W "I". III. ‘1? II” M. IWI/IIIIL1I1IIIII‘5I'IIII‘I' IIIRI‘H M II. *I’III II1I1I1I I III" 5. MIR-11' 11 1". furs 1.8+: 1.1;- .,.,, I .II:' IIIIIIIIIIII 1I11I I .1 . .II.... :I‘",III I... II'I'II WI. IIIII 11 73'I .‘.II IIIIIII II" II' I III IIIIIIIII .1 11. .IIIII. .1 .II 1.. I- .I' III. I- N I 1...I11 I ' IIIIIIIIIIIII IIIHIII MIIIIII I IIIII1I.1.IIII II 1111111I'IIIII1I'III' "III I II .. I III I. 1. I II "III. IIIII..I.1:1‘11II.I 1,11I. II" “III III" III 1. III‘“ . III III III "IIII’III “III. I... III III III III. IIII1II1'I'I’1'II IIIII1II.I'1II {1121. I1I1II.1.1.IIIII' III; .111 'II IIII II III1'III1I1.|. .11 IIIIIII 'I’IIIIIII'IIII III I IIIIIIIIII I III .. I§IIIII III .3 . . ‘ II Ir’II. .. IIIIIIIIII III II I.,.I ‘II‘II II IIIII I. II' |II I‘Ix. II Id“I I III“: I.;1 mII II“ I In; IIII“ II..'”II III III‘I"II IIIIIII 'II1II '111IIIII 1I‘1IIII 1111.11. . III11I IIII I.I,IIIIII III II 1II1II1I1111. .I11 I’II I 11.. II IIII I , II II III}: (I1:IIIII II“ 'I1II I1I1 II I -I.III¥ - .“1 IIIIIII‘II .II III 1II-11I,1. III I I IIII W I' III III IIIIIIIII IIl .‘ '~ II .I I III III III-III, IIIII 'III1-II.I 1111'II II III 1I1I ”911.?" "3‘5 IIWII INI I III 13' IIWfIV'J I14“ W1 III IIW HM” ‘f?&-'fi3 3| I: 1.1,1 .~ 1: 11II'I1'11'II ' 111.111 III1I1" I '1 " 1'1' :J, 4 WI I11IIIII|III 1-1... 111 I .11 11 1II1‘1 .1‘1 11III1II'I» II II II 1 'IIILII. ;- .. IIII‘II 1 I? 1. . ' I T I‘ . I‘I‘III'MV’ II. I .II. I.II-. "I ;~: . III-III 1 '.III I'. III .‘I . 1 "I.;.‘.‘II:.-lXIII-ItI'11‘I,.1'.'J'.‘.i'IIIIII‘, 11111111.1 III .. _ . .II'.’ - PI I‘11..1..III. III iIII‘ ‘1 .I I‘ TIII'IIIIIII .111111111 III I II III1 1'! I III." 5 1.13 I'III'. I I‘ IIII'I'1I1.. ' II III‘1.1 I I1I.1|I 1‘ ‘ ' 1 ”.I." _111I11 1 III'I11'. . 1.11 III » II :II“"‘ I . I II III I ‘ ,....I1;;.1I1.’I .:I‘1;: 11,.I.. .IIILIIIIII II; II .I.... ‘ "III.“ . I. ‘IVI'I "III IIII.II_‘*-I I II" IfII'i11'II‘I‘" I 1111.1 III 1111 III1 .1'111 III1‘ 11 I . 1 11”. 'I III II III 'I:|:1,1:II11I11I.I I111v11 11%;“. IhI'III I III.) II'1I . 1- IIII ' II I ‘ 11"11 ... I I. .I.. :II 111.7“); .31159m ‘jIIVl’I ' III . .. .1II... .. . I I 111 II I III '. 1 I. I 1.71 IIIIIIJI‘IJIfII‘ WI I1| 111 111.1II1I'1II1I'111111i11.111 I... 1. - II‘ .' , .‘III I I I IIII I.’.I‘III' '1. I‘11'1'I.1.1. 1 11.1 .I - ,, III1I.III'I‘ II, IIIIII'III" . ._ .III-t. . ., 1111111.. IIII \IIIIIIII'I' .- ., II III' III ’ II‘IIIIIIIIII - III II IIIII ‘I IIII I I . I; '1II|.I‘ 11.! , 1‘ 1. 1 , ”111111111111I1I.11111111;II1‘II IIII II‘ I I' I 'III” II 1‘ II II 11 1111' II.’ I .I .'I|I'III II III‘II‘III II II' IIIIII . I 11 11... I .II'I 1IIIIIIIII 'I. . fIIIIIIIIII'II'IL 1IIIIIII IIIIII. I II11111 IIIIQIIIIII [Lil'w'1 I ""5"" LIBRARY "1.8m University This is to certify that the thesis entitled Intergenerational Occupational Mobility of Elementary School Principals in the Middle United States presented by Patsy Robinson Hashey has been accepted towards fulfillment of the requirements for Ph . D . degee in Admini S nation and Higher Education 49;qu warm, Mail: professor Date Mal/1,,Icl7c1 0-7639 Stan: OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. © Copyright by PATSY ROBINSON HASHEY 1979 INTERGENERATIONAL OCCUPATIONAL MOBILITY OF MALE AND FEMALE ELEMENTARY SCHOOL PRINCIPALS IN THE MIDDLE UNITED STATES by Patsy Robinson Hashey A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Administration and Higher Education 1979 ABSTRACT INTERGENERATIONAL OCCUPATIONAL MOBILITY OF MALE AND FEMALE ELEMENTARY SCHOOL PRINCIPALS IN THE MIDDLE UNITED STATES by Patsy Robinson Hashey The study was conducted for the purpose of deter- mining the patterns and processes of intergenerational occupational mobility among elementary school principals in the middle United States. During the 1976-77 school year, approximately 6800 individuals were members of the National Association of Elementary School Principals (NAESP) in the middle United States. It was determined that a representa- tive sample of elementary school principals in the middle United States would contain at least 606 respondents. A sample of 977 members was obtained from the NAESP; data were collected from 697 elementary school principals (78.86% of the eligible respondents) by a researcher developed mail out questionnaire. Six related research questions were formulated for examination, primarily by the third edition of the Automatic Interaction Detector (AIDS) - a computer program designed especially for complex questions in the social sciences. The research questions were as follows: Patsy Robinson Hashey What is the pattern of intergenerational occupational mobility (as measured by the SEI) for elementary school principals in the middle United States from background characteristics? Do the patterns of intergenerational occupav tional mobility (as measured by the SEI) differ for male and female elementary school principals in the middle United States from background characteristics? Will the pattern of intergenerational occupa- tional mobility (as measured by the SEI) from background and intervening characteristics for elementary school principals in the middle United States be replicated by a cross— validation sample? Do the patterns of intergenerational occupa- tional mobility (as measured by the SEI) differ for male and female elementary school principals in the middle United States from background and intervening characteristics? What is the process of intergenerational occupav tional mobility for elementary school principals in the middle United States? Does the process of intergenerational occupav tional mobility differ for male and female elementary school principals in the middle United States? Patsy Robinson Hashey Results of the analyses indicated that elementary school principals in the middle United States were upwardly mobile from father's occupation. No major differences in the pattern of mobility were detected except that men from blue collar origins were more mobile than women in the same category. Of the 24 variables examined, only father's occupational category, father's education and, for subjects from farm, deceased, and unemployed father's origins, whom respondent lived with at age 16 exhibited importance for the pattern of mobility among elementary school principals in the sample. It was concluded that at the time the major- ity of the sample became elementary school principals (19605 and 19705), the position in the middle United States was open regardless of origin status. The process of mobility appeared to be different for men than women. The process of mobility was identified as follows: the mother did not work outside the home, the parents lived tOgether and had more than one child, male elementary school principals were teachers for 10 or less years, married with one to three children, 35 years of age or younger with a master's degree or higher at first princi- palship. For women the process of mobility was described as follows: the parents lived together and had more than one child, at first principalship women were between 26 and 45 years of age, held a master's degree or higher, and were teachers at the elementary school level. Patsy Robinson Hashey The study was successfully cross-validated by a 20 percent sample, with a 95 percent confidence interval about the means. DEDICATION It is to my parents, Patricia and Edward Robinson, whom I dedicate this dissertation. Without their reinforcement, and the background characteristics they provided for me, I would not have written a dissertation. ii ACKNOWLEDGEMENTS The author wishes to express sincere appreciation to the National Association of Elementary School Principals for supplying the names and addresses of a sample of elementary school principals in the middle United States. Special gratitude is extended to those principals who used their precious time to respond to the Survey of Elementary School Principals; their efforts made this dissertation possible. I am deeply indebted to the dissertation committee, especially for their patience in assisting me 600 miles from campus. To Dr. Stanley Hecker, committee chairman, I thank for helping me attain my goal; appreciation is also extended to Dr. Samuel Moore for his critical wit, to Dr. Glen Cooper for sitting-in after the death of Dr. VernOn Hicks, and to Dr. Philip Marcus for his patience with my neophyte approach to sociology. Computer assistance was provided by Fong Chan and John Yuen. Without their determination and the resources at Southern Illinois University Carbondale, I would not have persisted. iii Moral support, incentives, and distractions were kindly contributed, unknowingly at times, by J.R. Special appreciation is also extended to Brett, for ignoring his mother and living on self-made peanut butter sandwiches for months. iv TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES. Chapter 1. INTRODUCTION. The Problem . . . . . . . Purpose . . . . . . . . Significance and Need for the Study Research Questions. Basic Assumptions . . Definition of Terms Delimitations . . . . . Limitations . . . . . . . . . . . . . Summary . . . . . . . . . . . . . 2. REVIEW OF RELATED LITERATURE. . . . . . . . . MEASURING THE STATUS AND PRESTIGE OF OCCUPATIONS . INTERGENERATIONAL OCCUPATIONAL MOBILITY . METHODS OF DETERMINING PATTERNS AND PROCESS . Assessing Patterns of Mobility. Assessing Patterns and Processes of Mobility Difficulties in Measurement Page ix xii 10 11 12 13 15 25 32 32 35 38 Chapter 3. 4. MOBILITY VARIABLES SOME CHARACTERISTICS OF ELEMENTARY SCHOOL PRINCIPALS. THE AUTOMATIC INTERACTION DETECTOR . SUMMARY. METHODOLOGY. Sample and Data Collection . Instrumentation. Development of the Instrument. Descripter Variables Coding Occupation. Independent Variables. Dependent Variables. Design and Statistical Procedures. Summary. RESULTS AND DISCUSSION . Results. Research Question 1. Research Question 2. Research Question Research Question Research Question GUI->04 Research Question Review and Discussion of Significant Findings vi 47 57 65 71 73 73 75 75 77 78 81 89 91 99 100 100 100 107 116 119 125 127 129 Observations . . . . . . . . . Summary. . . . . . . . . . . . . Chapter 5. SUMMARY, LIMITATIONS, CONCLUSIONS, AND RECOMMENDATIONS. SUMMARY. Purpose. Research Questions Methodology. Findings LIMITATIONS. CONCLUSIONS. Importance to Education. RECOMMENDATIONS. APPENDICES A. B. Socioeconomic Index. Survey of Elementary School Principals Follow-up Postcard Frequency Distributions of Descripter Characteristics. Frequency Distributions of Background Characteristics. Frequency Distributions of Intervening Characteristics. vii 142 151 152 152 152 152 153 155 160 163 164 165 167 167 187 193 194 198 201 G. Definition of Categories of Elementary School Principal's Descripter Characteristics. . . . . . . . . . . 203 LIST OF REFERENCES . . . . . . . . . . . . . 206 viii la Table 10. 11. 12. LIST OF TABLES SEI and Prestige Scores of Elementary and Secondary School Teachers and Administrators . . . . . . . . . Definition of Categories of Elementary School Principals' Background Characteristics Definition of Categories of Elementary School Principals' Background Characteristics Correlations Between Study Variables Variation in Father's SEI Scores Explained By Respondent's Background Characteristics. The Pattern of Mobility from Background Characteristics, Final Groups in Rank Order of Mobility. . . . . . . . . . Frequency of Employed Father's SEI Scores When Respondents Were 16 Years of Age. The Pattern of Mobility from Background Characteristics for Men and Women, Final Groups in Rank Order of Mobility . Sex Variation in Father's SEI Scores Explained By Respondent's Background Characteristics Frequency of Employed Father's SEI Scores When Male Respondents Were 16 Years of Age. . . . Frequency of Employed Father's SEI Scores When Female Respondents Were 16 Years of Age. The Pattern of Mobility from Background and Intervening Characteristics of the Study Sample, Final Groups in Rank Order of Mobility . ix Page 25 84 87 95 101 103 106 109 112 114 115 118 Table 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Page Variation in Father's SEI Scores Explained By Respondent's Background and Intervening Characteristics for the Study Sample. 121 Cross-validation: Study Sample Mean Difference Confidence Intervals, Final Groups in Rank Order of Mobility. 121 The Pattern of Mobility for Men and Women from Background and Intervening Characteristics, Final Groups in Rank Order by Their Mean Differences Variation in Father's SEI Scores Explained By Intervening Characteristics of Men and Women. Variation Explained by Background, Intervening, and Descripter Variables for Men, Women, and Total Sample. . . . Variance Explained by Background Variables for Men and Women on Each Group (Minimum Group Size Reduced to Two) . . . . . . . . . Percent Men and Women From Each State in the Sample. . . . . . . . SEI Scores for Some Public School Professional Positions Were You a Teacher Before Becoming an Elementary Principal? At What Level/levels Did You Teach? Number of Years as an Elementary School Principal . . . . . . . Highest Earned College Degree Area of Specialization (highest degree held). Number of Schools Currently Under Direction . . . . . . . . . . . Age at First Principalship. 123 124 135 140 143 162 194 194 194 194 195 195 195 Table Page 28. Total Enrollment in the School/schools Under Direction. . . . . . . . . . . . . . . 196 29. Total School System Enrollment . . . . . . . 196 30. Regular Salary for the 1976-77 School Year . 196 31. How Many Months Are You On Contract? . . . . 197 32. State of Employment. . . . . . . . . . . . . 197 33. Father's Occupational Category . . . . . . . 198 34. Mother's Occupational Category . . . . . . . 198 35. Highest Level of Education Reached by Your Father. . . . . . . . . . . . . 198 36. Highest Level of Education Reached by Your Mother . . . . . . . . . . . 199 37. Sex. . . . . . . . . . . . . . . . . . . . . 199 38. Age. . . . . . . . . . . . . . . . . . . . . 199 39. At the Age of 16 Did You Live With . . . . . 200 40. Brothers and Sisters . . . . . . . . . . . . 200 41. Marital Status . . . . . . . . . . . . . . . 201 42. Do You Have Children . . . . . . . . . . . . 201 43. If You Have Children, How Many. . . . . . . 201 44. Within Which Age Range Does Your Youngest Child Fall . . . . . . . . . . . . . . . . . 201 45. How Many Years Did You Teach . . . . . . . . 202 46. Highest Earned College Degree At First Principalship. . . . . . . . . . . . . . . . 202 47. Size of City/town of Current Employment. . . 202 48. Definition of Categories of Elementary School Principals' Descripter Characteristics . . . 203 xi LIST OF FIGURES Figure Page 1. The Pattern of Mobility from Background Characteristics for Elementary School Principals . . . . . . . . . . . . . . . . . . . 102 2. The Pattern of Mobility for Men and Women from Background Characteristics of Elementary School Principals . . . . . . . . 108 3. The Pattern of Mobility from Background and Intervening Characteristics for Elementary School Principals (Study Sample). . . 118 4. Cross-validation Sample (20%) . . . . . . . . . 120 5. The Pattern of Mobility for Men and Women from Background and Intervening Characteristics to the Elementary School Principalship . . . . . . . . . . . . . . 122 6. A Pr0posed Model for the Study of the Mobility Among Elementary School Principals by Origin Strata . . . . . . . . . . 138 xii Chapter 1 INTRODUCTION The American occupational structure is stratified into a hierarchy of occupational groups.1 Within each major group of the occupational hierarchy, e.g., pro- fessional, clerical, laborers, work roles are further stratified. Varying levels of achieved education and income by societal members have contributed to the strati- 3 explains that: fied occupation structure.2 Sorokin If the members of a society are differentiated into various occcupational groups, and some of the occupations are regarded as more honorable than others, if the members of an occupational group are divided into bosses of different authority and into members who are subordinated to the bosses, the group is occupationally stratified.... The process of reaching a specific stratum in the occupational hierarchy and the subsequent interaction with- in that stratum, define the degree to which one has achieved occupational mobility.4 The amount of mobility 1Judah Matras, Social Inequality, Stratification, and Mobilit (Englewood Cliffs, New Jersey: Prentice-Hall, nc., , p. 110. 2Albert J. Reiss, Jr. and others, Occupations and Social Status (New York: Free Press of Glencoe, 1961), pp. 84-85. 3Pitirim Sorokin, Social Mobility (New York: Harper and Brothers, 1927), p. 11. 4Peter M. Blau and Otis Dudley Duncan, The American Occupational Structure (New York: John W1ley 8 Sons, Inc., 1967), p. 76. 2 found within a society reflects the Openness of the society, e.g., equal opportunity.5 The social origins of labor force participants are a major determinant of occupational mobility between occupation groups. Comprehensive research on the social origins of labor force participants has been conducted by Blau and Duncan6 who state that: The chances of occupational achievement are limited by the status ascribed to a man as the result of the family into which he was born. Indeed, a stable society is hardly conceivable that does not ascribe to every child a status in some kinship group, which is responsible for rearing and socializing him, and which, therefore, strongly influences his motivation to achieve, his qualifications for achievement, and hence his chances for success. Knowledge of the occupational structure and the conditions governing a person's chances of moving up the occupational hierarchy and achieving economic success is "...essential for understanding modern society and, particularly, its stratified character."7 The intent of the study presented here was to focus on a specific work role within a major occupation group. The study was a descriptive investigation of the intergenerational occupational mobility of male and female elementary school principals in the middle United States SAage B. S¢rensen, "Growth in Occupational Achieve- ment: Social Mobility or Investment in Human Capital," in Social Indicator Models, ed. Kenneth C. Land and Seymour Spilerman,(New York: Russell Sage Foundation, 1975), p. 336. 6Blau and Duncan, op.cit., p. 207 7Ibid., pp. vii-viii. 3 who were members of the National Association of Elementary School Principals during the 1976-77 school year. A survey of elementary school principals was conducted and the data were analyzed in an attempt to determine the similarities and differences in the patterns and processes of intergenerational occupational mobility among holders of that position. Statement of the Problem The problem is that the patterns and processes of intergenerational occupational mobility of male and female elementary school principals have not been studied. The source of the problem is two fold. First, occupational strata are broad - covering a variety of jobs with varying prestige and status accorded the individuals who fill the job roles within each stratum. As a result data gathered for the purpose of analyzing occupational strata supply only scant knowledge of individuals who work at specific jobs within stratum. Second, most studies of the American occupational structure have excluded the female worker.8 Therefore the 8Peter Y. DeJong, Milton J. Brawer, and Stanley S. Robin, "Patterns of Female Intergenerational Occupational Mobility: A Comparison with Male Patterns of Intergenerational Occupational Mobility," American Socio- logical Review, Vol. 36, (December, 1971), p. 1033. 4 question of whether the patterns and processes of inter- generational occupational mobility are the same for men and women remains unresolved. To further compound these problems, the similar- ities and differences in intergenerational occupational mobility of elementary school principals have been infrequently investigated in the sociology of work.9 Previous attempts have not only been rare, but have been very rudimentary in nature. Purpose It was the purpose of this study to define the patterns and processes of intergenerational occupational mobility among elementary school principals in the middle United States who were members of the National Association of Elementary School Principals during the 1976-77 school year. Significance and Need for the Study The movement of individuals from similar social backgrounds may disperse into many occupations or become concentrated in only a few.10 Studies of the United States 9Neal Gross and Anne E. Trask, Sex Factor and the Management of Schools (New York: John WiIey 8 Sons, 1976), p. 20. 10Blau and Duncan, op.cit., p. 42. labor force cover the entire occupational structure in order to assess determinants of patterns and processes of intergenerational occupational mobility to occupational categories, "...not the individuals composing them."11 In order to determine if a person's chances are limited or enhanced by their background characteristics, it was deemed necessary to view the occupational structure from the standpoint of individuals rather than from large occupa- tional categories. The elementary school principalship offers a unique and challenging area of study for at least two reasons. First, since little is known of the patterns of intergenerational occupational mobility to the elementary school principalship, recruitment for the position may be from a narrow or a wide base of origins. If it is found that elementary school principals in the middle United States experience upward mobility from their origins, it may indicate that these principals aspire to their posi- tions but if downward mobility is revealed, they may be settling for that position instead of aspiring to it. In either case of vertical mobility, it is possible that patterns of intergenerational occupational mobility from their occupation origins exist, and it is also possible that the patterns may be the same or different among holders of that position. 111bid., p. 23. 6 If school principals originate from a wide range of origins, this finding would lend more credibility to the notion that personal competence and the desire to perform as an elementary school principal are more influential in career selection than occupational origins. If, however, elementary school principals are from similar occupational origins, one alternative explanation might be that compe- tence and desire for that position are developed early in life as a result of familial influence. Second, the process of intergenerational occupa- tional mobility may be similar or different, i.e., back- ground and intervening characteristics salient to the process of attainment may vary among elementary school principals in the middle United States. If no differences are detected, it might be assumed that background and intervening factors influence elementary school principals similarly in the process of occupational attainment. It is possible however that the process of intergenerational occupational mobility to the elementary school principal- ship in the middle United States is different among holders of the position. In other words, comprehensive study of individuals who become elementary school principals in the middle United States may help clarify theories of career choice and individual qualification for entry into a specific occupation. In addition, the study of patterns and processes of intergenerational mobility of elementary school principals in the middle United States may shed 1'1 ge IN 81 7 light on the similarities and differences in the inter- generational occupational mobility of elementary school principals in the United States and in other occupational areas as well. Research Questions The problem of the present study points to the fact that there is only scant knowledge of the patterns and processes of intergenerational occupational mobility of elementary school principals. To begin to fill the void, it was determined that background characteristics and some intervening characteristics should be investigated in an attempt to define occupational flow to the elementary school princi— palship. While conclusive evidence should not be drawn from the results of one study, the primary objective here was to begin to determine some salient patterns and processes of intergenerational occupational mobility among elementary school principals in the middle United States. Six research questions were posed for the present study. They were as follow: Research Question 1: What is the pattern of intergenera- tional occupational mobility (as measured by the SEI) for elementary school principals in the middle United States from background charac- tics? Re: Re Research Question Research Question Research Question Research Question Research Question 8 Do the patterns of intergenerational occupational mobility (as measured by the SEI) differ for male and female elementary school principals in the middle United States from background characteristics? Will the pattern of intergenerational occupational mobility (as measured by the SEI) from background characteris- tics and intervening characteristics for elementary school principals in the middle United States be repli- cated by a cross-validation sample? Do the patterns of intergenerational occupational mobility (as measured by the SEI) differ for male and female elementary school principals in the middle United States from background and intervening charac- teristics? What is the process of intergenera- tional occupational mobility for elementary school principals in the middle United States? Does the process of intergenerational occupational mobility differ for male and female elementary school princi- pals in the middle United States? Basic Assumptions This research was based on the assumptions that: l. Intergenerational occupational mobility studies of the American labor force are limited by; a) attentiveness to major occupational categories, and b) the exclusion of women, both as subjects and as mother's of subjects 2. Identification of population subgroups and analysis of their background characteristics and intervening charac- teristics would indicate patterns and processes of intergenerational occupational mobility 9 3. Intergenerational occupational mobility is only partly due to background characteristics and intervening characteristics. For example, pressure to leave origin status and availability of positions in any destination status are factors influencing mobility Definitions of Terms The following definitions were used for this study: Background characteristics - A group of origin factors, measured by occupational origins (father's SEI), father's attained level of education, mother's attained level of education, parent(s) respondent resided with at age of 16, number of siblings and sibling placement, respondent's sex, and respond- ent's age. Intergenerational occupational mobility - Movement from background characteristics to the offsprings' own career destination. Intervening characteristics - Include the following cha- racteristics of respondents in the study: marital status, presence of children, number of children, age of youngest child, number of years a teacher, highest earned college degree at first principal- ship, size community of employment. EMTM 10 Middle United States - Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, Wisconsin.12 Occupational Origins - Socioeconomic status of respondent's father as measured by Duncan's15 Socioeconomic Index (SEI) when the respondent was 16 years of age. Parents who are not employed for wages do not, according to the SEI, carry socioeconomic status. Patterns of intergenerational occupational mobility - The flow in the relationship between occupational origins and occupational outcomes in terms of status. Patterns may be described in terms of distance and direction. Processes of intergenerational occupational mobility - The paths individuals follow to the elementary school principalship. Process may be described in explanatory terms, i.e., how and why. Delimitations Two major delimitations for the study were iden- tified and are as follows: 12NEA Research Division, Elementary School Princi- palship in 1968 (Washington, D.C.: Department of Elementa- ry School Principals, National Education Association, 1968), p. 7. 13Robert M. Hauser and David L. Featherman, The Process of Stratification: Trends and Analysis (New York: AEademic Press, 1977), pp. 320-329. 11 Data were collected from elementary school principals in the middle United States who were members of the National Association of Elementary School Principals during the 1976-77 school year. Occupational status was measured by the Socioeconomic Index set forth by Duncan. Limitations The following limitations were recognized for the study: 1. Intergenerational occupational mobility is only one component of social mobility Occupational status is only one aspect of occupational attainment The socioeconomic status scores obtained can not be directly compared with scores obtained through other scales, e.g., prestige There was no attempt to exhaust the list of variables relevant to background and intergenerational occupa- tional mobility, i.e., race, religion No attempt was made to collect data from elementary school principals who were nonmembers of the National Association of Elementary School Principals, nor was an attempt made to solicit information from junior high or senior high school principals, or other groups who may have been members of the Association 0 (r an me Sm Te: 12 Summary The researcher's purpose was to investigate through observation and description, the patterns and processes of intergenerational occupational mobility of elementary school principals in the middle United States. The data were collected from elementary school principals in the middle United States who were members of the National Association of Elementary School Principals during the 1976—77 school year. Chapter 2 will be a review of the literature on mea- suring the status and prestige of occupations, intergener- ational occupational mobility and traditional methods of determining the patterns and processes of, through which the basic concepts and premises of the study were esta— blished. The variables pertinent to the topic were identified in the literature and a recent computer innova- tion was presented for its possible utility to the study. Chapter 3 is a report of the sampling procedures of a description of the statistical methods employed for the study. Detailed information was supplied to explain the development of the research instrument, and the coding of the dependent variable. Chapter 4 will be the presentation of the data and analyses employed, findings of the study, and the develop- ment of a proposed model for future study. Chapter 5 is a summary of the study which includes the conclusions and recommendations. Chapter 2 REVIEW OF RELATED LITERATURE The study of social inequality, i.e., the unequal distribution of goods and services, rights and obligations, power and prestige is generally referred to as social 1 Social stratification is the set of rules stratification. and processes by which individuals of a given population attain incumbency in the roles and positions of the hier- archically superposed classes we call occupational 2 categories. Competition for status, and demand for par— ticipation in certain types of occupations create and per— petuate a stratified occupational structure.3 The occupational structure in the United States is the foundation of the stratification system. Blau and Duncan4 state that: 1James Littlejohn, Social Stratification (London: George Allen 8 Unwin Ltd., 1972), p. 9. 