatlon of \ 1%, Chapter II, "Social Class Criteria." 36 (Docupations into a seven-point prestige scale and its rela- txionship to the job satisfaction index of this study sub— srtantiate the first hypothesis of the thesis.1 TABLE 5 ASSOCIATION BETWEEN WARNER'S OCCUPATIONAL RATING AND JOB SATISFACTION INDEX ' it Warner's Job Satisfaction Index C)<:cupational Rating‘ l3 2 1 0 Total Stratum l 5 5 - - g Stratum 2 7 1 1 15 Stratum 5 2 6 4 1 l2 Stratum 4 5 5 11 2 l9 Stratum 5 4 9 10 8 51 Stratum 6 2 5 4 4 15 Stratum 7 l - 6 5 10 4 Total 22 29 58 19 108 12 a 12.650 .Ol)p).OOl C = .47 ¥ \ ‘Rows 1-5, 4 and 5, 6 and 7 were combined in compu- t51t3i.on of the Chi-square. "Columns 5 and 2, and l and 0 were collapsed in com- putation of the Chi-square. \ x 0:? 1Hughes suggests that an occupation is a combination wora price tag and calling card, thus summing up in a few Pa (153 the economic and status position which an occupation pspr‘esents. Everett C. Hughes, "Work and the Self,‘ Soc1al g echolo at the Crossroads, ed. John Rohrer and Muzafer 1’ , ew—York: Harper and Brothers, 1951, p. 515. “I . 57 Edwards' Occupational Groups. The rating scheme preposed and used by Alba Edwards in the U. S. Bureau of Census classificationsl is somewhat different than the pre- ceding scheme of Warner. Edwards placed men of the same occupation into one group rather than cutting across occu- jpations as Warner did. The groupings are ranked hierarchi— <3ally3 however, from professionals on the tOp to unskilled wcxrkers and attendants on the bottom. It was expected, tnierefore, that the association of job satisfaction to this Itxnking technique would also support the hypothesis that the higher the prestige of the job, the. higher the job sat- isfaction. Table 6 demonstrates the same heavy concentration of Satzisfaction in the top levels of occupations. Twenty-five 0f ‘the first group scored three and two, as contrasted with thiirteen scoring one and zero. The distribution of scores for"the lower group shows almost the opposite finding, with tWenty-two respondents expressing dissatisfaction and only ten- exxpressing satisfaction. The two middle groups are agaiill almost evenly distributed, although weighted on the \ 1Edwards, loc._cit. sifj_ 2In Edwards' scheme, each group is a separate clas— for clation. Compatible groups were combined in this study vMVP!)I‘esentation purposes and because of the few cases in- (floureeci in some of the categories when running Edwards . €511 occupational groups against the job satisfaction in- 58 TABLE 6 ASSOCIATION BETWEEN EDWARDS' OCCUPATIONAL GROUPS AND JOB SATISFACTION INDEX Job Satisfaction Index‘ Occupational Group Total 5 2 l 0 Professionals, PrOprietors, Officials and Managers 15 12 9 4 58 Office Clerks and Sales Personnel 2 2 4 2 10 Foreman and Skilled Manual Workers 5 7 12 4 28 Semi-skilled, Unskilled Domestic and Attendants 2 8 l5 9 32 Total 22 29 58 19 108 7(2 = 8.956 .05)p) .02 c = .58 L: *Columns 5 and 2, and l and 0 were collapsed in computation of the Chi-square. J‘Ssatisfaction side. The statistical Significance is ex- hibited by the Chi-square in which the probability of the two Variables being related falls between the five and two per Cent levels. The C exhibits a moderate association be- twee1'1 job satisfaction and Edwards' rating- The hYPOtheSis '1 . 8 again borne out by this finding, although not as dramat1- Ca 113? as with Warner's scale. _‘ _.-.....i__.._ __..~ . _ - 39 One of the possible explanations for the difference obtained from the two methods may lie in the fact that War- ner's technique seems to be more sensitive to social strat— ification than Edwards'. For example, a minister with col- lege education is placed in the top level of Warner's groups while an untrained minister is placed in the second group. He treats the case of college professor-high school teacher-grade school teacher similarly, but in Edwards' Classification, all of these would be placed in the same category. Hatt's Situs Classification. In addition to the above two ways of classifying occupations, a third approach was also investigated which, although a rough approximation of Prestige, emphasized the occupational groupings even more than Edwards' system and de-emphasized the stratifica- tlon aspects.1 Because many researchers have criticized the usual prestige ranking methods2 the writer investigated the technique of situs classification for this study. This cone ept of situs will be briefly reviewed. As Hatt at- tempted to scale occupations by prestige, he found that a Scale could not be obtained for some of the occupations. \ 1Hatt, loc. cit. See 2For a review and critique of ranking occupations, , Theodore Caplow, "Measurement of Occupational Status," S 0 M 3.1: Work, Minneapolis: University of Minnesota 38. 54, pp. 50-58. 40 He then suggested a new hypothesis, that sub-groupings of occupations would scale even though the whole series of co- cupations would not scale on a single continuum. He there— fore proposed an eight-occupational situs system, and with- in each cf these placed several occupational families.1 Our sample had no respondents falling in three of the eight situs categories. l Hatt did not establish the situs system in strict ranks. However, close inspection of them discloses a very rough approximation of a ranking system. For example, pro- fessional and business situs are placed above the situs of manual work and service. The real emphasis is, however, on ranking within the situs; for instance, it must be recog- nized that the manual work category contains such diverse occupations as airplane pilot (a skilled technician) to skilled and unskilled factory workers. Unfortunately, Hatt's numerous categories and the few cases of the sample made it impossible to use the situs system completely. The political, recreation and aesthetics, and military cate- gories had no representatives in the sample. Table 7 presents the distribution of satisfied and dissatisfied respondents in relationship to their placement in the situs categories. The professionals,again, are 1Hatt, 10c. cit. 41 shown on the satisfied side of the dichotomy. The business situs, however, now contains all pr0priet0rs, managers, and white collar workers. The differences are not as clearly evident,for twenty-one express satisfaction in either two or three dimensions, while eighteen express less satisfac- .f tion. The manual-work and service categories now manifest TI larger differences. Thirty-four 0f the manual workers and ; four of the service workers are found on the less satisfied A side as contrasted to twenty-one satisfied manual laborers a and one satisfied service worker. TABLE 7 ASSOCIATION BETWEEN OCCUPATIONAL SITUS AND JOB SATISFACTION INDEX Job Satisfaction Index“ Occupational Total Situs‘ 5 2 1 0 Professional 2 4 1 - 7 Business 10 ll 15 5 59 ABriculture”* 2. - - -, 2 Manual Work 7 14 £3. 13 55 SeI'vice l - 3 1 5 Total 22 29 38 19 108 \32 = 5.082 .05) p) .02 C = .33 , ‘The first two and last two rows were collapsed in computation of the Chi-square. *‘Columns 5 and 2, and l and 0 were collapsed in com- Put ation of the Chi-square. ***This group was dropped out of the Chi-square compu- tatiOn. 42 Statistically, the null hypothesis can be rejected. Job satisfaction may be associated with the higher situs, for the probability of the Chi-square falls between the five and two per cent levels, and a moderate association ‘between the two variables is demonstrated by the C of .55. The first guiding hypothesis for this chapter is also sup— ported by the findings in Table 7. Occupational Mobility Implicit in most of the studies reviewed was the notion that upward occupational mobility within and between generations should be reflected in higher job satisfaction. Suoh a position may be derived from the first hypothesis; if people holding jobs with high prestige have high job Satisfaction scores, than high job satisfaction scores shOuld be exhibited also by those who experienced upward mobility.1 The number of jobs workers have held should al- so be related to job satisfaction. There is some evidence that professional people, for example, experience work sta- \ of 1One problem which will not be touched here is that 121 ‘riewing the variable of job satisfaction and its rela- Woonship to social variables in adynamic manner. In other. 0&3]? a, several of the respondents are employed in hierarchi— 93:1 positions and have only climbed one or two steps and ThpeCt to go up a few more before ending their work career. nee? may be satisfied with their present job as one being ulp(3§%ssary for the next step up. It will not, however, show s1; Jen the job satisfaction scores by the nature of the in- ent which forces them to focus their answers on the O‘DED for which they are next in line. 43 bility with fewer trial jobs than do unskilled workers who may shift from job to job throughout their work history. This, too, would support the general prestige hypothesis that the higher the position on the occupational ladder, the fewer jobs held, and the higher the job satisfaction scores. Therefore, five facets of mobility will be inves— 12igated in reference to their relationship to job satisfac- tiicn, They are: number of jobs held, occupational mobil- jxty, generational occupational mobility, situs mobility, and generational situs mobility. Number of Jobs Held. The empirical studies on job satisfaction show conflicting results regarding the rela- tziconship between number of jobs held and job satisfaction. A study done by the Standard Register Company concluded: Employees who had had many jobs elsewhere were more favorably disposed than employees who had worked only at this company or who had had only one or two jobs elsewhere. (331 ‘the other hand, Eckerman's and Kerr's studies refute this finding. Eckerman stated that the "grievers" (dissat- isfied workers) held more jobs and had worked longer than the "non-grievers."3 Kerr found a significant correlation \ 1Delbert 0. Miller and William H. Form, Industrial SElSLiiglggy, New York: Harper & Brothers, 1951, pp. 706-711. Yo) 2Standard Register Company, "How to Find Out What ten]: Workers Think about. You," Factor. Management and Main- “¥EE§3332, 106, August, 1948, pp. 81-91. 