OCCUPAle MOBILITY. JOB SATISFACTION. .AND SOCIAL REFERENCES Thesis {or flu Dogma 95 M. A. MICHIGAH STATE UNIVERSITY James Arthur Geachwender £959 mum L [B R A R Y Michigan Sm: University OCCUPATIONAL MOBILITY, JOB SATISFACTION, AND SOCIAL REFERENCES By James Arthur Geschwender AN ABSTRACT Submitted to the College of Science and Arts of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Sociology and Anthropology 1959 Approved 42%égQézézéfi ’:;%f::;é}7¢v4// James Arthur Geschwender ABSTRACT This study attempted to analyze from the framework of reference group theory the relationship between vertical inter-generational occupational mobility and job satisfaction. Interview responses of 545 male manual workers were analyzed to test five major hypotheses: 1. There will be a low positive association between job satisfaction and occupational achievement relative to parent's occupational aspirations for the respondent, i.e., the higher the son's occupational achievement relative to his parents' aspirations for him, the higher will be his Job satisfaction. 2. There will be a moderate positive relationship between Job satisfaction and son's occupation relative to father's occupation, i.e., the higher the son's occupation relative to father's, the higher will be his Job satisfaction. 3. There will be a positive relationship between Job satisfaction and occupational level relative to brother's occupational level. 3a. In those cases where the respondent has two brothers, hypothesis three will be modified as follows: Job satisfaction will vary from low to high in the following order: The occupational level of both brothers being higher than that of the respondent; the occupational level of one brother being higher, and that of the second lower, than that of the respondent; and the occupational level of both brothers being lower than that of the respondent. James Arthur Geschwender 3b. The relative age of the brother will alter the relationship predicted in hypothesis three in the following manner: (1) Job satisfaction will be greater when an older brother has achieved a higher occupational status than when a younger brother has done so, (2) job satisfaction will be less when a younger brother has achieved a lower occupational status than it will be when an older brother has done so. 4. There will be a positive relationship between Generational Occupational Mobility Score (SONS) z scores and job satisfaction. 5. There will be a positive relationship between Job satisfaction and such standard variables as (a) age, (b) marital status, (c) number of children, (d) occupational prestige level, (e) job tenure, (f) income, and (g) education. It was found necessary in the testing of hypothesis four to make some minor revisions in the Generational Occupa- tional MObility Score (GOMS) as formulated by Tumin and Feld- man. The findings yielded some slight support for hypothesis one though it was noted that over four fifths of the respond- ents could not state their parents' occupational aspirations for them. Hypothesis two and three were supported while the two sub-hypotheses for hypothesis three were refuted. The findings also supported hypotheses four which was further analyzed by father's occupation and occupation of destination. Each of the groups classified according to father's occupa- tion exhibited a positive relationship between job satisfaction James Arthur Geschwender and vertical inter-generational occupational mobility while this was not the case when the sample was classified accord- ing to occupation of destination. Job satisfaction was found to be positively related to age, marital status, number of children, occupational prestige level, Job tenure, and income and negatively related to education. The above findings were analyzed as to their implica- tions for reference group theory. It was concluded that working class males, lacking the middle class belief in »opportunities for upward mobility, tend to evaluate their Jobs by using as social references other male members of their family of orientation, their peers, or other members of their occupational situs. This selection of social references results in a positive association between verti- cal occupational mobility and Job satisfaction. However, those members of the working class who come into sufficient contact with the middle class belief in opportunities for upward mobility tend to shift their social references to the middle class and thus exhibit a negative relationship between vertical occupational mobility and Job satisfaction. OCCUPATIONAL MOBILITY, JOB SATISFACTION, AND SOCIAL REFERENCES BY James Arthur Geschwender A THESIS Submitted to the College of Science and Arts of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the.degree of MASTER OF ARTS Department of Sociology and AnthrOpology 1959 ACKNOWLEDGMENTS I wish to eXpress my indebtedness to my parents, Elmer and Susie Geschwender, whose moral and financial support made it possible for me to pursue my education on the under- graduate and graduate levels. I also want to express my gratitude to Professor William H. Form for allowing me to conduct a secondary analysis of data that he had cOllected and for the generous giving of his time and advice which made this thesis possible. Finally, I wish to indicate my indebtedness to the members of my guidance committee, Dr. Richard Adams, Dr. William Faunce, Dr. Charles R. Hoffer, and Dr. Charles P. Loomis, for the aid and advice which they have given me at various stages of my graduate training. 11 TABLES OF CONTENTS CHAPTER PAGE I. THE STATEMENT OF THE PROBLEM AND THE REVIEW OF RELEVANT FINDINGS . . . . . . . . . . . . . Purposes . . . . . . . . . . . . . . . . . . . Research Design. . . . . . . . . . . . . . . . Theoretical orientation. . . . . . . . . . .4 UT-P-Wl-‘H Assumptions. . . . . . . . . . . . . . . . . . Hypotheses and pertinent findings in the literature.... ..... ................... . 7 II. THE MEASUREMENT OF INTER-GENERATIONAL OCCUPATIONAL MOBILITY. . . . . . . . . . . . . 18 III. THE MEASUREMENT OF JOB SATISFACTION. . . . . . . 43 IV. OCCUPATIONAL MOBILITY, JOB SATISFACTION, AND SOCIAL REFERENCES: AN EMPERICAL EXAMINATION 59 Job Satisfaction and Occupational Achieve- ment Relative to Parents' Occupational Aspirations for the Respondent . . . . . . . 59 A Job Satisfaction and Occupational Level of Sons Relative to That of Fathers . . . . . . 61 Job Satisfaction and Occupational Level Relative to That of Brothers . . . . . . . . 62 Job Satisfaction and Occupation Level Relative to That of Two Brothers . . . . . . 64 Job Satisfaction and Relative Age of Brother . 66 Job Satisfaction and Vertical Occupational Itobility.................. 67 TABLE OF CONTENTS (continued) CHAPTER Job Satisfaction Job Satisfaction Job Satisfaction Job Satisfaction Level. . . . . Job Satisfaction Job Satisfaction Job Satisfaction and Age . . . . . . . . . and Marital Status. . . . and Number of Children. . and Occupational Prestige and Hourly Wages. . . . . and Job Tenure. . . . . . and Education . . . . . . V. SUI'MARY AND GENERAL CONCLUSIONS. 0 o o o o 0 General Conclusions. . . . . . . . . . . . Scope, validity, and limitations . . . . Impolications for further research . . . LITERATURE CITED. . . . . iv PAGE 78 79 80 80 82 84 86 88 95 102 105 108 TABLE 9. 10. ll. 12. 13. 14. 15. LIST OF TABLES Mean Job Satisfaction Scores for Reapondents Grouped According to Their Occupational Achieve- ment Relative to Their Parents' Occupational Aspirations for Them . . . . . . . . . . . . . . Job Satisfaction and Occupational Level of Sons Relative to Occupational Level of Fathers. . . . Job Satisfaction and Occupational Level Relative to Occupational Level of Brothers. . . . . . . . Job Satisfaction and Occupational Level Relative to That of Two Brothers. . . . . . . . . . . . . Job Satisfaction and Relative Age of Brother, Controlling for Relative Occupational Levels . . Job Satisfaction and Vertical Occupational Mobility as Measured by the Revised Generational Occupational MObility Score. . . . . . . .-. . . Job Satisfaction and Vertical Occupational MObility as Measured by the Revised Generational Occupational Mobility Score by Occupation of Destination. 0 0 O O O O O O O O O O O O O O O 0 Job Satisfaction and Occupational Mobility as Measured by the Revised Generational-Mobility Score by Father's Occupation . . . . . . . . . . Job Satisfaction and Age . . . . . . . . . . . . Job Satisfaction and Marital Status. . . . . . . Job Satisfaction and Number of Children. . . . . Job Satisfaction and Occupational Prestige Level Job Satisfaction and Hourly Wages. . . . . . . . Job Satisfaction and Job Tenure. . . . . . . . . Job Satisfaction and Education . . . . . . . . . PAGE 60 62 64 65 66 68 69 74 79 8O 81 81 82 85 86 CHAPTER I THE STATEMENT OF THE PROBLEM AND THE REVIEW OF RELEVANT FINDINGS The purpose Of this thesis is to examine those rela- tionships that may or may not exist between vertical occupa- tional mobility and Job satisfaction among working class males. Caplow says: Vertical mobility is a movement of the indivi- dual upward or downward, with a gain or loss in social rank. This may occur in several different ways: 1. The simplest kind of vertical mobility is a change of occupation which involves a change in social position; as when a waiter becomes a busi- nessman, or an unsuccessful accountant goes to work in a factory. 2. A different form of mobility involves a promotion or demotion within an occupational group; as when a naval officer receives command of a ship, or a locomotive fireman becomes an engineer. 3. Another form of vertical mobility is incidental to ageing. Each occupational level displays a characteristic career curve; and in addition, the mere accumulation of seniority represents a significant change in status. 4. A fourth type of vertical mobility is the change in occupational assignment from one genera- tion to another, usually studied as the correla- tion between the occupations of fathers and sons. There is still another kind of vertical mobility which is not included in this discussion. It involves the ascent or descent of an entire 2 occupational group - the increasing dignity of nurses or the diminishing prestige of midwives may serve equally well as examples. This thesis will only be concerned with Caplow's fourth category of vertical mobility, that of vertical inter- generational occupational mobility. According to Lipset and Bendix, ”The ultimate reason for our interest in this sub- Ject (social mobility) is the study of the consequences of social mobility.”2 This thesis is focussing on one of the possible consequences of social mobility, Job satisfaction. The examination of this relationship will be conducted among working class males. This limitation is Justified because it is maintained that working class males constitute a seperate and distinct behavioral category from middle class males. This belief is supported by Lipset and Bendix when they list five differences between the two groups: 1. Most male nonmanual occupations have more prestige than most manual occupations, even skilled 01168. 2. Among males, white collar positions gener- ally lead to higher income than manual employment. 3. Nonmanual positions, in general, require more education than manual positions. 4. Holders of nonmanual positions, even low paid white collar Jobs, are more likely than manual workers to think of themselves as members 1 Theodore Caplow, The Sociolo 2; Work (Minneapolis: University of Minnesota Press, 1953), pp. 59-65. zseymour Lipset, and Reinhard Bendix, Social Mobilit lg Industrial Societies (Berkly: University of CaIifoFfiIa Press, 1959), p. . 3 of the middle class and to act out middle-class roles in their consumption patterns. 5. Low-level nonmanual workers are more likely to have political attitudes which re- semble those of the upper middle class than those of the manual working class. They further state, ”Consequently, we believe that using the break between manual and nonmanual occupations as an indicator of high and low occupational status is Justified whenever a dichotomous division of males in urban occupations is used."4 Research Design This study will consist of a secondary analysis of data collected by Professor William H. Form in 1953 as a study of workers in Lansing, Michigan. The respondents interviewed were distributed among the following occupational categories: 62 unskilled; 326 semi-skilled, 146 skilled and foremen; and 11 sales, clerical, and related occupations - a total of 545 respondents. The sample was drawn from the FOLKS DIRECTORY FOR LANSING.’ Every 10th name was drawn and any were rejected who were not working class males, 1.9. defined as 'One who works with his hands'. Eleven cases were retained that are usually categorized as 'sales, clerical, and related occupa- tions' as it was felt that their occupations did meet the criterion of the definition of 'working class male'. 41bid., p. 16. 3Ibid., pp. 14-16. 4 Information was garnered in interviews from the re- spondents with respect to their occupation, Job satisfaction, father's occupation, occupation of all brothers, age of all brothers, parents occupational aspirations for the responp dent, number of brothers in the labor force, respondents age, education, marital status, number of children, Job tenure, and income. Theoretical Orientation This thesis is developed from a framework of 'reference group theory' as suggested by Herbert Hyman and explicated by Robert K. Merton.5 An underlying assumption is that there may not exist any universally agreed upon definition of the significance of any given social position. The incumbant of any position selects certain portions of the 'social reality' and uses those portions as reference points.in de- fining his position. It is in terms of these reference 'points (reference groups) that he defines his expected be- havior, his values, his attitudes, and his evaluation of his position and of himself. It is assumed that this does not occur at random, but is influenced by at least two major factors. ‘Eiggg, there are certain specifiable structural influences that lead an incumbent of a given position who possesses certain specifiable 5Robert K. Merton, and Alice S. Rossi, ”Contributions _To The Theory of Reference Group Behavior”, printed in Mere ton, Robert K. Social Theory and Social Structure, (Glenooe: The Free Press, 1959). 5 social attributes to select certain predictable reference groups. Secondly, this occurs within an institutional frame- work that provides us with a loosely defined ordinal ranking of positions in terms of prestige. It is, therefore, the task of the sociologist to dis- cover the significant reference groups that are associated with any given social position and set of social character- istics. This may be done entirely in structural terms with- out reference to individual motivations. In fact, it must so be done if we are to be able to predict to different situations in which different actors, having different indi- vidual motivations, are involved. Assumptions Merton states that, "Some similarity in status attri- butes must be perceived to hold bewteen the individual and the reference group for the comparison to occur at all".6 These perceived similarities may be in terms of position of origin, present position, anticipated future position, and/ or relational ties among other possibilities. Allison Davis7 indicates that members of the working class are not motivated in terms of opportunities for 'social climbing’. Instead, they derive their satisfactions from immediate and present situations. He also indicates that the members of this class have relatively strong nuclear family ties. 61b1d., p. 242. 7Allison Davis, "Werk Habits and Class Position” Modern American Society, (New Ybrk: Rinehart and 00., 1951). 6 It may be inferred from this that the working class males involved in this study will use as their social refer- ences groups that are perceived to have some attribute in common with themselves. As a result of the lack of motiva- tions for 'social climbing', it is unlikely that they will perceive themselves as having shared attributes with white collar groups. It is much more likely that they will use as reference groups others who share their position of origin, i.e. immediate kin, peers, or the occupational class of which they are a member. It follows from reference group theory that there may not exist an absolute scale for evaluating the gratifi- cations and deprivations associated with any given position. Instead, there are probably relative evaluations of gratifi- cations and deprivations which are determined by the rela- tionship that exists between those of a given position and those associated with the corresponding reference groups. Merton indicated this when he said, "The degree of dissatisfaction with their lot is often less among people in severely depressed social strata in a relatively rigid social system, than among those strata that are apparantly better off in a more mobile social system".8 Evidence seems to indicate that, while there is con- siderable mobility within the working class, the line between it and the white collar stratum is rather rigidly drawn. 8Merton, 9-20 git-I0, p. 268. 7 For example, Form and Miller have found that, "There is a tendency for the children of white-collar persons to inherit their father's occupation or climb above it. Children of manual workers inherit their father's occupation or fall below it."9 From this it may be inferred that adult-socialized working class males may not use white-collar workers as a reference group but would more likely use groups from within their own occupational stratum. Hypotheses and Pertinent Findings in the Literature 1. There will be a low positive association between Job satisfaction and occupational achieve- ment relative to parent's occupational aspirations for the respondent, i.e., the higher the son’s occupational achievement relative to his parents aspirations for that achievement, the higher the job satisfaction. This is hypothesized because it is expected that the family will be a significant reference group for working class males thus making a person's occupational achievement relative to his parents occupational aspirations for him a determinant of his Job satisfaction. This association is not expected to be strong due to the ambiguity of the concept 9w1111am H. Form, and Delbert c. Miller, ”Occupational Career Patterns As A Sociological Instrument", American Journal of Sociology, 54:328-29. 8 'parents aspirations'. This ambiguity can probably be best illustrated by a comment made by Howard Ehrlich in the course of a conversation bearing on this problem. It was felt that a parent was likely to say something to the effect of, ”Gee, I'd like my son to be a doctor. I expect him to be a sales- man. He'll probably work in a factory like his old man." The first alternative may be a form of desire, the second an expectation of low probability, and the third being more probable - an anticipation considering all circumstances. The present researcher suspects that either the first or second alternative listed above would be reported by most respondents as their parents aspirations for them while the third alternative would have had the greatest amount of effect in determining the individuals own aspirations. 2. There will be a moderate positive relation- ship between job satisfaction and son's occupation relative to father's occupation, i.e., the higher the son's occupation relative to father's, the higher the Job satisfaction. In terms of the theoretical orientation of this study, it is expected that father's occupation would be an important reference point in determining the Job satisfaction among working class males due to the strength of the nuclear family ties in this class. This relationship is not expected to be as strong as it might be due to the intrinsic weakness of the comparison of son's occupation to that of his father as 9 a measure of occupational mobility. This weakness results from the fact that the occupational structure itself is constantly changing. 3. There will be a positive relationship between Job satisfaction and occupational level relative to brother's occupational level, i.e., those persons who have achieved an occupational level higher than that achieved by their brothers will be more satisfied with their Jobs than will those whose occupational achievement is less than that of their brothers. The same theoretical Justifications that apply to hypothesis two also apply here. But the limitation that applied there does not apply here as the occupations of the respondents and those of their brothers come from the same occupational structure. 3a. In those cases where the respondents have two brothers, hypothesis three will be modified as follows: Job satisfaction will vary from low to high in the following order: the occupational level of both brothers being higher than than of the respondent; the occupational level of one brother being higher, and that of the second being lewer, than that of the respondent; and the occupational levels of both brothers being lower than that of the respondent. 10 All brothers, being members of the immediate family, are assumed to constitute important reference groups for working class males. Thus, it follows that the relationship hypothesized in hypothesis three will be reinforced when the direction of the differences in the occupational levels of the respondents and each of their brothers are the same and weakened when the directions are opposite to one another. 3b. The relative age of the brother in compari- son to that of the respondent will alter the rela- tionship predicted in the third hypothesis in the following manner: (1) Job satisfaction will be greater when an older brother has achieved a higher occupational status than it will be when a younger brother has done so; (2) Job satisfaction will be less when a younger brother has achieved a lower occupational status than it will be when an older brother has done so. Age is expected to be a relatively important factor in this relationship because it is an important factor in career patterns. Status tends to increase with age simply because certain rewards tend to accrue to those with increased tenure, i.e., wage increases, longer paid vacations, more paid holidays, certain types of promotions. Ybunger brothers will tend to be less far along in their careers and older brothers correspondingly farther along in their careers. This might tend to alter the relationship resulting from a simple comparison of occupational levels. ll 4. There will be a positive relationship between Generational Occupational Mbbility Score 2 scores and Job satisfaction. This would follow as a simple consequence of the fact that all of the groups whose occupational levels are used in computing the z scores are important as reference groups for working class males, i.e., nuclear family members, peers, and members of the same occupational stratum or origin. The first four hypotheses deal, in one manner or another, with the relationship between occupational mobility and Job satisfaction. An examination of the literature re- veals an extreme dearth of studies relating these two vari- ables. The only two that the present researcher could locate, Trier10 and Eicherll, both reported positive associa- tions between inter-generational occupational mobility and Job satisfaction. 5. In order to compare the predictive power of the GOMS with other previously used predictors of job satisfaction, it is hypothesized that there will be a positive relationship between Job satisfaction and such standard variables as (a) age, (b) marital status, (c) number of children, lOH. E. Trier,’Ubb Satisfaction and Occupational Status" (unpublished Master's.thesis, Michigan State University, East Lansing, 1954): P- 33- 11Joann Eicher, "Job Satisfaction: Its Relation To Occupational, Stratification, And Community Variables" (unpublished Master's thesis, Michigan State University, East Lansing, 1956), pp. 48-50. 12 (d) occupational skill level, (e) wages, (f) tenure, and (g) education. All of the variables listed in hypothesis five have a long and hallowed tradition of use in job satisfaction studies as the review below will demonstrate. Herzberg and others12 have published an excellent summary of all Job attitude studies published prior to the summer of 1955. The present researcher selected and examined a small nonrandom sample of the studies reviewed by Herzberg and came to the conclusion that no major bias would result if the summaries by Herzberg and associates were relied upon rather than retreading ground previously covered. Consequently the summaries by Herzberg will form the basis of much of the review of the literature given below. (a) Job satisfaction - marital status Herzberg cited eight studies in this area, three which found that married employees liked their jobs more than single employees, one that found that married employees like their Jobs less than single employees, and eight that found that Job satisfaction was unrelated to marital status.13 The three studies that reported a positive associaidon were 12Frederick Herzberg, Bernard Mausner, Richard Peter- son, Dora Capwell, Job Attitudes: Review‘gf Research And 0 inions (Pittsburg: Psychological Service of Pittsberg, 1557). lBIbide, pp. 23-240 15 those of Harris,14 Chase,15 and Inglow‘.16 Only Eckermanl7 reported a negative association; and the eight reporting no association were an American Vocational Association Study,18 Brown and Neitzal,19 Cole,2O Fryer,21 Jurgenson,22 Kolstao‘,23 Kornhauser and Sharp?!1L and KcClusky and Strayer.25 14Frank Harris, "The Quantification of an Industrial Employee Survey," Journal of Applied Psychology, 1949, 15F. S. Chase, "Factors for Satisfaction in Teaching," Phi Delta Kappan, 1951. 