2Pitirim Sorokin, Social Mobility (New York: Harper 8 Brothers, 1927), p. 11. Judah Matras, Social InequalityJ Stratification, and Mobility (Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1975), p. 12. . 3Burton Wright II, John Weiss, and Charles M. Unkovic, Perspective: An Introduction to Sociology (Hillsdale, IIlinois: The Dryden Press, 1975), p. 19. 4Peter M. Blau and Otis Dudley Duncan, The American Occupational Structure (New York: John W11ey 6 Sons, Inc., 1967), p. 7. 13 14 The occupational structure . . . not only constitutes an important foundation for the main dimensions of social stratification but also serves as the connecting link between dif- ferent institutions and spheres of social life, and therein lies its great significance. Stratification systems are present in human society because humans differ in genetic endowment and because without a hierarchical arrangement in society as a whole and in organizations within each society, the accomplishment of societal goals would be impossible.5 Sorokin6 recognized three basic forms of social stratification: economic, political, and occupational. The study of economic and political stratification pre- supposes a thorough knowledge of the occupational hier~ archy - the connecting link between different institutions and spheres of social life.7 The primary thrust of this review of selected literature will be to view patterns and processes of intergenerational occupational mobility of the holders of a single occupation, i.e., the elementary school principal- ship. The intragenerational interaction necessary to attain an elementary school principalship will not be considered. First, the United States occupational structure will be defined within the concept of measuring the status 5Wright, et.al., op.cit., p. 172. 6Sorokin, op.cit., p. 12. 7Blau and Duncan, op.cit., p. 7. 15 and prestige of occupations. Patterns and processes of intergenerational occupational mobility of the labor force will be reviewed paying special attention to findings specific to the elementary school principalship when available. Finally some descriptive characteristics of elementary school principals will be examined. Measuring the Status and Prestige of Occupations Traditionally the study of occupational mobility focuses on the status or prestige accorded individuals in the various strata of the occupational hierarchy. Researchers have sought to rank occupations according to some scale, to assess the interaction between occupational groups, and to determine the factors which stratify the occupational structure and operationalize the interaction.8 The standard approach has been to rank occupations in one of two ways: 1) by a socioeconomic status scale, or 2) a prestige scale.g When ranking occupations, a 8Donald J. Treiman, "Problems of Concept and Measurement in the Comparative Study of Occupational Mobility," Social Science Research, Vol. 4, (1975), p. 185. 9Donald J. Treiman and Kermit Terrell, "Sex and the Process of Status Attainment: A Comparison of Working Women and Men", American Sociologigal Review, Vol. 40, (April, 1975), pp. 175-176. [C/JO 16 socioeconomic status scale represents a composite index of education and income for each occupation rated.10 Both measures of socioeconomic status (income and educa- tion) are aspects of occupational status since education is a basis for entry into many occupations, and for most people income is derived from occupation.11 Prestige ratings represent a more subjective evaluation of the relative social standing of occupations.12 The first attempt at rating the prestige of occupations was by George S. Counts in 1925.13 Respondents were asked to rank-order their perceptions of the prestige of a list of 45 occupational titles. Counts instructed raters to place the number one behind the occupation which was most ”looked up to", the number two behind their second choice in that respect, and so on until they reached the number 45, i.e., the occupation with the lowest perceived social standing.14 A number of studies were conducted in the 1930's and 1940's - all similar in method to those of Counts. All the early studies have been 10Blau and Duncan, 0p.cit., p. 118. 11Albert J. Reiss, Jr., Occupations and Social Status (New York: The Free Press of Glencoe, Inc., 1961), 84. p. 12Blau and Duncan, op.cit., p. 119. 13Reiss, op.cit., p. 2. 14George s. Counts, "The Social Status of Occupations: A Problem in Vocational Guidance", The Social Review, Vol. 33, (January, 1927), p. 17. H’t—l l7 criticized for questionable rating techniques, limited numbers of occupational titles, and representativeness of the occupations rated.15 More SOphisticated methods were employed when in 1947 North and Hatt conducted the first national study of the prestige of occupations.16 A scale was developed by asking respondents to rate their opinion of the general social standing of 88 occupational titles according to the following statements:17 Excellent standing Good standing Average standing Somewhat below average standing Poor standing WAUJNH O O O O 0 When rated the mean occupational prestige scores ranged from a low of 33 for shoe-Shiner to a high of 96 for United States Supreme Court Justice. At about the same time, A.M. Edwards ranked the United States p0pulation (using 1940 census data) according to their socioeconomic status, by first distinguishing between white collar/blue collar occupations, then according to education, income, and relative prestige.18 His grouping 15Ronald M. Pavalko, Sociology of Occupations and Professions (Itasca, Illinois: F.E. Peacock Publishers, Inc., 1971), pp. 132-133. 16Ibid., p. 133. 17Matras, op.cit., pp. 112-113. 18Matras, op.cit., p. 97. (h 18 of occupations led to the major groups used by the Bureau of the Census since 1940. Later investigations reinforced Edwards' method of ranking - the education and income of occupational incumbents have been found to correlate with the public prestige of their work and have frequently been used as indicators of status.19 Following the work of Edwards, and the North-Hatt study, Duncan devised a Socioeconomic Index (SEI) for 446 detailed occupational titles,20 by measuring the relation- ship between North-Hatt prestige ratings and the socio- economic (education and income) characteristics of occupa- tions using a prediction equation.21 Socioeconomic Index scores as a measure of occupational socioeconomic status, cannot be compared directly with prestige ratings. Instead, the scale, represented by two—digit numbers ranging from 0 to 96, is purported to be an estimate of the prestige of occupations.22 The SEI allowed for expansion of previous methods of research - especially to the process of mobility (to be discussed later in this review of literature). 19Marie R. Haug and Harold A. Widdison, "Dimen— sions of Occupational Prestige", Sociology of Work and Occupations, Vol. 2, No. 1, (February, 1975), p. 4. 20Blau and Duncan, op.cit., p. 121. 21Reiss, op.cit., p. 114. 22Blau and Duncan, op.cit., p. 120. A1 (I) 19 Although most mobility analysts today incorporate education as a research variable, Duncan utilized an average level of education attained by all individuals in an occupation in the SEI prediction equation. He reasoned that despite the apparent weighting of education, not all individuals in a given occupation have attained the same level of educa- tion.23 The 1947 North-Hatt study was replicated in 1963 by Hodge, Siegel, and Rossi in an effort to explore sta- bility and change in occupational prestige during the 16 year period. The outgrowth of the replication was a third survey of occupational prestige launched in 1964 which ultimately provided a set of scores ranging from 9.3 to 81.5 for all 1960 census occupational categories.24 Prestige ratings have been shown to be close to invariant with respect to composition and size of the sample of raters, the form of the rating scale, the inter- pretation of "general standing" by raters, and the 25 When comparing the stability of occu- passage of time. pational prestige and occupational status over time, Nam and Powers found a lag between changes in status and 23Blau and Duncan, Ibid., p. 125. 24Pau1 M. Siegel, "Prestige in the American Occupational Structure", (Unpublished doctoral dissertation, University of Chicago, 1971), Chapter II, pp. 29-30. 25Blau and Duncan, 0p.cit., p. 119. ( lid ‘6... .t 20 prestige.26 Public perceptions of the prestige of occupations apparently do not reflect changes in the income and education associated with occupations.27 Blau and Duncan28 state that: The higher order of reliability and stability evidenced by prestige ratings would command their use in problems requiring social distance scaling of the occupations pursued by a general sample of the working force, but for one fact: ratings have hitherto been available only for relatively small numbers of occupation titles. Since the development of the SEI, prestige ratings have become available for a large number of occupational titles and therefore one might assume that studies of occupational mobility would now employ prestige scores rather than socioeconomic. However, two long standing issues remain unresolved in the study of status attainment: 1) do prestige scales and socioeconomic scales actually measure what they purport to measure, and more important, 2) are the two distinctly different. Featherman and Hauser29 in a recent study concluded that occupational stratification in the United States is based primarily on socioeconomic status. Occupational 26Pavalko, op.cit., p. 140. 27Ibid. 28Blau and Duncan, op.cit., p. 119. 29David L. Featherman and Robert M. Hauser, "Prestige or Socioeconomic Scales in the Study of Occupational Achievement?", Sociological Methods and Research Vol. 4, No. 4, (May, 1976), p. 419. S | ti all 21 prestige represents many salient occupational characteris- tics30 while socioeconomic status is an objective measure of education and income which takes the "general social standing" of an occupation into account.31 The SEI is seen by Featherman and Hauser32 not as an estimate of occupa- tional prestige - rather "...that prestige is an error- prone proxy for socioeconomic status . . . ." As evidence, a Featherman and Hauser comparison of SEI and prestige distributions for the major occupational categories in the United States indicated a wider range of SEI scores than prestige scores. When the scores were normalized to a common percentile metric, sample variances were more similar than in their raw form, and differences in the scales become even larger. Treiman33 maintains that prestige mobility is one thing, and mobility among occupa- tions classified according to education and income is another. He states further that:34 We will not . . . make much progress in our understanding of the means by which advantage is transmitted from one generation to the next if we continue to construct status scales which combine, . . . in a confused way, diverse status attributes. 301bid., p. 404. 31Blau and Duncan, op.cit., pp. 124-126. 32Featherman and Hauser, op.cit., p. 406. 33Treiman, op.cit., p. 201. 34Ibid. (I) (I) 11 ar 0C an \ 22 Further, he asserts that if it is father's education and income we want to study, then that data should be collected separately rather than relying on a SEI score for that inference.35 Featherman and Hauser36 on the other hand, state that occupational prestige scores represent a ". congeries of salient dimensions or occupational characteristics," while SEI scores represent education and income. "Whatever it is that prestige scores scale . . . it is substantively different from socioeconomic status."37 Grasmick adhers to the idea that prestige scores and SEI scores are both measures of an occupational prestige 38 Structure, a structure that is (at least) two-dimensional in nature - "economic return" and "value to society."39 Although socioeconomic status and prestige scores are not interchangeable there is a close correlation between occupational prestige and occupational socioeconomic status, and either may be indexed by a quantitative score that has convenient properties for statistical analysis and model 351bid. 36Featherman and Hauser, 0p.cit., p. 404. 37Ibid., p. 405. 38Harold G. Grasmick, "The Occupational Prestige Structure: A Multi-dimensional Scaling Approach," The Sociological Quarterly, 17 (Winter, 1976), pp. 90-108. 391bid., pp. 100-102. co: pr: 5c in th mo SL p1 23 construction.40 Featherman and Hauser conclude that the primary differences in socioeconomic status and prestige scales are substantive, and the process of stratification in the United States is socioeconomic and not prestige.41 Until contradictory evidence comes in, it appears that the SEI may be of more value in the study of occupational mobility than any existing prestige scale. Even though prestige scores have remained highly reliable and stable, the use of the SEI allows for greater interpretation of the obtained results in terms of what is actually mea- sured.42 However, Featherman and Hauser suggest at the present time, that;43 One is best advised to use a scale for occupations which most accurately captures the features of occupations having force for the social process one is studying. In instances of occupational mobility . . . socioeconomic dimensions and socioeconomic scores for occupa- tions are the more central, and therefore are preferable over prestige scores. An additional concern remains in the measurement of occupational status and prestige. Both socioeconomic status and prestige ratings are questionable concerning their validity for measuring the occupational mobility of women, since in each case scores were computed on the 40Otis Dudley Duncan, David L. Featherman and Beverly Duncan, Socioeconomic Background and Achievement (New York: Seminar Press, 1972), p. 6. 41Featherman and Hauser, op.cit., p. 418. 421bid., p. 405. 43Ibid., p. 406. 24 44 characteristics of the male labor force. Treiman and 145 could not find enough information to decide Terrel definitively whether occupational status scales derived from characteristics of the male labor force are appro- priate for measuring the status of women. The evidence strongly suggests however that the prestige hierarchy and the socioeconomic hierarchy are clearly invariant with respect to sex so that the occupational mobility of men and women can be compared by means of a single occupational scale.46 It is possible among holders of a single occupa- tion, that in true fact, one sex may be perceived as having more status or prestige in that position than the opposite sex. Using elementary and secondary school teachers and administrators for purposes of illustration, the status and prestige scores (see Table 1) within each measure vary slightly when the scores are adjusted to include women. The pattern might lead one to conclude that women enjoy slightly higher (although perhaps not appreciably) prestige and status than men as teachers while, as school administrators, men enjoy more prestige than women although no more status. 44McKee J. McClendon, "The Occupational Status Attainment Processes of Males and Females," American Sociological Review, Vol. 41 (February, 1976), p. 53. 4STreiman and Terrell, op.cit., p. 176. 461bid. 25 Table l: SEI and Prestige Scores47 of Elementary and Secondary School Teachers and Administrators SEI Prestige Position men men/women* men men/women Elementary school teachers 71.2 71.4 58.9 59.2 Secondary school teachers 70.2 70.5 59.8 60.1 School administrators- elementary G secon- dary 71.7 71.7 61.7 61.6 *separate status and prestige scores where not available for women The purpose of assigning status or prestige scores to occupations is to supply a quantitative measure of intergenerational occupational mobility from occupational origins to destination. In the following section we will see how the concept of mobility evolved and some purposes for researching the tOpic. Intergenerational Occupational Mobility The traditional concept of social mobility was defined by Sorokin in 1927, when he gave an account of the 47Robert M. Hauser and David L. Featherman, The Process of Stratification: Trends and Analyses (New York: AcademicflPress, 1977), pp. 321-322. ma? £01 100' ir. bl. 26 main factors responsible for it, and classified various forms of mobility.48 Social mobility is defined as movements of individuals (intragenerational) or families (intergenerational) between social positions, whether economic, occupational, or political. Early analyses utilized local samples to measure intergenerational occupational mobility from father to son but since World War II, sociologists have put greater empha- sis on national surveys of intergenerational occupational mobility.49 The traditional thrust of mobility study is to define the amount and direction of occupational mobility from occupational origin (represented by father's occupa- tion) to son's occupation, in search of patterns of mobili- ty. Occupational origins, as measureable resources ascribed at birth and transmitted from one generation to the next, play an important role in determining the social opportunities one will experience through a lifetime. Hence, individual achievement relies to a great extent on the occupational origins of individuals.50 It is that 48Aage Bdttger S¢rensen, "Models of Social Mobility," Social Science Research, Vol. 4, (1975b), pp. 66-67. 49Harrison C. White, Chains of Opportunity: System Models of Mobility in Organizations (Cambfidge, Massachusetts: Harvard University Press, 1970), pp. 2-3. 50Blau and Duncan, op.cit., p. 207. sta Sté or be' Du: Th of fr all th to th 27 status that establishes the overall social and financial status of the family.51 A distinction is made as to whether sons of an occupational origin typically experience vertical (upward or downward change in category) or horizontal (movement between jobs at the same level) mobility. Blau and Duncan52 state that: The outflow of manpower from a given origin may disperse to supply many different career destinations or become concentrated to supply primarily a few. Correspondingly, the inflow of manpower into a given destination may be recruited from a wide base of different origins or largely from a narrow base of a few origins. The results of such study yield indicators of the amount of openness in a society, and the chances an individual from a category of origin has of experiencing movement, and how far he can hOpe to go.53 It is possible then to assess occupational mobility from father to son, determine the pattern of mobility experienced by the son, and come to some conclusions about the occupational structure in the United States. In the early years of mobility research, 51Treiman and Terrell, op.cit., p. 177. Joan Acker, "Women and Social Stratification: A Case of Intellectual Sexism," American Journal of Sociology, Vol. 78, No. 4, p. 937. 52Blau and Duncan, op.cit., p. 42. 53Peter Y. DeJong, Milton J. Brawer and Stanley S. Robin, "Patterns of Female Intergenerational Occupa- tional Mobility," American Sociological Review, Vol. 36, (December, 1971), p. 1034. 28 the nature of our society may have been simplistic enough to make the assessment of mobility direction and distance an informative pursuit. Also, statistical and mathematical procedures were not advanced enough to allow more SOphis- ticated analyses. The Sorokin model for assessing sources of indi— vidual differences in the process of mobility54 became known as the process approach. This approach (often referred to in the literature as status attainment*) to occupational mobility study is closely related but differ- ent than the traditional, in that the model decomposes the concept of occupational mobility into its major com- ponents.55 .The emphasis is on the degree to which the occupational status of a person is dependent upon that person's background characteristics and the degree to which occupational status is explained by the person's own experiences or characteristics that intervene between back- ground and destination statuses.S6 54S¢rensen, op.cit., p. 72. 55Blau and Duncan, 0p.cit., p. 195. 56William H. Sewell and Robert M. Hauser, Education Occupation and Earnin ngs (New York: Academic Press, 1975), p. 3. *The term status attainment is used most commonly in the literature in reference to the process of intergenerational occupational mobility. However, as S¢rensen explains, status attainment as a concept includes a plethora of cha- racteristics which are rarely studied concurrently by mobility analysts. st; SO ge ea HIE me 0i 29 In toto, the characteristics contributing to status attainment probably include all of the following: 1. occupational resources, e.g., education, race, and social origin 2. occupational achievement, e.g., achievement motivation, aspirations, and intelligence 3. occupational preferences, e.g., special skills and competencies 4. personal constraints, e.g., age, sex, and marital status Of the above, "occupational resources” and "per- sonal constraints" are most frequently analyzed in inter- generational occupational mobility studies,S7 unlike the early mobility model which measured only occupational move- ment from father to son. Through the study of the process of status attain— ment, Sorensen58 reports that it is possible to determine: 1) the chances an individual has for entering certain occupational levels; 2) the effect of various individual characteristics such as education; 3) mobility as a system characteristic; and 4) the study of individual variations in the distance and direction of mobility. The empiric question for such research is ". . . what if anything about socioeconomic background represents favorable or unfavorable 57Sprensen, op.cit., pp. 67-68. 581bid., p. 71. 30 conditions for achievement, and how do these conditions exercise their influence?"59 Instead of focusing on the relative importance of separate socioeconomic background factors, attention is on how the causes combine to produce the end result - an individual's occupational status.60 Until recently, the intergenerational occupational mobility of women was routinely excluded from such research efforts ". . . on the grounds that their experiences were too complicated for analysis."61 As recently as 1972, Duncan, Featherman and Duncan62 assume male and female mobility to be quite distinct so excluded women from their sample ". . . to make the investigation manageable." Rosenfeld63 described the most common reasons for exclusion as problems involved in studying women's as compared to men's occupational histories, the lack of data on women's occupational movement, and the feeling that women are only marginal workers. For these reasons also, mother's educa- tion and occupation were not included as origin statuses 59Duncan, Featherman and Duncan, op.cit., p. 4. 60Blau and Duncan, 0p.cit., p. 202. 61Treiman and Terrell, 0p.cit., p. 174. 62Duncan, Featherman and Duncan, op.cit., p. 15. 63Rachel Rosenfeld, "Women's Intergenerational Occupational Mobility," (University of Wisconsin-Madison: Center for Demography and Ecology, CDE Working Paper 75-28, 1975), p. 1. 31 in the status attainment model,64 nor mother's occupation in the mobility model. Since the benchmark study by Blau and Duncan, an increased number of women have moved into the labor force, which has caused a few to wonder if the patterns and pro- cesses of mobility are the same for men and women.65 Some are beginning to recognize this exclusion of women as a serious limitation to understanding female occupational mobility, and occupational mobility over the entire labor force.66 The concept of intergenerational occupational mobility has broadened since Sorokin defined the main factors involved in social mobility. We have seen a move away from local samples to national surveys designed to measure the pattern of intergenerational occupational mobi- lity. With the advancement of statistical methods and 64Treiman and Terrell, 0p.cit., pp. 174-200. Sewell and Hauser, op.cit., p. 5. 6SDeJong et al., op.cit., p. 1033. 66Ibid. Rosenfeld, op.cit., p. 1. McClendon, op.cit., p. 52. Ivan D. Chase, "A Comparison of Men's and Women's Intergenerational Mobility in the United States," American Sociological Review, Vol. 40 (August, 1975), p. 483. Treiman and Terrell, op.cit., p. 174. Acker, 0p.cit., p. 943. 1’11 l~—~l F! ‘1'}! (n [mm 32 sophisticated computer programs, researchers are now able to ask "why" do patterns of mobility form and ”how" do the phenomena occur. The methods used to determine the patterns and processes of intergenerational mobility vary somewhat depending upon the researcher's orientation. For example, mathematicians tend more toward stochastic models of mobility while sociologists carry on empirical research 67 The emphasis here is of course leading to causal models. on the latter although the former will be drawn from occa- sionally. Methods of Determining Pattern and Process Traditional Methods of Mobility Analysis - Assessing Patterns of Mobility: Intergenerational occupational mobility study is an inquiry into the importance of occupational origins for the purpose of measuring the distance and direction of movement between an individual's occupational origins and current occupational status.68 In the early days of mobility research, occupations were grouped into a limited number of categories, 67Raymond Boudon, Mathematical Structures of Social Mobility (San Francisco: Jossey—Bass, Inc., Puinshers, 1973), pp. 4-6. 68Blau and Duncan, op.cit., pp. 401-418. Andrea Tyree and Judith Treas, "The Occupational and Marital Mobility of Women," American Sociological Review, Vol. 39, No. 3 (June, 1974), pp. 293-302. 33 whether major occupational groups or simply white collar/ blue collar, farm, or manual/nonmanual, farm, and turnover tables were constructed.69 More recently, occupations have been ranked by some scale (either socioeconomic or prestige) and categorized by major occupational groups established by the United States Bureau of the Census.70 The categorized information is arranged in either a turnover table or a transition matrix, that is, a table displaying the cross-classification of father's and son's occupations,71 with rows representing father's occupation at some specified pointin the son's life and columns representing son's current occupation.72 A turnover table represents father/son pairs in either raw numbers or in pr0portions obtained by dividing each value in the table by the total number in the sample pOpulation. A transition matrix differs in that all elements in the table are divided by their corresponding row totals and therefore all row totals equal one. Through tables of this type, the patterns of outflow from occupationl origin to desti- nation are revealed. Turnover and transition matrices of this nature only indicate the direction of mobility from origin status. 69Treiman, op.cit., p. 185. 7OBlau and Duncan, op.cit., p. 26. 71Boudon, 0p.cit., pp. 7—9. 72Treiman, op.cit., p. 185. 34 Of equal interest to most researchers is the actual distance of mobility. Distance can be assessed by examining the relative proportion of men from the same origins who end up in a certain occupation category, that is, a ratio measuring the extent to which mobility from one generation to another surpasses or falls short of chance.73 The ratio in its simplist form can be expressed as:74 observed mobility Social distance mobility = expected mobility The ratio, sometimes termed the "index of association" or "social distance mobility ratio," is expressed by a value less than or greater than one, with a value of 1.0 indicating that observed mobility is equal to that expected on the assumption of statistical independence.75 A measure of mobility which can indicate distance and direction is the index of dissimilarity which measures how much more concentrated the destinations of individuals from a given occupational origin are than those of all persons in the sample, or what proportion from a given origin would have to change their occupation for their destination to equal that of the total pOpulation.76 73Blau and Duncan, op.cit., p. 35. 74Natalie Rogoff, Recent Trends in Occupational Mobility (Glencoe, Illinois: TheTFree Press, 1953), p. 43. 7SBlau and Duncan, op.cit., p. 35. 76Blau and Duncan, op.cit., pp. 43, 67. 35 Values are calculated by summing the percent differences of the same (I) sign. If father's occupation exerts no influence, and if the occupations of sons from a given origin are the same as the entire population, then the index value will be zero. If all individuals from a given origin are concentrated in a single occupation, the index will be close to 100.0. The methods presented here have been those most commonly used by sociologists to measure the patterns of relationship between the occupational origin and destina- tion of men in the labor force. Other procedures have been used - some being early forms of the methods reviewed here, while others have been explored for their possible theoretical value in improving upon the most common methods.77 Occupational Attainment Models - AssessingPatterns and Processes of Mobility: When analyzing patterns of mobility as a separate function, one is most concerned with relations among occupational groups within the occupational structure. Study of the process of mobility does not preclude analysis of distance and direction - it simply restricts pattern analysis to characteristics of individuals rather than to characteristics of 77One is advised to see Boudon, op.cit., for further information. 36 78 Within this framework pattern occupational groups. analysis provides a means of assessing the process of mobility, that is, the link between an individual's back- ground characteristics and occupational destination.79 When studying patterns and processes together, less rigorous and as we shall see shortly, probably more relia- ble methods are employed to study patterns of intergenera- tional occupational mobility. The procedure for studying patterns and processes together incorporates three basic components: 1) comparison of frequency distributions, 2) measures of association, and 3) tests of statistical significance.80 Frequency distributions allow one to deter- mine distance and direction of mobility for example, by scoring respondent's occupational origin, and respondent's current occupation on either a prestige or socioeconomic status scale, and substracting father's occupational status (Y) from respondent's (X).81 The observed distance (X - Y) reveals at the same time, the direction of mobility, i.e., the remainder identifies the son as upwardly mobile (a positive value), downwardly mobile (a negative value), or immobile (a remainder of zero). Using this method, 78Sdrensen (1975b), 0p.cit., p. 71. 79Featherman and Hauser, op.cit., p. xx. 80Matras, op.cit., p. 378. 81Blau and Duncan, op.cit., p. 152. 37 one can determine which groups, e.g., sex, educational, experience similar patterns of mobility. This method provides the researcher with a useful summary statement that is free of assumptions, taking into account the actual form of a distribution in a way that measures of associa- tion do not. Blau and Duncan state that:82 analysis of mobility distributions along the lines set forth here is useful in checking conclusions reached by other means and possibly in expressing those conclusions in a fashion that some readers may find more 1nterest1ng. The process of attaining occupational positions and factors that influence patterns of occupational mobility are analyzed for their relationship to background charac- teristics.83 Sorensen84 states that the most recent innovation in mobility research for determining the process of intergenerational occupational mobility is the use of regression to create causal models. This technique, path analysis, originated by Wright but adopted for use in mobility study by Duncan,85 was used extensively by Blau and Duncan to describe and measure occupational attainment. 