3121 3A. C. Eckerman, "An Analysis of Grievances and Ag- ea‘hed Employees in a Machine Shop and Foundry," Journal between high turnover rates and dissatisfaction.1 Since these studies were administered to particular working popu- lations (mainly factory workers), generalizations from them may not apply to a cross-section of occupations. The ex- pected results for this sample were in accord with the lat- 2 which suggested that higher prestige jobs 3 ter two studies are dominated by those with fewer previous jobs. The range for the sample was from one to eleven, with the mean falling at 4.4. The mean number of jobs for each of the scores of the index in Table 8 varies from 5.8 to 4.8, indicating that the more satisfied worker has had fewer jobs, although the difference is not great. - The biggest differences are found among the extreme- ly mobikeworkers. Thus, among those holding seven or more jobs, only three of them score three,and fifteen respond- ents score one and zero. The Chi-square test of association 2f Applied Psychology, 52, June, 1948, pp. 255-269. lWillard A. Kerr, "0n the Validity and Reliability of the Job Satisfaction Tear Ballot," Journal 2; Applied Psychology, 52, Aug., 1948, pp. 275-281. 2The study by Standard Register was more of a lay study than scientific investigation. Its purpose was to point out that companies with superior working conditions should "advertise" their advantages. ‘ 3Miller and Form, gpg. git. Although this is es- pecially true of professionals and 0f pr0prietors' sons who take over their father's business,a notable exception is that of the corporation climber, the junior executive who begins at a much lower position in the company than he later expects to attain. 45 TABLE 8 ASSOCIATION BETWEEN NUMBER OF JOBS HELD AND JOB SATISFACTION INDEX Job Satisfaction Index’ Number of Total Jobs Held 3 2 1 O l or 2 7 4 6 4 21 5 - 6 l2 18 21 ll 62 7 or more 5 7 ll 4 25 Total 22 29 58 19 108 Iflean.number of jobs 5.8 4.7 4.8 4.4 4.4 2:2 =- 7.2I+0 .05)p>.O2 t = .56 ‘Columns 5 and 2, and l and 0 were collapsed in com- ;nitation of the Chi-square. 1E3 moderately significant, and the null hypothesis cannot 'bee accepted. The probability of a Chi-square of such mag— iniytude occurring falls between the five and two per cent levels of significance, and the coefficient of contingency el'v’ipzresses a moderate association between the variables. T1lus the hypothesis, that the less the mobility, measured b5? ‘the number of jobs, the greater the job satisfaction, can be accepted. Occupational Mobilipy. The literature reviewed as background material for this study did not deal with the relationship of job satisfaction to occupational mobility. 46 With job prestige seeming to be a major factor in job sat- isfaction, it was posited that those persons exhibiting up ward mobility would also be more satisfied with their jobs than those exhibiting no mobility, while those exhibiting downward mobility would be least satisfied. The measure of occupational status used was the seven-point occupational prestige scale developed in Cold- water for use in the 130.1 The job history for each indi- vidual was classified in one of six possible mobility pat- terns: 1) Continuously upward, 2) Irregularly upward, 5) None, 4) Irregularities, no mobility, 5) Irregularly downward, 6) Continuously downward. There were so few in- dividuals in some of the six original mobility patterns, that it was necessary to collapse them into three in order to present a meaningful table and to compute the Chi- square. The three categories include an "upward mobile" group combining "irregularly up" and "up"; a "non-mobile" category combining "none" and "irregulare-no mobility"; and a "downward mobile" category combining "irregularly down" and "down." Table 9 demonstrates a definite trend for substan- tiation of the hypothesis. Thus, all downwardly mobile workers scored either zero or one in the index. Differ- 1Supra, Chapter II,'€ocial Class Criteria." 47 ences between the upward mobile and non-mobile workers are also in the expected direction, with relatively more re- spondents in the upwardly mobile group scoring three and two and expressing satisfaction. TABLE 9 ASSOCIATION BETWEEN OCCUPATIONAL MOBILITY AND JOB SATISFACTION INDEX 4. 4 m — u “2““ _.____.____ L m 11 .— Job Satisfaction Index‘ H Occupational Mobility 5 2 1 o TOtal Upward mobile 15 2O 19 10 (62) Non-mobile 9 9 14 6 (58) Downward mobile - - 5 2 (7) Total (22) (29) (58) (18)** (107)** 2:?- = 7.696 .1o> p>.05 U = .58 ‘— -—_v — j .— —r I J _: W J j ;— *Columns 5 and 2, and l and 0 were collapsed in com- putation of the Chi-square. ‘*One respondent's mobility could not be determined, for his work history was incomplete. Statistically, the null hypothesis cannot be reject— ed, for the probability of Chi-square was less than ten per cent and no less than five per cent. The Chi-square is approaching significance. However, since this is an ex- ploratory study with gross data, more refined studies may substantiate the direction of the results suggested here. 48 The coefficient of contingency of .58 demonstrates a moder- ate association between the variables of occupational mo- bility and job satisfaction. Generational Occupational Mobility. The previous study by Trier had demonstrated that those workers with higher occupational status than their fathers were more 1 In addition, it seemed that if the cultural satisfied. value of striving "up" is prevalent in American society, those persons who have climbed up from their father's posi- tion would exhibit job satisfaction. Since it was demon- strated that occupational mobility closely approached the significant level of relationship with job satisfaction, would another measure of mobility, that of comparing the son's occupation with his father's at age 40, approach the same level of significance? To compare the generational mobility of the son to his father's occupation, the respondents were asked to name their father's occupation at age forty (an age at which a person is usually in his stable occupation).2 Tables 9 and 10 reveal an interesting contrast in regard to downward mo— bility within and between generations. While Table 9 re- veals only seven cases of downward mobility within the oc- lTrier, 92. cit., p. 53. 2Miller and Form, gp. cit., p. 700. 49 TABLE 10 ASSOCIATION BETWEEN GENERATIONAL OCCUPATIONAL MOBILITY AND JOB SATISFACTION INDEX Job Satisfaction Index“ Occupggggizlifiggility 3 2 1 0 Total Upward 5 5 7 3 20 Upward, Same as Father 5 7 l - 11 None, Same as Father 4 4 5 l 12 Not different, Same as Father 5 5 9 l 16 Downward, Same as Father 5 2 6 5 l4 Downward A 4 5 10 8 25 Total 22 24 “‘ 56 "' 16 "‘ 98 *“ 762:8.060 .02>p>.01 Us .39 ’For computation of the Chi~square, the following rows were combined: 1 and 2, 5 and 4, 5 and 6. “Columns 5 and 2, and l and 0 were collapsed in com- putation of the Chi-square. “‘Mobility could not be determined for the rest of the respondents, as their fathers' occupations were not given. cupational careers, Table 10 shows thirty-nine cases of downward mobility when the son's occupation is compared to the father's. The respondents with upward generational mo- bility (rows one and two of the table) have twenty satis— fied with their jobs in either two or three dimensions, and SO eleven less satisfied.1 Those exhibiting no mobility show an even distribution of satisfaction and dissatisfaction, while for the downwardly mobile,twelve express satisfaction and twenty-seven,dissatisfaction. Table 10 demonstrates a statistically significant association between generational occupational mobility and job satisfaction, for the probability of the Chi-square falls between the two and one per cent levels. The C of .59, in addition, points to a moderate association between the variables. In regard to generational occupational mo- bility, it can be concluded that the expected results were verified for the sample; there is a significant association between job satisfaction and the comparison of a son's oc- cupation to his father's if the son has exhibited upward mobility. Situs Mobilit . In addition to charting mobility by Warner's method, Hatt's situs classification2 was utilized. This was done because Hatt filt that a move from one situs to another was psychologically more important than movement within a situs; thus, intra—situs mobility involves less risk concerning prestige and security, while inter-situs 1Again, we may have the possibility of bureaucratic risers who are not at their desired or expected position, who score one and zero. 2Hatt, loc. cit. 51 movement is a bigger risk. Hatt suggested that those who have moved between situses and returned will probably be more insecure than those who have not. Situs changes in- volve the acquisition of unfamiliar skills, becoming accus- tomed to new working conditions, and entertaining greater risks. A person not changing situs should reveal greatest security, while those changing situs and returning to pre- vious situs probably indicate an inability to adjust to the new conditions. Those changing and remaining in a new situs should indicate some security (although less than the no- changeperson), for they have apparently adjusted to a new work environment.1 There have been no previous empirical studies relat— ing situs mobility and job satisfaction. This study ex- pected to find that those persons not changing situs or those who changed and did not return would be the most sat- isfied with their jobs. The persons changing situses and then returning to the original situs would be dissatisfied. No clear cut pattern emerges in Table 11. The anal- ysis is hampered somewhat because'of the few cases involved in the categories of those who have changed situs and then returned. However, it can be noted that those not changing situs have a larger margin satisfied than those who changed lIbid. 