33:127-132. l6sail Inglow, "Job Satisfaction of Ldberal Arts Graduates," Journal pf Applied Psychology, 1951, 35:175-181. 17A. C. Eckerman, ”An Analysis of Grievances and Aggrieved Employees in a Machine ShOp and Foundary," Journal .2: Applied Psychology, 1948, 32:255-69. 18American Vocational Association, Committee on Re- search Publications, ”Factors the Satisfaction of Home Eco- nomics Teachers,” AVA, Research Bulletin #3, Washington, De Ce ’ 19480 19C. G. Brown and Betty Neitzal, "Communication, Supervision, and Morale," Journal g; Applied ngchology, 1952, 36:86-91. 2ORemson Cole, "J. A. Survey of Employee Attitudes," Public Opinion Quarterly, 1940, 4:497-206. 21Douglas Fryer, "Industrial Dissatisfaction," Ip- dustrial Payphology, 1926, 1:25-29. 22Clifford Jurgensen, "Selected Factors Which Influence Job Preferences," Journal 93 Applied Psychology, 1947, 31:553-564. 23Arthur Kolstad, "Employee Attitudes in a Department Store," Journal 23 Applied Psychology, 1938, 22:470-479. 24Arthur Kornhauser, and Agnes Sharp, ”Employee Atti- tudes: Suggestions From a Study in a Factory,” Personnel Journal, 1932, 10:393-404. 25H.Y.C. KcClusky, and Floyd Strayer, "Reactions of 14 (b) Job satisfaction - number of children Herzberg reported no studies that directly measured the relationship between Job satisfaction and the number of children but he did report two on the relationship between job satisfaction and number of dependents.26 These reported findings opposite to one another; Stagner and Britton27 found that men with three or more dependents tend to be more satisfied with their jobs than men with smaller families, and oaaon28 found that married men with no children liked their jobs better than married men with children. (c) Job Satisfaction - occupational prestige level Of the eighteen studies that Herzberg reviewed29 only one, that by Fryer,30 failed to report a higher level of Job satisfaction at higher occupational levels. Some of the representative studies reporting positive associations Teachers to the Teaching Situation - A Study in Job Satis- factions," School Review, 1940, 48:612-23. 26Herzberg, pp, cit., p. 24. '27Ross Stagner, J. W..Rich, and R. H. Britton "Job .Attitudes: Defense werkers,” Personnel Journal, 1941 20:90-97 28Hasly, Cason, "General Curves and Conditions of IPeeling,” Journal pg Applied Psychology, 1931, 15:126-48. 29Herzberg, pp. cit., p. 20. 30 Fryer, pp. cit. 15 are: Centers,31 two Fortune surveys,32 Katz,33 and Mann.34 One study not reported by Herzberg, that of Eicher,35 also reported a positive association. (d) Job satisfaction - Job tenure Herzberg reviewed seventeen studies in this area.36 The seven which were wide range studies reported an early period of high morale dropping off to a low period followed by a resurgance to a period of rising Job satisfaction. All of the others, except Ash37 and Stott,38 analysed short range periods that fit in with the general trend. 31D. H. Centers, "Motivational Aspects of Occupational Stratification," Journal'pg Social Psychology, 1948, 28: 187-2170 32Fortune Surveys, Fortune, 1938, 17:83-84, 86-88, 91-92 and 1947, 25: pp. 5-6, 10-11. 33Donald Katz, "Motivational and Mbrale in Industry," Current Trends 1p Industrial Psychology, (Pittsburg: Univer- sity of Pittsburg Press, 1959 . pp. 1 5-171. 34Floyd Mann, "A Study of Work Satisfactions as a Function of the Discrepency Between Inferred Aspirations and Achievement,”‘Qissertation Abstracts, 1953, 13:902. 35E10her, 22’ 2&2. ppe 35-42e 36Herzberg, pp, cit. pp. 11-12. 37Phillip Ash, "The SRA Employee Inventory - A Statistical Analysis,” Personnel Psychology, 1954, 7:337-364. 38M.B.A. Stott, "A Preliminary Examination In The Occupational Analysis of Secretarial Work,” Hum. Fact., Londe’ 1938, 9:249‘258e 16 (e) Job satisfaction - income Herzberg39 reported four studies, Centers and Cantril,40 42 and Thomson,43 which discovered that as Inlow,41 Miller, income goes up job satisfaction goes up. One study reported a drop in satisfaction for the middle group of earners, #4 and one study (Survey Research Center)45 reported Benge, that the amount of money earned was less important in deter- mining a worker's morale than whether or not he thought his pay rate was fair. (f) Job satisfaction - age Herzberg summarized twenty-three studies in this area.45 They all showed a consistent trend. In general, morale is high with young workers, it tends to go down until it reaches a low point in middle or late twenties or early thirties, and then it shows a steady rise with age. 39Herzberg, pp. cit., pp. 21-22. 4ORichard Centers, and Hardly Cantril, "Income Satis- faction and Income Aspiration,” Journal p3 Abnormal Social Psychology, 1946, 41:64-69. 4llnlow, pp. cit. 42Delbert C. Miller, ”Economic Factors in the Mbrale of College-Trained Adults," American Journal p§,Sociology, 1947. 47:139-156. 43William Thomson, ”Eleven Years After Graduation," Occupations, 1939, 17:709-714. 44Eugene Benge, "How to Learn What Workers Think of Job and Boss,” Factory Management Maintenance, 1944, 102(5), lOl'lOAe 45Survey Research Center, ”Effective Mbrale,” Part 3, Fortune, 2(8), 46-50. 1+6Herzberg, pp, p;3., pp. 5-8. 17 (g) Job satisfaction — education Herzberg summarized thirteen studies in this area with conflicting results.47 Five studies showed no difference in job satisfaction as education varied, three studies showed an increase in morale as education increased, and five studies showed a decrease in morale as education increased. Both Eicher48 and Trier49 found a small, but insignifi- cant, positive association between job satisfaction and edu- cation. 47Ibid., pp. 15-16. 48Eicher, pp, p;3,, p. 73. 4'9Trier, pp.'p;p., p. 34. CHAPTER II THE MEASUREMENT OF INTER-GENERATIONAL OCCUPATIONAL MOBILITY The purpose of this chapter is to describe and dis- cuss some of the more important methods that have been used to measure inter-generational occupational mobility. No attempt will be made to review the findings of such studies nor to review all of the studies of inter-generational occupa- tional mobility. Only enough studies will be included as is necessary to illustrate the major methods that have been used to measure inter-generational occupational mobility. This study is not concerned with intra-generational occupa- tional mobility per se. The first method is that used by Davidson and Ander- son.1 Using the Alba Edwards scale of occupations2 as their classifying device, they constructed two basic charts in which they distributed the level of the son's occupation by 1Percy Davidson, and H. Dewey Anderson, Occupational Mobility lp pp American Community, (Stanford: Stanford n vers ty Press, 1957). 2Alba Edwards, "A Social-Economic Grouping of the- Gainful Workers in the United States,“ Journal p; the American Statistical Association, 1933. XXVIIX:577-87. 18 19 the level of father's occupation and the level of father's occupation by the level of son's occupation. Davidson and Anderson computed what they called the transmission of status from father to son by computing, for each occupational level of fathers, the percentage of sons on the same level or adja- cent levels. They then reversed this procedure; for each level of sons they computed the percentage of fathers on the same or adjacent levels, in order to measure the occupational origins of the sample. This was carried one step farther in order to measure the three generational transmission of status; the occupational levels of the fathers were distri- buted by the occupational levels of the grandfathers. The direction and amount of vertical mobility were determined by computing the percentage of sons at each occupational level having fathers in each of the occupational levels ranked above or below that of the son in terms of prestige. Trends in occupational mobility were measured by comparing the amount of occupational inheritance from grandfather to father with that from father to son. There was also an attempt to describe career patterns at various levels but this will not be discussed because no attempt was made to control for the occupation of fathers and thus it is not relevant as a measure of inter-generational occupational mobility. Centers analyzed the data in his study in a manner similar to that of Davidson and Anderson but used different occupational categories.3 The categories he used were: 3Richard Centers, "Occupational Mobility of Urban 20 large business, professional, small business, white collar, skilled manual, semi-skilled manual, and unskilled. Eicher also used a related technique of measuring inter-generational occupational mobility.4 She also used the Alba Edwards Scale of Occupations but in measuring vertical mobility she compared the respondents present occupation to that of his father at age forty (an age at which Miller and Form state that a man is usually in his stable occupation).5 Upward mobility is defined as those cases in which the re- spondent's occupation is on a higher level than that of his father at age forty and downward mobility as those cases in which the respondent's occupation is on a lower level. Sorokin's study was concerned with a different problem but the technique he used to measure inter-generational occupational mobility is of interest here.6 He was not con- cerned with measuring the entire range of inter-generational occupational mobility but simply with determining the social origins of American men of wealth and any trends in changes in these origins. He measured this simply by analyzing the principaloccupations of the fathers of these men of wealth. Occupational Strata," American Sociolozical Review, 1948, 13:197-203. . 4Joann Eicher, "Job Satisfaction: Its Relation to Occupational, Community, and Stratification, Variables" (unpublished Master's thesis, Michigan State University, East Lansing, 1956), pp. 49-50, 53-55. 5Delbert c. miller, and William H. Form, Industrial Sociology, (New York: Harper and Brothers, 1951), p. 755. 6Pitirin, Sorokin, "American hillionaires and Multi- Killionaires," Social Forces, 1925, 3:627-40. 21 This is, again, a simple variant of the method used by Davidson and Anderson. Form and Miller have used a more sophisticated variant of the career pattern analysis presented by Davidson and Anderson.7 They analyzed a work career by breaking it down into segments according to the following classificatory scheme: initial work period, all part or full time jobs that an individual holds up to the time that he completes his formal education; trial work periods, those jobs that an individual holds while he is shopping around for his per- manent job - classified in terms of jobs held for less than three years; and the stable work period, any job at which an individual remains for three years or more. The various modal combinations of the initial, trial, and stable work periods were plotted and classified as being either "secure" or "insecure" career patterns. The reason that this type of career analysis is discussed in a chapter dealing with inter- generational occupational mobility is that Form and Miller analyzed the various modal career patterns in terms of the occupational levels of the fathers of the respondents. All of the above studies share the characteristics of measuring occupational mobility by comparing the occupa- tional levels of fathers and sons without making any attempt to control for possible changes that might have taken place in the occupational structure itself during the course of a 7William H. Form, and Delbert C. Miller, "Occupational Career Patterns as a Sociological Instrument,” American Journal p; Sociology, 54:317-29. 