82Ibid., p. 153. 83Blau and Duncan, Ibid., pp. 115-117. 84Sdrensen, op.cit., p. 72. 85Otis Dudley Duncan, "Path Analysis: Sociologi- cal Examples," The American Journal of Sociology, Vol. 72, No. 1, (July, 1966), p. 2. 38 The model is a recursive sequence of regressive equations formulated to interpret the process of mobility as opposed to discovering the causes of that process.86 Blau and Duncan87 assumed a causal ordering from the temporal order of the data, i.e., that father's occupation influences respondent's education, respondent's first job and current job, respondent's education effects his first job, and so on. Recent intergenerational occupational mobility/ status attainment literature does not show evidence of widespread use of path analysis, although the use of the correlation and regression as a method is extensive. The reasons for this, and some difficulties with other statis- tical methods and theoretical issues outlined here are discussed below. Difficulties in Measuripg_the Pattern and Process of Mobiligy: Traditional analysis measures patterns of mobility without decomposing the movement between father's and son's statuses into its constituent elements, thereby hindering the understanding of how vertical circulation among the statuses is facilitated or limited by events and 86Ibid., p. 1. 87Blau and Duncan, 0p.cit., pp. 168-171. 39 conditions in one's past and throughout the life cycle.88 It is difficult to interpret the data and identify patterns that may exist, but more important, pattern analysis ignores the process of mobility. Matrix data describes the proportion or number of people who were in an occupation at the time of data collection, among those who were in that occupation at a previous time. The matrix actually indicates the condi- tional probability of going from one state to the next.89 An apparent difficulty is that different rates of fertility among occupational groups will lead to an over-representa- tion of fathers of the more fertile social categories.90 The numerical values reported in mobility tables are not comparable from one study to the next since values vary according to the number of occupational groups used for values presented in a matrix depends on the marginals, and marginal distributions differ from pOpulation to 88David L. Featherman and Robert M. Hauser, "Design for a Replicate Study of Social Mobility in the United States," in Social Indicator Models, ed. Kenneth C. Land and Seymour Spilerman, (New York: Russell Sage Foundation, 1975), p. 222. 89Boudon, op.cit., p. 41. 90Ibid., p. 10. 40 population.91 Tyree and Treas92 adjusted male and female matrices to identical marginal totals to allow for compara- bility, but even with this procedure it was difficult to interpret what differences in cell entries actually meant.93 In fact, reanalysis shows that Tyree and Treas overestimated differences because of interpretation diffi- culties.94 Further attempts at improving the interpreta- bility of matrices, for example Duncan's method of "simultaneous adjustment," have met with little success.95 Probably the greatest value of pattern analysis is intercohort comparison of data within a given sample, and to establish a framework for further analysis. Boudon96 suggests that we consider mobility matrices as containing valid information on mobility but to interpret cautiously measures of association with the tables, or to use other methods. Analysis of patterns does not reveal the causes or consequences of differences in the distance and direction 91Natalie Rogoff Ramspy, "Patterns of Female Inter- generational Occupational Mobility: A Comment," American Sociolo ical Review, Vol. 38, No. 6, (December, 1 pp. 806- 7. , 92Tyree and Treas, op.cit., p. 295. 93Chase, op.cit., p. 485. 94Hauser and Featherman, op.cit., p. 193. 95S¢rensen, 0p.cit., p. 81 96Boudon, op.cit., p. 11. 41 of mobility, nor tell us how advantage or disadvantage is transmitted from one generation to the next.97 Boudon98 states that: Empirical research in the field of mobility has been overwhelmingly oriented towards a description rather than an explanation of the mobility processes. . . . the most interesting problem . . . is to know how and why they [people] are different rather than to know to what extent they are different. The attainment model focuses on the degree to which the status of the son depends upon the statuses of the father, and on variables that intervene between origin and destination to explain the paternal effect on off- springs' achievement.99 The most common procedures for determining the process of intergenerational occupational mobility, that is, the relationships and the effects of those relationships among the variables, are the correla- tion and regression techniques. The classical scientific research design calls for measurement of a characteristic of interest (dependent variable) on similar subjects, the manipulation of charac- teristics (independent variables) on one group of subjects, and remeasurement of the dependent variable on both groups. The differences between manipulated and nonmanipulated subjects may allow one to predict the causal effects of 97Blau and Duncan, op.cit., p. 152. 98Boudon, 0p.cit., p. 140. 99Featherman and Hauser (1975), 0p.cit., p. 222. 42 the independent variables. It cannot be assumed that a variable which is found to be a predictor of a phenomena is a causer, although the idea of causation implies that it is possible to predict an outcome.100 Nor is it possible to determine the causal order of variables, i.e., whether X causes Y or Y causes X, from this method of analysis.101 Often, variables other than those under consideration are responsible for the observed association.102 McNeil, Kelly and McNei1103 state that: .only a tight logical analysis can tease out the causative variables. Manipulation of the prOposed causative variables is a necessary step in determination of causality. In the social sciences, and in particular, in mobility research it is difficult to identify manipulatable charac- teristics of individuals. Human behavior is so complex that the effect on one variable may interact with another variable.104 Sonquist reports that interaction appears with such regularity in sociological research that 10”Keith A. McNeil, Francis J. Kelly and Judy T. McNeil, TestingResearch Hypotheses UsingMultiple Linear Re ression (Carbondale, Illinois: Southern Illinois n1vers1ty Press, 1975), p. 453. 101Gene V. Glass and Julian C. Stanley, Statistical Methods in Education and Psycholo (Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 19 , p. 121. 1021bid. 103McNeil, Kelly and McNeil, 0p.cit., p. 315. 104McNeil, Kelly and McNeil, Ibid., pp. 9-10. 43 nonlinearity and interaction of the data seems to be the rule, while additivity and linearity seem to be the exception.105 Regression analysis assumes that the underlying relationships among the variables are linear and additive in the absence of interaction. This implies that each bivari- ate relationship between the dependent variables and the independent variables is linear and that the combined ef- fects of the independent variables are additive.106 In the social sciences we cannot make these assumptions. For exam— ple, it seems unlikely that the relationship between marital status and number of years in the labor force would be the same for men and women. Certain factors may interact with gender to produce varied results. In such cases, the usual multiple regression equation would yield inaccurate esti- mates of the dependent variable.107 Before investigating the main effects of the variables, one should test for inter- action. McNeil, Kelly and McNeil108 state however that: The discovery of interaction should not be treated as a negative finding, but as an important finding in and of itself. 105John A. Sonquist, Multivariate Model Building: The Validation of a Search Strategy (Ann Arbor, Michigan: Institute for Social Research, 1970), p. 30. 106Norman H. Nie and others, Statistical Package for the Social ScienCes (New York: McGraw-Hill Book Company, 1970), p. 368. 107Nie et al., Ibid., p. 373. 108McNeil, Kelly and McNeil, op.cit., p. 391. 44 To avoid merely describing patterns of mobility and deter- mining statistical differences between groups, and to enable consideration of several mobility determinants simultaneously, Blau and Duncan109 assumed linearity and additivity in order to use regression and construct a causal model of mobility. Blau and Duncan admit the possibility of interaction effects but they are not sensi- tive to them ". . . on the supposition that interactions could be neglected when they were not explicit in the formulation of the classificatory [ordinal] variables them- selves."110 To the issue of statistical violation Blau and Duncan111 state that: With some techniques we clearly go well beyond the point where the requisite assumptions can be at all rigorously justified. This venture, however, will--to the extent possible--be counter- poised by alternative treatments of the same data, avoiding at least some of the questionable assump- tions. Causal analysis is probably the easiest way to introduce a sufficient number of intervening variables to explain the mobility process.112 However, at best we see a model based upon the somewhat "idealized assumption" of temporal order 109Blau and Duncan, op.cit., pP~ 116. 143- 11°Ibid., p. 132. 1111bid., pp. 116-117. 112Boudon, op.cit., p. 74. 45 from father's occupation and education (when son was 16 years of age), to son's education, to son's first job, to son's 1962 occupation.113 In retrospect, it was not clear to Blau and Duncan if respondent's had finished their attained level of education prior to first job, or if education had intervened between first job and 1962 occupa- tion.114 To this degree the temporal order of the model confuses even the developers of the model. At best, what we probably see is a quasi-causal model based on what may be felonious assumptions. Despite the statistical violations, Duncan stated that regression analysis is a straightforward and effective method of measuring the dependence of son's status upon his level of origin,115 and that path analysis makes the rationale for a set of regressions explicit.116 However, instead of demonstrating causality through a path diagram, a researcher may create a spurious model that demonstrates his own preconceived ideas rather than an actual 113Blau and Duncan, 0p.cit., pp. 166-168. 114Ibid., p. 166. 115Otis Dudley Duncan, "Methodological Issues in the Analysis of Social Mobility," in Social Structure and Mobility in Economic Development, ed. Neil J1 Smelser and SEymour Martin Lipset, (Chicago: Aldine Publishing Company, 1966), p. 96. 116Ibid., p. 7. 46 representation of reality. Nie et a1.117 state that: Path . . . is a method for tracing out the implications of a set of causal assumptions which the researcher is willing to impose upon a system of relationships. . . the incorporation of ambiguous assumptions in a model leads to ambigui- ties in interpretation of results. When a researcher decides to demonstrate causality in the variables, he needs to clearly establish causal relation- ships a priori. Since the benchmark study by Blau and Duncan, mobility researchers have assumed a causal effect of background factors on occupational attainment based on the temporal ordering of these factors.118 However, serious objection to these methods have been raised. For example, 119 states: Boudon When dealing with intergenerational mobility . . perhaps the crucial problem here is to develOp, so to speak, a systems analysis approach, i.e., to construct a formal theory including the intervening variables, the interaction of which is essential to the explanation of the mobility processes . . . . Up until now, studies in social mobility have been confronted with a difficult dilemma. Either the models include a sufficient number of intervening variables, but use general statistical instruments, the syntax of which is necessarily poor (for instance, the syntax of causal analysis where the only possible type of statement has the form: the variable X has an influence on the variable Y), or they use more SOphisticated mathematical models but exclude a number of intervening variables which are essential for the explanation of the mobility processes. 117Nie, et al., op.cit., p. 387. 118Matras, op.cit., p. 386. 119Boudon, op.cit., pp. 74-75. 47 According to Blalock120 social scientists should integrate nonadditive and/or nonlinear models with causal models which, thus far, have been confined to linear additive models. In the case of the elementary school principal we do not yet know what factors are causers or even predictors of the pattern and process of intergenerational occupational mobility among holders of that position. Considering the nature of the questions posed for this study and the violation of statistical assumptions required to use the standard measures in mobility analysis, alternative methods were sought for this study. Mobility Variables Since regression was introduced as a means of assessing the patterns and processes of intergenerational occupational mobility the dependent variable changed from the distance between father's and son's occupational levels, to the occupational level attained by the son.121 According 120H. M. Blalock, "Indirect Measurement in Social Science: Some Nonadditive Models," in Quantative Sociolo : International Perspectives on Mathematical and Statistica Madelin , ed. H.M. Blalock and others, (New York: Academic Press, Inc., 1975), pp. 377, 368. 121Sdrensen (1975b), op.cit., p. 72. 48 to Duncan and Hodge122 this was made feasible by the deve- lopment of an objective measure of mobility, that is, the Socioeconomic Index. However, the SEI lacks the properties of a true interval scale, which is a requisite for the dependent variable in the regression equation. Duncan and Hodgelz3 reason that despite its shortcomings, the SEI is a more appropriate measure of mobility than classifica- tion in the heterogeneous major occupational groups used in the past. Two limitations of the SEI were identified by Duncan and Hodge. First, is the difficulty of measuring non-farm and farm occupations on the same scale, and second, it is necessary to ignore variations in the time and places occupations were pursued. On the first point they recom- mend that: Users . . . hold these [farm] values suspect and to confine the main part of the analysis to men with non-farm origins. On the second issue we are advised to ignore "spatio- temporal" differences in occupational status. Most mobility variables are scaled on an ordinal rather than interval level. However, within each category of ordinal variables Blau and Duncan assume a linear 122Otis Dudley Duncan and Robert W. Hodge, "Educa- tion and Occupational Mobility: A Regression Analysis," The American Journal of Sociology, Vol. 68, No. 6, (May, 1963), p. 644. 123Ibid., p. 631. 49 relationship between the dependent and independent variables. They state that ". . . the assumption of linearity is usually close enough to the truth, where we require it, to make regression worthwhile."124 Duncan125 states further that: If one must perforce consider a variable for which only ordinal measurement can be claimed, what damage is done in assigning numbers to the various grades of that scale and henceforth mani- pulating those numbers as if they arose from measurements on an interval scale? In college, for example, instructors grade students on the ordinal scale, A, B, C, D, F, and the registrars assign to these grades the numbers 4, 3, 2, l, 0, respectively, in order to compute the "grade-point average." Clearly, such assignments are arbitrary. One might equally well use the numbers 16, 9, 4, 1, 0 in computing grade-point averages, unless, through convention or habituation, students and faculty come to feel that the difference between an A and a B is equal to the difference between a C and a D, and so on. Little notice is taken of causal factors in the traditional model of mobility; in fact, the only variable of interest is generally occupation. The causal model originated by Blau and Duncan includes background charac- teristics, e.g., father's occupation and education, and intervening variables, e.g., size of community during youth, presence of parents, all peculiar to a respondent.126 124Blau and Duncan, 0p.cit., p. 146. 125Otis Dudley Duncan, Introduction to Structural Equation Models (New York: Academic Press, 1975), p. 159. 126Blau and Duncan, op.cit., p. 197. 50 Of the causal model Duncan, Featherman and Duncan state that:127 . . one of the attractive features of the type of model investigated here is that it makes explicit both the direct and the indirect effects of causal variables on dependent variables and allows for the possibility that one variable may be 'dependent' with respect to its antecedents in a causal scheme but 'causal' with respect to sub- sequent variables. It should be noted that causal variables whether background characteristics or intervening variables, differ to some extent depending on the emphasis of the researcher. For example, Treiman and Terrell, credited with the first attempt at comparing the status attainment processes of men and women, added mother's education and occupation (when available), race and sex as background characteristics.128 In a separate analysis on employed women and their husbands, they included hours worked per year, number and age of children, and percent of years worked as intervening and outcome variables. McClendon's129 basic model of the status attainment processes of men and women consisted of father's occupation and education, and mother's education as origin factors in combination with number of siblings, and respondent's age to formulate the socioeconomic 127Duncan, Featherman and Duncan, op.cit., p. 23. 128Treiman and Terrell, 0p.cit., p. 181. 129McClendon, op.cit., p. 56. 51 background factors. Generally age has not been considered 130 reasoned that older 3 background status but McClendon workers would have lower levels of education than younger workers, more years in the labor force, and therefore age should be considered a background variable. An extended model for women incorporated marital status, number of chil- dren, and full time vs. part time worker as intervening factors.131 Although Rosenfeld132 studied only the intergenera- tional occupational mobility of women, and therefore did not include parental education as a variable, she found that both father's and mother's occupations are significant dimen- sions of women's occupational mobility. Rosenfeld133 suggests that women's occupational mobility cannot be studied exactly as men's and ". . . in particular, that in studying women's intergenerational occupational mobility, mother's occupation should be considered as an origin status." Rosenfeld134 states further that: . . . with respect to both men and women . when the mother works outside the home, father's occupation alone is not a sufficient indicator of social position of the family. Within any family 13OIbid. 131Ibid.. p. 61. 132Rosenfeld, op.Cit.. P- 17‘ 133Ibid., p. 2. 134Ibid., p. 3. 52 the father and mother may differ in social position as represented by occupation. Occupations of both the mother and father, then, might be needed to reflect the family's general social standing and life style and to indicate the occupation - relevant benefits provided by it to the next generation. Falk and Cosby135 studied the process of status attainment and identified mother's and father's education and occupa- tion as two of the more critical contingencies affecting the occupational choice and status attainment of women. Treiman and Terrell136 in their pioneering study of status attainment determined that it is no longer tenable to assume that the social status of married women is deter- mined by that of their husbands. Treiman and Terrell137 state that: The fact of the matter is that we do not yet know how the process of status attainment operates for women, especially in comparison with men, because there has been virtually no systematic work on the topic to date. New approaches are needed for further research on the patterns and processes of female intergenerational occupa- tional mobility.138 135William W. Falk and Arthur B. Cosby, "Women and the Status Attainment Process: A Working Paper," A paper presented at the Annual Meeting of the Rural Sociological Society (Montreal, Quebec, August, 1974), ERIC abstract ED097237. 136Treiman and Terrell, op.cit., pp. 174, 176. 137Ibid., p. 174. 138Falk and Cosby, op.cit. 53 It has been demonstrated that research variables in studies of the patterns and processes of intergenera- tional occupational mobility vary somewhat through the literature. What follows are the background characteristic variables and intervening variables identified for the present study of the patterns and processes of mobility for elementary school principals. Although the studies cited* below may not have used the precise verbage found here, the intent of the variables was the same. The variables are as follows: 1. Mother's occupational category139 2. Father's occupational category140 3. Mother's attained level of education141 139Treiman and Terrell (1975), op.cit., p. 179. Rosenfeld, 0p.cit., p. 18. 140Blau and Duncan, op.cit., p. 446. Chase, op.cit., p. 491. Betz, op.cit., p. 4. 141Treiman and Terrell (1975), op.cit., p. 179. McClendon (1976), op.cit., p. 56. *A citation under either background or intervening variables should not be taken to mean that the noted author used that variable as that category of variables. 54 4. Father's attained level of education142 5. Sex of respondent143 6. Respondent's age144 7. Parent/parents respondent resided with at age 16145 142Featherman and Hauser (1976), op.cit., p. 419. Treiman and Terrell (1975), op.cit., p. 179. Duncan, Featherman and Duncan, 0p.cit., p. 39. Blau and Duncan, 0p.cit., p. 449. McClendon (1976), 0p.cit., p. 56. 143DeJong, Brawer and Robin, 0p.cit., p. 1039. Tyree and Treas, 0p.cit., p. 297. McClendon (1976), op.cit., p. 56. Treiman and Terrell (1975), op.cit., p. 179. 144Duncan and Hodge, op.cit., p. 663. Rogoff, op.cit., p. 19. Rosenfeld, op.cit., p. 18. Duncan, Featherman and Duncan, 0p.cit., p. 17. Blau and Duncan, 0p.cit., p. 178. Betz, op.cit., p. 5. NBA Research Division, Elementary School Princi- palship in 1968 (Washington, D.C.: National Education Association, Department of Elementary School Principals', 1968), p. 10. McClendon (1976), op.cit., p. 56. 145Blau and Duncan, op.cit., p. 447. 55 8. Number of siblings and sibling placement146 Intervening variables 9. Current marital status147 10. Children (yes/no)148 11. Number of children149 146Duncan, Featherman and Duncan, 0p.cit., p. 39. Blau and Duncan, op.cit., p. 446. Christopher Jencks and others, Inequality: A Reassessment of the Effect of Family and SchEoling in America (New York: Harper 6 Row, Publishers, 1972), p. 321. McClendon (1976), op.cit., p. 56. 147Chase, op.cit., p. 491. Treiman and Terrell (1975), op.cit., p. 187. Duncan, Featherman and Duncan, op.cit., p. 13. Blau and Duncan, op.cit., p. 448. NBA 1968, op.cit., p. 12. McClendon (1976), op.cit., p. 62. 148Duncan, Featherman and Duncan, op.cit., p. 13. 149Treiman and Terrell (1975), 0p.cit., p. 187. Duncan, Featherman and Duncan, op.cit., p. 13. Blau and Duncan, op.cit., p. 382. Wendy Carolyn Wolf, Occupgtional Attainments of Married Women: Do Career Contingencies Matter? (University of Wisconsifi-Madison: TCenter fOr Demography and Ecology, CDE Working Paper 76-3, 1976), p. 27. Janet McIntosh, "Differences Between Women Teachers Who Do and Who Do Not Seek Promotion", The Journal of Educational Administration, Vol. 12, No. 2, (October, 1974), p. 34. 56 12. Age of youngest child150 13. Teacher (yes/no)151 14. Number of years a teacher152 15. Highest earned degree at first principalship 16. Size community of employ153 The background characteristics and intervening variables identified for the present study were elaborated upon in Chapter 3. By careful examination of the variables stated above, it may be possible to determine patterns and processes of intergenerational occupational mobility among elementary school principals in the middle United States. 150Treiman and Terrell (1975), op.cit., p. 195. Wolf, op.cit., p. 26. 151NEA 1968, op.cit., p. 13. Betz, op.cit., p. 4. 152NEA 1968, op.cit., p. 20. 153NEA 1968, op.cit., p. 91. 57 Some Characteristics of Elementary School Principals This review of the literature has shown that until very recently most research on intergenerational occupa- tional mobility of the United States labor force has concen- trated on the male worker, rather than studying the entire labor force or comparing men and women. In studies of the elementary school principalship conducted since 1952, there has been a tendency to compare the characteristics and capa- bilities of the male and female.154 We do not find comprehensive studies of the patterns and processes of intergenerational occupational mobility among elementary school principals however. In fact, little is known of the background characteristics and intergenerational occupa- tional mobility of public school administrators.155 One study surfaces from the literature in this respect. Gross and Trask156 conducted a national cross- section survey during the 1960-1961 school year of 189 elementary school principals in 41 large city school 154Joan D. Meskin, "The Performance of Women School Administrators - A Review of the Literature," Administra- tor's Notebook, Midwest Administration Center, The Univer- sity of ChiCago, Vol. 23, No. l, 1974, p. l. 155Neal Gross and Anne E. Trask, Sex Factor and the Management of Schools, (New York: John Wiley 8 Sons, 1976), p. 20. 1561bid., p. 12. 58 systems. Although background characteristics constituted only a small portion of their inquiry, Gross and Trask157 indicated that factors Operating early in an elementary school principals' life cycle may effect occupational per- formance, the functioning and productivity with their organization, and their orientations and responses to work. Within the past ten years, there appears to be only one published national survey of characteristics of elementary school principals.158 Although not a mobility study it was significant to the conception of the present study. It was estimated that between 45,000 and 50,000 persons in the United States held positions where they exercised the basic functions of the elementary school principalship.159 During the 1976-1977 school year approxi- mately 25,000 of those elementary school principals were members of the National Association of Elementary School Principals (NAESP), and about 6800 performed their princi- palship duties in the middle United States.160 The NBA found that in 1968, 77.6 percent of the elementary school 157Ibid., pp. 20-21. 158NEA 1968, op.cit. 159Ibid., p. 6. 160Edward Keller, Telephone communication, Deputy Executive Director, National Association of Elementary School Principals, May 25, 1976. 59 principals in their sample were men.161 By 1972-1973 the percent male elementary school principals increased to 80.4.162 For reporting purposes in this section of the literature review, gender will be used as a predictor varia- ble due to the disparity in numbers of men and women in the elementary school principalship, and because it is an independent variable in the proposed analysis. Gross and Trask164 found that a substantial prOportion of urban elementary principals had experienced upward occupational mobility but that a larger proportion of men achieved higher status through the principalship than women, i.e., the father's of female principals display a slightly higher occupational distribution than father's of urban male elementary school principals. In addition, these same men stated "upward mobility" more often than women as a reason for deciding to become a principal.164 Comparison of age cohorts reveals only two departures from the above - more women between the ages of 46 and 55 than men had father's in blue-collar jobs, and in the 25 to 45 age group more women were from farm origins. Betz165 in a study of the 162HEW, The Condition of Education (Washington D.C.: National Center fEr Education StatiEtics, Education Division, 1975), p. 173. 163Gross and Trask, op.cit., p. 25. 164Gross and Trask, 0p.cit., p. 75. 165Betz, op.cit., p. 6. 60 rate of intergenerational mobility of public school teachers during the 1960-1961 school year concluded that while white- collar origins (measured by father's occupation) were over- represented in all age groups, there were proportionately more female school teachers from white-collar origins, and more men from blue-collar origins in the public schools. Occupational inheritance was higher among women than men from mother's occupation - nearly one-half of the employed mothers of urban female principals were teachers while one-fourth of the employed mothers of urban male principals were teachers.166 In a Canadian study of female elementary school teachers, McIntosh167 found that of those women who had applied for promotion, 42.9 percent had working mothers while only 28.2 percent not applying for promotion had working mothers. Working mothers of teachers who had applied for promotion tended to be employed in semi- professional or managerial occupations (53.3%). White found that among female teachers having had a working mother was associated with a high commitment to the pro- fession.168 166Gross and Trask, op.cit., p. 27. 167McIntosh, op.cit., p. 31. 168K. White, cited in "Parental Influences on Women's Career Development," Janet Sorensen and Carol Jean Winters, p. 39, in Emer in Women: Career Analysis and Outlook, ed. Samuel H. 051pow, (Columbus, Ohio: CharIEs E. Merr111 Publishing Company, 1975). 61 The level of education completed by the fathers of urban female principals was higher than that obtained by the fathers of male principals but there were no sex differences in the level of education attained by their mothers.169 Female elementary school principals are clearly older than their male counterparts. The median age of male elementary school principals in 1968 was 43 years compared to 56 years of age for women.170 In fact, 70.9 percent of these male principals were under the age of 50, while 76.5 percent of the women were over 50 years of age. The same condition is seen for age at first principalship. Over half (58%) of all urban female elementary school principals were over 40 years of age at their first principalship while 67 percent of all urban male elementary school principals were 40 years of age or less.171 Despite these differences the median number of years total experience does not vary significantly between men and women.172 Nearly 66 percent of all elementary school princi- pals in the middle United States held the position of 169Gross and Trask, 0p.cit., p. 29. 170NBA 1968, op.cit., p. 10. 171Gross and Trask, 0p.cit., p. 51. 17ZIbid., p. 52 and NBA, p. 21. 