52 and returned. Those who have changed and not returned show a trend toward an equal distribution but weighted on the dissatisfied side of the index. TABLE 11 ASSOCIATION BETWEEN SITUS MOBILITY AND JOB SATISFACTION INDEX Job Satisfaction Index** Situs Mobility‘ Total 5 2 l 0 Changed twice, No Return 4 12 ll 8 55 Changed once, No Return 5 4 8 4 21 Changed once, Return 1 5 l - 5 Changed twice, Return 2 l 6 4 15 None 10 9 12 5 34 Total 22 29 58 19 108 Z2 a 1.660 .50)p>.30 H IM‘ 1% *The 5rd and 4th rows were.combined in computation 0f the Chi-square. "'Columns 5 and 2, and 1 and 0 were collapsed in com- Dutation of the Chi-square. The statistical computation for the data did not Support the hypothesis. The probability of the Chi-square -flills between the fifty and thirty per cent levels of sig- 53 nificance; the null hypothesis cannot be rejected. A pos- sible explanation for this lack of significance may be that situs mobility as presented above does not involve the pres- tige which Warner's technique measures. If there were enough cases to separate the change by particular situs in addition to the above categories, trends might be apparent. In addition, holding age constant might also prove useful in measuring this type of relationship, as older men have had more opportunity to make a situs change than the young- er workers. Generational Situs Mobility. Previous research on job satisfaction had not been concerned with a variable such as generational situs mobility. However, it was ex- pected that a situs change-no return of the son from his father's situs which involved a prestige change would be accompanied by job satisfaction, while the workers who changed and returned would be most dissatisfied. The data as shown in Table 12 do not bear out the hypothesis. There were so few cases contained in the cate- gories of changing situs and returning, that these two rows had to be drOpped out of the computation. Although small numbers are involved in the comparison of the no-change Category against the change—no return categories, differ- ences suggest that those not changing had a larger propor- tion.of cases (sixty-two per cent) than those changing 54 situs (forty-one per cent and forty-three per cent, respec- tively). TABLE 12 ASSOCIATION BETWEEN GENERATIONAL SITUS MOBILITY AND JOB SATISFACTION INDEX Generational Job Satisfaction Index“ . Total Situs Mobility" 3 2 1 0 Changed situs IO 8 18 8 44 Changed situs (was in father's) 5 12 10 9 54 Changed and returned (was in father's) - 1 1 - 2 Changed and returned to father's 1 l l - 5 No change 8 5 7 1 21 Total 22 27*" 57“ 18** 104’" 12 = 2.663 .3o>p>.2o * ‘Rows three and four were drOpped out of the Chi- square computation. HColumns 5 and 2, and 1 and 0 were collapsed in com— putation of the Chi-square. **'Generational situs mobility could not be determined for the remaining respondents. The p of the Chi-square computation is not statis- tically significant according to our standards, for it falls between the thirty and fifty per cent levels. The null hypothesis cannot be rejected. The significance is, 55 however, closer to showing a trend than in the previous table. This might suggest, as did the relationship of gen- erational mobility, that a son's making a situs change dif- ferent from his father's is of more importance to his job satisfaction than his own situs mobility. This is only a conjecture for future research, however, since the sample involved did not permit a rigorous test of association.1 Summary. The five tables investigating facets of . occupational mobility permit tentative acceptance of the hypothesis,that the less the mobility the greater the job satisfaction. To ascertain whether or not upward mobility is directly related to job satisfaction will take further research. It can be stated, however, that the occupational position of the son in comparison to his father's job seems to be significantly associated with job satisfaction. Summary of Chapter Table 15 summarizes the relationships found between the eight variables investigated in this chapter and the job satisfaction index. The first three variables exhibit- ed a significant association with the job satisfaction in- dex, thereby upholding the first hypothesis that there will 1This refers to the two categories which were drOpped When computing the Chi-square. 56 TABLE 15 SUMMARY OF TESTS OF ASSOCIATION BETWEEN OCCUPATION AND JOB SATISFACTION Level of Degree of Variable Significance Association (74 ) (U) Warner's occupational rating p<.Ol .47 Edwards' occupational groups p (.05 .58 Hatt ' s occupational situs p < .05 . 55 Number of jobs held p < .05 .56 Occupational mobility 1)(.1£> .58 Generational occupational mobility p < .02 . 39 Situs mobility p < .50 —-‘ Generational situs mobility p ( . 5O -—"‘ "Probabilities of the Chi-square were too high to make calculation of the coefficients of contingency mean- ingful. be a direct relationship between high prestige occupations and job satisfaction. The last five variables did not val- . idate the second hypothesis as conclusively, however. CHAPTER IV JOB SATISFACTION: ITS RELATIONSHIP TO SOCIAL CLASS AND INDICATORS» OF SOCIAL STATUS It was suggested in Chapter I that satisfaction with a job seemed to be correlated with the general status posi- tion of the person. The purpose of this chapter is to test more systematically the third guiding hypothesis: the high- er the status position of the person, the higher the job satisfaction. The variables investigated will be: 1) War- ner's social class, including its components of a) occupa- tional ranking, b) source of income, and c) house type; 2) Edwards' social-economic groups; 3) class identification; 4) education; and 5) amount of income. The two variables, Warner's social class and Ed- wards' social-economic groups,measure the general status position in different ways with a different emphasis. They were included to provide a comparison between two different methods. Class identification was added in order to ascer- tain the discrepancies between the person's objective social position (in terms of being rated "scientifically") and the position he attributes to himself. Income and education, important aspects of social position, were also investigat- ed because of their relationship with occupational posi- 58 tions. Income is usually the result of a man's work, and education often provides an entrée into an occupation. Social Class Social Class.1 While the job satisfaction studies reviewed did not use any measures of social class, factors related to or an integral part of social class were often studied. Although there were varying findings about the association of amount of income and job satisfaction,2 the trend of the research pointed to a direct relationship be- tween the two variables. The occupational position of the respondent plays an important part in determining social position,5 thus the findings relating job satisfaction and occupational prestige would suggest that the same associa- tion would exist in social class. As a matter of fact, most of the studies reviewed used the occupational position of the respondent as an indicator of his class position. Table 14 depicts a high proportion of the upper and upper-middle classes showing job satisfaction and no dis- satisfaction. This may have been a function of the small 1Social class denotes placement into classes by War- ner's Index of Status characteristics as used in the larger study. 2 chapter. 3 This will be discussed more fully at the end of the Supra, Chapter II, "Social Class Criteria." 59 sample involved, but it did suggest a trend worth noting. The lower-middle to lower-lower classes showed a concentra- tion in the lower scores of the satisfaction index. The significance of the two variables was demonstrated by the Chi-square computation, in which the probability of the Chi-square was less than .001. The null hypothesis may be considered untenable. In addition, the C of .51 indicates a fairly strong degree of correlation. TABLE 14 ASSOCIATION BETWEEN SOCIAL CLASS AND JOB SATISFACTION Job Satisfaction Index‘* Social Class" Total 3 2 1 0 Upper 4 2 - - 6 Upper-Middle 7 5 5 - 15 Lower-Middle 5 8 l5 4 28 Upper-Lower 6 ll 15 ll 41 Lower-Lower 2 5 9 4 18 Total 22 29 58 19 108 762 . 15.536 p<.001 U =- .51 ‘The top two and bottom two rows were combined in computing the Chi-square. “Columns 5 and 2, and l and 0 were collapsed in com- putation of the Chi-square. 60 In view this relationship, it must be remembered that social class in Warner's terms is the result of an In- dex of Status Characteristics score.l For this study, three characteristics comprised the 1802: occupational rating, source of income, and house type. When viewing these separately, it is realized that each constitutes a rough index of stratification. Since the relationship be- tween social class and the job satisfaction index has been demonstrated, it will be expected that each variable is significantly associated with the job satisfaction index. Each of the component variables will be related to the job satisfaction index to determine their relative degree of sensitivity to job satisfaction. Occupational rating. This characteristic has al— ready been discussed in Chapter III.5 The probability of association of the two variables fell between the .01 and .001 level of significance, and the C of .45 indicated the degree of association. Source of Income. Warner included source of income in his Index instead of amount of income, for he discovered that the former was more meaningful in determination of so- 1Warner, £2 51., pp. cit., Chapter X. 2Supra, Chapter II, "Social Class Criteria." 5Supra, Chapter III, "Warner's occupational rating." 61 cial class and also more easily and accurately found out than actual amount earned. He again used a seven-point scale (as he had in occupational rating) ranging the income sources from high to low, including: 1) Inherited Wealth, 2) Earned Wealth, 5) Profits and fees, 4) Salary, 5) Wages, 6) Private relief, and 7) Public Relief and non-respectable income.1 Table 15 only includes those categories which were represented by the sample in Coldwater, thus category 1 and 7 are not indicated. TABLE 15 ASSOCIATION BETWEEN SOURCE OF INCOM AND JOB SATISFACTION INDEX j _ -: __-: Job Satisfaction Index“ - r Source of Income‘ Total 5 2 l O Earned Wealth 8 5 - — ll Profits and Fees 4 7 5 l 17 Salary 3 8 12 7 50 Wages and Private Relief 6 10 19 10 45 Total 21 28 56 18 105*" 12 =- 18.158 p<.001 U :- .52 ‘The categories, in some cases, included averaged scores in which the respondent received income from two sources, such as earned wealth 32d profits and fees, or profits and fees and salary. "‘Columns 5 and 2, and l and 0 were collapsed in compu- tation of the Chi-square. **‘Source of income information was not available for remaining respondents, although it had been obtained and computed in the social class rating. 1Warner, 33 $1., gp. cit., p. 158. 