22 generation. Many of the occupational shifts noted in the above studies may have been the result of the expansion of job opportunities at certain levels and the contraction of opportunities at other levels over the time span considered. The failure to control for these changes does not in any way negate the findings in any of the above cited studies - it merely suggests that their findings are partial. They could 'be more complete. All of the findings presented in these .studies still hold - the controlling for shifts in the occupa- ‘tional structure would merely help to explain those findings. Both the studies by Davidson and Anderson and by (Senters indicated an awareness of this problem but did not directly deal with it. In a discussion of their findings Davidson and Anderson state, In recent industrial history, fundamental changes have occured which force masses of workers from older occupations into expanding categories, probably without regard to fine individual differences in competency or skill and without essentially altering their rela- tive status or economic well being. The older occupational situation was made up of a rather limited number of well estab- lished professions, some clerks and semi- skilled workers, and a preponderance of farmers and proprietors, skilled, and un- skilled workers. From this older arrange- ment a heavy flow of workers has moved to- ward new professions and into expanding clerical and semi-skilled occupations, in- volving a sharp reduction particularly in the proportion of farmers and farm hands among the gainfully employed. The effect of these changes upon occupational status is not clear with respect either to gradations of competency or to their economic and social implications. 8Davidson and Anderson, pp..cit., pp. 162-3. Also 23 The present researcher would tend to disagree with the portion of the above statement that implies that histori- cal occupational shifts were made without essentially alter- ing the relative status or economic well being of workers. Occupational categories which have contracted have not com- pletely died out, 9.5., there are still farmers. As long as these categories remain and some of the sons of farmers are farmers and some are members of other expanding categories «of occupations, such as the semi-skilled workers; if there :is a difference in status and/or economic well being between :farmers and semi-skilled workers today, then there has been can essential change. Centers indicates his awareness of the problem with tile following statement, "The determinants of these (mobility) Irelations can only be guessed at, of course, but the two main inactors would probably be found in the changing requirements cut the productive economy for workers at various levels, and irl the differing numbers supplied to the population by the fklthers of the various occupational strata, and from which tile labor force can be ultimately drawn."9 Ginsberg used essentially the same method of measur- 131£3 occupational mobility used by Davidson and Anderson, but at an earlier date, with a different classification of E 1N1r“tially quoted in Natalie Rogoff, Recent Trends ;p_Occupa- 33125121 Mobilit , (Glencoe: The Free Press, 1953), p. 21. 9Centers, pp, cit., also quoted in Rogoff, _p. cit., p. 22. 24 occupations.10 But he also presented data on the distri- bution of occupations which enabled one to allow for the demand factor mentioned by Rogoff. He limited himself to an impressionistic, rather than a statistical, treatment of this data. Rogoff indicates that any procedures which simply compare occupations of fathers and sons without controlling for the changes in the occupational structure over time are inadequate and she further states that, ”The source of the éiifficulty is that no account is taken of the total number (pf positions available in each occupational class. Movement :lnto and out of each occupational class needs to be considered ins relation to this availability or demand factor."11 Rogoff presented a sophisticated statistical technique fxor controlling for changes in the occupational structure in irlter-generational occupational mobility studies. This tech- ldixrue, formulated by Herbert Goldhamer,l2 was designed to snake it possible to measure social distance mobility when tile variation in the demand factor is held constant or con- trwolled for. Social distance mobility is defined as total r‘-'10b;'l.lity divided by the demand factor. The definition was exdlressed in standard notation as:13 10Morris Ginsberg, "Interchan e Between Social Classes," ChV1xbter IX of Studies in Sociolo London: Methuen and C°mbany. 1932). pp. 150-7 K llRogoff, pp. cit., p. 29. 12 ' 13 Rogoff, pp, cit., pp. 28—33. Ibid., pp. 31-32. 25 let X13 : number of sons moving from father's occupational class i to occupational class 3 R1 = number of fathers in occupational class i 01 = number of positions (sons) in occupational class 3 N = total number of positions (or sons or fathers) Total mobility: Xi , the proportion of sons with 1 R1 fathers who move to j Demand factor = 21, the proportion of total positions N available to class j Social distance mobility = X11 21: X11? $1N of sons R1 Rogoff has this to say with regard to the above nota- srtional expression of the 'social distance mobility' measure: In the last form, the denominator of the defi- nition of social distance mobility corresponds exactly to the definition of expected values in conventional contingency analysis. The whole expression is then seen as the ratio of the actual cell value to its expected value. In this form social distance mobility is defined as the ratio between actual mobility and the amount of mobility we would expect where there is no relation between the son's occupational class and the occupational class of his father. The expected mobility values represent the amount of movement that would occur if the social dis- tance factors did not operate, in other words, if only availfibility factors influenced occupational movement. 1“Ibid., p. 32. 26 This measure has the advantage of enabling one to make valid comparisons between amounts of mobility in two different occupational sturucures, or within the same struc- ture at two different points in time, without having to be concerned with the possibility that the differences observed are the product of differing amounts of availability of cer- tain positions in the two different structures. It controls for both of the factors that Centers thought would be the major determinants of mobility, the demand for workers at the various occupational levels and the supply coming from each level. There is one comment that must be made regarding the notational definition of social distance mobility, that is, with respect to the definition of R1. Rogoff states that R1 is to equal the number of fathers in occupational class i and this is used to indicate the amount of the labor supply available in that class. This is obviously a poor measure of potential labor supply as one father could easily have more than one son. It seems likely that this measure, if applied literally, would result in underestimating the poten- tial labor supply. It would be much more precise to let R1 equal the total number of sons of fathers in occupational class i. This change would not make any difference in studies cofiducted along the lines of Rogoff's where each son listed the occupation of his father and all fathers were treated as being separate and distinct persons regardless of whether or not the sample included any brothers or half brothers, 27 i.e. if two brothers were included in the sample each would name the occupation of his father and the resulting informa- tion was treated as if this represented two fathers. How- ever, it is quite conceivable that this could make a differ- in the results of a study that was designed differently. ence Rogoff used the above described tool to compute the social distance rates from each father's occupational class to each of the son's occupational classes. This basic statistic was then used in order to compute the average inability out of each class of origin, the average mobility :1nto each class of destination, the ratio of the average inability out of each class to the average mobility into that <31ass, the amount of occupational inheritance for each occupa- tional class, and the ratios of occupational inheritance to out-mobility and to in-mobility for each occupational class. All of the above cited characteristics measured are cllaracteristics of the occupational structure and not of 1rndividuals. The technique of measuring social distance mobility will not aid one in determining whether a particular lruiividual is upwardly mobile, downwardly mobile or immobile. Tumin and Feldman listed five major problems raised by the use of current measures of occupational mobility.15 1. .Wide differences might occur between the occupational distributions of fathers and sons without a significant change in the distribution of power, prestige, or property. 15Melvin Tumin, and Arnold Feldman, ”Theory and Measure- Tent, of Occupational Mobility," American Sociological Review, 957. 22:281-288. pp. 282-3. 28 2. Certain 'goodness of fit' techniques, i.e. that presented by Rogoff, do not allow for the drawing of inferences regarding the extent to which groups that are alleged to be mobile perceive themselves as mobile. Inasmuch as peer groups are used as reference groups as well as parental groups an objectively upwardly mobile person may perceive himself as being downwardly mobile. They suggest that a measure of mobility might do well to include elements of both parental and peer position. 3. Most techniques do not allow for the measurement of mobility in such a manner that individuals can be scored and groups of re- spondents sorted on the basis of their mobility scores. 4. The occupational experiences of sons of the highest rated and of the lowest rated groups are not adequately handled by most measures as, by definition, it is not possible for the sons of the highest rated groups to be upwardly mobile or the sons of the lowest rated groups p to be downwardly mobile. 5. Most measures fail to account for birth order, birth spacing, and number of siblings. This is relevant because the mobility experiences of younger and older sons could be different as a result of the possibility that the younger sons may be able to count on aid in educational achieve- ments from an older working brother. In order to compensate for the above problems they devised a measure of inter-generational occupational mobility which was called the 'Generational Occupational Mobility Score' and abbreviated 'GOMS'. A summary of the computational Procedure is as follows:]6 A. Raw gay; 1. Occupation of ego 2. Occupation of ego's father 16Ib1de, pp. 283-840 3. 29 Occupation of all of ego's brothers Procedure 1. Divide the sample into the different father's occupational groups. For each of the father's occupational groups in turn, compute: (a) The mean occupational score (x) of 31;. their sons, i.e.,the mean occupational score of all respondents and brothers whose fathers share a common occupation. (b) The standard deviation (0') of this distribution, i.e., the 0‘ of the occupa- tional distribution of all respondents and brothers whose fathers have the same occupation. For each respondent compute the deviation of his occupational score from the E'occupational score of all respondents and brothers whose fathers have the same occupation, i.e., the deviation of the respondents' occupational score from the mean computed in 2(a) above. Divide the deviation computed in 3 above by the 0‘ (computed in 2(b) above) of that dis-' tribution. The resulting z score is the GOMS. Ten is added to make all scores positive. The procedure may be stated as follows: GOMS=-f-XW+10.=Z to- 30 where 3? = mean occupational score of all respondents and brothers of occupa- tionally similar fathers. x = respondent's individual occupational score. 0' = standard deviation of the distribution of occupational scores of all respondents and their brothers within the same fathers occupational group. There are two problems with the above formulation as it now stands. The first is with regard to whether or not ‘bhe GOMS measures'fintimidiitpumxrmito measure and the second is whether ten is added prior to, or after the standard de- triation is divided into the deviation of the respondents occupational score from the mean occupational score of all ruespondents and their brothers whose fathers have the same 0 c cupation. The problem of whether or not the GOMS measures what it; purports to measure is a rather serious one. The computa- tion of the GOMS by the method suggested by Tumin and Feld- lnaml yields some rather unusual results. Those individuals whc>.have experienced the greatest amount of occupational mObility, in the positive sense of rising in status, receive lower: scores than do those who have experienced less mobility. This can be illustrated by taking a hypothetical case of two brothers, one a professional and one an agricultural day laborer, whose father was a clerk or office worker and 31 computing their 2 scores. Inasmuch as they are brothers, the mean occupational score for all sons of occupationally similar fathers and the standard deviation of that distri- bution would be the same for both respondents, 6.19 and 2.72 respectively.17 The only relevant bit of data that would differ for them would be their own occupational scores. The professional would receive an occupational score of 11 and the agricultural day laborer would receive a score of 1.18 Thus, the formulae for computing their z scores would be GOMS professional = é:l§.:§%%§1119, or.