62 elementary classroom teacher just prior to their first 173 It is not uncommon for an elementary principalship. elementary school principal to have as many as nineteen years experience as an elementary teacher prior to first principalship.174 Men in the principalship have fewer mean years in the classroom however than women. The mean years of teaching experience among women in urban principals was 175 15.9 years compared with 9.2 years for men. Nationally, women average 15 years as classroom teachers - a full ten years more than men.176 Elementary school principals tend to be a highly educated occupational group. The majority hold at least a master's degree with only slight variation with reSpect to gender, or geographical location.177 Principals in the middle United States found the highest rate of master's degrees of all four sections of the country - 84.1 percent of all elementary school principals sampled had a master's degree. In the middle United States 6.1 percent had a six year degree and 1.6 percent had a doctor's. It may be interesting to note that although a reported 70 percent of 173NEA 1968, op.cit., p. 13. 174Ibid., p. 20. 175Gross and Trask, op.cit., p. 45. 176NEA 1968, op.cit., p. 20. 177Gross and Trask, op.cit., p. 52. and NBA 1968, op.cit., p. 21. 63 all doctorates in education are granted to men,178 among elementary school principals in the United States, nearly equal numbers of men and women had that degree in 1968.179 The notion that women earn less money for equal work in the labor force has been well documented elsewhere. Featherman and Hauser180 conclude that women earn propor- tionately less for equal work and equal occupations. This phenomenon may not hold true however in the case of elemen- tary school principals. The median salary of female principals ($11,000) was slightly higher in 1968 than for male principals ($10,100). This may be due to the concen- tration of female elementary school principals in urban schools which tend to offer higher salaries than rural systems.181 A recent salary survey shows that the national mean salary of elementary school principals has risen to 178NEA 1968, op.cit., p. 24. 179Patricia Cayo Sexton, Women in Education (Bloomington, Indiana: Phi Delta Kappa Educational Foundation, 1976), p. 79. 180David L. Featherman and Robert M. Hauser, "Sexual Inequalities and Socioeconomic Achievement in the U.S., 1962-1973," American Sociological Review, Vol. 41, (June, 1976), p. 129. 181NEA 1968, op.cit., p. 129. 64 $22,132.182 For their salaries, the majority of elementary school principals work between ten and eleven months.183 The majority of the women (63%) in the urban sample were never married with only 37 percent currently or ever married.184 Men on the other hand displayed an over- whelming tendency to be married (92%) with only 5 percent never married. We do not know, especially when speaking of elemen- tary school principals, whether the variables included in current studies are causal effects or simply correlates of occupational status. It seems worth exploring this dilemma before assuming we know the causers by analyzing variables which may be predictors of the process of intergenerational occupational mobility to the elementary school principalship. In essence, what is required is an interactive model to explain similarities and/or differences in the distance and direction of intergenerational occupational mobility, via the observed processes of mobility. A technique has been identified which may allow the deve10pment of such a model. 182William L. Pharis and Edward P. Keller, "Bucks, Benefits, and Bargaining: The BIG Picture," The National Elementary Principal, Vol. 57, No. 3, (March,_1978), pJT25. 133NBA 1968, 0p.cit., p. 39. 184Gross and Trask, op.cit., p. 23. 65 The Automatic Interaction Detector The Automatic Interaction Detector (AID) is a computer program deve10ped ". . . in rebellion against the restrictive assumptions of conventional multivariate tech- niques and the cumbersome inconvenience of ransacking sets of data in other ways. . . ."185 Reichardt and Schmeikal186 report that the AID method allows a researcher to "look beneath the surface of data" in order to expose social processes; with conventional statistical methods, we observe only the end-product of those processes. The AID procedure is appropriate when the problem in data analysis ". . . is to determine which of the variables are related to the phenomenon in question (through what conditions and through what intervening processes) but may not necessarily involve the exact testing of specific hypotheses."187 The AID is a special regression method which uses the basic principles of analysis of variance188 - examining 185John A. Sonquist, Elizabeth Lauh Baker and James N. Morgan, Searching for Structure (Ann Arbor, Michi- gan: Institute for Social Research, 1970), p. vii. 186Robert Reichardt and B. Schmeikal, "Theoretical Considerations and Simulation Models Related to the Model of Sonquist and Morgan," pp. 451-465, in Blalock, op.cit., p. 465. 187Sonquist, Baker and Morgan, 0p.cit., p. 1. 188G. Bonelli, "Tree-Analysis -- The Method by Sonquist and Morgan," pp. 465-472, in Blalock, op.cit., pp. 465-466. 66 a full data set using repeated one-way ANOVA, in search of predictors that account for variance in the dependent 189 The AID algorithm performs a series of binary variable. splits by locating and partitioning the predictor which reduces the variance of the dependent variable the most - continuing to less and less stable predictors on smaller and smaller mutually exclusive subgroups.190 Each split of an initial group will produce greater homogeneity within each subgroup. At the same time a split produces two mutually heterogeneous groups. The basic question according to Sonquist, Baker and Morgan191 is as follows: . what dichotomous split on which single predictor variable will give us a maximum improve- ment in our ability to predict values of the dependent variable? Certain conditions must be applied to the data and data analysis to enable accurate interpretation of the pro- gram output. First, it is assumed that the continuous dependent variable has few if any extreme cases, although should they occur, the program has provision to handle them. Predictor variables may be a combination of independent variables and intervening variables but should be single 189Sonquist, Baker and Morgan, op.cit., pp. l-lS. 190Sonquist, Baker and Morgan, 0p.cit., pp. 2, l6. 19116id., p. 2. 67 dimension classifications, scaled as nominal and/or ordinal.192 The AID algorithm uses degrees of freedom very quickly and therefore it is necessary to use samples of 500193 to 1000.194 The amount of variance which must be explained by a split should be some prestated fraction of the original variance around the variables mean. Sonquist, Baker and Morgan195 indicate this fraction is often .006 or 0.6 per- cent, while Bonelli196 states that for a partition to supply additional explanation the fraction should be greater than one to two percent. Bonelli also specifies that the variance of any given subgroups should be greater than one to two percent of the original variance, otherwise the sub- group and its parent are fairly homogeneous. It is also advisable to set a minimum number of cases allowable within each subgroup to keep the standard error at a minimum. 197 Sonquist, Baker and Morgan suggest setting this number 19ZIbid. 193B. Bolton, Personal communication, June 18, 1963 [sic] , in "A Methodology for the Development of Empirically Based Differential Service Patterns for Clients in Rehabi- litation Facilities," Jerome R. Lorenz, (Doctor's disserta- tion, University of Wisconsin-Madison, 1973), Dissertation Abstract International, 34 (10), 5171B, (University Micro- film No. 74-3533), p. 50. 194Sonquist, Baker and Morgan, op.cit., p. 3. 195Sonquist, Baker and Morgan, op.cit., p. 16. 196Bonelli, op.cit., p. 471. 197Sonquist, Baker and Morgan, 0p.cit., p. 16. 68 at twenty-five while Bonelli198 indicates ten to twenty. One might also limit the total number of possible splits to avoid generating so many subgroups that interpretation becomes difficult.199 It is not necessary to use each of the three safeguards, yet one or more should be employed.200 When any one or a combination of the above criterion have been reached the partitioning process "automatically" ceases for that subgroup. A unique feature of the AID is that variables are not described in relation to something else, hence one has a set of subgroups whose characteristics are clearly defined by the dependent variable through simple statistics (mean, standard deviation).201 The results of the AID are displayed pictorially in a tree structure which make the variable splits (the inter- active prOperties of the independent variables) and the interpretation of processes explicit.202 The predicted value of the dependent variable for any individual is the mean of his final group. The configuration of the output 193Bonelli, op.cit., p. 471. 199Sonquist, Baker and Morgan, 0p.cit., p. 17. Bonelli, op.cit., p. 471. 200Sonquist, Baker and Morgan, op.cit., pp. 16-17. ZOlIbid., p. 2. 202Reichardt and Schmeikal, op.cit., p. 451. 69 tree can assist the analyst in data interpretation, i.e., whether the predictors are additive or interactive.203 The researcher has the flexibility to prespecify ways in which the data are handled. By rank ordering the sequence of various types of independent variables, the researcher is able to determine linearity or non-linear- ity.204 For example, Sonquist, Baker and Morgan205 state that: One can introduce a set of basic background factors, remove their influence by calculating for each individual his deviation from the average of the final group to which he belongs, reassemble the full data set and analyze these residuals using another set of predictors. Since this process assumes no interaction between stages, one may want to introduce some of the initial predictors at the second stage. Of import to any analysis is the issue of consis- tent or stable results, given similar data. Sonquist206 suggests several methods for examining the stability of an AID analysis, i.e., reviewing the explanation power of the variables and the tree structure, and the shapes of the effects of the predictor in various parts of the tree. The researcher can examine the total amount of variance explained by the tree structure; examine the amount of 203Sonquist, Baker and Morgan, op.cit., pp. 49-50. 204Ibid., p. 46. 20516id., p. 19. 206Sonquist, op.cit., pp. 87-89. 70 variation explained by each split; review the ordering of the Splits; or examine the composition of the final groups. A researcher would probably use more than one of the above techniques to compare two samples since although the order of the splits may vary for two samples, the final groups may still prove them similar. Sonquist207 also suggests exact replication of the analysis by dividing the sample in half from the onset or, the most stringent test, cross- validation. The cross-validation can be accomplished by selecting a random sample of the full sample and retaining it for later use - at which time the cross-validation sample is forced to reproduce the AID splits obtained in the study sample.208 Sonquist did not provide a method for comparing the results of the sample other than visual examination. Lorenz209 therefore proposed placing a confidence interval around the p0pu1ation mean (derived from sample means with a pooled estimate of the variance) in order to be more confident of: l) the reliability of the original AID results, and 2) to enable the researcher to predict varia- bles significant to the outcome in question. Although this writer finds no instance of this algorithms use for study of intergenerational occupational mobility, the AID has been used within the field of 207Ibid., p. 90. 208Lorenz, op.cit., pp. 70-71. 209Ibid.. pp. 71-72. 71 rehabilitation to assess patterns of client characteristics to predict client outcomes, e.g., job placement.210 Lorenz211 concludes that the AID has potential for problem finding and hypothesis generating because it enables the researcher to construct, in a systematic way, inductive models based on sample data. Summary The variables relevant to this study were identi- fied from an extensive review of sociological literature pertaining to the social origins and intergenerational occupational mobility of the United States labor force, and from the intervening and background characteristics of elementary school principals. This literature review esta- blished that few studies of the United States labor force have included gender as a stratifying variable. In addition, the typical study views broad occupational categories rather than individuals within specific roles. It was concluded that by studying a single occupation, it could be deter- mined if incumbents experience similar patterns and pro- cesses of intergenerational occupational mobility to that position. For the purposes of this study it was assumed that the background and intervening characteristics of both ZlOIbid., p. 9. 211Ibid., pp. 121-123. 72 male and female elementary school principals could be examined vis-a-vis a similar set of variables, but that methods of data analysis should be sufficiently flexible to allow for differences to surface, should they exist. An algorithm was identified that exhibits such flexibility. This, and other procedures will be expanded upon in Chapter 3. Chapter 3 METHODOLOGY The Sample and Data Collection The sample consisted of 697 elementary school principals in the middle United States who were members of the National Association of Elementary School Principals (NAESP) during the 1976-77 school year. During that school year, approximately 6800 NAESP members performed as elemen- tary school principals in the middle United States. The sample was identified from a NAESP membership list (listed alphabetically by zip code) maintained on computer. Since that computer was not programmed to select subjects by simple random sampling techniques, a modification known as systematic selection1 was employed to identify a representa- tive sample of the population. It was determined that a representative sample* of elementary school principals in 1Donald P. Warwick and Charles A. Lininger, The Sample Survey: Theoyy and Practice (New York: McGraw-Hill Book Company, 1975), pp. 101-103. *The formula was: n = Z2 452 + 22 N where: N = total population = 6800 Z = 2.58, 0<= .01 E = error term = .05 n = representative sample size 73 74 the middle United States would contain at least 606* respondents. The NAESP sells membership mailing labels in lots of lOOO--therefore, a membership list equal to or just under 1000 was requested. Following a random start, every seventh unit on the middle United States membership list was identified for the sample. The result was 977 computer printed labels. A survey instrument (see Appendix B) was mailed during June, 1977, to 977 elementary school principals who were 1976-77 members of the NAESP in the middle United States. A follow-up postcard (see Appendix C) was mailed the following September to non-respondents. Of the original sampling frame, 14 surveys were returned "address unknown" which left 963 possible respondents. Responding in some way to the survey were 829 individuals (84.85%)--of which 78 indicated they were not elementary school principals**, 40 stated that they preferred not to respond. The number who were not heard from was 148. Therefore, the adjusted *Formula obtained from Maryellen McSweeney, Class Lecture, Education 967, Advanced Research Methods in Educa- tion, Michigan State University, East Lansing, Spring, 1976. **Those members of the NAESP who were not elemen- tary school principals were retired, deceased, or had job titles such as media specialist, university professor, superintendent of schools, junior high school principal, or teacher. 75 sample size was determined to be 963 less the non-elementary school principals or N = 885. In total, 583 instruments were returned on the first mailing and 114 on the second, yielding n = 697 or 78.76 percent of the adjusted sample size; 71.34 percent of the original mail out. Instrumentation The Survey of Elementary School Principals (see Appendix B) was developed in absence of a pre-existing instrument for collecting data pertinent to the measurement of patterns and processes of intergenerational occupational mobility among elementary school principals. This section explains the development of that research instrument, and defines the variables of the present study. Development of the Instrument The research instrument was prepared by: 1) review- ing related literature to identify those variables which reportedly enable one to measure patterns and processes of intergenerational occupational mobility, 2) reviewing related literature to identify variables Specifically related to the elementary school principalship, and 3) exploring various approaches to stating survey questions. The instru- ment was designed so that respondents needed only to check Q/) the appropriate response category to each question. The first draft was reviewed by selected university faculty from the disciplines of sociology, educational administration, and educational psychology for clarity, 76 accuracy, and relevancy. After slight modification, the instrument was reviewed by a computer consultant to deter- mine if the format was conducive to efficient transcription to computer scan sheets. The result was a five page (twenty- three item) questionnaire, divided into three general areas: personal characteristics, work experience, and parental information. Only one response to each question was possible with the exception of "level/levels of past teaching experi- ence" (Item 8-B). No pre-existing indicators of reliability or validity were available since the Survey of Elementary School Princi- pals was a new instrument. However, the questions solicited only descriptive, categorical information and were patterned after questions from tested instruments: two studies of elementary school principals2 and the Occupational Changes in a Generation survey3 (part of the Bureau of Census' 1962 Current Population Survey) were used as models for question 2NEA Research Division, Elementary School Principal- ship in 1968, (Washington, D.C.: National Education Associa- tion, Department of Elementary School Principals). Neal Gross and Anne E. Trask, Sex Factor and the Manggement of Schools (New York: John Wiley 8 Sons, 1976). 3Peter M. Blau and Otis Dudley Duncan, The American Occupational Structure (New York: John Wiley 3 San, Inc. I967), pp. 445—449. 77 preparation. Therefore, the concerns of reliability and validity were not judged to be a serious issue. Duncan's Socioeconomic Index (SEI), used to measure the dependent variable, that is, distance and direction of mobility from father's occupation when the reSpondent was about 16 years of age, is a widely standardized scale.4 Tests of validity reveal correlations of approximately .75 for adult son's report of father's occupation.S An added precaution (double-coding) was taken in converting father's occupational title to scale scores and will be discussed later in this chapter. Descripter Variables The descripter variables of the study were not identified for statistical analysis but for population description. Frequency distributions are presented in Appendix D. The descripter variables of this study were: past teacher, level/levels of teaching, number of years as an elementary school principal, highest earned college degree, area of specialization (highest earned degree), number of schools under direction, total enrollment under direction, total school system enrollment, salary for 1976-77 school 4Robert M. Hauser and David L. Featherman, The Process of Stratification: Trends and Analysis (New York: Academic Press, 1977), p. 53. 51bid., p. 57. 78 year, number of months under contract, state of employment, and age at first principalship. Coding Occupation Two standard methods for scoring occupations (prestige and socioeconomic status scales) were reported in Chapter 2 of this study. It was found that: a) the pres- tige and SEI scales were computed on the characteristics of the male labor force, b) the prestige and the status hier- archies are nearly invariant with respect to sex, c) the SEI represents a composite index of education and income, taking prestige into account, d) the prestige scale measures general social standing, e) there is a close correlation between occupational prestige and occupational socioeconomic status, f) both prestige and status can be indexed by a quan- titative score, g) the SEI offers a wider range of scores than the prestige scale, and h) the process of stratification in the United States is socioeconomic and not prestige. In view of the above, the following propositions were offered: Proposition 1: If the prestige hierarchy and the socio- economic status hierarchy are nearly invariant with respect to sex, and Proposition 2: If the Socioeconomic Index takes the prestige of an occupation into account, and Proposition 3: If the Socioeconomic Index offers a wider range of scores than the prestige scale, 79 Then it would follow that: Similarities and differences in the distance of intergenerational occupational mobility among elementary school principals may be more easily detected using the SEI than a prestige scale. It was determined therefore to measure the occupational status rather than the occupational prestige of occupations. Occupational status was measured by the SEI deve- loped by Otis Duncan and updated by Hauser and Featherman6 to the 1970 Census occupational codes (see Appendix A). Since the statistical algorithm used in this study does not handle decimal places in the dependent variable efficient1y7, SEI scores were rounded to the nearest whole number for ease in scoring and interpretation. Occupations were double-coded - a practice underscored by Treiman8 in order to minimize coding error and arbitrary scoring judgements, that is, scored by two coders working in isolation. The two sets of scores were then compared and reconciled when disagreement was evident. It was therefore assumed that coding reliability of father's SEI score was maximized. 6Hauser and Featherman, op.cit., pp. 320-329. 7John A. Sonquist, Elizabeth Lauh Baker, and James N. Morgan, Searchin for Structure (Ann Arbor, Michigan: Survey Research Center, 19 4), p. 55. 8Donald J. Treiman, "Problems of Concept and Measure- ment in the Comparative Study of Occupational Mobility", Social Science Research, Vol. 4, (1975), p. 197. 80 If a respondent did not specify father's occupa- tion title on the survey instrument but did provide the occupational category, the median score for occupation category as specified on the instrument was assigned. The median scores were as follows: Occupational Category Median Score a) professional or scientific 71 b) managerial or executive 56 c) clerical or sales 50 d) skilled craftsman or foreman 33 e) unskilled worker 11 In the case where a parent was not employed for wages when the respondent was 16 years of age, the parent was coded "00"; deceased parents were coded "99", and part- time workers were coded "98". In the case of small business ownership the parent was coded "62 - managers - administra- tors, not elsewhere classified" rather than as a worker in a specified business since it was assumed that ownership would confer more status than merely working at the place of business. When a specified occupation could not be located in the "Occupational Classification System" (see Appendix A), the Dictionary of Occupational Titles9 was 9United States Department of Labor, Definition of Titles, Vol. 1 of Dictionayy of Occupational Titles, lWashington, D.C.: Government Printing Office, 1965). 81 used to gain enough information on that occupation to identify an appropriate title. An example of the coding process was provided as a preface to Appendix A. Independent Variables The independent variables were identified in the literature review section of this study. These variables were categorized as background characteristics and inter- vening characteristics. The statistical procedure identi- fied for the study necessitates strict adherence to a set of criteria (discussed more fully under Design and Statis- tical Procedures of this chapter). In accordance with those criteria, frequency distributions for each original independent variable were calculated (see Appendixes E and F) for recoding purposes. The criteria for recoding the independent variables were as follows: 1) a functional yet limited number of variable categories (usually three to five but rarely more than seven are acceptable),10 2) 20 percent or more of the sample represented in one classification of any given variablell, or a minimum of 10John A. Sonquist, Multivariate Model Building: The Validation of a Search Strategy (Ann Arbor, Michigan: Ifistitute f6? Social Research, 1970), p. 192. 111bid., p. 204. 82 approximately 50 cases per classification or morelz, and 3) an awareness of correlations between study variables.13 It was recommended by Rosenfeld, and Falk and Cosby that mother's occupation be included in the measure- ment of intergenerational occupational mobility.14 It was not possible to use mother's occupation as a dependent variable with accuracy, however, for three reasons: 1) an acceptable method for combining the effects of father's and mother's occupations has not been identified, 2) the frequency distribution of mother's occupation as measured by the SEI (see Appendix B; Table 34) was trimodal and in violation of AID3 criteria for the dependent variable,15 and 3) there were a high proportion of mothers in the 12Frank M. Andrews. James N. Morgan, and John A. Sonquist, Multiple Classification Analysis: A Report on a Computer Program for Multiple Regression Using Categorical Piedictors (Ann Arbor, Michigan: Institute for Social Research, 1967), p. 79. 138onquist, 0p.cit., p. 78. 14Rachel Rosenfeld, Women's Integgenerational Occupational Mobilipy (UniverEity of Wisconsin-Madison: Center for Demography and Ecology, CDE Working Paper 75-28, 1975), p. 2. William W. Falk and Arthur B. Cosby, "Women and the Status Attainment Process: A Working Paper", (a paper presented at the Annual Meeting of the Rural Sociological Society, Montreal, Quebec, August, 1974) ERIC abstract ED097237. lsSonquist, 0p.cit., p. 197. 83 sample classified as "homemakers." Instead, mother's occupation was incorporated as an independent variable, specifically a background characteristic which it was felt, acknowledged the possibility that mother's occupation played a role in the intergenerational occupational mobility of labor force participants. Background characteristics were identified as: father's occupational category, mother's occupational category, father's attained level of education, mother's attained level of education, respondent's age, respondent's sex, parent(s) respondent resided with at the age of 16, and number of siblings and sibling placement. The data were gathered using the Survey of Elementary School Princi- pals developed in the previous section and presented in Appendix B. The raw data were coded on computer scane sheets as shown in Table 2 under "Original Categories". For analysis purposes, variable categories were recoded by computer program as shown in Table 2. Intervening characteristics were identified as characteristics particular to each respondent assumed to occur since 16 years of age. They are as follows: current marital status, children and number of children, number of years a teacher, age of youngest child, highest earned college degree at first principalship, and size of community of employment. The raw data were coded on computer scane sheets as shown in Table 3, and recoded by computer program for analysis. 84 Table 2: Definition of Categories of Elementary School Principals' Background Characteristics Recoded Definition of variable Original Categories Categories Recoded Categories N ather's education Grade 8 or less 1 Grade 8 or less 311 Some high school 2 Some high school 239 or high school Completed high graduate school Some college, 3 Some college, 147 technical or college graduate special training or above after high school Bachelor degree Master degree Doctorate or professional degree Mother ' 5 education Grade 8 or less 1 Grade 8 or less 201 Some high school 2 Some high school 330 or high school Completed high graduate school Some college, 3 Some college, 166 technical or college graduate special training or above after high school Bachelor degree jMaster degree Doctorate or professional degree Table 2 (cont'd) 85 Recoded Definition of variable Original Categories Categories Recoded Categeries N Sex Male 1 Male 536 Female 2 Female 161 Age 25 years or under 1 35 years or under 123* 26 - 35 years 36 - 45 years 2 36 - 45 years 239 46 - 55 years 3 46 - 55 years 223 56 years or over 4 56 years or over 96* Residence at age 16 Both parents 1 With both parents 6151”"I l Mbther only 2 Not with both 82* 5 parents § Father only Neither parent Number siblings An only child 1 Only child 57* and sibling placement Oldest with l - 3 2 Oldest child 217 Oldest with 4 or more siblings *Represents less than 20% of the sample in a given classifi- cation of a variable or 50 which ever is less, and therefore a poten- tial source of analysis difficulty. **Represents more than 80% of the sample in a given classification of a variable and therefore a potential source of analysis difficulty. Table 2 (cont'd) 86 Recoded Definition of variable Original Categories Categories Recoded Categories N INumber Youngest with l - 3 3 Youngest child 170 siblings and Youngest with 4 or sibling more siblings placement cont'd ( ) Nfiddle with 2 or 3 4 hfiddle child 253 siblings Nfiddle with 4 or more siblings Father's Professional or 1 White collar 218 occupa- scientific tional category Managerial or executive Clerical or sales Skilled craftsman 2 Blue collar 319 or foreman Unskilled worker Farmer 3 Farm and 150 unemployed Unemployed 'Mother ' 5 Professional or 1 Employed 197 occupa- scientific tional category Managerial or executive Clerical or sales Skilled craftsman or fereman Unskilled worker Farmer Hememaker 2 unemployed 491 Table 3: Definition of Categories of Elementary School Principals' 87 Intervening Characteristics Recoded Definition of lvariable Original Categoires Categeries Recoded Categories N iMarital .Married - living with 1 Currently 570 status spouse married Married - separated 2 Net currently 127‘"i , from spouse married "Widowed Divorced Never married lChildren No 1 No children 132*' Yes: lNumber of l - 3 children 2 1 - 3 children 454 children Yes: 3 4 or more 104*I 4 - 6 children children Yes: 'MOre than 6 children Number 0 - 1 years 1 0 - 5 years 216 years a teacher 2 - 5 years 6 - 10 years 2 6 - 10 years 283 11 - 15 3 11 or more years 194 16 or more years *Represents less than 20% of the sample in a given classifi- cation of a variable or 50 which ever is less, and therefore a poten- tial source of analysis difficulty. Table 3 (cont'd) 88 Recoded Definition of variable Original Categories Categories Recoded Categories N 'ghest No college degree 1 Bachelor or less 133 earned degree at .Associate first iprincipal- Bachelor ship Master 2 Master 532 Education 3 Education 32* Specialist specialist or doctorate Doctorate Size Rural - non-farm 1 Rural or small 297 community town of employ Rural - farm Small town (2,500- 19,999) Small city (20,000- 2 City 248 49,999) Medium city (50,000 - 249,999) Large city (250,000- 3 Large city or 152 or more) suburb of SUburb of a large city Age of under 6 years 1 under 6 years 90* youngest child 6 - 18 years 2 6 - 18 years 247 Over 18 years 3 Over 18 years 142 *Represents less than 20% of the sample in a given classifi- cation of a variable or 50 which ever is less, and therefore a potential source of analysis difficulty. 