62 All of those with earned wealth income scored three and two in the job satisfaction index, while the category of wages and private relief included twenty-nine of forty- five reapondents, who indicated dissatisfaction. The middle groups who earned profits and salary show more even distributions, but the heaviest concentration in the prof- its and fees category falls on the satisfied side of the in— dex, while for salary falls on the dissatisfied side. An extremely significant relationship (p of less than .001) was obtained between source of income and job satisfaction, indicating that this variable is even more sen- sitive than the previous variable of occupational rating. The association between them is also exhibited by the CI of .52. House Type. Warner stated that the house in which a person lives is related to his social status, in that it is a part of his wealth and reflects his social position. He concluded that the criteria for judging a house and assign- ing it a score from his seven-point scale should be focused on size and condition. His final house-type scale includes the following categories: 1) Excellent, 2) Very good, 5) Good, 4) Average, 5) Fair, 6) Poor, and 7) Very poor.1 Inspection of Table 16 discloses that those respond- ents living in excellent and very good homes signified job 1Ibid., p. 143. 65 satisfaction to a large degree. Those living in good and in average types of dwellings were more heavily concentrat- ed on the dissatisfied side of the index rather than on the satisfied, but the difference was not too great (twenty— four and nineteen respectively). Those persons living in lower housing categories, however, indicated more dissatis— faction, for twenty-eight scored one (u? zero, and seven- teen scored three or two. TABLE 16 ASSOCIATION BETWEEN HOUSE TYPE AND JOB SATISFACTION INDEX Job Satisfaction Index‘ House Type Total 5 2 1 _ 0 Excellent and Very good 9 4 2 - 15 Good and Average 7 l2 l6 8 45 Fair, Poor, and Very poor 5 12 18 10 45 Total 21 28 56 18 105‘" 942 a 11.119 .01>p >.001 U . .45 W L _: ‘.:- ‘Columns 5 and 2, and 1 and 0 were collapsed in com- putation of the Chi-square. *‘House type information was not available for five respondents, although it had been obtained and computed in the social class rating. 64 The statistical computation bore out the hypothesized results, for the association between the two variables was found to be significant, the probability falling between the .01 and .001 levels. The C of .45 indicated the degree of association. To sum up the findings for the relationship between job satisfaction, social class and the characteristics com- prising social class, significant associations were obtained which bore out the hypothesis that the social position of the person will be directly related to job satisfaction. The most significant associations found (p less than .001) were those of: 1) social class itself, and 2) source of income. These results definitely suggest that the prestige element is somehow reflected in the satisfaction a man has in his job. The occupational prestige involved, the type of house lived in, and source of his income all compose a tnype of hierarchy of status which is associated with his job satisfaction. These elements are important to this vazdable of job satisfaction, even though the social class 1hierarchical concept of Warner's was not completely appli- cable for the sample.1 —k , 1Gregory P. Stone and William H. Form, "Instabilities 1n.:Status: The Problem of Hierarchy in the Community Study Of' Status Arrangements," American Sociological Review, 18, fiPIIil, 1955, pp. 149-162. The status contest between the 1c3'<.‘.a.1ites" and the "cosmopolites" prevented a unidimension- 65 Other Social Position Variables Social—Economic Groups. Another way of testing the hypothesis that high social position will be directly re- lated to job satisfaction was found in using Edwards' social-economic groups. These groupings are based on the occupational groups used in the census and used in this thesis in measuring occupational prestige in Chapter III. Edwards stated that in addition to the occupational posi- tion, certain occupations were logical combinations as to their social position, for their economic return from their 1 occupation provided the basis for their way of life. In other words, a factory worker making 35,000 per year and a high school teacher making 35,000 per year would spend their incomes differently. Edwards' groupings, however, indicate a prestige scale, thus it was decided to investi- gate the relationship of this variable to job satisfaction. _ al ranking in Warner's terms. These groups represented the established, sedate way of life of the town (the "local- ites") and the more urbane life of the city (the "cosmOpol- ites"). However, the objective status characteristics seem to transcend the status contest when related to the job satisfaction index. For example, in occupational prestige, if a man is a "localite" professional or a "cosmopolite" manager, his objective rating is the significant item of association with job satisfaction. 1Alba M. Edwards, Sixteenth Census of the United States, Comparative Occupation Statistics for the United States, United States Government Printing Office, I945, m. 66 Table 17 shows the same trend of relationship be- tween job satisfaction and the social-economic groups as obtained with social class. There are not as sharp breaks in the distributions as there were for social class. For example, the social-economic groups of professionals and of pr0prietors, managers and officials have ten respondents who signify less satisfaction with their job, while on the upper and upper-middle class ratings by Warner, only three respondents indicated dissatisfaction. The trend, however, is in the same direction as in social class, for twenty- four persons in these same two categories of professionals and proprietors, managers and officials do indicate satis— faction,and the lowest category, that of unskilled workers, indicates eleven of the twelve respondents signifying dis- satisfaction. Again, the middle status groups have respond— ents in all the categories with a tendency toward an indica— tion of dissatisfaction rather than satisfaction. The statistical relationship is significant, since the probability that the variables were not related was less than one per cent. The null hypothesis can be reject- ed. The guiding hypothesis for this chapter can be consid- ered supported as Edwards' social-economic groupings are significantly associated with job satisfaction, the "high" groupings with "high" satisfaction and the "low" ones with "low" satisfaction. As mentioned above, neither the sta- 67 TABLE 17 ASSOCIATION BETWEEN EDWARDS' SOCIAL-ECONOMIC GROUPS AND JOB SATISFACTION INDEX Job Satisfaction IndexH Edwards' Social-Economic Total Groups‘ 5 2 1 O l. Professionals 1 4 l - 6 2. Proprietors, Managers, and Officials ll 8 8 l 28 5. Clerks and kindred workers 5 2 7 4 l6 4. Skilled workers and foremen 4 8 l2 5 27 5. Semiskilled workers 2 7 5 7 19 6. Unskilled workers 1 - 7 4 12 Total 22 29 58 19 108 7,2 = 11.961 .Ol)p) .001 c .. .42 ‘To‘p two and bottom two rows collapsed in computation of the Chi-square. ‘*Columns 5 and 2, and l and 0 were collapsed in com- putation of the Chi-square. tistical significance nor the trend of association is as dramatic as the social class variable. However, since Ed- wards' scale is a more easily obtained index, it may be used as a substitute for social class for general purposes. Class identification. The literature reviewed did not indicate any substantive findings which had investigat- ed the subjective class identification of the respondent and 68 its relationship to job satisfaction. For this study, how- ever, it was hypothesized that those identifying themselves correctly1 would indicate higher satisfaction if they were from "higher" social classes and lower if from "lower" classes, while those identifying upward would be unhappy with their jobs, at least in some aspects. These would be the respondents who are aspiring and who would like to im- prove their social and occupational positions. Because of the cultural value attached to "getting ahead," it was ex- pected that the workers identifying "down" would also show dissatisfaction. The above statements are supported in Table 18. Those cases in the upper and middle classes who estimated their position correctly had few persons scoring zero, but the percentage of those scoring one and zero was 40 per cent. For those in the lewer classes who estimated their position correctly, the percentages are higher. Sixty per cent were dissatisfied,while only 40 per cent of the former classes had signified dissatisfaction. Fifty-nine per cent of the upper and middle classes were satisfied as contrast- ed to forty per cent of the lower classes. The workers who identified upwards had five persons who were satisfied with l"Correctly" means that the subjective rating by the respondent matched the objective social class rating (the Index of Status Characteristics score). 69 TABLE 18 ASSOCIATION BETWEEN CLASS IDENTIFICATION AND JOB SATISFACTION INDEX Own Social Class Job Satisfaction Index" Estimate in Relation Total to ISO Rating 5 2 l 0 Same (Upper & Middle) 9 10 9 4 52 Same (U per-Lower & Lower- Lower) 5 ll 14 10 40 Identified Up 5 2 10 5 18 Identified Down 5 4 5 l 15 Total 22 27"" 56"" 18"" 105"" 7.2 a 7.985 .05>p>.o2 U .. .37 "Columns 5 and 2, and l and 0 were collapsed in com- putation of the Chi-square. ""The remaining reapondents were not able to place themselves in a social class. their work, thirteen who were not; those workers identify- ing down had nine persons satisfied and four who were not. The expected results were verified except for the persons A in the "downward" identification category. The Chi-square was of moderate significance, the probability of the two variables being independent fell between the five per cent and two per cent levels; thus, the null hypothesis was re- jected. The C computed for the table also supports the as- sociation between the two variables. 