§al%:%§ll + 10 by the revision suggested in the second part of this paper, and GOMS agricultural day laborer = 2:22557%_:_19, or éaég7gl + 10. When evaluated this becomes GOMS professional = 1.91, or 8.23, and GOMS agricultural day laborer = 5.95, or 11.91. With either method of computation, the professional would clearly receive a lower 2 score than would the agri- cultural day laborer. This means that those members of the group of all sons of occupationally similar fathers who have achieved the highest occupational positions would be receiving lower z scores than would those who have achieved lower occupational scores. There is nothing intrinsically wrong with this result as long as one is aware of it when he does his analysis and consistantly interprets the lower scores as indicating greater upward mobility and the higher scores as indicating less 17Ibid.. Table 1. p... 285-: 18.1.1219." 13- 285-- 32 upward mobility. It does not appear that Tumin and Feldman are aware of this. They consistently interpret the higher scores as indicating greater upward mobility than the lower scores. They specifically state that, "All scores above ten represent over achievement; below ten under achievement", and they further define under achievement in the following manner, ”...under achievers--i.e., men in those occupations are at present located in lower positions than the average position now held by those with whom they started."19 This problem is very easily resolved if the computa- tional fromula of the cows is revised to read sous = x -'i + 10 U or X -'X + 10 if the change suggested in the second part of this gaper is taken into account. This would then change the z scores of the above mentioned hypothetical brothers to GOMS professional = 5.44, or 11.77, and GOMS agricultural day laborer = 1.77. or 8.09. This would have the effect of giving the professional a higher z score than his brother who is an agricultural day laborer. In general, this proposed revision would have the effect of assigning a higher z score to those members of the group of all sons of occupationally similar fathers who have achieved higher occupations than it would to those who have achieved lower occupations. There is one possibility that I have not yet accounted for, that is, that Tumin and Foldman did not assign values to the occupational categories in the manner that I have assumed, 191bid.; p. 286 33 i.e., Professional, 11; Semi-professional, 10; ...Agricultural day laborers, l; but instead assigned them in this order, Professionals, 1; Semi-professional, 2; ...Agricultural day laborers, 11. If this were the case then those individuals who achieved the greatest amount of upward mobility would be receiving the highest z scores. They do not clearly indicate which order of assignment of values they used. The only relevant statement that they made is, "All occupations were arbitrarily classified into eleven categories and assigned numerical ratings of one to eleven, starting with agricul- tural day laborers at the bottom and ending with professionals at the top."20 This could possible be interpreted in either manner. Nevertheless, the data that they present in their article clearly indicates their order of assigning rankings. In Table 1,21 they present the mean occupational scores of all sons of occupationally similar fathers for each of the eleven occupational categories of fathers' occupation. They are as follows: Professional, 8.06; Semi-professional, 6.71; Owners of business, 5.81; Managers and white collar sales, 5.25; Owners and managers of farms, 4.44; Clerks and office workers, 6.19; Skilled labor, 4.72; Semi-skilled labor, 4.36;Service workers, 4.50; Unskilled labor, 3.91; and Agricultural day laborers, 3.19. In general, it can be seen that as one moves down the level of occupations from the most prestigeful to the least, 291919.. p. 285. 21Ibid., p. 285. 34 the mean occupational score of all sons decreases. If Puerto Rico is like most other societies, in that the sons of men in the highest ranked occupations have the best oppor- tunity of achieving high occupations, then the only manner that this result could be achieved is if the occupational categories were ranked from professional as 11 to agriculr tural day laborer as 1. Thus, the foregoing analysis is correct and the proposed revision is in order. This proposed revision would have a dual advantage. Epppp, it would bring the computation of the GOMS into line with the manner in which it was interpreted by Tumin and Feldman. Second, it would be heuristically valuable. Socio- logists are accustomed to interpreting higher scores from an index designed to measure vertical occupational mobility as indicating greater upward mobility. It appears to be simpler to revise the formula than it would be to revise the thoughtways of sociologists. The second problem with the computation of the GOMS is less serious but, nevertheless, important. This is the question of whether or not ten should be added prior to, or after, the diviSion of the distribution of occupational scores into the deviation of the individuals occupational score from the mean occupational score of all sons of occupa- rionally similar fathers. The verbal description of the com- putational procedure of the GOES and its symbolic representa- tion appear to contradict one another on this count. The verbal description (see step 4) seems to indicate that the 35 division takes place prior to the adding of ten to the scores. However, the symbolic representation clearly indi- cates that ten is added prior to the division. It appears that the standardization process that this division represents is only meaningful if the division takes place prior to the adding of ten to the scores. If the ten is added first then 2 scores will not be comparable when taken from two different father's occupation categories. We would be adding ten divided by the standard deviation to each score. This would yield a different value everytime the standard deviation of the distribution involved was different. The 2 scores would be strictly comparable as standard devia- tion units above or below the mean if the division takes place prior to the adding of ten. In a discussion of their Table 2 quoted above, compar- ing adjacent occupational classes with respect to occupational mobility scores, they stated that all scores above ten indi- cate over achievement and all scores under ten indicate under achievement. The only way in which it would be possible to use ten as the cutting point to distinquish between over achievement and under achievement is if ten were added to the scores after the standard deviation of the-occupational distribution was divided into the deviation of the indivi- dual's occupational score from the mean occupational score of all some of occupationally similar fathers. This could be illustrated with the case of the two brothers mentioned in the hypothetical case earlier. The 36 mean occupational score of that distribution was 6.19, the distribution of all sons of clerks and office workers, the z score of the professional son would be 5.22, ll_:_gs1%_:;h9, if ten is added prior to the devision. This would have to be interpreted as under achievement by Tumin and Feldman. But this would be an incorrect interpretation. The pro- fessionals are ranked at the top of the occupational hier- archy with a ranking of 11. This is clearly over achieve- ment with respect to a mean of 6.19. If we make our de- vision prior to the adding of ten, the z score of the pro- fessional becomes 11.77, ll_%_$%l2.+ 10, a score correctly indicating over achievement. .The z score of the agricultural day laborer would be 8.19. l—g—géla + 10, correctly indicat- ing under achievement. . It is thus evident that if one wishes the computation of the z score of the GOMS to be consistent with the dis- cussion of findings that Tumin and Feldman present in their article that the standardization, i.e., the division of the standard deviation of the distribution of occupational scores into the deviation of the individuals occupational score from the mean occupational score of all sons having occupationally similar fathers, must take place prior to the adding on of ten in order to make all scores positive. The fact that Tumin and Feldman interpreted their GOMS scores in a manner contrary to that which is indicated by the manner in which they said it should be computed.ppy not invalidate their findings in the article under discussion. 37 There are certain indications that they actually computed their 2 scores in the manner indicated by revisions proposed here rather than the manner in which they said they had. In Table 2 they indicate that the mean mobility score for ag- ricultural day laborers is 9.04, thus indicating under achieve- ment.22 This under achievement is quite possible. However, such a value, 9.04, could not be arrived at as the mean mobility score for agricultural day laborers by the formula that they present. 'It could be arrived at through the use of the revision suggested here. Not having the data that Tumin and Feldman used, it is not possible to determine if the proposed revision of their formula would actually yield a value 9.04, but it may be concluded that the formula that they suggest could not. It is possible that their actual computations were made in accord with he revised formulation suggested here. In conclusion it is proposed that the GOES should be sgnabolized as follows: GOI-‘IS = as; + 10 = 2 CT where X respondents individual occupational score ? : mean occupational score of all respondents and brothers whose fathers have the same occupation CT’: standard deviation of the distribution of occupa- tional scores of all respondents and brothers whose fathers have the same occupation 22Ibid., p. 286. 38 The computational procedure would be same as that proposed by Tumin and Feldman except for step three, where the deviation of the individuals occupational score from the mean computed in step 2(a) would be arrived at by subtract- ing the mean from the individual score, and in step 4, where the division of the standard deviation computed in step 2(b) into the deviation computed in step 3 would take place prior to adding ten to the scores. Tumin and Feldman listed five possible advantages and two possible disadvantages of the GOMS as a measure of inter- generational occupational mobility. They are:23 Advantages 1. Minor changes in occupational position that are inconsequential are less likely to bias the measurements as mobility is measured relative to the average achievement of all those whose fathers had the same occupational rank. 2. Both fathers' and peers' statuses are built into the measure. 3. Individuals can be assigned mobility scores and sorted in terms of them. 4. The full experience of the sons of the highest and the lowest ranked groups can be accounted for as they may be called upwardly or downwardly mobile if their achievement is greater or less than that of the average for the sons of their group. 5. Birth order, birth spacing, and number of siblings are controlled for by randomizing their effect by building their achievement into the GOD/ES. eslbido, p. 284-850 39 Disadvantages 1. Concrete distances moved by the respondent and his actual career history can not be calcu- lated due to the fact that the GOMS is calculated in standard scores. 