89 Dependent Variable The dependent variable used to determine the pattern of intergenerational occupational mobility in this study was identified as distance and direction from father's occupation as measured by the SEI, to the elemen- tary school principalship. The following mathematical procedures were employed to determine the value of the dependent variable for each respondent. The distance and direction of intergenerational occupational mobility from father's occupation16 to the elementary school principal- ship was measured by the following formula: R - Y = D where; Y = father's SEI score R = respondent's SEI score = 72 D = distance and direction of mobility Values of the dependent variable were positive (denoting upward mobility from father's occupation) or negative (meaning downward mobility from father's occupation) according to the formula, depending upon direction of mobility. If, for example, the father had been employed as a fireman at the time our respondent was 16 years of age, 16Blau and Duncan, op.cit., p. 152. 90 the SEI value for that occupation would be 37.0. The called for values would be substituted in the formula above as follows: 72 - 37 = +35 Therefore, that respondent's distance of mobility would be equal to 35 SEI points, and the direction of mobility would be upward from father's occupation as indicated by the positive (+) value. The AID3 computer program does not accommodate negative numbers however, so values - using the above formula, were recoded using FORTRAN by subtracting if positive or adding if negative the value from 100. Therefore, in the above example of a fireman, the respon- ent's distance and direction of mobility would be recorded as 135. If a respondent's father worked as a dentist (SEI = 96) the procedure would be as follows: 72 - 96 = -24 100 - 24 +76 This would indicate that the respondent had experienced downward mobility equivalent to 24 SEI points. The resulting distribution would have 100 as a midpoint if there was no mobility and a range of 76 - 172. Through the literature, occupational inheritance is often defined as a son inheriting his father's occupation category.17 For purposes of this study occupational inheritance or immobility was determined on the basis of a 17Hauser and Featherman, op.cit., p. 158. 91 specific SEI score, i.e., occupational inheritance was taken as zero mobility, D = 0 or recoded score = 100. SEI scores are estimates of the prestige of an occupation, derived from a composite index of income and education. Therefore the reader is cautioned against concluding that zero difference between respondent's and father's SEI scores necessarily mean the father worked as an elementary school principal, only that the father enjoyed the same level of socioeconomic status. In fact, SEI = 72 applies also to prekindergarten and kindergarten teachers, whole- sale and retail trade buyers, high school principals, and stock and bond salesmen (see Appendix A). Design and Statistical Procedures It was established that little is known of the background and intervening characteristics leading to the elementary school principalship. It would have been pre- sumptuous then, if not impossible, to develop statistically testable hypotheses and follow the established methods of labor force analysis of intergenerational occupational mobility without making assumptions about linearity and additivity of the data. Instead a statistica1 measure was sought that would not make causal assumptions about the data, and would also look for interactions among the vari- ables rather than ignore there existence. The procedure identified - the third edition of the Automatic Interaction Detector (AID3) was used to answer the research questions stated in Chapter follows: Research Question Research Question Research Question Research Question Research Question Research Question 92 The research questions were as What is the pattern of intergener- ational occupational mobility (as measured by the SEI) for elemen- tary school principals in the middle United States from back- ground characteristics? Do the patterns of intergenerational occupational mobility (as measured by the SEI) differ for male and female elementary school principals in the middle United States from background characteristics? Will the pattern of intergenerational occupational mobility (as measured by the SEI) from background charac- teristics and intervening character- istics for elementary school prin- cipals in the middle United States be replicated by a cross-validation sample? Do the patterns of intergenerational occupational mobility ( as measured by the SEI) differ for male and female elementary school principals in the middle United States from background and intervening charac- teristics? What is the process of intergener- ational occupational mobility for elementary school principals in the middle United States? Does the process of intergenerational occupational mobility differ for male and female elementary school princi- pals in the middle United States? The use of the AID3 necessitates controlling several factors during the computer run to avoid misinterpretation of the data. In addition, one must be aware of correlations between study variables since their presence may mask the importance of some variables. The criteria for AID3 use 93 are: l) calculation of correlations between study variables,18 2) data that are not badly skewed,19 3) not more than seven categories within each independent variable, and generally only three to five,20 4) unidimensional categories within each independent variable,21 and 5) too few (less than 50) cases within a variable class.22 Violations of one or more of the criteria may result in "loss of competition", i.e., a variable not being used in the splitting process, or a large sampling error, thereby causing misinterpretation of the data.23 In addition; 6) extreme cases or bimodalities in the dependent variable, and 7) small sample size can cause spuriousness.24 Although sample size was sufficiently large (greater than 500 cases) for data sets of less than 1000 cases, controls must be placed on the search process. Finally, it was recommended 18Sonquist, Baker, and Morgan, op.cit., pp. 11-15. 19Ibid., p. 50. 20Sonquist, 0p.cit., p. 192. 21Sonquist, Baker, and Morgan, op.cit., p. 3. 22Andrews, Morgan, and Sonquist, op.cit., p. 79. 23Sonquist, op.cit., p. 78. 24Sonquist, Baker, and Morgan, op.cit., p. 3. 94 that the stability and predictive power of the AID3 analysis be examined through cross-validation of the sample.25 To avoid misinterpretation of the data the following precautions were taken: a) the correlations between study variables were computed (see Table 4), b) frequency distributions were calculated for each variable (see Appendixes E and F) to assess the balance in cell frequencies. Original independent variable categories were collapsed and redefined when necessary to attain maxi- mum balance, and to limit the number of categories within each variable. Also; c) the amount of variance explained by a binary split was prestated at .006 or .6 percent,26 d) an allowable minimum group size (n = 25) before a split could occur was specified,27 and e) 20 percent of the orig- inal sample were randomly selected for cross-validation of of the sample results. The specific procedures followed to answer each research question were as follows: Research Question 1: What is the pattern of intergener- ational occupational mobility (as measured by the SEI) for elementary school principals in the middle United States from background charac- teristics? 25Sonquist, op.cit., pp. 89-90. 26Sonquist, Baker, and Morgan, 0p.cit., p. 16. 271610. SIS 8:?th .350 6x35 umozofl ou. amen—aw: Scum vac—5." mm: «a 09.82— .Cowouuu 3.838580 how @333» couuo 85330-300 gwuawozm mo.a Ac Ho.u a «a ~cc.n a «cc was 30 9.558.. c; 8.88. . 8.8 8o .1658 m; 8.88.- ...88. 8538 .5 «:38.- «emwoa. 333:. 993ng m9 .883.- .888: 8.82.- 8.88. SE .1658: NS «ccwov~.- «cenoomf «gnome: .58.- «3.38. 0:3 9.553 :> 88: 888: 88: S8: 88: 88. sxusaeg .88. 8.8:.- 88: 88.- 88. .88. .88. E... a...“ 888 86.3: 8 28. 28. .88. .888. .88. .28.- ..88. 88. 85.3. as.» . 8 .88.- 85 .88. 88.- 88.- 88. 882. .88.- ..88. 6:6 «mouse» a: S 88.- 88.- 88. 88. 888.- 88: 8.8:. .88.- 8.88. .882. 58:6 . 8 888.- 88.- 88.- :8.- 88. 88. .88. 88.- .88. 8.28. .888. 58:6 8 2.8.. owmof vmmo. cmmof $8.- 88.. .2382. «83.- «gamma. «acumen. «gnome. 23:8. maumuw Hana-”Sm .5 828. .88. 88. 88. 8.88.- 8.8:.- .:8. .88.- .888. 8:2. 88. .888. .828. a... 8 88.- 88.- 88. 88.- 88. 88. 8.88. .88.- ...88. 8.88. 8.88. .888. .888. 8.32. x8 ~> e3 8; e; 2> N; :> e; a> 8 S 8 m> e> n> 9 3338.5 >38 :60:qu 36332.88 3 038. Research Question 2: 96 Procedures a) b) The AID3 was used to: 1. determine the pattern of intergenerational occupa- tional mobility from back- ground characteristics for elementary school principals in the sample, and 2. examine the variance explained by background characteristic variables. Frequencies of employed father's SEI scores were calculated for the total sample and presented in a histogram. Do the patterns of intergenerational occupational mobility (as measured by the SEI) differ for male and female elementary school principals in the middle United States from background characteristics? Procedures a) b) C) The AIDS was used to determine the pattern of mobility for men and women from background charac- teristics by: l. Forcing the AID3 to split first on the sex variable, and 2. Visually examining the AIDS splits and the amount of variance explained. Frequencies of employed father's SEI scores were calculated sep- arately for men and women in the sample and presented in a histo- gram, A Z-test of male and female respondent's means of father's SEI scores was calculated. 97 Research Question 3: Will the pattern of intergenerational occupational mobility (as measured by the SEI) from background character- istics for elementary school princi- pals in the middle United States be replicated by a cross-validation sample? Procedures 3) A 20 percent cross-validation sample was randomly selected from the total sample, b) The AIDS splits obtained in the remaining study sample were duplicated on the cross-validation sample, and c) The resulting end group means from the cross-validation sample were compared with those of the study sample end group means using the formulae (see p.93) developed for cross-validation analysis by Lorenz-Z8 Research Question 4: Do the patterns of intergenerational occupational mobility (as measured by the SEI) differ for male and female elementary school principals in the middle United States from background and intervening characteristics? Procedures The AID3 was used to determine the pattern of mobility for men and women from background and intervening characteristics by: a) Forcing the AIDS to split first on the sex variable, and 28Jerome R. Lorenz, "A Methodology for the Development of Empirically Based Differential Service Patterns in Rehabilitation Facilities," (Doctor's disserta- tion, University of Wisconsin-Madison, 1973), Dissertation Abstracts International, 1974, §£(10), 5171B (University Microfilms No. 7Z-3533i. The formulae were as when and where Upper limit Lower limit SSss,cv ’ d + NCV XSS td ssSS ssCV 98 follows: ).(55 + tdSEss,cv ss,cv 3555 + ssCV Nss + ch ‘ 2 l NCV = mean of the study sample end group = value of t with d degrees of freedom = sum of squares for the study sample = sum of squares for the cross- validation sample = number of subjects in study sample = number of subjects in cross- validation sample = pooled estimate of standard error 99 b) Visually examining the AID3 splits and the amount of variance explained. Research Question 5: What is the process of intergener- ational occupational mobility for elementary school principals in the middle United States? Procedures Frequencies of respondent character- istics were examined to determine process. Seventy percent response to an item category was arbitrarily set as an acceptable limit for process identification. Research Question 6: Does the process of intergenerational occupational mobility differ for male and female elementary school princi- pals in the middle United States? Procedures Frequencies of respondent character- istics were examined separately for men and women to determine male and female processes of mobility. Seven- ty percent response to an item category was arbitrarily set as an acceptable limit for process identi- fication. Summary The methodology for sample selection and data collection were set forth in this chapter. The variables employed and means of coding each were described, and the research questions were presented, and the statistical pro- cedures were outlined which included AID3 controls for data misinterpretation. Chapter 4 RESULTS AND DISCUSSION The researcher's purpose for this study was to define the patterns and processes of intergenerational occupational mobility among elementary school principals in the middle United States who were members of the National Association of Elementary School Principals during the 1976-77 school year. The procedures used in analyzing the data were delineated in the preceding chapter. In this chapter, the results of the analyses were presented and dis- cussed in the order of the research questions. Results The data were analyzed by answering six research questions. The questions and results of the analyses were as follows: Research Question 1: What is the pattern of inter- generational occupational mobility (as measured by the SEI) for elementary school principals in the middle United States from background charac- teristics? To answer the above question, the AIDS was used to determine the pattern of intergenerational occupational mobility and examine the variance in father's SEI scores explained by the background characteristics of elementary school principals in the study. The variances explained by background characteristic variables were presented in 100 101 Table 5. Father's occupational category explained the greatest amount of variance among the background variables (39.9%), followed by father's education (18.7%). The least significant variable in this respect was sex with only 0.5 percent of the variance in father's SEI scores explained by that background characteristic. Table 5. Variation in Father's SEI Scores Explained by Respondent's Background Characteristics Background Characteristic Percent Variation 1. Father's occupational category 39.9 2. Father's education 18.7 3. Mother's education 9.4 4. Mother's occupational category 2.3 5. Lived with at 16 2.2 . Age 1.4 . Siblings 1.1 . Sex 0.5 Total variation explained 47.8% The pattern of mobility from background character- istics to the elementary school principalship was presented in Figure 1; an explanation of each end group was offered (see Table 6). AID3 results indicate that father's occupa- tional category, father's education, and respondent's residence at age 16 [lived with] explained 47.8 percent of the criterion variance. By tracing the sequence of splits in the tree structure, we see that the pattern of mobility from back- ground characteristics for elementary school principals Figure 1: 102 The Pattern of Mobility from Background Characteristics for Elementary School Principals, Reducibility = .6; Minimum Group Size = 25 Lived with Xd = 51.77 -__.E2 Father's Occ Cat 3,2 x.d = 43.59 Father's Educ Xd = 39.49 2,3 EljFather's Occ Cat Rd = 33.34 . 1 2 EE] . 1 EilFather's Educ Xd = 10.83 :55! X nd= i—E = $7.01 = 24.73 = 43.24 = 34.23 = 16.09 112 103 Table 6: The Pattern of Mobility from Background Characteristics, Final Groups in Rank Order of Mobility roup Number MEan umber of Cases Mobility, Characteristics 8 134 57.01 Father was a farmer, deceased, or unemployed; at age 16 respondent lived with both parents 6 186 43.24 Father was a blue collar worker with an eighth grade education or less 7 133 34.23 Father was a blue collar worker with higher than am eighth grade education 9 26 24.73 Father was a farmer, deceased, or unemployed; at age 16 respondent did not live with both parents 10 112 16.09 Father was a white collar worker with a high school education or less 11 106 5.26 Father was a white collar worker with higher than a high school education 104 whose fathers were in the white collar occupational category differs from those whose fathers were in the blue collar, and farm, deceased father, or unemployed categories. Father's education was the only salient background variable for principals from white collar origins. If the white collar father did not attend beyond high school, the ele- mentary school principal experienced an average increase of 16.09 SEI points over the father. When the father did attend beyond high school however, the principal averaged only a 5.26 increase in SE1 points. Principals from blue collar origins were similarly effected except that when the father did not attend school beyond the eighth grade the average SEI increase was 43.24 points, while if the father attended school beyond the eighth grade, respondents gained only an average of 34.23 SEI points. Principals from farm origins or with father deceased or unemployed fathers were not similarly effected by father‘s education. In fact, beyond father's occupation- al category, residence at the age of 16 [lived with] was the only salient variable. Respondents from that origin category who lived with both parents experienced the [greatest amount of mobility among all groups (id = 57.01). When the respondent did not live with both parents, average mobility to the principalship was 24.73 SEI points. 105 An additional measure was utilized to examine the pattern of mobility from the background characteristics of elementary school principals. Frequency of employed father's scores, when respondents were 16 years of age, were calculated for elementary school principals in the sample and presented in a histogram (see Table 7). The most frequently observed SEI score was 14 (farmer), which some- what skewed the distribution negatively due to the relative- ly large number. The majority of father's SEI scores (60.1%) fell below the midpoint of the SEI range. The median of the distribution was 33. The mean father's SEI score was 38.66 with a stan- dard deviation of 25.23. When subtracted from the SEI for elementary school principals (72), mean difference (id) for the entire sample was +33.34, i.e., the average elemen- tary school principal in the study was upwardly mobile 33.34 points from father's SEI score. Few subjects experienced downward mobility from father's SEI score - in fact, in raw numbers only 30 feel into the downward mobility group. When farm and deceased fathers were omitted from the distribution, the mean father's SEI score was 42.04 (id = 29.96), with a standard deviation of 21.63. The median score for this group was 40 when farm (n = 145) and deceased (n = 29) fathers were omitted. 106 nwommouow mm “woxofigsmcs one: whoaumw my moouz Aoo.wmmwv monoom Hmm m.uo;umm 84mm 8 o. a N. E 0 mm o me be nm E m- 45 a; a. a $5 $4..“ $4 3. mm _ 2 NH 3 wwm 2 m m; 5H .23 $4 $4 on *3. 8 mum 3 $0 3 $5 in; N: 2.2 gm $.mm OH on om ow cm 00 on om om ooH OHH ONH omH ova omH com ona owH omH com OHN Number of Respondents om<.wo mpmo> 0H one: mucoccommom cos: mohoum Hmm m.Ho:umm vmongEm mo zucmscopm K «Es. 107 Research Question 2: Do the patterns of intergenerational occupational mobility (as measured by the SEI) differ for male and female elementary school principals in the middle United States from background characteristics? To determine the pattern of mobility from back- ground characteristics for men and women in the sample, the AID3 algorithm was forced to Split first on the sex variable (see Figure 2; end groups explained in Table 8). Visual examination indicates that for both men and women, the most important variable in the pattern from origin status was father's occupational category. The mean difference between respondent's and father's SEI scores were slightly greater for men (id = 34.29) than women (id = 30.17). As one would expect, sons and daughters of white collar fathers experienced very little mobility. Male elementary school principals whose fathers were employed in white collar occupations were strongly influenced only by father's education; when the father did not attend beyond high school mean mobility was 14.63 SEI points but only 5.60 when the father attended beyond high school. The pattern of mobility was more complicated for sons from blue collar, and farm, deceased father, or unemployed occupational origins than for individuals from white collar origins. The greatest amount of variance among the background variables for blue collar, and farm, deceased father, or unemployed origins was explained by son's residence at the age of 16 [lived with]. For those sons living with both parents, father's occupational category 108 Figure 2: The Pattern of Mobility for Men and Women from Background Characteristics of Elementary School Principals Father's Educ Rd = 41.76 2 Father's Occ Cat n - l Xd = 34.29 5 ather's Educ d = 9.94 _§___1Ez} id = 5.60 i] 81 xd = 33.34 Ra = 51.81 n = 42 Father's Occ Cat Xd = 40.89 ’ = 32.84 ,_2_ ,E-5Father's Occ Cat 11 g 57 Xd = 30.17 - = 19.05 1,2 n = 37 Father's Educ Xd = 13.05 3 ' _ X6 4.16 Table 8: 109 The Pattern of Mobility from Background Characteristics for Men and Women, Final Groups in Rank Order of Mobility |Group umber Number of Cases Mean Mobility Characteristics:MEN 10 96 57.51 Father was a farmer, deceased, or unemployed; son lived with both parents at age 16 12 144 45.27 Father was a blue collar worker with less than an eighth grade education when the son lived with both parents at age 16 13 92 36.26 Father was a blue collar worker with higher than an eighth grade education when the son lived with both parents at age 16 48 30.31 Father was a blue collar worker, or a farmer, deceased, or unemployed; the son did not live with both parents at age 16 16 75 14.63 Son's father was a white collar worker with no education beyond high school 17 81 5.60 Son‘s father was a white collar worker who was educated beyond high school Characteristics:WOMEN 14 42 51.81 Daughter's father was a farmer, deceased, or unemployed Table 8 (Continued) 110 roup Number ean umber of Cases Mobility Characteristics:WOMEN 15 57 32.84 Daughter's father was a blue collar worker 18 37 19.05 Daughter's father was a white collar worker with no education beyond high school 19 25 4.16 Daughter's father was a white collar worker who was educated beyond high school lll explained the largest portion of the variance among back- ground characteristics. Sons from farm, deceased father, or unemployed origins were, on the average, the most mobile among the total sample of elementary school principals (id = 57.51). Sons from blue collar origins who lived with both parents were however, split once more; fathers not exceeding an eighth grade education were in lower socio- economic positions than fathers who attended school beyond the eighth grade. When the son did not live with both parents, he experienced less mobility from father's SEI score to the elementary school principalship (id = 30.31) than sons who lived with both parents. The pattern of mobility for female elementary school principals whose father's worked in white collar occupations was very similar to their male counterparts. As with the male principal, the single important background Variable was father's education. Among women whose father's did not attend beyond high school, the mean amount of mobility was 19.05 to the elementary school principalship. Daughters whose father attended beyond high school experienced very little mobility (id = 4.16). Among women from blue collar, and farm, deceased father, or unemployed father's categories the pattern of mobility was somewhat different than for men from like origins. The two categories of women were, in fact, effected by little else than father's occupational category; mean differences between women's and father's SEI scores 112 indicate a 32.84 increase in SEI for women from blue collar origins, and a 51.81 point increase for women from farm, deceased father, or unemployed origins. Overall the variance explained by background characteristics was less for women than men; the total explained variance represented in Figure 2 was 47.8 percent. Variation explained by individual background characteristics (see Table 9) indicates that the mean variance for men and women in the sample differ by as much as 6.8 percent or as little as 0.1 percent. Table 9: Sex Variation in Father's SEI Scores Explained by Respondent's Background Characteristics Percent Variation Background Characteristics Men Women Difference 1. Father's occupational category 42.0 35.2 6.8 2. Father's education 18.4 19.7 1.3 3. Mother's education 10.1 7.0 3.1 4. Mother's occupational category 2.3 2.2 0.1 5. Age ‘ 1.4 5.5 4.1 6. Lived with at 16 1.8 4.0 2.2 7. Siblings 0.8 6.5 5.9 To add support for the patterns of mobility iden- tified, the frequencies of employed father's SEI scores were calculated separately for men and women in the sample and presented in two histograms (see Table 10 - women; Table 11 - men). 113 cm mvomwouow mam: mpoaumm my NmHn: mmw.HeuMMu mopouw Hmw m.go:umm mm om mm om) mm mm cm .3 ca MN bNI H -o..im . :w woo . 2.9m 3. , B “.1.“ a. 3 cm on ow om oo on ow om cod 0.: ONH OMH o: SH 03 o: o3 o3 com Number of Respondents mm< mo 3mm» 3 0.83 mucowconmom 325m 5:3 monoum Hmm 985mm 6983.5 mo 46:263.; OHN ”S man—mp. 114 hwommouow ON "wozofimsocs one: muogumm mO mam": nau.umujxv mmpoum Hmm m.ho;uwm .w on m Om e O mm Om N ON n O? HH w w NH *o.H. ”H NH *S.H OH .m WH.~ wm.~ ON . em on am m . «a 34 we b ca OO Oh Ow he a «O.NH OOH OHH ONH OMH ova EH o3 *o.om 08H oafi omfl cad com oHN Number of Respondents ow< we mhmo> ca who: mwcowcoammm magz_:o;3 mohoom Hmm m.po;pmm Ooxofimsm mo Aucoscohm "Ha manmh 115 The mean father's SEI score for women was 41.83 (id = 30.17) with a standard deviation of 26.05, while for men was 37.71 (id = 34.29) and a standard deviation of 24.95. Downward mobility was experienced by 23 male elementary school principals (4.29% of the male sample) and seven women (4.35% of the female sample). The Z-test, which revealed no significant differ- ence (Z = 1.32, p<.05), was used to test for a statistical difference between the male and female means of father's SEI scores. A difference between the two groups was noted however, in that 42.2 percent of the fathers of women as opposed to 63.3 percent of the fathers of men were below the midpoint of the distribution. The median were 40 and 33 respectively. Therefore it was determined appropriate to test male and female means for differences between segments of the continuum, as a post hoc procedure to Research Question 2. The Z-test of means for men (id = 62.06) and women (id = 58.95) whose father's worked in white collar occupa- tions revealed no statistically significant difference (Z = 1.26, p<.05). However, the same test of principals (im = 31.07; if = 39.16) with fathers in blue collar occupations indicated a statistically significant difference between men and women in that category (Z = 2.33, p<.05). There was no significant difference (2 = 0.02, p<.05) 116 between male (X = 20.25) and female (i = 20.19) principals from farm, deceased father, or unemployed father's origin as a category.* Research Question 3: Will the pattern of intergenerational occupational mobility (as measured by the SEI) from background and inter- vening characteristics for elementary school principals in the middle United States be replicated by a cross- validation sample? To answer Research Question 3, it was necessary to randomly select a cross-validation sample from the total sample (N = 697). The random cross-validation sample was selected by computer program, and contained 97 men and 42 women (nCV = 139) — precisely 19.94 percent of the total sample. An AID3 analysis was made on the background and intervening characteristics of the remaining 558 subjects (henceforth referred to as the study sample); the results were presented in a tree structure (see Figure 3), with end group explanations offered in Table 12. The criterion variance explained by respondent's background and inter- vening characteristics in the study sample were shown in Table 13. *Means in this category were higher than the SEI score for farmer, since deceased fathers were coded "99" to distinguish them from unemployed fathers coded "0". 