70 The explanation of the findings needs little comment except for the downward identifying category. It may be that those who identify down, do so on the basis of their job classification. They may have termed themselves as "working" class while they had been objectively classified as "middle class." In other words, their job may be the major factor in their identification. As a contrast, those who identified up may not use their job as the major factor in identification, but may rely for identification upon their reference groups or upon their desire to climb. These suggestions are in no way substantiated by the data at hand, but seem to be plausible explanations. Education. The category of education is often used as an index of social status, for often college degrees are indicative of the economic and social status of the parents. It is also important to this study because of its indica— tion of extensive training for an occupation. The litera- ture reviewed brought out conflicting statements in refer- ence to the association of job satisfaction and education. Trier had hypothesized that workers with less education ‘would be more satisfied, but he found that workers with "more education were slightly but insignificantly more satis- fied than those with lower education.1 Kessler discovered, 1Howard E. Trier, Job Satisfaction 223 Qggupational SiSatus, unpublished M. A. thesis, Department of Psychology, Mlxehigan State College, 1954, p- 54. 71 too, that the satisfied group of veterans which he studied had a higher level of formal education than the dissatis- fied group.1 Bullock's study, however, reported that the ex-employees (the dissatisfied) had a higher percentage of persons with some college background than the employees. The employees also had more technical training and more high school graduates than the ex-employees.2 This study posited that the higher the education of the respondent, the higher his expectations for the job and the greater his probability of getting better jobs. If "better jobs" is taken to mean higher occupational prestige and more pay, then it would be expected that high job satisfaction.scores would be associated with more education. The expected trend is borne out by the data shown in Table 19. In other words, for the sample, those twenty-two respondents who scored three are typified by 15.7 average JYears of schooling, those who scored two by 10.1 years, those scoring one by 9.7 and those scoring zero by 9.7 Years. The biggest contrast was between the group complet- ing eight or less years of school and the group completing Some college. Fifty-nine per cent of the first group indi- ——_ 1Milton S. Kessler, "Job Satisfaction of Veterans Reh.a.bilitated Under Public Law 16, " Personnel and Guidance :LJurnal, 55, October, 1954, pp. 78- 81. 2Robert P. Bullock, Sggigl Factors Related to Job E3a'tisfaction, Research Monograph NumBer 70, Bureau of-Bfisi- neass Research, Ohio State University, p. 89. 72 cated dissatisfaction, while only 56 per cent of the second group expressed dissatisfaction. TABLE 19 ASSOCIATION BETWEEN EDUCATION AND JOB SATISFACTION INDEX M Job Satisfaction Index"" Grades 8 or less 6 8 l5 7 54 9-11 5 7 11 5 24 12 6 7 8 7 28 15 or more 7 7 6 2 22 Total 22 29 58 19 108 Mean number of years 15.7 10.1 9.7 9.7 10.1 ’12 a 3.186 .SO>p).5O U - .23 —'___ n..— "The second and third rows were combined in computa- tion of the Chi-square. ""Columns 5 and 2, and 1 and 0 were collapsed in com- putation of the Chi-square. These trends were not large enough to show a statis- tical significance between the variables of education and Job satisfaction. The probability that the two variables ‘Were independent fell between the thirty and fifty per cent levels of significance, indicating no association. The C (Df .25 is low, but does signify an association between the variables. 75 There may be several explanations for this finding. One may conclude that for this sample, the variable of edu- cation was not significant in relationship to job satisfac— tion because longer training does not necessarily mean more satisfaction. On the other hand, it might be possible that with a larger sample and more categories education would be sensitive of job satisfaction. Although this table showed neither a direct nor an inverse relationship to job satis- faction, another possible explanation is that the more edu- cated persons are more dissatisfied if they are thwarted in finding better paying or higher prestige jobs, while the less educated may not be aware of the potentials of the job market. In conclusion, it is interesting to note that per- haps education may be sensitive only at the extremes. Amount of Income. Much of the literature reviewed was concerned with the relationship between amount of income and job satisfaction. The findings conflict, however, for some state that income is an important variable and others assert that it is low in importance. Typical of the ar- ticles suggesting that motives other than the pay check are influential is the statement by Mace, who says workers are satisfied if the job . . . contains a certain minimum of inherent interest, if it brings a reward accepted as apprOpriate and fair,_ 74 and if the work contributes to his self-respect and to a measure of recognition from his fellow men.1 Popular articles such as that found in Fortune also sub- stantiate the fact that money is not of prime importance. Their conclusion is that both bosses and laborers want se- curity and more than pay, and that they will find it through recognition of achievement and dignity of position.2 A study by Jurgensen reports that,out of thirteen hundred applicants in their sample, pay rated sixth out of ten fac- tors selected as important.3 0n the other hand, other research points to signifi- cant relationships between amount of pay and job satisfac- tion. HOppock noted that "there appears evidence of some sort of connection between earnings and job satisfaction." Perhaps, he reasoned, high income in the teaching profes- sion causes satisfaction, or that satisfaction causes the worker to do a better job and thus increase his earnings.4 These statements were made in reference to his study of satisfied and dissatisfied teachers, and he did not inves- 1C. A. Mace, "Satisfaction in Work," Occupational Psychology, 22, January, 1948, pp. 5—19. 2Anonymous, "What Makes the Boss Work?" Fortune, 57, April, 1948, p. 104. 5C. E. Jurgensen, "Selected Factors Which Infl%ence Job Preferences," Journal pf Applied Psychology, 51, ecem- ber, 1947, pp. 555-565. 4Robert Heppock, Job Satisfaction, New York: Harper, 1955. P- 158- 75 tigate the relationship in his New Hepe sample of all occu- pations. Two Fortune articles also point to a direct rela- tionship. The first one, which investigated a cross- section of veterans, concluded that lower economic status was increasingly found among more dissatisfied workers. The top income group had only 2 per cent who admitted dis- satisfaction, the upper-middle 8 per cent, lower-middle 16 per cent, and the poor, 21 per cent. "This suggests low pay is at least as great a factor as dull work . . . in "1 The second article causing job discontent among veterans. reported on a study of cross section of workers and their attitudes to their companies. Those who said their own company was as good to work for as any other emphasized the importance of high wages, and of those preferring other firms, forty-three per cent gave better wages as a reason.2 The most conclusive evidence comes from two studies by Cen- ters. The first study involved a cross-section sample of the total United States employed pOpulation eighteen years and older. One of his conclusions was that, Over one-half of the pOpulation is dissatisfied with its present income, and a large increase in income is generally desired by those who are dissatisfied.5 1"The Fortune Survey," Fortune, December, 1946, P- 14- 2"The Fortune Survey," Fortune, June, 1947. P- 5: 3Richard Centers and Hadley Cantril, "Income Satis— faction and Income Aspiration," Journal pf Abnormal and Social Psychology, 41, 1946, pp. 64-69. 76 To relate this finding to job satisfaction, one may look to Weber who emphasized that income fixes the life-chances of the bread-winner and his family.1 It seems plausible that those men having small incomes and being dissatisfied with them might also be dissatisfied with their occupations which provide the source of the income. The men with ample means, on the other hand, may look to their job as the source of providing an adequate return for their work and may be more satisfied. The second study by Centers focused on job satisfac- tion. This research concluded: Occupational groups vary in the frequency of their sat— isfaction and dissatisfaction with their pay. The higher occupational groups contain the largest per— centages of satisfied persons, while the lowest occupa- tional ranks contain the largest pr0portions of dissat- isfied persons.2 Therefore, if the studies which emphasize the direct relationship of high income with high job satisfaction are considered valid, the findings of the group of studies which de-emphasize income may be incorporated to a certain extent. When pay is adequate enough to insure a comfortable living (as in the case of the "boss" in the Fortune ar- 1Max Weber, "Class, Status, Party," From Max Weber: ESsa s in Sociolo , trans. and eds. Hans Gerth and C. Wright E1118, New Iork: Oxford Press, 1946, p. 182. 2Richard Centers, "Motivational Aspects of Occupa- tional Stratification," Journal p; Social Psychology, 28, NO‘Vember, 1948, p. 216. 77 ticle), then perhaps other factors are of importance. On the low income levels, however, it is expected that the worker will be discontented with the job which provides the low income for him. The expected results for this study are that income if a rough indication of social position and sets the stage for the workers whole style-of—life; thus, there will be a direct relationship between high in- come and high job satisfaction and low income and low job satisfaction. In analyzing Table 20, it can be noted that the mean income for each group declines as the satisfaction score decreases from the most satisfied to the least satisfied. Those scoring three on the job satisfaction index averaged 31600 more income than those scoring two. The group scoring two received 31500 more than those scoring one. However, these men earned only 3500 more than those scoring zero. The dramatic contrast lies in the almost doubled amount of income received by those most satisfied when com- pared with those least satisfied. The mean income for the complete sample is rather high at 34,912.00. It is pulled up by the nine cases fall- ing in the above $10,000 bracket. The median. is about 35886 with a range of under $1000 to over $10,000. The probability of the Chi-square falls between the -01 and .001 levels. The null hypothesis can be consid- 78 TABLE 20 ASSOCIATION BETWEEN AMOUNT OF INCOME AND JOB SATISFACTION INDEX @3233le a... “In..." To: 5 2 1 0 Under 31,000 - $1,999 1 1 5 4 9 32,000 - 32,999 5 2 13 4 22 33,000 - 34,999 8 20 16 8 52 85,000 - 39,999 5 5 5 5 16 810,000 and over 5 5 l - 9 Total 22 29 58 19 108 Mean"' 36,977 35,510 84,015 35,710 34,912 7C2 . 