2. Due to the fact that the GOMS is a stand- ard score mobility scores computed from differently dispersed distributions may not be strictly com- parable. Tumin and Feldman's GOMS appears to have all of the advantages that Rogoff's social distance mobility measure has in controlling for the demand at various occupational levels and the supply of labor from each level due to the fact that it measures mobility relative to all peers who started out at the same point. The major difference is that it measures mobility as a characteristic of an individual while Rogoff's technique measures mobility as a characteris- tic of an occupational structure. .It appears that they are :mutually supplementary techniques for the measurement of :1nter-generational occupational mobility. The final method‘of measuring inter-generational ocnzupational mobility to be discussed in this chapter is the aéée cohort method of Gerhard Lenski.24 He classified his Inesspondents into three occupational classifications, white CCJILlar, blue collar, and farming; and then broke them 24Gerhard Lenski, ”Trends in Inter-Generational gs cupational Mobility in the United States,” American £°1ological Review, 19.48, 23:514-23- 40 down into five age cohorts in terms of whether_the date of birth of the respondent was between 1853 - 1892, 1893 - 1902, 1903 - 1912, 1913 - 1922, or 1923 - 1932. He further dis- tributed each of the age cohorts in terms of the occupation of sons vis-a-vis that of fathers. The trends in vertical occupational mobility were then measured by comparing the differences between upward and downward mobility for each age cohort after the latter two cohorts had been corrected for the differential affect of intra-generational occupational mobility.- Upward mobility was defined as movement from either blue collar or farmer occupational categories into that of white collar. Downward mobility was defined as movement from the white collar category to either blue collar or farmer. Movement from blue collar to farmer or from farmer to blue collar was not defined as being either upward or downward mobility. Lenski's age-cohort method of measuring occupational mobility.is, like that of Rogoff, one which measures mobility as a characteristic of occupational structures rather than of individuals. The problem that this thesis is concerned with is an analysis of the relationship between job satisfaction and inter-generational occupational mobility. It is therefore desirable to select for use measures of mobility that will be both valid and compatible with reference group theory. The direct comparison of the occupational levels of sons and fathers in the manner of Davidson and Anderson was 41 selected for that very reason. Parents are often used as a reference group and this method of measuring occupational mobility has been used in the bulk of the earlier studies. The revised GOES was selected for a similar pair of reasons. It is a very sophisticated and fine device for measuring inter~generational occupational mobility and it does dovetail well with reference group theory as it includes as components in its construction two groups that are important as reference groups. A third, but much more indirect method, was chosen more for its relevance to reference group theory than for its value in measuring occupational mobility. The comparison of occupational levels of brothers does yield some inferences regarding mobility vis-a-vis the immediate kin group but it fails to yield any information regarding achievements rela- tive to non-kin groups. If a respondent reported an occupa- tional level below that of his brother one would never know if the respondent was downwardly mobile, stationary and his brother upwardly mobile, or upwardly mobile at a slower rate than his brother. The method of career pattern analysis presented by Form and Miller would have been quite appropriate and the results from such an analysis would have been very interest- ing but it could not be used in this study due to the lack of the requisite data. This is one drawback of secondary analyses. One often will find data missing that one would like to use in.the secondary analysis that was simply not pertinent to the original study and thus not gathered. 42 The methods of Rogoff and Lenski were not used as they were not applicable to the problem of this study. They both treat inter-generational occupational mobility as a characteristic of the occupational structure rather than as a characteristic of the individual. CHAPTER III THE MEASUREMENT OF JOB SATISFACTION The purpose of this chapter is to describe and to dis- cuss some of the major methods that have been used to measure Job satisfaction. It is not its intention to discuss the findings of the many Job satisfaction studies, nor to describe all of the methods that have been used to measure Job satis- faction at various times. The purpose rather is to select certain measurement devices that are deemed to be representa- tive of the various types of techniques that have been used to measure Job satisfaction and to discuss their relative merits. For the purpose of this study we will accept HOppock's definition of job satisfaction. ”(Job satisfaction) is any combination of physical, physiological, and environmental circumstances that cause a person truthfully to say, 'I am satisfied with my Job.”1 HOppock goes on to state, with regard to this definition, that, Obviously a person may be satisfied with one aspect of his Job and dissatisfied with another. lRobert Hoppock, Job Satisfaction, (New York, Harper and Brothers, 1935), p. 47. 43 44 Our definition assumes that it is possible for him to balance the specific satisfactions against the specific dissatisfactions and arrive at a composite satisfaction with the Job as a whole. We do not mean that an unskilled laborer will sit down with a pencil and paper and analyti- cally check the advantages of the Job against the disadvantages until he strikes a logical balance. What, we do mean is that, through some psychological or physiological mechanism, most of us acquire a sort of vaguely defined attitude toward our total job situation. A multitude of satisfactions and dissatisfactions may, we assume, play upon each other to produce the composite attitude reflected in the state- ment, '1 am satisfied with my job.'2 The assumptions included in the above statement are acceptable to the present researcher. There are two basic types of approaches to the measurement of Job satisfaction, the direct and the indirect. Eicher has this to say with regard to them, The measurement of Job satisfaction may be further divided into 'direct' and 'indirect' 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 directly questioned about how well he likes his Job, whether he has ever thought of changing his Job, and so on. The indirect 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 presumably will indigate present Job satisfaction or Job aspiration. In the above statement, she has summed up the differ- ence between them. Her own choice was to use the indirect 2Ibid., p. 48. 3Joanne Eicher, "Job Satisfaction: Its Relation to 1Occupational, Stratification, and Community variables” (Unpublished Master's thesis, Michigan State University, East Lansing, 1956), p. 3. 45 method in her Master's thesis. Eicher used a job satisfac- tion index which was constructed out of the following three questions:4 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? 3. Which of all the Jobs you have had, did you like best? A response to any of these questions which indicated a choice of the respondent's present Job was scored one, and a response indicating a choice of any other Job was scored zero. This resulted in a four point scale ranging from zero to three. It is doubtful whether these three questions may legitimately be assigned equal weights and added to yield a composite Job satisfaction score. A respondent who has only had one or two jobs would have a limited range to select from in replying to Question 3 and might conceivably select his present Job as the one he has liked best and still be quite dissatisfied with it. On the other hand an individual, having all the occupations in the world to select from, might easily express an unrealistic desire by answering Question 2 with an occupation other than his own while still being quite satisfied with his present Job. In any event, \ Ibido , p13. 23-24. 46 the mere fact that a respondent would be more likely to select his present Job in response to Question 3 than he would to Question 2 as a result of chance alone should cause one to doubt the advisability of assigning equal weights to the two questions and adding them together in order to get an index of Job satisfaction. Woods attempted to measure a concept that is closely related to Job satisfaction, morale.5 He defined morale as, "the liking or disliking, in some degree, the various situa- tions that arise in the process of working with others'.’.6 He indirectly measured morale with a scale covering seventeen areas. Each of these areas contained a number of possible responses of which the respondent was to select the one which best approximated his feelings.. The responses had previously been given scale values from zero to ten in terms of the amount of morale they presumably represented. The categories covered by the scale were: the meaningfulness of instructions given to employees, the wisdom of employees making decisions and assuming responsibility, the utility of employees making constructive suggestions, capabilities of the supervisors, essentialness of the work done, employees willingness to do favors'for employees, employee courtesy to employee, employee courtesy to the public, source and 5wa1ter Woods, "Employee Attitudes and Their Relation to Morale," Journal 3; Applied Psychology, 1944, 28:285-301. 6Ibid., p. 285. 47 opportunity for advancement, recognition of merit, and re- lated type categories. A much less complex indirect technique of measurement was that used by Fortune in two of its surveys.7 In 1947 they asked two questions: "1. For a person in your trade or occupation do you think your company is about as good a place there is to work, or do you think there are other places that are better?" and "2. If you could go back to the age of fifteen and start life over again, would you choose a different trade or occupation?" In 1938 they asked simply, "If you could go back to the age eighteen and start life over again, would you choose a different career or occupation?" Super used a three question indirect method of measur- ing job satisfaction.8 His questions were: "1. What do you think you might like to be?”, "2. Are you satisfied with your opportunity to eXpress yourself in your work?" and "3. If not, why?" Lyman used a much more complex indirect technique when she focused on the problem of job dissatisfaction rather than job satisfaction.9 The three questions that she used 7Fortune Quarterly Survey XI: Fogtune:l7, January, 1938, p. 86 and The Fortune Survey: Fortune:35, May, 1947,, p. 10. 8Donald Super, "Occupational Level and Job Satis- faction," Journal 2; Applied Psychology, 1939, 23:547-64. 9Elizabeth Lyman, "Occupational Differences in the varue Attached to Work," American Journal 33 Sociology, 1955 , 61 :138—150 . 48 were: "1. If you could start over, would you go into the same kind of work again, or what would you like to do? (probe: this is 'are there any regrets' question.)”, ”2. Suppose you could get the same pay no matter what kind of work you did. Of all the kinds of work you can think of, what would you like best?" "3. How would you feel about a son of yours going into your kind of work?" A respondent was classified as being satisfied if, and only if, he gave the following combination of responses: "1. He would go into the same type of work again and he expressed no regrets. 2. He would like best the same type of work that he has at present. 3. He would approve of his son going into his.present occupation or if he somehow indicates that everyone should follow his own interests." Any other combination of responses was classified as indicating dissatisfaction. Lyman's technique seems to suffer from the same pit- falls that most of the indirect measurement techniques do, plus a few more that are peculiar to hers. It seems quite likely that an American factory worker could and would say that he would be a little disappointed if his son also be- came a factory worker, and yet still be very satisfied with his present Job for himself. Analytically, it would appear Ublae a mistake to lump together all of those who answer one, two, or three questions with a job other than their 49 own as being dissatisfied. Even if one were to grant that these questions validly measure job satisfaction or dissatis- faction, it would seem that such a lumping process would tend to obscure the relationships that go with the degree of satisfaction or dissatisfaction that one has with his job. In general, this researcher would have to accept the conclusions that were drawn by Gonon in his review of the various techniques used to measure Job satisfaction. They are as follows:10 ‘ The review of the literature has lead this researcher to believe that: (l) the proper method of measuring Job satisfaction is by directly asking such questions as 'Are you sa isfied or dissatisfied with your present job?