117 Figure 3: The Pattern of Mobility from Background and Intervening Characteristics for Elementary School Principals (Study Sample) ' = 56.77 fid= 107 "Father's Occ Cat Xd = 45161 xa = 44.24 n = 137 Father's Educ X = 40.42 __E7jl.ived with d 3,2 Xd = 43.21 Xd - 34.80 n - 93 2 x = 26.31 —-E nd= 48 Father's Occ Cat [:1] xd = 33.11 ______E:} x = 15.59 1 2 0 d , n = 84 l 1 C3 _ather's Educ = 10.62 Xd "" 13.72 11 - 32 Mother's Educ Xd = 5.92 Xd_= 1.54 118 Table 12: The Pattern of Mbbility from Background and Intervening Characteristics of the Study Sample, Final Groups in Rank Order of Mobility Group Number Wiean er of Cases Mobili_ty Characteristics 6 106 56.77 Father was a farmer, deceased, or unemployed; respondent lived with bothH parents at age 16 8 137 44.24 Father was a blue collar worker with less than an eighth grade education; respondent lived with both parents at age 16 9 93 34.80 Father was a blue collar worker with higher than an eighth grade education; respondent lived with both parents at age 16 5 48 26.31 Father was a blue collar worker, or a farmer, deceased, or unemployed; respondent did not live with both parents at age 16 10 84 15.59 Father was a white collar worker who did not go beyond high school 12 32 13.72 Father was a white collar worker who ‘was educated beyond high school. Mother had higher than an eighth grade education but did not go beyond high school 13 57 1.54 Father was a white collar worker educated beyond high school. Mbther either went beyond high school or did not exceed eighth grade 119 The AID3 splits obtained for the study sample were replicated for the cross-validation sample by way of a card sorter and presented in Figure 4 with mean differences in principal's and father's SEI scores. A confidence interval was placed about the means using the Lorenz formulae pre- sented in Chapter 3. The end group means of the cross- validation sample were compared with end group means for the study sample (see Table 14). It was found, with 95 percent assurance, that each cross-validation sample end group mean feel within the confidence intervals about study sample end group means, i.e., the results of the study were successfully cross-validated. Research Question 4: Do the patterns of intergenerational occupational mobility (as measured by the SEI) differ for male and female elementary school principals in the middle United States from background and intervening characteristics? The AID3 was forced to split first on the sex variable to answer Research Question 4. The patterns of mobility from background and intervening characteristics for male and female elementary school principals in the sample (see Figure 5) were identical to that explained by background characteristics only (see Figure 2). The end groups were defined in Table 15. The overall variance explained by background and intervening characteristics was 47.8 percent (the same as was explained by background characteristics alone). Table 16 indicates that little variance is accounted for by the intervening variables for either men or women. The 120 Figure 4: Cross-validation Sample (20%) Father's Occ Cat Father's Educ Xd = 41.02 2 Lived'with 2 ——Ls‘1 l Jéather's Occ Cat d = 34.50 1 2 EE] 9 _1_EJFather ' s Educ X6 = 11.78 xd = 58.00 27 :5 II = 44.48 31 23><| Xd = 36.14 n = 22 X = 37.57 nd 14 X = 17.82 nd= Z8 Mother's Educ xd = 6.82 X =-1.43 nd= 7 121 Table 13: Variation in Father's SEI Scores Explained by Respondent's Background and Intervening Characteristics for the Study Sample Variable Percent 1. Sex 0.6 2. Age 0.7 3. Marital status 0.2 4. Number of children 0.3 5. Age of youngest child 0.2 6. Years of teaching 0.8 7. Degree at first principalship 1.3 8. Size city/town 0.5 9. Father's education 17.5 10. Mother's education 8.6 11. Father's occupational category 38.0 12. Mother's occupational category 2.7 D3. Lived with 2.6 14. Siblings 1.1 Total explained variation 47.8% Table 14: Cross-validation: Study Sample Mean Difference Confidence Intervals, Final Groups in Rank Order of Mobility End Study Sample Cross-val. Study Sample id Group X X Confidence Interval 6 56.77 58.00 53.40 < p > 60.14 8 44.24 44.48 37.52 < u > 50.96 9 34.80 36.14 26.79 < u > 42.81 5 26.31 37.57 6.26 < u > 46.36 10 15.59 17.82 9.62 < u > 21.56 12 13.72 -l.43 -3.80 < u > 31.24 13 1.54 4.10 -9.15 < u > 12 23 Figure 5: The Pattern of Mobility for Men and Women from Background and Intervening Characteristics to the Elementary School Principalship ather's Occ Cat = 46.31 Father's Educ Xd = 41.76 23 [:25 Father's Occ Cat 1 Xd = 34.29 ._fi_g_} 1,2 J: iather' s Educ d = 9.94 i—lfik Sex [IJXd = 33.34 14 3 [1 Father's Occ Cat 3,2 ’ XC1 = 40.89 2 15 2 Father's Occ t ‘ id = 30.17 18 1,2 1 a Father' 5 Educ ' xd = 13.05 3 9 Xd = 57.51 n=96 = 45.27 = 144 3&0 X = 36.26 nd= 92 16 = 30.31 b><| d = 14.63 4.16 :3 >4. ll 123 Table 15: The Pattern of Mobility for Men and Women from Background and Intervening Characteristics, Final Groups in Rank Order by Their Mean Differences roup Number Mean umber of Cases Mobility Characteristics:MEN 10 96 57.51 Father was a farmer, deceased, or unemployed; son lived with both parents at age 16 12 144 45.27 Father was a blue collar worker with less than an eighth grade education when the son lived with both parents at age 16 13 92 36.26 Father was a blue collar worker with higher than an eighth grade education when the son lived with both parents at age 16 7 48 30.31 Father was a blue collar worker, or a farmer, deceased, or unemployed; the son did not live with both parents at age 16 16 75 14.63 Son's father was a white collar worker with no education beyond high school 17 81 5.60 Son's father was a white collar worker who was educated beyond high school Table 15 (Continued) 124 GrOup *Number Mean Number of Cases Mobility Characteristics:WOMEN 14 42 51.81 Daughter's father was a farmer, deceased or unemployed 15 57 32.84 Daughter's father was a blue collar worker 18 37 19.05 Daughter's father was a white collar worker with no education beyond high school 19 25 4.16 Daughter's father was a white collar worker who was educated beyond high school Table 16: Variation in Father's SEI Scores Explained by Intervening Characteristics of Men and Women Percent Total Percent Intervening Characteristics Sample Men Women Difference Marital status 0.1 0.0 0.4 0.4 Number of children 0.4 0.7 4.2 3.5 Age of youngest child 0.3 0.4 4.5 4.1 Years of teaching 0.6 0.9 2.0 1.1 Degree first principalship 1.0 0.8 1.5 0.7 Size city/town 0.7 0.3 2.1 1.9 125 differences in male and female percents ranged from 4.1 percent to 0, with the greatest difference in intervening variables being shown for age of youngest child. That variable represented 4.5 percent of the variance for women while only 0.4 percent for men. Research Question 5: What is the process of intergenera- tional occupational mobility for elementary school principals in the middle United States? To identify the process of mobility for elementary school principals in the middle United States, frequency tables of respondent characteristics (presented in Appen- dixes D through F) were evaluated by the following criterion: no less than 70 percent of the sample were required to respond to an item category, or a combination of contiguous categories when appropriate. The results were presented in an assumed chronological order. Part of the process of mobility for elementary school principals in the sample appears to be having a mother who was not employed outside the home (see Appendix E; Table 34) when respondents were 16 years of age (70.45%). The sample, 76.90 percent of whom were male (see Appendix E; Table 37), tended not to be only children - in fact, 91.12 percent were raised with at least one other child (see Appendix B; Table 40), and lived with both (see Appendix B; Table 39) parents (88.24%). 1 Not surprising perhaps, 99.57 percent had experience as a teacher prior to their first principalship (see Appendix D; Table 21) although the level at which they 126 taught (see Appendix D; Table 22) was not a relevant indi- cator (according to the established criterion) of who became elementary school principals. Ordinarily a respon- dent did not teach (see Appendix F; Table 45) more than 10 years (71.60%) or exceed the age of 45 (see Appendix D; Table 27) (7.60%) at first principalship. Being married (see Appendix F; Table 41) and living with spouse (81.78%) and having (see Appendix F; Table 42) children (80.20%) were salient factors among the sample. Finally and perhaps most important in light of the analyses of mobility patterns, 80.78 percent of the respon— dents had a master's degree or higher (see Appendix F; Table 46) at the time they first assumed an elementary school principalship. In summary, the process of intergenerational occu- pational mobility for elementary school principals in the middle United States was to be male and reared by both parents.. The mother remained at home to care for more than one child. One must almost certainly have been a teacher, but for not more than 10 years, nor exceeded the age of 45 to have become an elementary school principal in the middle United States. Marriage and children were as common as having a master's degree or higher at first principalship. 127 Research Question 6: Does the process of intergenerational occupational mobility differ for male and female elementary school princi- pals in the middle United States? To answer the above question the data (presented in Appendixes D through F) were examined separately for men and women in the sample. The processes of mobility were identified by the following criterion: no less than 70 percent of the sample were required to respond to an item category, or a combination of contiguous categories. The results of these analyses were presented in an assumed chronological order. The Process of Mobility for Men: Mothers of male elementary school principals (see Appendix E; Table 34) were not employed outside the home (70.90%). Among the male sample, 88.43 percent lived with both parents (see Appendix B; Table 39) and 90.86 percent had brothers and sisters (see E; Table 40). As teachers (99.44%) for 10 or less (see Appendix F; Table 45) years (80.42%), men did not find teaching at the elementary school level (see Appendix D; Table 22) a prerequisite for an elementary school principalship. Marriage (92.16% living with spouse) and (see Appendix F; Tables 41 through 43) children (88.43%), although usually not more than three (71.46%), were very common. At first principalship, men (see Appendix D; Table 27) were likely to be 35 years of age or younger (76.68%) and possess (see Appendix F; Table 46) a master's degree or higher (82.46%). 128 In summary, according to the criterion established, age of youngest child and level taught were not part of the process of mobility for men in the sample. The process of intergenerational occupational mobility for male elementary school principals in the middle United States, i.e., how they reached the position, was as follows: 1. The mother did not work outside the home, 2. The parents lived together and had more than one child, 3. Experience as a teacher, 4. Male elementary school principals were teachers for 10 years or less, 5. Marriage and one to three children were common, 6. At first principalship, men were 35 years of age or younger, 7. Men held a master's degree or higher at first principalship. The Process of Mobility for Women: The majority (87.58%) of the women in the sample lived with both parents (see Appendix E; Table 39) and only 6.83 percent were an only child (see Appendix E; Table 40). Without exception (100.0%) the women were teachers (see Appendix D; Table 21) prior to becoming an elementary school principals, and 72.05 percent taught in an elementary school (see Appendix D; Table 22). Although marital status was not salient in this analysis for women, it may be interesting to note that nearly 52 percent were not currently married (see Appendix F; Table 41). 129 A full 75.78% percent were between the ages of 26 and 45 (see Appendix D; Table 27), and 75.15 percent had a master's degree or higher (see Appendix F; Table 46) at the time they first became elementary school principals. According to the criterion established, marital status, having children, age of youngest child, mother's occupational category, and years a teacher were not part of the process of mobility for female elementary school principals. The process of mobility for female elementary school principals in the middle United States was summarized as follows: 1. The parents lived together and had more than one child, 2. At first principalship, women were between 26 and 45 years of age, 3. Women held a master's degree or higher at first principalship, 4. A teacher at the elementary level. Review and Discussion of Significant Findings Answers to six research questions were sought through various statistical methods. It was found that the pattern of mobility among elementary school principals in the middle United States was one of upward mobility from father's occupation when respondent was 16 years of age; only 4.30 percent of the sample indicated downward mobility. Women in the total sample were slightly less mobile than men, although the overall differences in mean father's SEI scores 130 were not statistically significant. The most mobile segment of the sample was principals originating from the farm, deceased father, or unemployed father's category who lived with both parents (19.23% of the sample), and as one might expect, the least mobile group had white collar fathers who attended beyond high school - 15.21 percent of the sample. A statistical difference in mean white collar origins was not found between men and women, although a higher pr0portion of women (38.50%) than men (26.11%) found their origins in the stratum (percents derived from Appendix E; Table 33). A significant difference was detected among men and women from blue collar origins. Women in that category were from higher blue collar origins than their male counterparts. It may be interesting to note also that 35.40 percent of the women and 48.88 percent of the men were from blue collar origins (percents derived from Appendix E; Table 33). Men (23.69%) and women (31.06%) from farm, deceased father, or unemployed origins showed no difference in mean father's SEI scores (percents derived from Appendix E; Table 33). Of the eight variables identified as background characteristics only three, i.e., father's education, father's occupational category and whom respondent lived with at age 16, were important in the AID3 analyses of mobility pattern. Father's occupational category explained 39.9 percent of the criterion variance for the total sample. 131 Father's education was the only important variable for elementary school principals from white and blue collar origins, and the single salient variable for principals from farm, deceased father, or unemployed origins was whom they lived with at age 16. When the algorithm forced the sample to split first on the background variable sex, it was found that the same three variables remained salient in the AID3 analysis for men and women. However, only father's occupational cate- gory and father's education were indicated from background characteristics for women, while all three variables pre- vailed for men. Although at first review there appeared to be a difference in the pattern of male and female mobility, in all likelihood the patterns were approximately the same. Careful analysis of the AID3 output indicated that a split would have occurred for women from the blue collar, and farm, deceased or unemployed father's categories on the variable "lived with" had the group size been larger. Split of Group 14 (see Figure 5) would have produced two new groups - one with n = 38 and a second with n = 4. The algorithm was programmed for a minimum group size of 25 to prevent spurious results. It was concluded then that if there had been more women in the sample, the pattern of mobility for men and women as explained by the AID3 analyses would be approximately the same. 132 Figure 5: The Pattern of Mobility for Men and Women from Background and Intervening Characteristics to the Elementary School Principalship l'_': ._____+:4 3,2 LL % Father's Occ Cat Xd = 34.29 - —EL 1,2 1—[9 Lived with Xd = 44,29 —@ 3 16 Father's Occ Cat Xd = 46.31 hFather' s Educ id = 41.76 Father's Educ Xd = 9.94 a Father's Occ Cat X = 40.89 Xd = 57.51 n=96 L‘E’» Father's Occ Cat ‘ X = 30.17 ‘1 . 58 1,2 1 1 '9 Father's Educ d = 13.05 3 19 133 The cross-validation procedure produced an 80 percent study sample that traced patterns of mobility from background and intervening characteristics for elemen- tary school principals in order to test overall the stabil- ity of the AID3 results. The study sample was successfully cross-validated; therefore it was concluded that since the mean differences expressed in the end group of a random sample of respondents were within the limits established by the cross-validation formulae, the results of the AID3 analyses of the study sample are reasonably reliable pre- dictors of the population end group means. The results of the analysis for men and women from background and intervening variables taken together were the same as when background variables were analyzed sepa- rately. Therefore, the intervening variables identified for the study explained little if any variance in the pattern of mobility for male and female elementary school principals. As a post hoc search for variables significant to the pattern of mobility, the researcher included the descripter variables discussed earlier, in two additional AID3 analyses. The first analysis was for the background, intervening, and descripter variables of the total sample while the second analyzed male and female elementary school principals separately. Descripter variables were recoded (see Appendix G) as per the AID3 criteria. The first analysis revealed no 134 descripter characteristics in the AID3 tree structure; the addition of variables made no difference in the pattern of mobility. The total variance accounted for also remained the same - 47.8 percent. The second analysis produced no differences in male or female mobility using all three variable classifications from the results presented in Figure 2 (mobility from background characteristics alone), nor did the total criterion variance explained by the addition of descripter variables change. Table 17 was prepared to show the amount of criterion variance explained by each study variable. Using the 0.6 percent criteria of significance suggested for AID3 use, it is apparent that ten of 24 variables did not account for an acceptable level of variance for the total sample. A technique, developed by Pohlmann and Moorel, indicated no statistically significant differences between the overall variance accounted for by the gender variable, at the 95 percent level of confidence. Exploration of the frequency data revealed the following variables as having importance for the process of mobility for elementary school principals: 1. Sex 2. Whom respondent lived with at age 16 1John T. Pohlmann and James F. Moore, "Interval Estimation of the Population Squared Multiple Correlation", Multiple Linear Regtession Viewpoints, Volume 8, Number 1, pp. 18:31. 135 Table 17: Variation Explained by Background, Intervening, and Descripter Variables for Men, Women, and the Total Sample. Total Men Women Variable Sample n=536 n=161 Difference Background: Sex 0.5% --- --- --- Age 1.4 1.4% 5.5% 4.1% Father's Occ Cat 39.9 42.0 35.2 6.8 Mother's Occ Cat 2.3 2.3 2.2 0.1 Father's Education 18.7 18.4 19.7 1.3 Mother's Education 9.4 10.1 7.0 3.1 Lived With 2.2 1.8 4.0 2.2 Siblings 1.4 0.8 6.7 5.9 Intervening: Marital Status 0.1 0. 0.4 0.4 Number of Children 0.4 0. 4.2 3.5 Age of Youngest Child 0.3 0.4 4.5 4.1 Years a Teacher 0.6 0.9 2.0 1.1 Degree First Prin. 1.0 0.8 1.5 0.7 Size City/Town 0.7 0.3 2.1 1.8 Descripter: Current Education 0.0 0.0 0.0 0.0 Specialization 0.0 0.2 0.0 0.2 Level Taught 0.4 0.4 1.1 0.7 Years a Prin. 1.6 1.1 5.1 4.0 Age First Prin. 0.1 0.7 0.0 0.7 Number Schools 0.1 0.0 0.5 0.5 System Enrollment 0.0 0.2 0.9 0.7 Salary 1.6 2.3 0.9 1.4 Contract 0.6 0.6 1.1 0.5 State 1.4 1.7 0.7 1.0 Total Explained Variance 47.8% 49.8% 40.7% 9.1% 136 * 3. Mother's occupational category 4. Presence of siblings 5. Employment as a teacher * 6. Number of years as a teacher ** 7. Level taught 8. Age first principalship * 9. Marital status *10. Having children 11. Education first principalship *12. Number of children It appears that there is not just one process of mobility for elementary school principals in the middle United States. Although the processes for men and women are similar in several ways, they are dissimilar in as many others. Therefore the process of mobility should be studied separately for men and women in the future. One advantage of the AID3 is its ability to ferret out a plethora of information about the variables under study for the purposes of further research and theoretical model building - an advantage not so readily possible with some other statistica1 procedures. Two primary issues were addressed here: 1) the effect of some correlations among the variables, and 2) the identification of some interactions among the variables. Using that information, the researcher * Men only ** Women only 137 pulled together some considerations for develOping a model for further study of the patterns intergenerational occupational mobility of elementary school principals, with implications for the process of mobility. Since only background characteristics were indicated as important to the patterns of mobility from the AID3 analyses, this post hoc investigation was confined to the same (see Figure 6). Correlations among the background and intervening variables under study were presented in Table 4 of this dissertation. It was shown that no variables were more than somewhat correlated (r = f3 to t.6). Of those variables that fell within this range, all were what would ordinarily be expected, e.g., father's education was some- what correlated with father's occupational category (r = :5067), and marital status was correlated with having children (r = +.6671). The variance explained by father's education dropped as that explained by father's occupational category was used for the split which produced Groups 2 (men) and 3 (women), indicating the strength of the relation- ship (see Table 18). When Group 2 split into Groups 4 and 5, the variance explained by both variables dropped sharply. At the same juncture, the effect of whom the son lived with at age 16 (Group 4) nearly doubled. It was concluded then that for male elementary school principals from blue collar, and farm, deceased father, or unemployed origins, "lived with" interacts with father's occupational category and/or father's education. Moving a step farther, 138 Figure 6: A Proposed Model for the Study of the Mobility Among Elementary School Principals by Origin Strata .- -—----—. V..— .-o c....—- (:1\~1'\e 1mm 24 1 ’ White Collar Origin Women (FE V age] 4 ' MOBILITY (“‘1 FE \1 m > , EDBILITY ----------------~-------------------------------------------. ....... FOC l p mBILIrY age FE L we Blue Collar Origin Women m: D PDBILITY ( lived FE with 10C C .1 Farm, Deceased, and Unemployed Origin Men and Women FOG-father's occupational category, FEafather's education, 30C: mother's occupational category, .‘iEsmother's education. A straight line indicates a direct effect; a curved line indicates a correlative effect; a zig-zag indicates an interactive effect. 139 it was apparent that when Group 4 split into Groups 6 and 7, the variance explained by father's occupational category increased for Group 6 although not for Group 7. The split of Group 6 saw the variance of father's educa- tion double while the variance in father's occupational category was nil. This indicates an interactive effect between father's education, father's occupational category, and the "lived with" variable for sons from blue collar origins. By following this procedure, it was possible to conclude the following about mobility patterns among ele- mentary school principals: 1. For women from white collar origins, siblings appears* to interact with father's occupational category and father's education 2. For men from white collar origins, age inter- acts with father's education and father's occupational category 3. For men from blue collar origins, father's occupational category and father's education interact with the "lived with" variable *Note: For white collar women, split of Group 9 raised the variance of siblings (Group 15) indicating interaction, even though a split did not occur due to the small number. If the female sample size had been suffi- ciently large, Group 14 would have split on the "lived with" variable. Also, had the sample size been larger Group 15 would have split on father's education. There were also indications from GrouplS (see Table 18) that for blue collar women, age interacts with father's occupational category. All this suggests that the pattern of mobility is approximately, though not exactly the same for men and women. 140 m. o. O. m. N. - N. N. o. O. O. m. N. O.H e. m. O. mwcHHnHm m. O.H - H. - - H. - O. - H.H - - O.H e.N m.H N.N :qu Oo>HH H. H. H. O. e. e. H. O. H. N. O. m. O. m. m.H N.H m.N u uuo m.he:uoz - - m. - - H. - - - H. O.N -- O.m N.O H.N «.mN N.Om O 000 m.uo:pmm e. O. O. m. H. O. O. H. O. O. H. v. e. m.H m. v.m N.O usvm m.uo:poz H. H. O. - O. N. o. - N. e. o. O.H m. m.m m. O.HH «.mH oswm m.ho:umm H. N. H. N. m. NH m. o. o. m. m. m. e. w. m. N. a. owe. II IIIrlllilllvlllélllvlllrlllillllllllTllllllllHIII.IIl-lllrilll mo “mm OH OH OH OH mH NH mH NH m N N HH 0 m e N H dachu wouuHOopO H039 ow Oeuswem oNHm macho EOEHOHZO macho comm co :oEoz cam :02 wow moHan~m> Oczoumxumm Np wochmem mucmHhm> "OH mHnmN 141 4. For men from farm, deceased, or unemployed origins, father's occupational category interacts with the "lived with" variable. The same appears* to be true for women in the same category. 5. For women from blue collar origins, age appears* to interact with father's occupa- tional category. Mother's occupational category and mother's educa- tion were somewhat correlated (r = v.4485) with each other, and with father's occupational category and father's educa- tion (see Table 4). The variance explained by each dropped as Groups 2 and 3 were created. Neither mother's variable appeared to interact with any other variables. Variables not appearing in the AID3 (pattern) analysis were potential process variables since the variables indicated in an AID3 analysis are those which explain the largest portion of the criterion variance. Process varia- bles, i.e., those which indicated homogeneity among the subjects on a given variable, if examined with pattern *Note: For white collar women, split of Group 9 raised the variance of siblings (Group 15) indicating interaction, even though a split did not occur due to the small number. If the female sample size had been suffi- ciently large, Group 14 would have split on the "lived with" variable. Also, had the sample size been larger Group 15 would have split on father's education. There were also in- dications from Group 15 (see Table 18) that for blue collar women, age interacts with father's occupational category. All this suggests that the pattern of mobility is approxi- mately, though not exactly the same for men and women. 142 variables should also allow for theory development and model building. By studying pattern and process together, it may be possible to predict who will become an elementary school principal. Observations As a post hoc extension of the analysis of the process of mobility some additional observations were made on the study variables, by state of employment, to include those not identified as important process variables. In this section state refers to politically organized bodies with definite boundaries such as Illinois, Michigan, and Wisconsin. Although state as a variable explained only 1.40 percent of the variance in the pattern of mobility (1.70% for men and 0.70% for women), some differences were detected by state of employment. Interpretation of the data by state must be read with caution in that the sample was drawn to be representative of the middle United States rather than individual state. Although not a research question in the study, it is possible that the process of mobility differs to some extent by state of employment. Such will not be determined here; rather, some data were presented by state of employment for utility in future research. The proportion of women in the sample (23.10%) was slightly higher than the national average among elementary school principals; some variation by state was noted in the 143 proportion of men to women in the middle United States (see Table 19). The majority were employed east of the Mississi- ppi River with the largest single number being in Michigan. Table 19: Percent Men and Women From Each State in the Sample State Percent Men Percent Women N Illinois 68.75 31.25 80 Indiana 88.31 11.69 77 Iowa 75.51 24.49 49 Kansas 81.58 18.42 38 Michigan 75.33 24.67 150 Minnesota 77.55 22.45 49 Missouri 79.59 20.41 49 Nebraska 75.00 25.00 28 North Dakota 90.91 9.09 11 Ohio 72.73 27.27 88 South Dakota 72.73 27.27 11 Wisconsin 80.00 20.00 60 Total 690* *State of residence could not be determined for seven respondents Typically the respondent was principal of one elementary school (83.50%). However, when respondents directed two schools in Illinois, Missouri, and Nebraska the probability was greater that the respondent was a woman. In Michigan, Minnesota, North Dakota, and Ohio a man was more likely to direct two schools than a woman. It was not common to find an elementary school principal directing three schools except in Nebraska (17.86%). The majority of principals directed (indirectly at least) the activities of between 200 and 599 students (see Appendix D; Table 27). Schools with fewer than 200 144 students appeared to be nearly equally distributed between men and women in all states but Minnesota, Missouri, and South Dakota where the smallest schools were almost exclu- sively lead by women. At the same time, the largest enroll- ments (more than 600 students) were under the direction of men in Minnesota (26.53%) while distributed almost equally in all other states. Few differences occurred by state for respondents who directed between 200 and 599 students. However, in Illinois, Iowa, and Nebraska women were more likely to direct 200 to 399 students while more men than women were directing 400 to 599. Total school system enrollment for the majority (51.23%) was 3000 to 24,999 students. While the trend was toward an equal distribution on number of students under the direction of male and female elementary school principals in the middle United States, when viewed from the point of system enrollment the picture seemed to change. There were more women than men in large school systems 25,000 students or more - especially in Illinois (40.0%), Indiana (33.33%), Nebraska (42.86%), and Wisconsin (41.67%). In Indiana and Wisconsin there were twice as many men in systems with 3000 to 24,999 students while Kansas had twice as many women in this category. In the smaller systems of Illinois, Kansas, Nebraska, and Wisconsin (300 to 2999 students) men predominated two to one. These findings were consistent with those of size of community of employment. It was found that in Illinois, Kansas, Minnesota, and Missouri at least 14S twice as many women were employed in suburbs of large cities than men. In Wisconsin a woman was 10 times as likely to work in a large city while in Illinois and Missouri women were about twice as likely to work in a large city. In the medium cities (50,000 to 249,999) of Indiana, Iowa, Michigan, and Ohio proportionately twice as many women or more were found while men predominated in Illinois and Missouri. Small cities (20,000 to 49,999) were about equally represented except that in Kansas and Nebraska there were more women while in Minnesota, Ohio, and Wisconsin more men were found. In nearly every state there were pro- portionately more male elementary school principals in small towns (2500 to 19,999); this tendency was marked in Iowa, Kansas, Nebraska, and Wisconsin. There was little difference by state or sex of respondent in the distribu- tions in rural farm and rural nonfarm communities. A wider range in salary was noted among women than men. While men tended to be clustered between $16,000 and $23,999, women were more evenly dispersed among the salary categories. For example, 26.09 percent of the women earned less than $16,000 per year as an elementary school princi- pal while only 7.27 percent of the men fell into this category.* *Women in the following states averaged more women under $16,000 per year than the total female sample: Minnesota (36.36%), Missouri (50.0%), South Dakota (66.67%), and Iowa (50.0%). Men in Illinois (9.62%), South Dakota (12.5%), Nebraska (14.29%), Kansas (25.81%), Missouri (15.79%), and Minnesota (10.53%) averaged more men under $16,000 per year than the total male sample. 146 At the same time 23.69 percent of the men and 21.74 percent of the women earned $24,000 or more per year as elementary school principals. Salary differences in isolation of current level of education and number of years in a position are difficult to interpret since these two variables reportedly determine an individual's salary. A clear majority (75.38% of the men and 83.23% of the women) held a master's degree when the sample was drawn and it was not uncommon for respondents to hold an education specialist degree (17.36%).* Even at first principalship only 19.08 percent held less than a master's degree. Of those holding a master's degree at first principalship, 12.20 percent completed an education specialist degree and 1.58 percent a doctorate at the time the sample was drawn. It should also be noted that the yearly contracts of elementary school principals in the middle United States extended 10 to 11 months (73.89%) with little difference detected by state or gender. In general, women tended to have held their posi- tion as an elementary school principal fewer years than men (see Appendix D; Table 22), a factor which may help explain the salary differences noted earlier. However, *The highest rates were noted among men in Kansas (30.0%), Iowa (24.32%), Indiana (23.53%), Michigan (21.62%), Minnesota (21.05%), Missouri (20.51%), and Nebraska (45.0%); women in Minnesota (18.18%) and Michigan (24.32%). 