17.640 .01) p) .001 C = .45 ‘4— "The first three and last two rows were combined in computation of the Chi-square. ""Columns 5 and 2, and l and 0 were collapsed in com- putation of the Chi—square. """The mid-point used in computation of the mean for the first row was 3500 and for the last row, 315,000. ered untenable, and the expected results are further borne out by inspection of the 0 of .45, which is fairly high in its indication of association. The hypothesis that income is directly related to job satisfaction is borne out by the data. 79 Summary of Chapter The guiding hypothesis for this chapter was that the higher the status position of the person, the higher the job satisfaction. The relationships obtained for all but one of the variables were found to be significant. Table 21 summarizes the statistical significance found between satisfaction and the variables of social status. TABLE 21 SUMMARY OF TESTS OF ASSOCIATION BETWEEN SOCIAL POSITION AND JOB SATISFACTION So;;g;aggggus eiéiliiciic. ABZ§iiztiin (312) (B) Social Class p < .001 . 51 Occupational rating p ( .01 .47 Source of income p (.001 .52 House type p < .01 .45 Social-economic groups p( .01 .42 Class Identification p<.05 -57 Education p( .50 .25 Amount of income p< .01 .45 W Social class was extremely significant in its rela- tionship to job satisfaction, but the characteristic most sensitive to the job satisfaction index was source 80 of? income. Edwards' social-economic groups variable also upheld the hypothesis by exhibiting a significant associa- tion with job satisfaction. It was found that the respond~ ents who identified upward were more dissatisfied than those who subjectively rated their social class as the same as their objective ISC rating. The workers who identified downward were noted to be more satisfied than dissatisfied, and it was posited that this may reflect their identifica- tion with their job as a "working class" job. Education was not found to be significant in its association with job satisfaction, but the amount of income findings substan- tiated the hypothesis that the higher the income, the more job satisfaction. The guiding hypothesis for this chapter may be considered confirmed. CHAPTER V JOB SATISFACTION: ITS RELATIONSHIP TO COMMUNITY INVOLVEMENT Two possible theoretical positions guided the re— search in this chapter, which had as its purpose the test- ing of the hypothesis, that job satisfaction will be relat- ed to community involvement. First, those persons with low job satisfaction would exhibit high community involvement, or second, those persons with high job satisfaction would exhibit high community involvement. The low satisfaction- high community involvement was projected as including the mobile workers who express dissatisfaction because they are on their way up in the work world. These persons might very likely seek recognition and responsibility outside of their occupational worlds in order to attract attention from their superiors in their jobs. Another possibility was that low job satisfaction reSpondents might logically seek satisfaction outside their work world and become high- ly involved in community affairs. On the other hand, those exhibiting high satisfaction-high community involvement could be respondents whose community involvement reflects their social and occupational position. Thus, from the findings in Chapters III and IV, it might be posited that Persons with higher social positions and high prestige oc— 82 cupations who eXpressed high job satisfaction would partici- pate actively in and identify strongly with the community. There was very little in the literature reviewed to sub- stantiate either position. Therefore, this chapter was in- cluded as an exploratory venture in discovering possible relationships between the job satisfaction index and cer- tain selected indices of community involvement. The variables used as indices of community involve- ment were: 1) age, 2) size of community of socialization,1 5) newspaper readership, 4) number of organizational mem- berships, 5) church attendance, and 6) party membership. It was realized that these variables are not unambiguous, systematic indices of community integration, for they might reflect social position as well. However, they were used as rough clues to community participation which might explain divergences in job satisfaction scores for persons who otherwise have similar work environments. The Chi-square test of association was run between all of the above variables and the job satisfaction index. The results were generally non-conclusive and did not sup- port either of the possible expectations. Tables for each variable are not presented as_in the preceding chapters. However, each factor and its relationship to job satisfac- tion is discussed and described briefly. 1The writer submits the term "community of sociali- zation" as comparable to "birthplace," Specifically meaning tine place in which the respondent was.reared or socialized. 85 Community Variables Ag_. There were several non-conclusive and con- flicting results in the literature regarding the relation— ship of age to job satisfaction. It was expected, however, that older persons who held the same job for many years would be more satisfied than younger men in their first jobs. It was observed in Youth and ppp prld pf prk that high school graduates have high occupational aspirations but do not get the jobs they want when they first enter the work world.1 Hence they are more likely to be dissatisfied with their jobs than older workers. Hoppock's studies sup:\—— ported this finding. He found that satisfied teachers were an average of 7.5 years older than the dissatisfied, but in the New Hope study, the correlation between age and job satisfaction was only .21:.04? Trier's study concluded that the older workers were significantly more satisfied than the younger workers,3 but in Kessler's sample of vet- 1Social Research Service, Youth and the World pf Work, Michigan State College, 1949. 2Robert H0ppock, Job Satisfaction, New York: Harper and Brothers, 1955, p. 40. 3Howard E. Trier, Job Satisfaction and Occupational Status, unpublished M. A. thesis, Department of Psychoiogy, Michigan State College, 1954, p. 54. 84 arena the satisfied workers were eighteen months younger, -on the average, than the dissatisfied.1 The Coldwater sample consisted of experienced workers, all over the age of twenty. The range for the sample was twenty to seventy-nine with a mean age of 44.5. The men signifying satisfaction in all three dimensions were on the average 50.9 years old (7.6 years older than any of the other three groups). However, there was not enough devia- tion in the data to support the hypothesis that older age would be related with high job satisfaction, for the prob- ability of the Chi-square was between seventy and eighty per cent. The null hypothesis cannot be rejected. In ex- plaining this finding, it must be remembered that the very young workers were not included in the sample. Further re- search is needed to discover whether age and high job sat- isfaction are associated. Size of community of socialization. A plausible hy- pothesis would suggest that those people who grew up in a community of the same size as Coldwater might exhibit high job satisfaction. Contrariwise, those born and reared on a farm or in a large city might find it difficult to accommo- date to a town the size of Coldwater, and this resentment Inight be reflected in low job satisfaction. ‘ 1Milton S. Kessler, "Job Satisfaction of Veterans Eiehabilitated Under Public Law 16," Personnel and Guidance €Tourna1, 55, October, 1954, pp. 78—81. 85 The data available to test this were crude. The residence categories included: farm, small town, town 2,500-25,000, and city over 25,000. Forty-five respondents were reared in towns 2,500-25,000--towns approximately the same size as the place they were now residing in. The as- sociation between job satisfaction and community of social- ization was not significant for the Coldwater sample, for the probability of the Chi-square fell between the seventy and eighty per cent levels of significance. Newspaper readership. The category of newspaper readership was used in the larger study as an indication of local or non-local (urban) identification. It was particu- larly useful in explaining the status split in the town, for the "localites" were characterized by reading only the Coldwater or surrounding town papers while the "cosm0po— lites" read the metropolitan papers of Chicago or Detroit. For the purposes of this study, it was hypothesized that the "localites" perhaps would exhibit more satisfaction with their job in Coldwater than the "cosm0polites." The results of the Chi-square test were generally non-conclu- sive, for the probability of the two variables being asso- ciated fell between the fifty and thirty per cent levels of significance. Inspection of the table, however, showed a slight reversal of the expected findings. Of‘the thirty- four "localites," twenty were dissatisfied and fourteen 86 satisfied with their jobs, while of the seventy-four "cos- m0polites" thirty-seven expressed satisfaction and thirty- seven dissatisfaction. As suggested earlier, the job sat- isfaction of the person may not be reflected by his iden- tification with the community. The variables reflecting social position, i.e., income, house type, occupational prestige, may be of more importance to job satisfaction than place in the status structure of the community. Number of organizational memberships. The expected results for this table were based on the assumption that highly mobile people are also active in community organiza- tions. Particularly in the case of higher status men, luncheon clubs, lodges, and social organizations play an important part in their work lives. If they are mobile, they would not be particularly satisfied with their job, according to the job satisfaction index, for they would have indicated their preferences for other jobs in answer to at