‘ (2) there are no conflicts between the findings of job aspiration and Job satis- faction studies, because each of these studies is separate and distinct, using different questions that get at different variables and produce different results; (3) the incorpora- tion of Job aspiration questions in Job satis- faction scales is methodologically unsound, because it is possible for a dissatisfied person to desire to remain on his present job, and it is also possible for a satisfied person to desire a Job other than his present job; (4) attempts at imputing Job satisfaction to present Jppvresponses and attempts at im- puting Job dissatisfaction to other than present 192 responses to job aspiration questions are methodologically unsound for the same reasons given in (3) above; (5) it is proper methodology to study job aspiration and Job satisfaction separately; and (6) it is also proper method- ology to attempt to relate Job aspiration to Job satisfaction for the same sample in a research effort as long as the variables are prOperly labeled and kept separate and distinct. —— loPierre Gonon,-"The Relationship of Independence in the Work Situation to Job Acceptance: A Study of the Job Aspirations of Employed Adult Males" (Unpublished Master's thesis, Michigan State University, East Lansing, 1956) p. 45. 50 It is quite easy to visualize how a person could be satisfied with his present job and at the same time have aspirations for another job. This would especially be true if the jobs were heirarchically aligned in a normal career pattern; i.e., a medical intern who is satisfied with his position as an intern but aspires to be a doctor. The fact that responses can differ to questions about Job satisfaction and job aspiration within the same sample is well illustrated by a study conducted by William Form in Greenbelt, Maryland.11 He found that, "When members of the sample were asked what occupations they had most aspired to enter over one quarter, 26.6 per cent, stated that they wanted to go into business for themselves. Hbst of the remaining mentioned independent or dependent professional occupations and none mentioned skilled labor or office work. Only 17.3 per cent said they would like to enter the occupations they were in."12 He pointed out in a footnote, "Yet when asked directly how they felt about their present occupation, 45.8 per cent commented that they were happy in it, and 30.6 per cent were 'on the whole satisfied', 13.8 per cent were ambi- valent, 11.8 per cent were dissatisfied."13 This is a striking difference in response to two different types of questions that have at various times been used to measure Job satis— faction. The question is, ”Which is the more valid technique?" 11William Form, "Toward An Occupational Social Psychology," The Journal 9; Social Psychology, 1946, 25:85-99. 12Ibid., p. 86. 13Ibid., footnote, p. 86. 51 Another major problem with the indirect technique is that when one asks a respondent a question like, "Of all the Jobs in the world, what Job would you like the most?” or "If you could get the same pay regardless of the type of work you did, what type of work would you most like to do?”; one might be asking the respondent about things outside of his world of experience. It is possible that these are questions about a dream world that does not exist. In the world of the respondent's eXperience there may be real limitations to the Jobs available to him and different Jobs may be paid different amounts of wages. It is likely that when a respondent is asked questions about a dream world that his responses would also reflect that dream world and not his world of experience. Thus his answers to the questions might not be relevant to inferences one might wish to draw with respect to his actual behavior or his definition of the Job that he presently holds. As a result of the reasons stated above this researcher is inclined to agree with Hoppock's conclusion that, "Pending the development of a measuring instrument which is independent of rationalization and falsification, we are inclined to accept the subject's own estimate of his satisfaction as the best available criterion, and to regard a dissatisfied works man (one who says he is dissatisfied) as a dissatisfied work- man, whether his dissatisfaction is the result of rationaliza- tion, supervision, or indigestion."14 14Hoppock,_o_p. cit., p. 49. 52 It was therefore decided to use a direct method of ascertaining Job satisfaction in this study. There are two types of the direct method; the method of the 'whole' and the method of 'parts'. In the former method one asks a question regarding satisfaction with the Job as a whole, such as ”Are you satisfied or dissatisfied with your Job?”, while in the latter method one asks a series of questions re- garding satisfaction with components of the total Job situa- tion, i.e. supervision, wages, promotion possibilities, etc.. and from the responses to these one composes an index of total Job satisfaction. It is worthwhile examining a few of these methods that have been used in studies. Inasmuch as the Job satisfaction index scales COD! structed by Hoppock and by Bullock have either been used directly or as a model in the majority of Job satisfaction studies they will be discussed first. It is difficult to classify them as they each used a combination of the-in- direct, the direct whole, and the direct part methods. Hoppock's scale consisted of ten questions as follows:15 ”1. Check ONE of the following statements which best tells how well you like your Job.“ There were seven possible responses ranging from ”I hate” to ”I love it”. ”2. Check one of the following to show HOW MUCH OF THE TIME you feel satisfied with your Job". There were seven possible responses ranging from "all of the time" to "never". ”3. Check the ONE of the following which 15Ibid., p. 243. 53 best tells how you feel about changing your Job”. There were seven possible responses ranging from "I would quit this Job at once if I could get anything else to do.” to "I would not exchange my Job for any other.” "4. If you could.have your choice of all the {obs in the world which would you choose? check one)” There were three possible re- sponses: “your present Job”, "another Job in the same occupation”, and "a Job in another occupationt? ”5. Check one of the following to show how you think that you compare with other people” There were seven possible responses ranging ‘from "No one likes his Job better than I like mine” to "No one dislikes his Job more than I dislike mine." "6. Which gives you more satisfaction? (check one)” The two possible responses were ”Your Job" and "things you do in your spare time.” ”7. Have you ever thought seriously about changing your present Job?" "8. Have you ever declined an opportunity to change your present Job?" ”9. Are your feelings today a true sample of the way you usually feel about your Job?" ”10. On the line below, place five-check marks to show how well satisfied you were were with your last five Jobs. Use a separate'check mark for each Job. Draw a circle around the check mark which indicates yourapresent Job." The line referred to was broken down into eleven divisions with five statements ranging from "completely dissatisfied" to "completely satisfied. It is easy to see that questions 1, 2, 9, and ID are direct attempts to measure Job satisfaction, while questions 3, 4, 6, 7, and 8 are indirect attempts. Question 5 is a direct but relative attempt. The indirect questions all ‘suffer from the same weaknesses that were discussed earlier. Question 5 has a weakness all its own which will be discussed below. 54 Bullock modeled his scale after that of Hoppock. It consisted of the following ten items:16 ”1. Place a check mark in front of the statement which best tells how good a Job you.have.” There were five possible responses ranging from “The Job is an excellent one, very much above the average,” to "The Job is a very poor one, very much below the average.” "2. Place a check mark in front of the statement which best describes your feelings about your Job." There were five possible responses ranging from "I am very satisfied and happy on this Job.” to "I am very unhappy and dissatisfied on this Job.” "3. Check one of the following statements to show how much 2; the time you are satisfied with your 333}'—-There were five possible responses ranging from "most of the time” to "seldom.” ”4. Place a check mark in front of the statement which tells what kind of an organization it is to work for.” There were five possible responses ranging from ”It is an excellent organization to work for - one of the best I know of." to ”It is probably one of the poorest organizations to work for that I know of." "5. Place a check mark in front of the statement which best tells how your feelings compare with those of other people you know." There were five possible responses ranging from "I dislike my Job much more than most people dislike theirs” to "I likenmy Job much better than most people like theirs. "6. Place a check mark in front of the statement which best tells how you feel about the work you do on your Job." There were five possible re- sponses ranging from "The work I do is very un- pleasant. I dislike it.” to "The work is very enjoyable. I very much like to do the work called for on this Job." 16Robert Bullock, Social Factors Related _t_9_ Job Satisfaction, Monograph Number 75, fifireau oflfiasiness Re- search, (Columbus: The Ohio State University, 1952), pp. '10. 55 ”7. Check one of the following which best describes any general conditions which affect your work or comfort on this Job.” There were five possible responses ranging from "General working conditions are very bad.” to ”General working conditions are very good, much better than average for this kind of Job." ”8. Check one of the following statements which best tells how you feel about changing your Job," There were five possible responses ranging from ”I would quit this Job at once if I had anything else to do." to "I do not want to change Jobs even for more money because this is a good one." "9. Suppose you had a very good friend who is looking for a Job in your line of work and you know of a vacancy in this organization which your friend is qualified to fill. Would you:” There were five possible responses ranging from "Recommend this Job as a good one to apply for?” to "Try to discourage your friend from applying by telling the bad things about the Job?" "10. On the line below place a check mark to show how well satisfied you are with this Job." The line had five divisions and three statements distributed on it. They were "completely dis- satisfied," "about half and half," and "com- pletely satisfied." Questions 1, 2, 3, and 10 are attempting to ascertain Job satisfaction through the direct whole method. Questions 4, 6, and 7 are attempting it through the direct part method. Questions 8 and 9 are attempting it through the indirect method. Question 5 is a direct relative question, i.e. it is asking the respondent to evaluate the amount of his Job satisfaction relative to others that he knows. Questions 8 and 9 are subject to all of the short- comings of indirect questions that were discussed above. Gloria Cheek conducted a study in which she compared Bullock's index of job satisfaction with the Science Associates 56 Employee Inventory, a test of 78 dichotomous items of the agree - disagree type designed to measure Job satisfaction.l7 She discovered that Item 5 of Bullock, and consequently Item 5 of Hoppock, was neither valid nor reliable and that its removal improved the internal consistency of the scale. She therefore concluded that a person's satisfaction relative to others was not a good indication of his satisfaction. It is quite possible that a respondent may view himself as being more satisfied than his associates and still be dissatisfied or less satisfied than his associates and still be satisfied.18 Uhrbrock devised a direct part technique of measuring job satisfaction}9 He used a fifty item scale in which each item was designed to express varying shades of attitudes to- wards the company policies and practices. Several state- ments ranged from being extremely laudatory to being ex- tremely unfavorable towards the company policy or practice. The Fortune survey of 194620 and a study by Centers21 are typical examples of the direct-whole technique of measur- ing job satisfaction. The former asked the question, ”Are 17Glorla Cheek,’%.Psychometric Study of Two Indices of Job Satisfaction” (unpublished Master's thesis, Michigan State University, East Lansing, 1955). lBIbide' pp. 27-280 19Richard Uhrbrock, "Attitudes of 4430 Employees," Journal g£_Socia; Psychology, 1934, 5:365-377. 