147 such was not the case in Minnesota, Kansas, and Nebraska where 40 percent or more of the women queried had worked as an elementary school principal 15 or more years. It may be interesting to note that the women in those same states were not as highly salaried as the men even though twice as many men as women were elementary school principals 15 or more years. However, as was noted earlier, a dispropor- tionately high number of men in each of these states held an education specialist degree. Area of specialization for the highest degree held was generally supervision/educational administration (68.15%) or elementary education (21.66%); women were somewhat less likely to specialize in supervision (59.01%) than men (70.9%) but slightly more likely to concentrate in elemen- tary education (27.33%) than men (19.96%). Some differences were noted among the states in that 70 percent or more of the women specialized in supervision in Illinois, Iowa, Minnesota, and Nebraska; women in Kansas (71.43%) spe- cialized most often in elementary education. Men and women in Ohio tended to be the most evenly split between specia- lization in supervision and elementary education. Age of respondent in 1977 was more diverse for women than men with women being generally older than men - a factor perhaps not consistent with the findings on number of years a principal until we note that women were generally older at first principalship than men. It was found that 68.66 percent of the male sample was 26 to 35 148 years of age at first principalship while only 36.65 per- cent of the women were of the same age (see Appendix D; Table 26). Such consistency occurred among the states that it would almost appear that if a man had not become an elementary school principal by the age of 35 his chances diminished to about one in five; in Nebraska and North Dakota his prospects were even less - about one in ten. Women however beyond the age of 35 have a 50/50 chance or more of a first principalship - in fact, in Indiana, Missouri, and Wisconsin a woman had little chance of an elementary school principalship under the age of 35. As might be expected from the above findings, women taught more years prior to first principalship than men. Regardless of gender it was most common to teach at the elementary school level (68.15%) and/or in a junior high school (35.29%). Men in the sample (92.16%) were almost exclusively married and living with spouse while 51.55 percent of the women were not married. In fact, 34.78 percent of the women were never married as compared to 4.10 percent of the men. The rate of never married women_was even higher in Illinois (45.83%), Indiana (66.67%), Iowa (50.0%), and Minnesota (45.45%). Although having children was common among those who marry, men were more likely to have more than three chil- dren (16.79%) than women (8.79%). The fact that the children of women tended to be older in 1977 than the children of men is probably explained by the ages of male and female reSpond— ents. 149 Father's occupational category did not vary signi- ficantly by state from the overall sample (see Appendix E; Table 32), except that nearly all women in Wisconsin were from white collar origins (83.33%). In general we would assume that father's education would remain consistent with father's occupational category. In Illinois, for example, one and a half times more of the men than expected were from white collar origins. At the same time twice as many fathers as was expected had a college degree. Evaluating the sample in this manner it was found of men that in those states that produced a disprOportionate number of farm fathers for the sample, whether more or less than the 19.4 percent found in the male sample (Michigan 7.08%, Minne- sota 32.43%, Kansas 38.71%, Iowa 40.54%, Nebraska 57.14%, South Dakota 37.5%, and North Dakota 55.65%), father's education was no different than the overall sample except in Kansas where twice as many fathers had a high school diploma than was expected; North Dakota where only two- thirds as many as expected had less than a high school diploma; and in Nebraska where one and a half times as many as expected had a high school diploma. For the most part consistenCy was found for the fathers of women in terms of education and occupational category. In Kansas however women were one and a half times more likely to have farm origins (42.86%) than women in the total sample (25.47%) 150 yet no differences were detected in father's occupational category. Although mothers tended not to be employed outside the home (70.45%), they were more likely to have completed high school than fathers of elementary school principals in the middle United States (60.98% of the fathers did not have a high school diploma as compared to 48.93% of the mothers). Men in Michigan (59.82%), and women in Wisconsin (41.67%) and Kansas (57.14%) were some what less likely to have had an unemployed mother, while men and women in Iowa (81.63%), North Dakota (81.82%), Nebraska (85.19%), and Minnesota (85.11%) were somewhat more likely to have had an unemployed mother. Mothers who were employed outside the home were most often found in the white collar occupations (64.19% for men and 73.47% for women). Most principals lived with both parents at the age of 16 (88.24%) and one or more siblings (91.12%) regardless of gender or state of employment. The majority had one to three siblings with little difference in terms of sibling placement (oldest 31.14%, middle 24.39%, and youngest 35.59%) except in Nebraska, North Dakota, and South Dakota where men and women had four or more siblings a majority of the time; and in Kansas, Ohio, and Wisconsin where women were only children twice as often as was expected. Women in Minnesota, Nebraska, Illinois, and Wisconsin were middle children more often but most often the oldest in Kansas 151 and Indiana. Men in North Dakota and Wisconsin were middle children a majority of the time. Of the variables studied, some differences were detected by gender and by state of employment. It would appear that stratification by these two factors in future study of elementary school principals in the middle United States would be warranted. Summary The results of the statistical analyses were reported in this Chapter in the order of the research questions. The salient findings were reviewed and discussed, and some post hoc analyses were presented. The summary, conclusions, and recommendations of this study are presented in the final chapter. Chapter 5 SUMMARY, LIMITATIONS, CONCLUSIONS, AND RECOMMENDATIONS Summary A summary of the study is presented in this section through a review of the purpose, research questions, methodology, and findings of the study. Purpose The purpose of the study set forth by the researcher was to define the patterns and processes of intergenerational occupational mobility among elementary school principals in the middle United States who were members of the National Association of Elementary School Principals during the 1976- 77 school year. Knowledge of the distance and direction of mobility as well as identification of factors which in- fluence an individual to become an elementary school princi- pal were considered important in order to determine if an individual's occupational opportunities are limited or enhanced by accidents of birth and/or subsequent experiences. Research Questions In an attempt to fulfill the purpose of the study answers to the following research questions were sought: 1. What is the pattern of intergenerational occupational mobility (as measured by the SEI) for elementary school principals in the middle United States from background characteristics? 152 153 2. Do the patterns of intergenerational occupational mobility (as measured by the SEI) differ for male and female elementary school principals in the middle United States from background characteristics? 3. Will the pattern of intergenerational occupational mobility (as measured by the SEI) from background characteristics and intervening characteristics for ele- mentary school principals in the middle United States be replicated by a cross- validation sample? 4. Do the patterns of intergenerational occupational mobility (as measured by the SEI) differ for male and female ele- mentary school principals in the middle United States from background and inter- vening characteristics? 5. What is the process of intergenerational occupational mobility for elementary school principals in the middle United States? 6. Does the process of intergenerational occupational mobility differ for male and female elementary school principals in the middle United States? Methodology The sample was composed of 697 elementary school principals who were members of the National Association of Elementary School Principals during the 1976-77 school year, all of whom lived and worked in a 12 state area referred to as the middle United States. Data were collected during the summer and fall of 1977, using an instrument developed by the researcher named the Survey of Elementary School Principals. The data were transferred from the returned questionnaires to com- puter op scan sheets for use and storage on computer tape. 154 The dependent variable, pattern of mobility from father's occupation, was coded according to Duncan's Socioeconomic Index (SEI) by two independent coders. Upon completion, scores were compared, and rectified when necessary. The independent variables were defined in two major categories: background characteristics and inter- vening characteristics; a third category of variables used primarily for sample description, was labeled descrip- ters. ~The background characteristics identified for the study were: sex, age, father's occupational category, mother's occupational category, father's education, mother's education, whom respondent lived with at age 16, and siblings and sibling placement. The intervening variables were: marital status, children and number of children, number of years a teacher, highest earned college degree, degree at first principalship, size community of employ, and age of youngest child. The descripter characteristics were: level/levels of teaching, number years an elementary school principal, highest earned college degree, area of speciali- zation (highest degree), number of schools under direction, age at first principalship, total enrollment under direction, total school system enrollment, salary for the 1976-77 school year, number months under contract, and state of employment. The third edition of the Automatic Interaction Detector (AID3), a component of the OSIRIS package, was used as the primary method to evaluate the research questions. The Statistical Package for the Social Sciences (SPSS) 155 was used to supplement the AID3 when simple frequency distributions were required and to obtain the variable intercorrelations. The study was cross-validated by computing a 95 percent confidence interval about the study sample end group mean differences in order to determine if the cross- validation means were statistically different. Findings The major findings of the investigation were summarized for this section by presenting the results of the analysis of each research question. A proposed model for future study of the patterns and processes of mobility of male and female elementary school principals was gleaned from the findings. It was found that the average amount of mobility for elementary school principals in the sample was +33.34 SEI points from father's occupation, on a 96 point scale (+ indicates upward mobility); 4.3 percent of the total sample experienced downward mobility. The sample consisted of 31.28 percent principals from white collar origins, 45.77 percent from blue collar origins, and 20.95 percent from farm origins. The mobility patterns of male and female elementary school principals were compared with the following results: Male elementary school principals were slightly more mobile than their female counterparts although the difference in mean mobility was not statistically significant (34.29 and 156 and 30.17 respectively). Downward mobility was experienced by approximately equal percents of men (4.29) and women (4.35). More women (38.5%) than men (26.11%) found their origins in the white collar stratum but there was no significant difference in the amount of mobility experienced by each group (mean mobility was 13.05 and 9.95 SEI points respectively). Male elementary school principals from blue collar origins were significantly more mobile (40.93 SEI points) than women (32.84 SEI points), and a higher proportion of men (48.88%) found their origins among the blue collar than women (35.44%). No difference in distance of mobility was detected among male and female elementary school principals from farm, deceased father, or unemployed origins (mobility was 51.75 for men and 51.81 for women). The percents from farm origins, not including those with deceased or unemployed fathers varied slightly with 19.4 percent of the men and 25.47 percent of the women originating in that category. The results of the AID3 analyses indicated that the pattern of mobility for elementary school principals in the middle United States was dominated by father's occupational category, father's education, and whom respondent lived with at age 16. It seems remarkable that with 47.8 percent of the variance in father's occupation accounted for, all was attributed to background characteris- tics, with father's occupational category and father's 157 education explaining 39.9 percent and 18.7 percent respec- tively. It was found that these two variables were somewhat correlated (r= -.5067). The pattern of mobility appeared to be similar for men and women even though some differences were identified. It was found that the pattern of mobility for men and women from white collar origins was similar with father's education being the most salient variable. For men, age interacted with father's occupational category and educa- tion while for women, siblings appeared to interact with the father's variables. Women from blue collar and from farm, deceased and unemployed fathers were effected by little other than father's occupational category, although had sample size been larger, whom they lived with at age 16 probably would have produced an AID split. The age of blue collar origin women appears to interact with father's occupational category. Men who were not from white collar origins were dominated by whom they lived with at age 16 - a variable which interacted with father's occupational category and education. Those not living with both parents experienced the least amount of mobility. Men from blue collar origins who lived with both parents were influenced by father's education; men from farm, deceased and unemployed fathers who lived with both parents were effected by little else. For men from the latter origin category, father's occupational category interacted with the "lived with" variable. 158 The pattern of mobility for male and female elementary school principals in the middle United States could be represented as an additive model if, when interac- tions occurred among variables, the effects of the inter- acting variables could be combined at various points in the model. Since: 1) mother's occupational category and mother's education were correlated with father's occupa- tional category and education, 2) a technique was not available to combine the effects of father's and mother's occupational status', and 3) the mother's variables accounted for little of the overall criterion variance for either gender, it would not appear necessary to include mother's occupational category and education in the model. However, as was emphasized earlier, it is possible that in the future, one or both of the mother's variables will have more effect on the pattern of mobility of elementary school principals as more mothers become wage earners. Thus, the researcher recommends retaining the mother's variables in the model. The study sample was successfully cross-validated; the results of a random sample of respondents were within the limits established by the cross-validation formulae. Sex of respondent determined the process of mobility for elementary school principals; the conditions surrounding employment were more complicated for men than women. The process of mobility was identified for men as follows: 159 l. The mother did not work outside the home, 2. The parents lived together and had more than one child, 3. Experience as a teacher* for 10 years or less, 4. Married with one to three children, and 5. 35 years of age or younger with a master's degree at first principalship. The process of mobility for women was identified as follows: 1. The parents lived together and had more than one child, 2. Experience as an elementary teacher, and 3. At first principalship, women were between the ages** of 26 and 45 with a master's degree or higher. As important perhaps as what was found is what was not revealed through the study. In this country many pride in their perceived opportunity to excel; we often read and hear that individual success depends on individual effort. The results of this study reinforce such a generalization, at least in part. Since the pattern analysis indicated that elementary school principals in the middle United States were extremely mobile from father's occupational category and father's education, we must look to the process of mobility to determine how and why some individuals become elementary *It was more common for men than women to have had teaching experience other than at the elementary school level. **59.63% of the women were over the age of 35 at first principalship as compared to 23.14% of the men. 160 school principals. It was found for example that over the total sample nearly all subjects had been a teacher, the majority had a mother who did not work outside the home, most were male, and most had a master's degree or higher upon entry as a principal. It is possible that only one is an overriding factor in the process of mobility, e.g., attained level of education. However, the scope of this investigation did not include provision for such assessment. When variable correlations and interactions were studied, differences in mobility by origin strata for men and women were indicated. A pr0posed model for study was drawn from the findings which may have utility for future analyses of the patterns and processes of mobility of male and female elementary school principals. Limitations All research is somewhat hindered by a variety of limitations; this study was no exception. Although it is perhaps dangerous to study segments of human behavior in lieu of the holistic, in the social sciences the researcher is faced with the awesome responsiblity of keeping the data manageable. It was recognized that intergenerational occupational mobility is only one component of occupational attainment. No attempt was made to observe factors perti- nent to occupational attainment outside the area of inter- generational occupational mobility. An exhaustive list 161 of relevant variables was not studied; the exclusion of race as a background variable was perhaps the most serious omission. Jencks1 et al found that black men from equal origins with white men (as measured by father's occupation) averaged less mobility than white men. Therefore, it is possible that black men who become elementary school principals are from higher occupational origins than white men in the same position. The study was also limited by the lack of a tech- nique to combine the effects of father's and mother's occupational status'. Such a formula would enable the researcher to more realistically assess the socioeconomic status of the respondent's family, since the overall status of the family is likely to increase somewhat due to advantages provided by the second income. The Socioeconomic Index created a limitation in that the scores within the area of education were not necessarily consistent with what we might assume the public school hierarchy of positions to be (see Table 20). 1Christopher Jencks et a1, Inequality: A Reassessment of the Effect of Family and Schoolingin America, (New Yofk: Harper 5 Row Publishers, 1972), p. 190. 162 Table 20: SEI Scores For Some Public School Professional Positions Position SEI Scores* Adult education teachers 61.3 Secondary school teachers 70.2 Elementary school teachers 71.2 School administrators, elementary and secondary 71.7 Prekindergarten 6 kindergarten teachers 72.0 *Scores were rounded to the nearest whole number in the analysis since the AID3 algorithm was unable to handle decimals in the dependent variable In addition, it is possible that elementary school princi- pals vary in the amount of socioeconomic status they enjoy within their communities, or as compared to one another since variance in education and income among principals in the sample was noted. Finally, the study results may have been hindered by the disproportionate number of women in the sample. Although the researcher could have weighted the data for females, a larger sample size would probably be more informative. 163 Conclusions The conclusions of the study, drawn from the findings, were as follow: 1. Until future study contradicts these findings, one might assume that the patterns and processes of mobility for elementary school principals in the United States are similar to those in the middle United States. 2. The patterns of mobility among elementary school principals suggests that: 3. a) the occupational structure has remained relatively stable over time, since age was not a salient variable. Unless decided changes occur in the political and economic structures of the United States, one would expect this phenomenon to prevail, b) in terms of origin strata, equal oppor- tunity for entry into the position is a reality, based on the study variables. Certain variables e.g., race of respondent, were not included in this study. The processes of mobility for male and female elementary school principals suggest that differ- ences in recruitment practices exist for men and women . 164 Importance to Education: The study of intergenerational occupational mobility among elementary school principals provides information regarding the openness of that position within the American occupational structure. Since an individual's career occupies a dominant place in his/her life, interest in opportunity for the position are natural. The results of such study have implications for career guidance and recruitment to the position. Knowledge of the process of mobility to the ele- mentary school principalship can assist aspirants to the position in establishing factors which limit or enhance their opportunity for the position. It was found that the process of mobility was somewhat different for male and female elementary school principals, e.g., men were younger with less teaching experience than women, and proportionately there were few women in the sample. It was also found that women were older, taught more years, and were slightly more likely to have taught at the elementary school level than men. These factors may indicate that few women aspire to the elementary school principalship, or that women do not have the appropriate characteristics for the position. One might also express curiosity for an unmarried man's chances of becoming an elementary school principal since nearly all male respondents were married and living with spouse. In short, awareness of the characteristics of those who have successfully competed for an elementary school principalship 165 provides a baseline of information for knowledgeable career planning by aspirants to the position. The study of entry level characteristics of holders of an elementary school principalship may have implications for past recruitment by school administrators. A pertinent question might be, do men and women with these characteristics aSpire more often to the position than others with differing characteristics, or do individuals who hire elementary school principals seek men and women who have these characteristics? Finally, this analysis provides a baseline for continued study of intergenerational occupational mobility among elementary school principals in the middle United States. The large amount of mobility recognized among elementary school principals also presents us with a serious question. Did these individuals become elementary school principals to enhance their personal status rather than for more altruistic reasons? Recommendations Some recommendations were noted through the course of analyzing and summarizing the findings; several deserve final mention. The pattern of intergenerational occupational mobility from a single occupation would be enhanced by the ability to measure individual socioeconomic status enjoyed by holders of the position as determined by income and 166 education. This would allow the investigator to use the respondent's status as the dependent variable. Using the content of this study as an example, intergenerational occupational mobility to the elementary school principal- ship could be measured from father's occupation, and other variables of choice. Recognition of the process of intergenerational occupational mobility for elementary school principals is far from a reality. Such knowledge of this and other professions would be beneficial, not just to satisfy a researcher's curiosity but, for use in career planning and as a vocational guidance tool. A technique to combine the effects of father's occupation and mother's occupation would be valuable in determining family socioeconomic status - a factor that this researcher believes will become increasingly potent due to the increasing number of women who are becoming wage earners. It is with regret and chagrin that race was not included as a background variable to the study since it is possible that racial minorities and the white majority experience different processes of mobility. The AID3 has potential for generating testable hypotheses for further study; more remarkably, it has potential for theory development in the social sciences. 167 Coding the SEI To arrive at a code for a specified occupation, we use ”millwrights" for purposes of illustrations. The Occupation Code (this Appendix) must first be obtained from the Occupational Classification System* (this Appendix). That code is then checked against the Duncan Socioeconomic Index (SEI).* In our example, the Occupation Code for the occupation millwright is 502 (see p. 172 of this Appendix). Code 502 receives a SEI score of 31.0 (see p. 182 of this Appendix). *From: Robert M. Hauser and David L. Featherman, The Process of Stratification: Trends and Analysis (New York: Academic Press, 1977), pp. 309-329. APPENDICES APPENDIX A Socioeconomic Index 168 ...9:u_.5.: 9.: .9... ........$ 3 9......929 ..._::....:_.:.. 9.: .5. 9...... 9... 4:9. .........:. 9.... v.92... 9.9.; .u.............. ...:......:9 3.2.... 5.5:. 9.97:1... 5?... 9... 9.5:... .....:..:.... 9.: 2:... 17:..5 ..: 99.... .c: 3.. £97.... 5.3.90 :....:..:.:.. 9... 5...... 1......9. 91......" ...9.:..:..: 9:: 9.: 1.95... .......:..9. 2.: 9»: 9.....3 2.9.9952. :. .I.......:.....:. .9359... .5. 9.3.5... 3. ..9,.. .I.. 9...... It; . v.7.:::....r..7.::.. 1...: z.....,..._:-.. mm: 399...»...9 .:9:..::.:7: ..:: .:......:...:..< .5: 5.7.9.... .:...:...-..:.:.: 22:... I“: . 2.. . :24...— . . a a .5 r. 9.9 .. .I........:u:... mm: .99.: {21:99.7 .........:..U m...: 1.9.5.5.... 7......r we: 7.79....5. 7.....7. .5 9:23:53 2.: £99.33.....:.....::.... .1. 59:53.53... .....:...::C :5 7.99:3..9 3:37. Om: 17.—399.7 .9......::.v 1.9.2.3.... r.._......:.: .5: .:.._u.:........../. n... . . . . . .- . . . - £993.35. 1x. !..........J.... .......:......../ I: m...:.....€99< :x. v.........u.... ......7:..:. fl... £9933... ........:.....9 .5: .:...:.....m. m... £333.33 Cut—27...]. 9......» 2.9953... :10 :c 27)... ....:.. 1:: 9:92.... .2.......:..... 2.9.4.2.: ..3::.......: 2.9.1:... 1:931 €94.39. 23.1.9.9... 1.... .......3x .99... .z.a.........z .........m 29:25.. 1.23:... 1:: :21... 31.43.3922). 1.2.23.2..9zl— 27...........’ 1.9.1.2.. t—t.......=..um 27........., .........m 5.9... .z..9...3.$ £51.29: 29:13.... 1.94:5. 1:54.29: .....9.: i...:......:....... 1:: r.z.u......:.9... .115: 2.....77’: 9.............. .I................... 1.... a...........:....... .....u...:....:.. 52.31239. 1:: 2.x..u......:.99. 1:5... .113... r...’......... :— 1......2 25.3.5.9... 1:: 2.7.u.........9... 9......:.....:_ 1.9.5.... 5:191:39. 1:: 2.7.3.1....3... 519.. £11.52... ’93:: 19.97.39: . £51.33:— z.....:.9... 1:: .z....:..9_1 £92.27. .99... .w..9.......:..:... .11....— m: .2 :2 m... 2: a... m... :: 2.: 3...: ma: 2... 2... m2. .2. =2. .7: mr... 7r... .2: my... .1... :r: E... mt. v.2. E... 75:322....dt3/ 37.3.3.1... 9.............£. 1:: 15.19:. {5:97.23 w.a.9......:.... 3......9...:...C 271:92 m....9:.........uv a...........9:.._ 19....9. 1:: {7.1.51 .z..:.....’....... 2.9.1:... 52......9. 5.1.. 1:: ........7...... £3.12... 1:: 2.9.....:......... 7.597.... 1:: m......:.....C 5.9.: .z.a.....9.9.z. .:......:.>..... 1:: 9.1.. 2.95:5..7... 1.... 5:9,...2... 2.7.59.9, 9......7. 2.7.2:..er 3,129.... z......9.9... .:.......3.3..: 7.7.39.9. 99...... 1:: .....9....£.....< 2.2.2.9.... .:.::.:9...u< 31.2.9.9. 1.9.7.9.... 1:... 9......— 1::.9...:.:.m 2..........:.:9...:7. z..........9< a.z.......9..z .:9..::.9...:./. 1.3:...9 1:: a7..9...9.< w..:.......... £2.55 1:: £73.59... .z::..:.....a. 2.99.5.1. 7.9.1.... 292...... 1:: 1.99.3.1. 2.3.31: .c9:.9u::::. 2:3: 19:5...30'm:”.52.?” QEGQZE CZ< J.:fio>_:= 1:: sun—1.... :53: .2. 3.25... m: z.;._._.:u.....:_._ 2.. arias... 1::—.7. .’..E_=.3.....m v: 2.5.3.5..” 1:: iiizm C:— auu£.:...z =2..=u§_:.§ 1:: zat=391=595 n: w...v.,..EE=a 1:5 1.2537,:7. m5 7.5.525. .33.: .25....535 A7; NZ Ear—car: 3:: 53:13 3: 23.—.25. 5532...; :35. :— 2......u_v...2 3.: >.:fi?:== 1:: 83:3... :53... .fi..._.:..v._. 29:25 .2: 1359:? fi:_=:< :2 .c: 5.93:2. 3.37.335. 1:: 9292.5 .E.9._.:§.—. c: 2.5453 130:3 1:: 7.23:3. cr— ates—92:: 1:: mus—=5 £55.35. 25.532993: mm.— .fi:3< m: 2.5.3:...“ 152:9: 1:: 4:27.55. 5.22.? 3..— fiaigrzz... 1:... £73.: «£9.23 2.5.3:... ans—Fifi mm..— fc_...:==:... _::c=:.:._5 1:: 1:53:25 3.— I..:_u.5. 3.1 mm..— .ufié .v=§3..__u...—. n: 2.5.3.5. 2525:52 2:3: 2; 1555 15.55:: £3:_:.Eu::_ 1::. m: 2.5.3:... 3:53.:— :u_.::..— :2 2.3.3:»... 2:24: C.— fi.:_...=.: zazuzm cm— .fioazfizz 23...— ct $2.25. 55:2...“— mm— .,,_..z:_.x_:_m_ m:— fii—uzu. 25.33.... _=.::.._: 7:: 5.3.25 «.m— v._..::.::.3 95:... .._< 3: . 1.2.9.5. 97...: 1:: .555. .:< mm— a...—_r_ 9:..7:_< mm: .95.: .3235... 99:3,: 139% NN— 9.2.9.3. 5.5.4.3. mus-3.5m ~N— 1:: 35.92.52... 1:: .575; 3...}... .fi_.:..._=._ua.—. 5.2.25. .232: ON— .ué... .2.._E..._.:_u..: .5233. 1:: 3:52:33.“ m5— a...:_u:.: .I..a_:_:.=5m m: .I...¢..I....:..v. 3: 7.5.3:... 395.555 1:: amazmazm m: 2:33:39... _:....:__..:_.=_.=—/. cm— 2.5.3.5. ans—csohzm :— 1:392:33 3:52:35 twat—5.934 mm: 3.5.25. 3:13.an¢ 5.50: M: 2:53:29... 3:52:95 13:25.; «m4 £325.98 u—_t.§:_u=~ 3:92.40? 62¢ 755.55 mm.— vezczcooJxmzznga QMGQZE £50 Esseaa mm. 32 :202203 J:_ 1:: 2.2.3:}... {b.7211 12:57:— 31.4. 2.1,.— fi..::.:::.. 55.51:... 1.... 1.51.53. 5...... .131..J:{::_ 1.... 1:14:72 z._...$.:.:.::_ 1.... L::....::.:.m_ 131...; .5311? 1:: ,._.:_.:...7._2 1.5.. 13.5.. £43.: b.1533 1:59;: 1.... =:_ .a.....u..__:uv 5.9.: .I....fiikri. 152...; ....:.:.. 5 11.27. .I.—2.7.75. 132...; fizz—zany f....;......3::2 1.3.9 3.2:: £32... 1::: m:...§:3.$ 23:27.: 272.. .7...::m_._uv .. 1;... 21.”. :21; 1.... ...:.7..__.n :.::r...1rz 1::; 1.... 4.57. 1.31....- 1:: 5:91.: 2.3!. 1...: Iii/.517. ,._.:._._.$._.._::. 1.... £51.... .115”? «3:355; 1.5—11.:— 1:... 1.2.74.5: 13:51:55: a._.a.::.:.:.< 957.91}. 1:: 3:34: 9:731:14. wig/3.0.5 mmq93. 5.52.5: 5......5: u:::......_:2 5.72.5: .5..:.=.._._... 72.2.5.2. .5: 5755.5 55:2... .5352: ”5.7.757... fi...:....x.: 2.7.2:: "5...... .5: 5.7.9.3775: £3.55... .57....2: .33.: {5.75 .a.....:... 55.1. is». 9...... .5: 55:92.1...7. 3...: 7.5 .._....z.. {5:55. 