least one or two of the questions. Thus, it might be expected that low job satisfaction might be related to a high number of club memberships. This expectation was not borne out upon inspection of the data, as no consistent pattern of relationship de- veloped between the two variables. The range of memberships was from zero to more than seven, with the mean number of memberships for the sample falling at 2.0. In other words, the number of memberships for the satisfied and dissatisfied 87 respondents did not vary greatly. However, the variation was in a consistent direction. Those having a job satis- faction score of three average 2.6 organizations. Those with scores of two belonged to 2.0 organizations on the av- erage, and those with scores of one or zero belonged to 1.8 organizations. This suggests those who have more satisfac- tion also are more involved in the community. However, prob- ability of the Chi-square fell between the fifty and seven- ty per cent levels of significance. The null hypothesis could not be rejected. Church attendance. This variable is indicative of community involvement, for it represents the percentage of Sundays the respondent had been in church within the six months immediately preceding the interview. The expecta- tion was that highly mobile persons, those with high com- munity involvement, would use church attendance as another organization important to their pgppg-occupational position and that they would have low job satisfaction. The 105 re- Spondents who answered the question about church attendance inere placed into one of three categories: 1) twenty-four ‘were "regular" church-goers, having admitted attending churdi eighty per cent or more of the Sundays, 2) twenty-eight ‘were "occasional" church-goers, having stated they attended 14338 than eighty per cent to at least once, and 5) fifty- One did not attend church at all. 88 The expected results were not borne out, but the data showed a reverse ppppg toward substantiating a state- ment of high job satisfaction would be related to some church attendance, or that low job satisfaction would be related to no church attendance. For example, of the fifty-one non-attenders, thirty-one expressed dissatisfac- tion; of the twenty-eight occasional attenders, eleven ex- pressed dissatisfaction; and of the twenty-four regular church-goers, half expressed dissatisfaction. The split in the last category prevented any conclusive statement about direct relationships between church attendance and job sat- isfaction. The Uhi-square computed, too, indicates a trend toward significance but is not conclusive, for the probabil- ity was between the five and ten per cent levels. It may be noted that church attendance shows a trend toward sig- nificant association with high job satisfaction, even though the number of organizational memberships was not closely associated with job satisfaction. A possible ex- planation may be that intensity of involvement in one or- ganization (as indicated by church attendance) may be of more importance than the number of token memberships a per- son displays. On the other hand, church may figure more prominently in general community identification in smaller towns. 89 Party membership. This variable was included under 7the community involvement category because the sample seemed to reflect the political affiliation pattern of the community as a whole. Residents of Branch county and its county seat, Coldwater, are known to be predominantly Re- publican and Independent. However, the variable of polit- ical affiliation also might reflect social position, for such studies as The Pe0ple's Choice reported that political affiliation and income are highly related.1 In addition, Centers' results stated that political behavior and occupa- tional status are related. "The higher groups are charac- terized by much greater support of the traditionally con— servative Republican Party than is the case with the lower occupational strata."2 It was expected that this variable was interrelated with social position and that Republican and Independent affiliation would be associated with high job satisfaction. Forty respondents identified themselves as Repub- licans, thirty as Independents, and thirty-four as Demo— crats. The Republicans were divided equally between being satisfied and dissatisfied with their jobs, the Independ— 1Paul F. Lazersfeld, Bernard Berelson, and Hazel Gaudet, The Pe0ple's Choice, New York: Duell, Sloan and Pearce, 1944, pp. 16-27. 2Richard Centers, The Psychology of Social Classes, Princeton: Princeton University Press, I949, p. 2 . 90 ents had seventeen satisfied workers and thirty dissatis- fied, while the Democrats had twelve satisfied and twenty- two dissatisfied. This relationship showed a Chi-square which was not statistically significant when the index was dichotomized. The probability that the two variables were associated was between twenty and thirty per cent. However, because there were no Democrats who scored three in the in- dex, when a Chi-square was computed without dichotomizing, it was statistically significant at the .01 and .001 level. The C computed for this significant relationship was rather high for this study, .51. Apparently, Republicanism is somewhat associated with job satisfaction and community in- tegration. Summary of Chapter The guiding hypothesis for this chapter was very generally posited: job satisfaction will be related to community involvement. Two divergent explanations were submitted: 1) low job satisfaction will be related to high community involvement, and 2) high job satisfaction will be related to high community involvement. Neither was upheld. If the community participation of the reapondent is of importance in job satisfaction, it will take further re— search to discover which relationships are significant. 0n the basis of the findings in this chapter, the guiding hy- 91 pothesis, job satisfaction will be related to community in- volvement, cannot be verified. CHAPTER VI SUMMARY AND DISCUSSION Summary In this thesis, an attempt was made to explain dif- ferences in job satisfaction by an investigation of the so- cial background of the respondents. Specifically, this "so- cial background" consisted of factors of occupational pres- tige, occupational mobility, social position, and community involvement. Four focal hypotheses guided the study: 1. High job satisfaction will be directly related to high occupational prestige. 2. High job satisfaction will be directly related to upward occupational mobility. 5. High job satisfaction will be directly related to high social position. 4. Job satisfaction will be related to community involvement. A job satisfaction index was deve10ped from three questions probing job preference in order to measure roughly whether a respondent was "satisfied" or "dissatis- fied" with his present job. The hypothesis that high job satisfaction will be related to high occupational prestige was considered first. 95 Warner's technique for rating occupations, Edwards' occupa- tional groups, and Hatt's situs system of rating occupa- tions were employed as indicators of occupational prestige, and the findings exhibited positive relationships between all three ratings and the job satisfaction index. The second hypothesis was that high job satisfaction will be directly related to upward occupational mobility. Mobility was considered from five perspectives: 1) number of jobs held, 2) vertical occupational mobility, 5) genera- tional occupational mobility, 4) situs mobility, and 5) gen- erational situs mobility. Two of the variables, number of jobs held and generational occupational mobility, demon— strated a high degree of association with the job satisfac- tion index. A third variable, occupation mobility, ap- proached significance and demonstrated a trend toward up- holding the hypothesis. The situs variables of situs mo- bility and general situs mobility did not manifest any sig- nificant relationships with the index. Overall, it was felt that the second hypothesis, high job satisfaction will be related to upward occupational mobility, could be con- sidered empirically verified for the Coldwater sample. The third hypothesis was that job satisfaction will be directly related to high status position. General status position was measured by Warner's Index of Status Charac— teristics and Edwards' social-economic groups. The three 94 components of the Index of Status Characteristics were tested to detect their sensitivity to job satisfaction. In addition, class identification, education, and amount of income were included as social position variables. All but one variable, education, were found to exhibit statistical significance in relationship to the job satisfaction index. It may be that mass education in the lower grades in con- temporary America prevents using education as a sensitive index to social class. The hypothesis that high job satis- faction will be directly related to high social position, can be considered empirically verified by the data under scrutiny in this study. The final hypothesis, job satisfaction will be re- lated to community involvement, was of a more tentative na- ture. Two possible and divergent expectations were posited: 1) high community involvement will be related to low job satisfaction as a compensatory outlet for dissatisfaction with the job, or 2) high community involvement will be re- lated to high job satisfaction as an extension of high com- munity position. Six variables were included to investi- gate the hypothesis: age, size of community of socializa- tion, neWSpaper readership, number of associational member- ships, church attendance, and party membership. The results w e o e e o 1 were not s1gn1f10ant in five of the s1x cases. The one 1One of the five variables, party membership, pro- 95 variable which approached significance was church attend- ance.’ Eor future research, it was suggested that perhaps investigation of strategic affiliations in the community may be of more importance in relationship to job satisfac- tion than general community involvement factors. Neither the general hypothesis that job satisfaction will be relat- ed to community involvement nor either possible explanation (low job satisfaction-high community involvement or high job satisfaction-high community involvement) could be sub- stantiated. In sum, three of the four hypotheses were generally substantiated. It was realized that factors of the hy- potheses reflected each other to some degree. For example, occupational prestige and upward occupational mobility were found to be highly related to job satisfaction. Both of these variables are obviously part of the third hypothesis which dealt with social position. The fourth hypothesis investigating community involvement was also reflective of certain aspects of community position, for Warner's social class system is cOmmunity-based. vided a peculiar case. For further elaboration, see su ra, Chapter V. 96 Discussion The most general sociological conclusion drawn from this study of social variables and job satisfaction is that the rewards from life, both on and off the job, are related in a positive manner. It may be premised that such general factors as social prestige, high income, and independence are satisfying elements in life, and that overall acceptance of the job is the basic to this satisfaction.1 In terms of this study, job satisfaction was found to be positively re- lated to occupational prestige, upward occupational mobil- ity, and high social position. Those pe0ple in the sample with the most prestigious jobs, with more money, power, and independence are apparently more satisfied with their jobs. In addition, these same people apparently have higher status positions in the community. They have experienced less job mobility (or if mobile, have climbed up), live in better houses, and have had more education than those who are less satisfied with their occupation. To be sure, their posi— tion in the community may be set largely by their occupa- tional position; the two may be aSpects of the same thing. 1While this is a basic assumption to the whole study, it may be that this is a current sociological bias. Thus, persons in manual jobs may have values different from those in the professions. If such is the case, then dif- ferent variables would be important in investigating job satisfaction. 97 Most previous studies had investigated the individ- ual, psychological variables, the specific work conditions and environment, or the labor turnover aspects, while these data verified common observations which the sociologist, along with other members of society, may have taken for granted. They indicate the value of studying over—all job satisfaction from a sociological orientation. The sociologist will also be interested in the prob- lem areas which arose during the course of this study. Like most pg post facto studies, it was hampered because the re- search had not been designed to anticipate all of the rel- evant aspects of the research question. Therefore, certain recommendations can be made now as guidelines for future w research. The job satisfaction index should be refined. Per— haps, another question could be added to the series; e.g., "What job of all those you know would you most like your son to hold?" Another area of refinement might be to ex- plore further the items. Inspection of the questions used in the job satisfaction index indicated that respondents easily selected their own job in answer to question (a), "What job would you most like to have at your present place of work?" and/or question (b), "Which of those jobs you have held,did you like best?" This was not true for the ques- tion, "Of all the jobs you know of, what job would you most 98 like to have?" If respondents named their own jobs in an- swer to this question, perhaps their scores should be given extra rather than equal weight. Thus, a weighting system might improve the general sensitivity of the index. After refinement of the index is undertaken, com- parison of the two methods of measuring job satisfaction might be fruitfu1--to ascertain whether the job satisfac- tion index, as a representative of the indirect technique, and Bullock's job satisfaction scale, as a representative of the direct technique, investigate the same factor of over-all pleasure or displeasure with one's job. It has also been suggested that job satisfaction may be more accurately examined if the question is approached in terms of jobs which are in hierarchical work organiza- tions or in autonomous work situations. The social psychol- ogy of job selection may vary for peOple who work in large hierarchical structures as Opposed to those who work alone or for themselves. Perhaps these two work environments need different types of indices of job satisfaction. Still another problem is the investigation of com- munity involvement and identification in relationship to job satisfaction. This study was not able to state whether those with unsatisfactory occupations find compensating mechanisms off the job. The many sensitive non—job indices investigated here did not reflect any such relationship. Therefore, the research question arises, how do they ad- just, if indeed they do? 99 LITERATURE CITED Bullock, R. P., Social Factors Related to Job Satisfaction, Columbus, The Bureau of Business-Research, College of Commerce and Administration, The Ohio State Uni- versity, 1952. Caplow, T., The Sociology of Work, Minneapolis: University of Minnesota Press,—I954. Centers, R., "Motivational Aspects of Occupational Strati- fication," Journal pf Social Ps cholo , 28, 1948, pp. 187-217. , The Psychology pi Social Classes, Princeton: Princeton Univers1ty Press, . , and H. Cantril, "Income Satisfaction and Income Aspiration," Journal of Abnormal and Social Psychol- pgy, 41, 1946, pp. 64:69. H) Cheek, G. L., A Psychometric Studg 0 Two Indices 2; Job Satisfaction, unpublishe . . thesis, DEpartment of Psychology, Michigan State College, 1955. :4 Durkheim, E., Division pf Labor, translated by G. Simpson, Glencoe, Illinois: The Free Press, 1949. Eckerman, A. C., "An Analysis of Grievances and Aggrieved Employees in a Machine Shop and Foundry," Journal pf Applied Psychology, 52, June, 1948, pp. 255-269. Edwards, A. M., Alphabetical Index pf Occupations, Washing- ton: United States Government Printing Office, 1957. , Sixteenth Census of the United States, Comparative Occupation Statistics for the United States, Washing- ton: United States Covernment’Printing OTTice, 1940. Form, W. H., and G. P. Stone, The social significance of clothing in occupational life. Michigan State Col- lege Agricultural Experiment Station, East Lansing. Technical Bulletin 247, June, 1955. "The Fortune Quarterly Survey: XI," Fortune, 17, January, 1958, p. 86. 101 "The Fortune Survey," Fortune, 54, December, 1946, p. 14. "The Fortune Survey," Fortune, 55, June, 1947, pp. 5-6. Hagood, M. J., and D. 0. Price, Statistics for Sociologists, New York: Henry Holt and C01, 1952. Hatt, P. K., "Occupation and Social Stratification," Amer- ican Journal pf Sociology, 55, May, 1950, pp. 555- Holland, J. B., Attitudes Toward Minority Groups in Rela- tion to Rural SociaI Structure, unpuinshed_PE. D. dissertation, Department of Sociology and Anthro- pology, Michigan State College, 1950. Hoppock, R., Job Satisfaction, New York: Harper and Brothers, 1955. , "Job Satisfaction Researches," Occupations, see vols. 16.2 - 55, April, 1958; October, 1940; Febru- ary, 1945; April, 1945; April, 1948; December, 1948; December, 1949; October, 1950; May, 1951; May, 1952; September, 1955; September, 1954; May, 1955. Hughes, E. C., "Work and the Self," Social Ps cholo 23 the Crossroads, eds. J. H. Rohrer and M. Sher f, York: Harper and Brothers, 1951. Jurgensen, C. E., "Selected Factors Which Influence Job Preferences," Journal pf Applied Psychology, 21, 1947. pp. 55}- Kerr, W. A., "On the Validity and Reliability of the Job Satisfaction Tear Ballot, " Journal of Applied Psy- chology, 52, August, 1948, pp. 275-281. Kessler, M., "Job Satisfaction of Veterans Rehabilitated Under Public Law 16," Personnel and Guidance Journal, 55, 1954, pp. 78-81. Khan, L., An Exploratory Stud of the Relationships Amon Personali_y Character1st1cs, Work Situation, 305 Satisfaction and Abilit to Empathize in an Indus- trial Fiamework,unpu511sh—d M. A. thesis, Depart- ment of PsychoIogy, Michigan State University, 1955. Lazersfeld, P. F., B. Berelson, and H. Gaudet, The Peo le' 8 Choice, New York: Duell, Sloan and Pearce, I942. 102 Mace, C. A., "Satisfaction in Work," Occupational Psychol- pgy, 22, 1948, pp. 5-19. McCormick, T. C., Elementary Social Statistics, New York: McGraw-Hill, 1941. Miller, D. C., and W. H. Form, Industrial Sociology, New York: Harper and Brothers, 1951. Reynolds, L. G., and J. Shister, Job Horizons, New York: Harper and Brothers, 1949. Schaffer, R. H.,‘Uob Satisfaction as Related to Need Satis- faction in Work," Psychological Monographs, 67, No. 14, 1955. Schell, W. A., A Study of Empathic Abilit gpd pf Egg Validity of Some—Indices pf Job Satisfaction, un- published—M. A. thesis, Department of Psychology, Michigan State College, 1954. Social Research Service, Youth and the World pf Work, Michi- gan State College, 1949. Standard Register Co., "How to Find Out What Your Workers Think About You," Factory Management and Maintenance, 106, August, 1948, pp. 81-9i. Stone, G. P., and W. H. Form, Clothing inventories and preferences among rural and urban families. Michi- gan State College Agricultural Experiment Station, East Lansing. Technical Bulletin 246, March, 1955. , "Instabilities in Status: The Problem of Hierarchy in the Community Study of Status Arrangements," American Sociological Review, Vol. 18, April, 1955, pp. 149-162. Super, D. Em, "Occupational Level and Job Satisfaction," Journal pf Applied Psychology, 25, 1959, pp. 547-564. Trier, H. E., Job Satisfaction and Occupational Status, unpublished M. A. thesis, Department oi PsychoIogy, Michigan State College, 1954. Thomson, W. A., "Eleven Years after Graduation," Occupa— tions, 17, May, 1959, pp. 709-714. 105 Walker, J., and R. Marriott, "A Study of Some Attitudes to Factory Work," Occupational Psychology, 25, July, 1951, pp. l81~l91. Warner, W. L., M. Meeker, and L. Eells, Social Class 3p America, Chicago: Science Research Associates, 1545. Weber M., Max Weber: Essays ip Sociolo , translated and ’ edited by H. H. Gerth and CT‘WEE§%% Mills, New York: Oxford University Press, 1946. "What Makes the Boss Work," Fortune, 37, PP- 104‘107 EEEEEE! April, 1948. Woods, W. A., "Employee Attitudes and Their Relation to Morale," Journal 25 Applied Psychology, 28, 1944, l. D“ .' 11“.! l ' j.” 'q’- luaw ['35, 3hr “; t, aaDate Due Demco-293 HICHIGQN STATE UNIV. LIBRQRIES ||l|| ll 1 312931046 3306