20The Fortune Survey: Fortune, 1946, 34:Dec., p. 10. 21Richard Centers, ”Motivational Aspects of Occupa- tional Stratification," The Journal 2; Social Psychology, 1948, 28:187-217. 57 you reasonably well satisfied with your present job?" and the latter asked, "Are you satisfied or dissatisfied with your present job?" Thorndike used a slightly different variant of the direct whole method of measuring job satisfaction.22 He had skilled interviewers note the person's liking for his work on a seven point scale after a conversation designed to evoke a frank expression of his attitudes toward it. Hoppock has summed up excellently the relative merits of the direct whole and the direct part methods of measuring job satisfaction. Questions and statements regarding the job as a whole place upon the subject the responsibility for weighing all of the factors involved, according to their relative importance to him, and reaching a summary conclusion. The advantage of this is that the relative importance of the different aspects of a job, as determinants of job satis- faction, may vary greatly from individual to in- dividual. Consequently the summation of these in the hedonic tone of the subject, as he tries to express the degree of his satisfaction, may give to each a relative weight which is more accurate than the best regression equation that could possibly be computed from group data. The disadvantage is that temporary elevation or misery, occasioned by a single factor in the situation, may for the moment completely overshadow all other con- siderations and lead the subject to express a de- gree of satisfaction quite different from his average feelings.... The approach by means of questions regarding specific aspects of the job has the advantage of the stability above noted, and relieves the indivi- dual of the necessity of reaching a composite 22Edward Thorndike, "Workers Satisfaction,” Occupa- tions, 1935, 13:704-6. 58 opinion. In so doing, it places upon the in- vestigator the responsibility for assigning proper weights to each of the many aspects involved. Be- cause of the high probability that these weights differ greatly from one person to another; and the possibility that some single item may for one sub- ject outweigh all of the others, while a different item may be the all-important one for another sub- ject; it seems inevitable that any scoring kgy would be seriously misleading in some cases. It appears that Hoppock's analysis is essentially correct, except that it might be better to substitute the statement that "certain types of individuals would differ in the relative importance that they would give to the various items" for his statement ”that individuals would differ." As a result of the various conclusions discussed above, this researcher has decided to use the direct whole method of measuring job satisfaction in the present study. Inasmuch as the focus of this study is on the relation- ship between job satisfaction and occupational mobility, a variable which can be quantified, and because other com- parisons will be made with quantifiable variables, i.e. education, age, income, tenure; it was deemed desireable to use a question to measure job satisfaction that would die- criminate between degrees of satisfaction and not merely between satisfaction and dissatisfaction. The question selected to be used was, ”How do you like your job?" and the five alternative responses were: "very much", ”pretty good”, "average", ”not so good", and ”not at all”, 23H0pp00k, 22. Cite, pp. 272-740 CHAPTER IV OCCUPATIONAL MOBILITY, JOB SATISFACTION, AND SOCIAL REFERENCES: AN EMPIRICAL EXAMINATION This chapter consists of an empirical examination of the data collected, the testing of the hypotheses presented in Chapter I, and the relating of the findings to reference group theory. The mean job satisfaction scores presented in the findings below were arrived at by arbitrarily assigning weights to the response categories of the job satisfaction question, ”How do you like your job?", in the following manner: "very good” = ; "pretty good" = 4; "average” = 3; "not so good" = 2; and "not at all" = l. gypothesis 1. There Hilllgg'g Low Positive Association Between Job Satisfaction and Occupational Achievement Relative pg Parents' Occupational Aspirations For the Respondent The majority of the respondents in Table I reported no parental occupational aspirations for themselves. The number reporting parental aspirations was too small to per- mit any statistical analysis. It is worth noting, however, as an indication that might bear later study, that those who 59 60 had achieved an occupational level equal to their parents' occupational aspirations for them and those whose occupational achievement exceeded parents' aspirations had higher mean job satisfaction scores than those reporting no parental aspira- tions and those reporting parental aspirations higher than their present occupational levels. Those who were at the occupational levels which their parents aspired for them had higher mean job satisfaction scores than did those who ex- ceeded parental aspirations. The number of cases in these cells were too small to be significant. The most significant finding here is the large pro- portion of respondents who reported no parental aspirations. This could quite possibly indicate that parental occupational aspirations are not significant in determining the occupa- tional aspirations of working class males and such parental aspirations are not remembered or reported. TABLE I MEAN JOB SATISFACTION SCORES FOR RESPONDENTS GROUPED ACCORDING TO THEIR OCCUPATIONAL ACHIEVEMENT RELATIVE TO THEIR PARENTS' OCCUPATIONAL ASPIRATIONS FOR THEM Occupational AchIevement Relative To Parents' Occupational Aspira- Mean Job tions For Respondents Satisfaction Frequengy No aspirations 3.83 296 Occupational level lower than aspirations ' 3.88 66 Occupational level same as aspirations 4.50 8 Occupational level higher than aspirations 4.00 3 W 61 Hypothesis 2. There Will fig 2 Moderate Positive Relationship Between Job Satisfaction and Sons' Occupational Level Relative pp Fathers‘ Occupational Level The hypothesis is confirmed. The relationship is strong and in the expected direction. Thus in Table II the difference in mean job satisfaction scores between those sons who were upwardly nobile, i.e., those sons whose occul - tional level was higher than that of their fathers, and that of those who were downwardly mobile, i.e., those sons whose occupational level was lower than that of their fathe.s, was significant at the .0005 level. The stationary or immobile sons, i.e., those sons whose occupational level was the same as that of their fathers, had mean job satisfaction scores that were much higher than the downwardly mobile sons, 4.01 to 3.39, but only slightly lower than the upwardly mobile sons, 4.01 to 4.03. This might mean that job satisfaction will be seriously impaired if a working class male fails to achieve an occupational level comparable to that of his father but that once he has reached this level rising above it does not significantly increase his job satisfaction. 62 TABLE II JOB SATISFACTION AND OCCUPATIONAL LEVEL OF SONS RELATIVE TO OCCUPATIONAL LEVEL OF FATHERSl Eons' Occupations Relative Mean Fob Eatis- _ To Fathers' Occupations faction of Sons Frequency 1. Sons' occupation lower than fathers' occupa- tion 3.39 122 2. Sons' occupation the same as fathers' occupation . 4.01 101 3. Sons' occupation higher than fathers' occupation 4.03 77 t13 = 4.57. P< .0005, df, 197 hypothesislfi. There Will £3 a Positive Relationship Between Job Satisfaction and Occupational Level 9; Respondent Relative 32 Respondents' grothers' Occupational Level Hypothesis three is confirmed. As is illustrated in Table III, the direction of the relationship between Job satisfaction and the respondents' occupational level relative to that of their brothers’ occupational level is as expected. The difference between the mean Job satisfaction scores Of those whose occupational level was lower than that of their brothers and the scores Of those whose occupational levels 1Helen M. Walker and Joseph Lev, Statistical Inference, (New YOrk: Henry Holt and Company, 19535. p. 155. The source of the formula used in computing the significance of the differences between means in this table and wherever used in this thesis. 63 were the Same as their brothers was not statistically signifi- cant. However, the differences in mean Job satisfaction scores between each of these categories and the category of respondents whose occupational levels were higher than the occupational levels of their brothers was statistically significant. It is on this basis in addition to the general trend that the conclusion is drawn that the hypothesis is confirmed and that occupational mobility relative to one's brother is a significant determinant of Job satisfaction among working class males. It is worth noting that the mean Job satisfaction score of all those whose brother was in farming was 3.85 and that the mean job satisfaction score of all those whose brother was in a non-farming occupation considered to be non-comparable to that of the respondent was 3.90. Both of these scores closely approximate the mean job satisfaction score of those respondents whose brother was on the same occupational level as themselves, 3.88. These close rank- ings are quite consistant with reference group theory. If those that one would normally use as social references in judging one's occupational success are in occupational situses that are not comparable then it is difficult to imagine how a comparison could be made that would result in feelings of gratification or deprivation. This is directly analogous to the situation that occurs when one's social references are in the same occupational Situs as oneself. No com- parisons which result in feelings of gratification and deprivation can occur in either case. 64 TABLE III JOB SATISFACTION AND OCCUPATIONAL LEVEL RELATIVE TO OCCUPATIONAL LEVEL OF BROTHERS* OOOupational Level Relative Mean Job To That of Brother Satisfaction Frequency 1. Occupational level lower than that of brother 3.67 42 2. Occupational level same as that of borther 3.88 42 3. Occupational level higher than that of brother 4.33 18 *t1 = .88, P> .10, df, 82; t2} = 2.05, P‘Dil.11ty. We found no association whatsoever. This would mWEI-n that Job satisfaction is not in any way a function of upward mobility for unskilled workers. It could further indicate that they do not consider a job to be something to be satisfied with or dissatisfied with. Instead it may be merely a necessary condition in one's life and one's major Batistactions are derived from off-the-Job situations. ‘1 N ove is with reference to mobility into 5 m ,. }.J H 0 w CL 0 “”1 r7“ _cvw (occupation of destination. TLis raises certain questions tai_tfi; reference to the relationship between job satisfaction J arzc: u1ward LOblllby when occupation of origin, i.e., father's o C C Li“: Lion, 1:3 control l certain insi;_;li;ts into til: r:lationsl-iv rm- , . ,-4.'£A. .2“ -. o ‘, I 46“ t ( res nlent: are classiil:u in terns oi father 3 oczczxrjstion or go:ition of origin, there is a roierate positive ’2“ t H O H 7 P. H x (r s (" H .- S... (.4 '3 fl 1 O ("7‘ H I ‘ity and 300 satisfaction for f fgtficp'y accusations. Lo setter where one starts H u 4 O I...) O i 't. in life, upward mobility results in increased job satis- L o fhime‘tion in comparison with the amount of job satisfaction of c>tli<3rs who started life in the as: .0 position. The conjunctijn of the indiigs regsrdir; the relation- :Z;1.”\Tletweefi ‘01:;uwtisffc tio:::ILi upward motilitw rwfig:aongents are classified accordin.i to position of origin l..J the present researcher DJ (J srri ‘by rosition of is tination C-Q ‘ ‘- r 1 A. - -o '1 V, ‘ .w ,‘I ‘ "‘ ‘ \ ‘-‘ '." | w‘ V to cconclude that working clzss males is use their peer ”coupe, P91 :I‘ehccs in deternining their satisfaction with their occupa- ‘vat. If tLo" Ltve achieved LOPG upward sobility 1? their peers their level of job satisfaction will be COP1~ :3 31-01101 “1" higie'. This relationship will be modified 0 ‘ ‘1“‘K‘ "I1? +‘ o o . " . '~ . ‘ " " “ u--c position of 3a"tinution. If the position of destina- ikely to FJ ti xa v '1 - ‘ \ . ‘- ‘ - '1'" ' (V f ‘ ‘ '1 lei- ls. TY LIBRARIES STATE UNIVER MICHIGAN 3 IT? 1293 0306’ 2877