7:7. 7.5::zznulmzwzxpfi 25.27.... 272 ....... 9:37 1.... 1........... .........J. w........u.... 22.2.11; a............_ 1:... z............ :Em ....::..:..: 9:7. 19...??? I... ............... 1.............7. £157.... 1.... 1.1:... _:............_m 2.....7 1:... fry—.5: 1.7.... .5271... 1.... 1.3:...— a................. 21:71....— u.._....... .I........... 9...... 1.... ...:.7!.... z..:........: ...........1 .55.... a.................... 5.... 2.... 1.... ....._:.:_._ #5.... 2.... 1.... .I...._....._._ .I......_....:._.... .........z..._._ I......./.._._ ...........x..: 1.... 1.3:... :21... 1:... 2...... 3.2.2....321: 1.... t........u..........._.. 52.x. .3... £51.... 1.1:... 1.... 5.5.... a....u.........a._..._ 2.5.1.9....1: .....:..... v..:.:.....:...... 1.... .............I...... 2.35.... 1.3.7:...— 1.... (31...; .1...— 1.... 5.33:1: z.z.............q...... 3.5.... 55:7. .rQ................= ....1_:_/. 13...... .I...._._.._/. 3.13.3231. 1...... 1.... 5...: ..._.....J 2.3—:7. ................. 1:... .I................. 19.1.3.7 ...2 ...u:........... 1.... .I................. z..c......:....r.:/. 2.52.55»... 5...: .3... ..............7. C-mf‘ I“? .I.! .l nmm ....m. we... mam. .cm m3 m2. .2. 2...? =3 1:... 1.5.1.... 2.2.5.... 1.... E1...— .:._._..:... .......=3 1.5.... :5...— .r........_....... 1.... [5:3,]... .I.....,z........ 1.... 1.2.5.2.... 1_..._...:.....: 15...... .3... .r..._...:_....... .......::....... >925: .........._._:._ 5...... .................. 2......2: u.._,,,........... 3:: f...._.........._.. 92:23:... ...__._:......:< 51......5... ..._._.....:.:¢. ........._........ .13.. ..:.......:...< ...........< ...._.......u......... 1.... 451.....— .u......._.__....... :4. :...._......._... 1.... 1.1.2.397. 7............:.._.. 11.1337. 1.7.12.3..7. ......_...... ...:............._ z...........»...... ..>............... 1...... .33.... 9.1 1.... .1... !S._.......u....$ 1.... £91.»...— ......... .53....27... ..L._....._ 1.... a... .1511... 1.... .3313. £23.37... m............:.... 1.... £31.23... .323... .3...— a....s...uv 231...... 2.9.7:... 1::... 1.... 3.5.5....— c...=....=.::£ 1.... :0593...» 5.0.: ......E.......r.~ 1......_.....Ul.m:mzz..3 Quiz—7...:— DZr. Zn..7.w._L<:U ct... mxv 3.... 5.... Mr... .5. :5. mu... 3... 5. 3... m2. 7.... 2.3. m... .5 cm... mmw in... mm... mm... :m... E... m... 2... 2... m... 2... 11:9 .33.... .......O 174 25...... ...:..= 1.... £91....» 5...... .m................ 91.5.... iris...” 2.1:... 1.... 5.7.3. .z......:_:m .I....E... 1.... fir...— .25....3. a............._.. .4235... 1.... 42:21..— 5.2.1:». w...>........._.. 2.1.5.... .55.... .m......:..:.m z........1_:m 19./.222... 0:32.... 3.5.2:...25 firs—.2... 1.... 1.....5...r. 2.9.32.1. 21....._._....1 1.... 1.3.1.1. 5......7... 1.... 2......3: a................ 27...... u......:...,_ 1.... 4......— .......: .z.............: 2.1.2.... .529...— z...._........_: ...._._....... "5:1... 1.... 2:...— w...>:......._: .:._._..=... 3:15.”. 3.915.923 5...... 5.2 a...»........... 2.1.2.... .52....— z....._.:.$ 5.....3: .._._._.:..:...._.. .I............ 1........:...::...: .I....:_..._ 02.13.: 1.... 2...... .95... £5225... 1.... 1.54:... 3.... 5...... .7....:......:. 1.... [9.1: 55.22.: 3.3.7. ......... .1............_: :33. (5.517. z......._._ 1....7. 31.... 1...... .a.............. .197. u...........1::..: £545.... 1.... 5...... =57. 3.2.52.1...3... .8... .m....._....... 1.... 5...... :57. 5...... £91222... 3.1.5.... 21 1.... .I.—2......— 1.....: £6.25: 3...... 7...». 2:». Nb: _ h: 3...... cs: ms: 3.... i... We: _ .2 cc: cm... an... m... . m: cm: m. : I... 2 c m... _ E E... . c. .c 5...». 3...: 5.... 2... 3...: cm: :2... 1.... .92.... 7.3.5.. .z......_...... 1.... z...._...... .5212.— .J.._.:......_.:...:. .I....ti 1.... z.......... .1.....1........ 5.3:... .I.... 1.... 2.31:... .35.”. .I........:._ 1.... ............._.....z ......................~ 1::...— 1.... .t...1...£ .1521... .151... 2......2 1.21..— 1.... 2...:27... :3: .32 .12... 2.5212 .I............ .29.... .r...fi....../.......7. 1.... .I.....4.....$....A_ .......: .2..>_........_: 3......3 .I....zy..... 1.... 2.........._ 5.2.1:... u.._.................:. I..........:.... 1.... .z................ {bi—9...”. u...>...>..7. 3......5. 1.... .:......1:. ....:...._..._U a............... 3.1.5.... 1.... 3.21.32 ..........13:.. 1.... 2.35.:— y............:.:(. 1.1....3 ..:_..._...£.. 1.... a....r..x_.;.< F:C..mZ<:.—. .7532”.— .mm..:._.<=m_..3 5...... 1......5. L... ..: .I....__:..... ...:..:.._ ......... .m............. 13.1....— 1._.. ......7,.._....“. z...._........_.... 1.5.9:? .:./. 5...... .m...._......._._.. .1... 1.5.9....» .I........../........... {2.25.123 .31.... .._1 1.... 1...... (54...: .._1 1.... 1...... 18...... :1... . 1.5.17. 1.... ...........: ....:._._.._....—. ................. 1.... 5.7.2.2.... 2.3.1.3.... 1......_...:Ul..m: 21:35 amt—2 .I. .2 2 ./.< KEEP—£51. mm: em: mm: NW». 3.... :9. m3. v . m. m. _ .c m . .c — .c c _ m. ms: 2.: ms... C. m3... :5 vmm mmm. 11:... .5... .2. .......C 17S 1......,_//.. _.......A_ 1.1.42.5 .9112? £21.;— ——_:.—..I_...._ 2...}...— .._.5(. .55.: «LL45: .. ..:...7 1::..— 1......n’2 I...4:..>’ .__.._.._..J. ~u.....r..._m..ya _..:.u_ z...._.......$._,..2 1::—.723; .....::_ 3.5;; £432; 527:: 1253...: 1.4.3.: 9.3...7 1......— z::.z.7 1:: 1::..2 :.:::$::_u 1.... z...._......; 1_:._.7:.:_ .....::_ .._.._I...z1.:.:. 1.... r1._.:_....:_..::_... 1.71:: 9.3.3:. "2.2.3:; 243225....— .,._._.’../_:._ 37.”.— .mzzxzcyz why—9:21. _..:.:_.__.....:.7 .I...:.._..._ 9:13 :.:..._ 1.1—.55 1::... 1.1.7:. [3:21.— :..:..._ 1.1....” .52.: .z.....=x1._ :21..— :.::...2.. :23; 74.—74.3.3..— _/.:<.._ 27.x. «av—H.325»... 74;...— ..._..u.::.:_ 5:...— 7..:.:.: 1.... 1...: E; I..::.:.._ 51..”5/(7 7.7.: 27.4. fizn.__/~_/...._ 1.....:”.._ _r.__ru...*z ...1L z....3._.._ l......._..__...7:/. .7. : ....::.7...:_..,:.s 17:23:; 1::-.22.... 12.. 1::—1.1.5 .._....._.../ «7.. .9: _m.. E: .I.... I: 2: m... .I.. E: 1:? MC? _:3 . m1 rmx max 3...: ~31 :2 min :7.“ :...... E... 1.....7::r..~. 1.52.3; 4.31. a....._._:._.1::.$ 1.... .:..:7:... ._..::...:_:.:._ 7......_...’...’ 1.... .....:_...=:_I.u..:._ 5...... if. .I....zrrlz1..::_u 1:: 1.........1.:.C t.........::.. 3.4.2.2.; I...__:...__ 1......2: 1:: 21:9...— ._...._....r 1.. _..... .r....a..._r_.A 1.2.7:. .1............... if. .z....._:._.._ ..:=.:...7::; !..::.:_ .n.........:_..: 2......— Jf. .a...1........3 1::...4 7:4... .7. m... .74.. .51... z. :— a: _ 1.1.1 4.2:. 1::—1.1;.1::I...,:.__.._...._./:._. :..::_.:_ 3’ 1.5.1:: :..:...4....._ 12.3.11: fizz—5...... 3.24:..— ..:.. .15.... 35.4.1...— J..:..:.._ ...:_:. 5.5.2.57. 7:92.22... .52.... .5... 1.... z: 4:...— :.::.:::. 1:... :.::.a.....:_..2 .75.... 1.: :21: .:.::.:.::. 1:: 1.395155 1.5.1.1 5.: :.::_.:.=... 1.... :.:::.:= (3:1...5: 1.11.3.2, .37. (33......2: £.:.:...:...7.7. 1.5327. 2.: .I.::....:_: 2:44;. 1.5.9:? 15.52:...75. .a..1......fi... 9:32.; 5...... .I................ 251....” _...._..:__:H.l._.::.7.z,..:._. 1:.sz .xn.3_.5..._n._._: 3... m. ... It 3:... hi“. :1 7...... .3. m. h I h r.— ... m. ... _ I. :_ ... :2... m5... 2.... v.3... :K. m3: :5 m3: 3:: :3 .45..» 55:: .......C 176 1.539.: 9.: .3... 59...? z. 9.52...." :::...5::..: a... 5. 91.5 .5. 5.3 19.2.... 9.: v.95. 9.2.... 4.....2195... 559...... 5:51 1:5... 55.5.5.5 5...... .5. "5.5.: 12.153... :5 2...... 5.1.5.5.... :: 9.2:. .3: :1 £9.55 52.5.9 5:51.23.— a... 53.... 7:57.... 3.3....» 1.55:: :5 .z: 2.7:. 1......5... :5 a... 33...: 2.32.; :. 50.15.55 135...: :5 9.2.51. 3 1.7.: m. 31.... 2...... u 1.55:.....|r.........:.: 1.5—.75... .51.... c7... 7.31:... 2.21.... 1:: 21.5.... 5.71:... .53 195.5..:.l1.:.....’.::.. ......_...:_ .3... .r...;.......$ .5353. E... 5.3.5.5 9......7. 9.2.55... 1.55:..:.|...::..::.. 5:... 1:: 5.52.... 5:... :2... .31.: 5.5.7 ..........5 .ma 1.5.....::.I!....u::::. 5...... 1:: 1.55:... ccr 52:32::: .5: 55.53.: .I....TD Mum: 1......5::....:::.. .193: 5.25:1“.— ca... z............. .35...» mm: 1.53:..:l.:.......5.:_: ....5:..:_... 1275...... .cml . 1.5.3.5... .53.... .3... .m..x..5.....’.::.. 3mm 1.53:..:l....:7::._. 5.55 .7......:.S..C $3.». 325.57.: .5351. 15.5.... mi. 1.53:._:l.,...:.5.$ 13.51. .5: 2.57:5”. txm 22.23.33.225 .5: 5.42......,...... .72. 1.55:..:..Iz........:.$ .......5.4 .5: ..........U can 5......5: 525...”. MI. 1.53....:.lz..:..:.$ 5...:m cam 17:32.5... 3:3... 52.. 5.31:... ....:.. 1...... m... 12.33::lzz... 2.92.. .z..:.:..,.:.:.1: 1:: 1.95:2:7. cwm 5......112... .2. 1:55.... 5.5.5.. 3.15... «5.31:. 1:: 5:125: :2. '53....3 1.....5... 1:: ._:.5:...... ._:::.5......:.. ca. 1.5.1.... on: . .. . .. . ‘ 2...... a. 5: 25...... 3:14: . . . ..m......:........... y..........:. _< . E . E .. .. : a. .3 u .. .. 93.1.? 15.7.... 1:51:33. v.2. 52.5.3.3... .57. z.._._.....:.;:: 2... 2.5.2.5.. .z... :::...:...: ..,..._..._.......< m... _ . . . . 1.3512259? .:......< 3... 1.5.3.5.. .5}... 15......7 1.... 11.. 7 .3. 2...: 53.5.3 .53.? 1257......— 1..:....,...:._ .55.... .7.7.z...,.1::.w. fir: . . . . I . . .z 7:... . 1::. .I.. x... 7:: . .. . 2.35: 1:52.... cc: . . .. . . . . .. . . . 4.. .. 3.... 55.53.: 1:: .5251... .z..1.: 5.7....7. mm: 1.51.12... 3.5.... 3.....3 .xa fry—31:: .2... #1... 1.3.77.5:— ...:.:.:. .z...3..:..$ 9:3 1...... 1.4.: . . a . . . . . . . 1.9.5.5. .1152 Mama 1......2. 1.... £1.57 mt: 5.7.5: 5?. .51.: .115: mma .r...k............ 1:: .55....25— 35 . . a...._:.z..:.. 1:: 7.1.5.17. 9.5 1.:5.5:U.|QAC..mmDC.. 91:0 3.35.913, .2... 1.5.2.; mm... “hf—.323...— .u /.m .mm—m2~_c.$ SUP/ZMW git: .:........:.=. 9... 5.53.... F... .553 1970 (It-Inns «wt-"patina. code Duncan 8 HI" Mule sum-s T0101 .su 01's 177 Sic-gel ( 1965 NOHC) l’rvsh'uv" Mule sum-s Tutu] sum's 001 002 (Kr) 004 005 006 010 01 l 01:2 01:1 01 1 015 020 021 022 03.1 021 03", 036 0:30 0:11 0.12 03:) 0“»! 0.15 0:16 0.12 01.3 044 015 051 05'} 054 055 0.56 061 062 063 061 065 71 072 074 075 076 080 081 082 76.8 85.2 65.0 65.0 65.0 87.0 89.9 84.0 81.0 8.55 80.2 815. 1 8.5.0 81.0 87.0 860 8:10 180 8.1.0 9.1.0 92.3 60.0 74.6 81.0 80.0 81.0 80.0 62.0 80.0 79.4 80.0 80.0 77.2 65.7 83.6 75.0 96.0 79.0 81.3 92.1 58.0 78.0 39.0 44.3 59.9 48.0 48.0 60.0 76.9 85.3 81.4 58 .9 55.9 66.7 50.6 50.6 50.6 71.1 67.2 7.8 69.4 55.6 62.1 58.4 61.6 57.1 50.6 67.0 51.0 51.0 5.1.0 75.7 7.5.1 51.6 59.6 55.4 65.0 55.4 55.8 47.0 67.7 67.1 67.2 73.8 64.8 50.8 55.8 60.0 73.6 62.0 60.3 81.2 36.7 59.7 52.1 60.1 40.5 61.0 61.0 51.6 56.0 65.1 55.2 38.5 178 Duncan: SEI" Siam-1 (1965 NOIKD I’rcs‘ligc‘ 1970 Ccmns occupant)" uulc Mala.- .wurcs Total scurvs Male sum-s Tnml scores 083 48.0 61.0 084 48.0 61.0 085 52.2 55.2 49.8 51.1 086 52.0 69.0 090 56.7 57.1 55.0 54.6 091 74.4 74.3 53.6 53.5 093 81.0 71.4 095 65.0 50.6 096 81.0 65.6 100 64.0 2.4 101 67.0 48.6 102 84.0 78.3 1113 54.0 78.3 104 84.0 78.3 105 84.0 78.3 110 84.0 78.3 111 84.0 78.3 112 84.0 78.3 113 84.0 78.3 114 84.0 78.3 115 84.0 78.3 116 84.0 78.3 120 84.0 78.3 121 84.0 78.3 122 84.0 78.3 123 53.2 55.6 46.8 48.6 124 64.0 53.2 125 84.0 78.3 126 84.0 78.3 78.3 130 84.0 78.3 132 84.0 78.3 133 84.0 78.3 131 81.0 - 78.3 135 84.0 78.3 140 84.0 78.3 141 61.3 ' 64.3 44.3 43.9 142 71.2 71.4 58.9 59.2 143 72.0 56.1 14 1 70.2 70.5 59.8 60.1 145 62.3 57.7 44.2 44.9 150 62.0 47.2 151 2.0 47.0 152 67.0 56.1 153 62.0 51.6 51.5 154 64.1 64.0 49.5 49.4 155 62.0 47.0 161 45.4 , 53.1 162 62.0 ' 47.0 179 Duncan SEI“ Siogvl (1965 NORC) l’rt-stiuc' 11170 (In-Inns (K'cuputiun cmlc- .\1;|1(‘ \'( mes T0101 sum-s Mule s‘um's 'l'nlul wuws 163 79.0 70.1 164 69.0 42.8 165 60.8 60.3 51.7 51.9 ' 1711 48.0 48.2 171 69.0 42.8 172 62.0 47.0 173 62.0 47.2 174 65.0 50,6 175 1511.11 . 55.11 180 59.4 60.2 51.8 52.1 161 70.0 59.8 182 45.0 37.6 183 70.5 70.4 56.5 184 82.0 51.2 165 52.0 411.11 1911 67.11 511.2 101 511.0 411.5 192 82.0 56.7 1.93 65.0 511.6 1114 40.2 45.4 38.0 41.2 193 65.0 50.6 ‘-201 61.2 58.8 50.9 53.8 2112 79.5 811.11 011.1 ' 611.0 203 50.5 50.3 43.0 42.9 205 72. 1 50.0 210 74.0 48.8 21 1 511.11 52.2 212 74.1 56.9 63.8 59.4 213 57.6 57.5 39.6 215 66.7 66.6 42,3 2 16 32.0 ' 38.3 2211 75.1 73.1 ‘ 57.6 56.0 221 49.9 56.7 2'22 57.3 1511.5 1511.7 1511.3 22.1 50.8 (511.11 48.1 43.4 224 61.3 60.8 58.4 58.3 223 74.7 74.8 46.4 226 53.2 40.11 2:50 37.6 :18. I 38.7 38.0 2.311 711.0 1111.3 48.5 47.3 233 74.7 74.6 54.2 2515 77.9 77. 1 711.6 69.6 2411 71.7 01.7 01.0 215 112.11 111.7 511.3 511.7 260 66. 1 42.-'3 2111 411.11 111.11 262 3.5.0 :38. I 28.3 510.6 261 08.8 12.9 18.6 20.2 C. _ O .0-. — 180 Duncan 8121" Sit-gel ( 1965 NUIKI) Prestige" 1970 Cvnsus ou'llpulimn cm1v .\1.|1v st'mc's 101.11 wcm's 3101(- sum's '1'0101 sum-s 265 00.0 40.8 266 27.0 15.4 270 02.0 44.11 271 72.3 72.4 51.2 51.3 281 65.0 . 49.1 282 611.9 39.9 283 39.0 28.7 28.6 284 39.0 28.6 285 52.7 52.2 35.8 35.4 301 52.0 51.8 49.5 49.1) 303 44.11 36.2 305 50.8 50.9 47.3 47.4 3111 44.11 43.9 31.4 31.11 312 43.6 43.8 35.8 36.0 313 43.3 42.5 28.4 27.9 314 44.11 36.2 315 39.9 411.1 33.3 33.4 320 44.11 36.2 321 59.2 56.2 42.9 41.2 323 . 43.7 43.6 36.0 325 44 .0 31.4 30.8 326 62.1 47.6 330 44.0 43.1 40.4 39.7 331 53.0 42.3 332 43.11 43.3 35.1 35.5 333 28.2 19.4 19.3 334 44.11 36.2 ’ 341 44.9 43.7 44.3 342 45.0 44.9 34:3 45.0 44.9 5141 45.0 44.9 345 45.11 44.9 350 45.0 44.9 355 45.11 44.9 360 44.0 41.2 41.3 361 44.7 44.6 42.3 42.4 362 44.11 36.2 363 67:81 43.1) 364 44.11 37.1 38.9 3711 61.11 45.8 371 61.0 45.8 372 61.9 61.0 46.5 45.8 374 24.2 24. 1 29.9 75 43.7 43.9 35.8 36.0 376 61.0 43.3 381 44.11 43.9 25.11 25.2 382 63.2 62.4 49.3 48.8 383 . 22.0 29.8 1131 Duncan 8151" firm-111965 .\'1 1111'.) I’m-011:0" 19711 (Imusus 0111111311100 (“0111- Muh- xcm‘cs '1'0101 \(‘Itnw .\1.111- sum-s '1‘0101 sum-s 381 47.11 43.5 365 45.11 411.4 3181 511.8 35.4 391 61.11 41.3 392 41.9 41.8 35.5 35.-1 391 43.7 36.2 36.5 395 41.11 36.2 4111 21.6 32.5 4112 21.9 21.7 34.0 311.8 40.1 10.0 33.5 4111 ‘).2 11 311,11 4115 39.11 38.11 31,3 31.1 4111 27.11 35.7 41 1 32.11 411.6 412 19.7 32.3 -1 13 22 :1 318.1 -1 15 18.11 519.7 319.0 401 :01» 400 4211 12.0 313 421 19.11 31.6 422 52.11 35.11 12:1 40.0 40.8 42-1 2 1.11 38.7 423 .100 : 7.4 420 45.0 01.0 ' 4311 44.11 49.2 431 : 7.0 40.8 4.3.3 49.0 39.2 43-1 55.0 36.11 435 47.11 41.2 436 22.8 31.5 440 17.3 17.1 31.1 31.8 441 49.7 49.5 45.3 45.3 442 23.0 22.8 35.5 4411 17.8 29.1 444 39.5 33.4 35.2 32.3 4-15 25.2 25.3 26.7 446 21 7 35.3 4511 22.1 22.5 '11 11 452 41 2 41.11 31 3 453 36.4 37 5 37 5 454 11 5 46.4 455 57 8 50.8 456 45 0 36.2 461 32.9 47.7 4112 11 11 411.6 4711 ' 27.0 36.7 471 48.11 48.2 182 Duncan SE1" Sicgc1(1965 NURC‘.) Prestige“ 1970 Ccmus occupation code Male sum-s T0101 scores Male scorn Total scores 472 19.0 36.7 473 19.0 36.7 474 25.0 411.8 480 27.0 32.6 481 26.6 32.8 462 27.11 32.6 483 10.0 311.4 464 35.9 353.8 485 36.11 35.0 486 211.5 35.6 491 34 .0 40.8 492 26.5 32.8 4st 27.0 32.6 5111 19.0 25.2 502 31.0 40.3 503 12.0 39.1 504 33.0 40.8 505 43.0 33.9 5116 39.11 51.4 510 16.4 29.-9 29.8 51 1 29.0 40.8 512 13.7 13.5 27.7 27.5 514 43.11 38.7 38.7 515 63.11 61.1 40.1 39.7 516 38.0 32.0 520 25.0 33.2 521 29.0 ‘ 40.8 522 34.0 40.6 523 33.0 40.8 25 50.0 38.8 530 46.3 45.6 39.3 39.1 531 40.0 40.8 533 22.0 36.0 531 15.1 31.5 535 - 33.0 36.8 536 33.0 40.8 540 34.0 35.5 542 12.0 1 1.9 32.6 543 16.9 16.8 30.7 515 45.2 34.9 32.9 546 24.0 31.7 - 550 33.7 33.4 35.6 551 22.0 21.4 34.0 552 48.8 39.1 554 49.0 39.2 5011 28.2 38.4 561 49.2 42.3 562 41.0 40.8 111711 (3”qu Duncan SICI" 183 510201115111...) N‘111(:‘1’|l'\1121"' (u't'llpulinlt l'Htlt' \lult' \(‘Hl’t‘\ ‘I'ulal srmrs \I.|1(- u-nu-s Tut.“ sum-s 36.1 21.1 21 2 29.9 571 514.3 .116 40.8 572 39.0 40,8 37.") 23.7 23.3 42.1 41.7 m] 32.0 28.-1 602 17.2 27.3 NH “,0 32.1 NH 115.4 16.1 23.3 3515 1,113 25.),” {V1.1 Iiln l9.) 16.1 36.1 MI 17.6 169 21.!) 20.7 612 18.35 13.9 25.3 28.3 61‘} 23.0 22.1 311.7 31.0 611 21.6 26.2 61°) 21.3 ”21.-1 36.4 36.1 620 12.0 23.0 621 16.7 18.. 23.1 23.3 622 18.1 32.9 62.1 17.9 2137 624 17.0 1615 32.9 32.6 623 12.2 13.7 23.6 24.5 (526 2‘1." 312.?) mu 15.0 18.2 6:11 28.6 32.0 6.111 16.-l 16.1 2.3.6 2.1.3 6'11 16.0 19.4 635 111.8 19.7 30.3 30.2 636 46.0 33.1 6-10 16.5 16.5 2631 611 17.6 17.3 7.5 27.3 612 15.0 24.-2 643 18.0 18.1 19.5 614 18.1 29.0 26.8 615 12.1 1.- 3.3.9 33.5 630 21.8 22.0 31.7 31.8 6.51 21.9 19.0 632 21.5 21.6 31.9 31.9 65.1 21.0 20.9 31. 1 656 19.4 111.5 30.4 30.3 660 20.1 20.6 31.6 31.5 661 16.0 33.7 662 04.9 27.6 66'} 18.2 25.2 ’31 661 (19.2 31.6 ms 23.8 24.-1 35.-1 36. 1 666 16.6 31.7 670 03. 1 03.3 28.9 671 21.0 29.-1 - fi 9“ n.. 184 Dunc-mu 51".1" $11'fl1’1(19‘15 N1 11“ I1 I'm-dim” 1970 Census nu'upaliun (‘mlc Madc- \um's 'I'uml scum-s “J“. wem's 101:11 wares 672 03.8 01.1 28.2 673 05.9 25.0 25.1 674 06.1 08.6 28.3 680 24.0 40.1 681 19.6 21.7 32.0 33.6 690 19.0 28.-1 28.5 692 19.3 19.3 29.3 694 19.2 18.8 29.1 28.9 695 19.2 19.5 29.1 701 24.0 36.8 703 24.0 21.0 32.4 704 32.5 28.0 705 31.0 28.2 28.3 706 16.8 16.8 28.4 28.-1 710 03.0 27.2 711 18.8 22.0 712 42.0 31.7 713 44.0 32.8 71-1 10.0 21.5 715 15.1 32.1 7-10 16.9 17.5 28.7 30.3 750 07.2 23.0 751 07.1 07.1 71.4 75" 10.6 30.3 773 08.7 08.9 18.8 ° 19.0 7'54 06.0 06.0 17.53 755 10.9 22.1 760 1 1.0 24.4 761 04.1 25.9 762 16.7 17.3 20.6 211.7 76.") 08.0 12.2 764 08.6 08.6 18.5 18.5 770 08:; 20,3 780 08.2 08.2 19.1 785 08.3 08.3 17.5 801 14.0 40.7 802 36.0 43.7 821 20.0 35.0 822 06,3 18.9 18.8 823 17.0 18.-1 18.5 824 22.0 26.8 901 13.-1 11.6 16.6 1-1. 1 902 07.8 09.8 18.-1 17.4 903 12.7 12.5 19.5 19.3 910 19.0 19.9 911 11.0 14.4 912 15.0 26.4 913 1 1.0 21.8 185 Duncan 8121" 811110111965 NOIKD l'n-stiul'" 1970 (701181" 011111101100 r0111- \1;|11' \t‘nn‘s T01..| \( 1111's .\|.|10 sum-\- T0141 wm‘t‘s 9| 1 17.0 15.1 1115 1611 2113 mm 1|.0 :09 20.8 1121 38.11 4:3 922 23.0 29 4 26.3 40.5 92.1 51.0 45.1 921 .' 7.0 2.1.3 925 13.7 151.5 36.8 35.-1 926 22.0 -1 1.9 911 11.0 16.1 9.12 19.1 19,3 15.6 161 9:13 26,3 288 21.7 22.5 111-1 (17.8 117.11 17.5 9515 7.0 37.9 910 80.0 22 1 1111 118.0 1111.3 1112 28.2 24.11 24.1 94:3 10.0 20.9 911 17.0 33.2 91') :1 | .0 10.8 11.511 .11 .11 36.-1 952 26.0 14.1 95.3 “.5.0 14.9 931 11.11 14.4 960 17.9 17.5 25.3 24 5 961 . 7.0 43.8 962 18.2 22.2 22.3 . 963 21.0 45.8 961 40.5 40.4 47.7 965 34.0 55.0 980 07.0 22.6 22.9 981 (17.0 18.0 982 10.7 18.6 20.1 24.7 98.3 12.0 17.6 984 07.0 18.0 052 80.0 67.2 07:1 58.0 36.7 (1112 81.11 71.4 (194 81.0 71.4 131 72.0 533.2 156 53.0 46.8 280 49.4 35.4 311 44.0 31.4 475 27.0 40.8 775 07.9 17.5 782 07.6 19.1 795 07.9 - 17.5 186 Duncan 5151" Sit-110111905 .\'()11(I1 1’l'(°\‘11u('" 1970 (It-mus occupatinn cmlu Mule: sum-s Tuml SCHH'S 31.110 sum-s Tulul scurm 790 07.9 17.5 805 17.0 18.4 882 00.3 13.9 " '1'1uwc stun-5114111111w mulr culnpmilmn ”1 11w Hnlmn 1.:1mr 1011-1-1"Iu.11(- sum-C). sun'cs refit-dim: 11H.- cnmpmltinn "1' mm! pvrwm in 11w civn1i.m lulu" {uru' un- m.u’1u'(1 "10ml 50010.5” ("11) in tlmw cases where the: Lum- (111101‘ 1mm "male 50mm." X0 scuu'x amwalr 1m (1 u1c~ 580 (Armed Furcvs) or 995 (.\nt uwvrtaim'd). APPENDIX B Survey of Elementary School Principals 187 33.3325 335 53:32 sawumuzcm mo mamaaou bhmfi mCSh WEE—02.”?— JOOIUw >m<._.2m_>_m._m “.0 >w>m5m 62901 Patsy Hashey 1105 West Willow Street Carbondale, Illinois staple here zosmm xmumm be) .zfiwuwu m .msosmcocm mum momcoamwu HH< ”achHWficwww mum mumv use» ~50» no ucoum c« m« u“ mauzs so a“ vcmm can so: mwsu 0v so» u.c03 .mEHu uaoa mo mwuscfis on no cmu uaonm magnumu aaws mash .wwumwumunam afiummuw on Haw: newumuoaoou use} can mamaaucfium Hoozum zumucwsmam mo :Owumfiuomm< Hmcofiumz may a“ nunmuoneos scum vmuuoamm wumz no» .ouwmccofiumoav vmmoaucw 0:» cu wcqvcoammu a“ mam: use» kuc H .uwxuo: hwacmum uommwuoum mo uuamvwsw may umvca zuwmuw> Ian: ouuum cawasuqz um acuumuumwaqav< Hmcowuwuavm a“ oumuouuov u no mauxuoa .Homwo .mwocaaaH .mamvconumo .uouuum soflflfiz ”was mead we “mean: magma an H 1138 ~mcaafimu mg on: "mam: mmqmz ”mam: mA: 603:: 1.- 333. «3.3.! 2.9330 .6 ~26 go on . .0 3-3 1... 3.: 1.0 2.2 1... 3:31... :25 he nu 1.. 3w< .n can! .- 33m .N <93 .2209”: on .a 00> 1.. 52:06:.— uoozun raucoauqo e. no; 0.3 .n aeolian: n .cdunwlunox ooumov 6:30.38.— .3 3333.. 625300 .u 3.36 uouma: 60330.0 .0 00.306 032.03 603000.. 1.0 ~93: 235 030- 9350.3 «:02.» .3 332203 500:3 0300 16 «09.00 :3... 60003.03 1.... «oozun nu... 0.30 1.... 000." no 2.30 063» ....q :05: 25> .3 69.03.. 603-260 «0 #023 00232. amazon—z" 1.9—gm .33 van: a «a 9:530 . .u 88 3 000.02 :3... 3.... 1.. 000.0% 3 000.00 :33 52.2.1... 02...... 8 000.2 .33 :2... 1... 80.2 8 000.0 .56. :2315 a: .10..» 1..— auqueoc Jana» 1.. nU 3:093:60 ucouuao uo 8.50133 .0 003 m 23.32.. 0:01: .350 1.0 2 1... 2 1d 0. 1... 0 1.. «uuquucou no :0» 0: 05:0... .33. 30: 000.03 1... 2.0.2» 3 000:...» 1.. 08.2» 3 000.03 1.0 8......» 3 000...; 1... 30.2... 3 000.20 1.. 00...; 3 000.0» 1... 0006» :2.» 003 1.: undo» door—on :12: 05 new .33: audio... 0.3.. no 8°.nN 16 . 30.: 3 000... 1.. 80.0 3 8n 1... con :22 on».— 1.- 350.335 noun». ~85. .300... «:9...— So 026 16 03...... 00.. 3 000 16 5...... Sn 3 000 1... 0:93 08 .36.... 1.- 2.3332. .305 .36.... 38:00:02.: ...u .3 nan-:85 «30h 0.3.. .3 n 1.0 ~ 1... a 23393:. .30» .369... 3.2.93.5 floccu- 00 30!: I .2 33398 1.0 0.3.32..." .3332.» 1.0 ~38... 1... 03.2.0: 1.9 A930» 3 3.32.2 1... 03.06 .0033 o: 1.- .ndgnaoauuauun uuuau u. cannon ououuou vacuum «noznua an on an 00 00 at an on n.” 3 00 1... 006:: no.5 .0 6 .u .2 .3 an 1.- 333335%. an»: an .2 9:0» :0. .3 3 1.9 an.» «a 3 n 1.9 :00» n :15 anon 1.- 32.3500 ”oeso- Daueoguo an on :10» no 021:: 0.3. no 31.. 2 3 21.. 3 3 0 02.3.20... .350 1.. n 8 ~ hung—.0630 1...— u 3 a £09..» :0» 36 0.0-o» has! .5: "2:02.- 033... .0050 14 0333 1.x :3; ~39..- 1.n .0 6 .0 2.33 so» 36 325:1): a...) u< .n In» an 8 1... no» 1.: 3032:».— .Cnucoluuo .3 93802. :0qu 389.3 I so» out: .2 .2 .3 .GA .9 .< 8% is .0.— 191 0000... 00:000.. 1.0 3:0 00:00. 1.0 300 00:00... 1... 0000000 000.. 1 .. 2.0... 2... so. 3.. 0. .o 00. 2.0 0. .2 000.0330; 1.n - 0050.0 1.0 00x00... 00:00.00... 1.0 00003.0 .3 03000000 000:... 1.0 000.0 00 3000000 1.0 030.000... 00 00000000.... 1|... 000000000. 00 000000000000 .0 "00000000 3.3.0.0300 0.00.000: 200000 3.000000 0000...: 3 0:00. .0.) :0» 00.... 0.. 00:00... 000» 30 0.00... 00 3.0.. 00:?0000000000 0.0030: .3 .00 >0~00000 . u 00.00.. 1.0 000.00... 00:00.00: 1.0 3.00000 00 :3000000 003.... 1.0 00a... 00 00000000 1.0 03030000 00 3000000.... 1... 030000000 00 «0000000000.. 1.. 00.3300 0000. 00 c .00.... 320 02.3... 0 1.» ”P300000 000000000000 0.00:0...— 005300 n 00 N :03 3.30 0.2.3... 0 1.0 200000 33000 00.003 3 000... 000... 03:30 0030 000... 00 .0 :00... 320 0000.30» 0:0 1.0 .3." 00..) 00 00.0000 0:05 03 1.33 00 .30.. 00070000000000 0.00.00... .ou 00:00.... 0030 n 3 .— 000... 30.00 0000.50» 0:0 1... 000000 0.0300300.— 00 00.30000 0030000 1.0 003300 009.000 0000 00 a :0.) 320 3030 0:0 1.0 000000 0000.: 0030000 1.0 0003...... 009.00» n 3 a :00: 30:0 00030 0:0 1... 000000 030.00.. 00.50000 1.0 320 .200 0. 1.0 30:00 :00... 0000. 0:05.00 «.0000. 00 .00003000 .000300 030 1.0 .00» 000: H0300 :02 3003.000 1.0 20000000 0.3» >0 00000... 0003050 .3030. .30.— 030 1.5 .3. .0000000 00:30 000300010000 0. :0... .03.. 0:0» 3 00>: 0..) . 0..." 00 0.0.0. 0.0.0. 1.. 000330 02.00 000. 003000 .003: 00000000 0.... 00:. 003 009.0 . 3. .3 .30. 0000.0. 00. 0005000.. 00:00. 000000. 0... 002.0000 .2 «00.000. 000» up 00:00.0 0000.00.00 00 30..- 003300 .2 192 Pat Hashey 1105 West Willow Street Carbondale, Illinois 62901 APPENDIX C Follow-up Postcard 193' "AhuHovnm madeqv uczpo .o vacuum» HHHD H oco ccOm :0» HH Hommw AH .decconudo pap .vuwwconpmozd ha vm>Hmovu uo>oc\umoH H .c vacuum 30HHH3 any: mOHH hogan: .H huuum moquccoHpmvsv hw>uam ou vacancy on poc uououa H .o chuocHun Hoosou hudvcoaoHo ad hom:0H oc\uo: an H candoon uncanny no: ch H .p vacancy ch H .u "HHds as» :« wade mHnu noun and mac gonzo ondoam . vmnmk v. mmmrmk HHOm swan SHHHoc madam” . .. anacosappa. Hr ammop “awnuuwnnmul .I I!" o 1.., .l'ct . ‘ . 2.37:. ,‘I- » I.-. II mall. -:I d . ms 22.99.»? mg 0p pH cuspmu can came omnowpud mg» on unoawmu on opscHs a mxdp so» vH=o3 .noHuprmmmHo 9.5m as 90 coHpmHQEoo map you pcdofimwcmwm mud «paw use» .owohoomu no: we: omcoamwu use» van deQHocHum hhwpcmEva mo hm>hsm cmHHao gosssm ude oquccoHpmmza a 50» cwHHds H .afiocaaaH .oauoaopuao .pooupm soaaflz moHH .hvnmdm .H hmvcm an H woe umpEoewm ”Hm - '\\_ APPENDIX D Frequency Distributions of Descripter Characteristics Table 21: were You A Teacher Before Becoming An Elementary Principal? 194 Teacher .Male Female Total n=536 n=161 =697 Yes 533 (99.44%) 161 (100%) 694 (99.57%) No 2 ( 0.37%) 0 (0%) 2 ( 0.29%) Total 535 (76.76%) 161 (23.10%) 696 (99.86%) Table 22: At What Level/levels Did You Teach Leve1(s) of Male Female Total Teaching* n=536 n=161 N=697 Elem. 360 (67.17%) 116 (72.05%) 475 (68.15%) Jr. High 185 (34.52%) 47 (29.19%) 232 (35.29%) Sr. High 105 (19.59%) 13 ( 8.08%) 118 (16.93%) College 10 ( 1.87%) 7 ( 4.35%) 17 ( 2.44%) Other 8 ( 1.49%) 6 ( 3.73%) 14 ( 2.01%) Table 23: Number of Years As An Elementary School Principal Years a Phle Female Total Principal n=536 n=161 N=697 5 years 101 (18.84%) 53 (32.92%) 154 (22.10%) 5-14 years 284 (52.99%) 77 (47.83%) 361 (51.79%) 15 or more 151 (28.17%) 31 (19.26%) 182 (26.11%) Total N = 697 536 (76.90%) 161 (23.10%) 697 (100.0%) Table 24: Highest Earned College Degree Highest Male Female Total n=536 n=161 Né697 No college 0 0 0 Associate 0 0 0 Bachelor 7 ( 1.31%) 4 ( 2.48%) 11 ( 1.58%) Master 404 (75.38%) 134 (83.23%) 538 (77.19%) Educ Spec 102 (19.03%) 19 (11.80%) 121 (17.36%) Doctorate 19 ( 3.55%) 2 ( 1.24%) 21 ( 3.01%) Total 532 (76.33%) 159 (22.81%) 691 (99.14%) *Respondents were allowed to check more than one category, therefore, percents do not equal 100. 195 Table 25: Area of Specialization (highest degree held) Area Mhle Female Total n=536 n=161 N=697 Elem Educ 107 (19.96%) 44 (27.33%) 151 (21.66%) Sec Educ 6 ( 1.12%) 1 ( 0.62%) 7 ( 1.00%) Supervision/ 380 (70.90%) 95 (59.01%) 475 (68.15%) Educ Admin Counseling 10 ( 1.87%) 7 ( 4.35%) 17 ( 2.44%) A.Content Area 7 ( 1.31%) 3 ( 1.86%) 10 ( 1.44%) Other 22 ( 4.10%) 9 ( 5.59%) 31 ( 4.45%) Tbtal 532 (76.33%) 159 (22.81%) 691 (99.14%) Table 26: NUmber of Schools Currently under Your Direction Schools Male Female Total n=536 n=161 Né697 1 444 (82.84%) 129 (86.34%) 583 (83.64%) 2 65 (12.13%) 19 (11.80%) 84 (12.05%) 3 or more 25 ( 4.66%) 2 (1.24%) 27 ( 3.87%) TOtal 534 (76.61%) 160 (22.96%) 694 (99.57%) Table 27: Age At First Principalship Age Male Female Total n=536 n=161 N=697 25 or less 43 ( 8.02%) 5 (_3.11%) 48 ( 6.89%) 26 - 35 368 (68.66%) 59 (36.65%) 427 (61.26%) 36 - 45 104 (19.40%) 63 (39.13%) 167 (23.96%) 46 - 55 19 ( 3.55%) 31 (19.26%) 50 ( 7.17%) 56 or more 1 ( 0.19%) 2 ( 1.24%) 3 ( 0.43%) Total 535 (76.76%) 160 (22.96%) 695 (99.71%) 196 Table 28: Total Enrollment In The School/schools Under Direction Enrollment Mle Female Total n=536 n=161 N2697 Uhder 200 30 ( 5.60%) 16 ( 9.94%) 46 ( 6.60%) 200 - 399 176 (32.84%) 65 (40.37%) 241 (34.58%) 400 - 599 225 (41.98%) 48 (29.81%) 273 (39.17%) Over 600 103 (19.22%) 30 (18.63%) 133 (19.08%) Total 534 (76.61%) 159 (22.81%) 693 (99.43%) Table 29: Total School System Enrollment Enrollment IMale Female Total n=536 nel6l N-697 Under 300 5 ( 1.83%) 7 ( 4.35%) 12 ( 1.72%) 300 - 2,999 192 (35.82%) 44 (27.33%) 236 (33.86%) 3,000 - 24,999 281 (52.43%) 76 (47.21%) 357 (51.23%) 25,000 or more 55 (10.26%) 31 (19.26%) 86 (12.34%) _ Total 533 (76.47%) 158 (22.67%) 691 (99.14%) Table 30: Regular Salary For The 1976-77 School Year Salary Male Female Total n=536 n=161 N=697 Under $8,000 0 20 (12.42%) 20 ( 2.87%) $8,000 - 11,999 5 ( 0.93%) 7 ( 4.35%) 12 ( 1.72%) $12,000 - 15,999 34 ( 6.34%) 15 ( 9.32%) 49 ( 7.03%) $16,000 - 19,999 168 (31.34%) 39 (24.22%) 207 (29.70%) $20,000 — 23,999 197 (36.72%) 44 (27.33%) 241 (34.58%) $24,000 - 27,999 105 (19.59%) 25 (15.55%) 130 (18.65%) $28,000 or more 22 ( 4.10%) 10 ( 6.21%) 32 ( 4.59%) Total 531 (76.18%) 160 (22.96%) 691 (99.14%) 197 Table 31: How Many Months Are You On Contract? Months Male Female Total n-536 n=161 Na697 9 10 ( 1.87%) 8 ( 4.97%) 18 ( 2.58%) 10 265 (49.44%) 88 (54.66%) 353 (50.65%) 11 132 (24.63%) 30 (18.63%) 162 (23.24%) 12 97 (18.10%) 27 (16.77%) 124 (17.79%) Other 29 ( 5.41%) 8 ( 4.97%) 37 ( 5.31%) Total 533 (76.47%) 161 (23.10%) 694 (99.57%) Table 32: State of Employment State Male Female Total Original n=536 n=161 N=697 N2977 Illinois 55 (10.26%) 25 (15.63%) 80 (11.48%) 125 (12.79%) Indiana 68 (12.69%) 9 ( 5.63%) 77 (11.05%) 111 (11.36%) Iowa 37 ( 6.90%) 12 ( 7.50%) 49 ( 7.03%) 77 ( 7.88%) Kansas 31 ( 5.78%) 7 ( 4.38%) 38 ( 5.45%) 49 ( 5.02%) Nfichigan 113 (21.08%) 37 (23.13%) 150 (21.52%) 212 (21.70%) Nfinnesota 38 ( 7.09%) 11 ( 6.88%) 49 ( 7.03%) 68 ( 6.96%) Nfissouri 39 ( 7.28%) 10 ( 6.25%) 49 ( 7.03%) 61 ( 6.24%) Nebraska 21 ( 3.92%) 7 ( 4.38$) 28 ( 4.02%) 30 ( 3.07%) Nerth Dakota 10 ( 1.87%) 1 ( 0.63%) 11 ( 1.58%) 17 ( 1.74%) Ohio 64 (11.94%) 24 (15.00%) 88 (12.63%) 134 (13.72%) South Dakota 8 ( 1.49%) 3 ( 1.88%) 11 (1.58%) 16 ( 1.64%) ‘Wisconsin 48 ( 8.96%) 12 ( 7.59%) 60 ( 8.61%) 77 ( 7.88%) Total 532 (76.33%) 158 (22.67%) 690 (99.00%) 977 (100.0%) APPENDIX B Frequency Distributions of Background Characteristics 198 Table 33: Father's Occupational Category Category Male Female Total n=536 nel6l N=697 Professional 48 ( 8.96%) 12 ( 7.45%) 60 ( 8.61%) Managerial 67 (12.50%) 29 (18.01%) 96 (13.77%) Clerical 41 ( 7.65%) 21 (13.04%) 62 ( 8.90%) Skilled 172 (32.09%) 38 (23.60%) 210 (30.13%) Unskilled 90 (16.79%) 19 (11.80%) 109 (15.64%) Farmer 104 (19.40%) 41 (25.47%) 145 (20.80%) Unemployed 3 ( 0.56%) 0 3 ( 0.43%) Tbtal 525 (75.32%) 160 (22.96%) 685 (98.28%) Table 34: MOther's Occupational Category Category .Male Female Total n=536 n=161 =697 Professional 43 ( 8.02%) 13 ( 8.08%) 56 ( 8.03%) Managerial 11 ( 2.05%) 4 ( 2.48%) 15 ( 2.15%) Clerical 41 ( 7.65%) 19 (11.80%) 60 ( 8.61%) Skilled 13 ( 2.43%) 0 l3 ( 1.87%) Unskilled 30 (p 5.60%) 10 ( 6.21%) 40 ( 5.74%) Farmer 10 ( 1.87%) 3 ( 1.86%) 13 ( 1.87%) Unemployed 380 (70.90%) 111 (68.95%) 491 (70.45%) Tbtal 528 (75.75%) 160 (22.96%) 688 (98.71%) Table 35: Higher Level of Education Reached By Your Father Father's Educ. JMale Female Total n=536 n=161 Ne697 Grade 8 or less 246 (45.90%) 65 (40.37%) 311 (44.62%) Some High School High School Some College Bachelor Master Doctor/Prof Total 88 (16.42%) 91 (16.98%) 60 (11.19%) 20 ( 3.73%) 14 ( 2.61%) 15 ( 2.80%) 534 (76.61%) 26 (16.15%) 34 (21.21%) 21 (13.04%) 10 ( 6.21%) ( 1.24%) ( 1.24%) 160 (22.96%) NN 114 (16.36%) 125 (17.93%) 81 (11.62%) 30 ( 4.30%) 16 ( 2.30%) 17 ( 2.44%) 694 (99.57%) 199 Table 36: Highest Level of Education Reached By Your Mother Mother ' s Educ Male Female Total n=536 n=161 N=697 Grade 8 or less 155 (28.92%) 46 (28.57%) 201 (28.84%) Some High School 112 (20.90%) 28 (17.39%) 140 (20.09%) High School 147 (27.43%) 43 (26.71%) 190 (27.26%) Some College 73 (13.62%) 30 (18.63%) 103 (14.78%) Bachelor 37 ( 6.90%) 9 ( 5.59%) 46 ( 6.60%) IMaster 7 ( 1.31%) 3 ( 1.86%) 10 ( 1.43%) Doctorate/Prof 1 ( 0.19%) 1 ( 0.62%) 2 ( 0.29%) Tbtal 532 (76.33%) 160 (22.96%) 692 (99.29%) Table 37: Sex Sex IMale n=536 Female n=161 Total N=697 536 (76.90%) 161 (23.10%) 697 (100.00%) Table 38: .Age Age Male Female Total n=536 n=161 N=697 25 or less 2 ( 0.37%) 0 2 ( 0.29%) 26 - 35 101 (18.84%) 20 (12.42%) 121 (17.36%) 36 - 45 196 (36.57%) 43 (26.71%) 239 (34.29%) 46 - 55 169 (31.53%) 54 (33.54%) 223 (31.99%) 56 or more 57 (10.63%) 39 (24.22%) 96 (13.77%) Total 525 (75.32%) 156 (22.38%) 681 (97.70%) 200 Table 39: .At The Age of 16 Did YOu Live With Residence ‘Male Female Total n=536 n=161 N=697 Both 474 (88.43%) 141 (87.58%) 615 (88.24%) Mother Only 38 ( 7.01%) 15 ( 9.32%) 53 ( 7.60%) Father Only 14 ( 2.61%) 0 14 ( 2.01%) Neither 7 ( 1.31%) 5 ( 3.11%) 12 ( 1.72%) Total 533 (76.47%) 161 (23.10%) 694 (99.57%) Table 40: Brothers and Sisters Number Siblings Male Female Total Sibling Placement n=536 n=161 Nh697 Only Child 46 ( 8.58%) 11 (_6.83%) 57 ( 8.18%) Oldest of 1-3 138 (25.75%) 46 (28.57%) 184 (26.40%) Oldest of 4 or more 26 ( 4.85%) 7 ( 4.35%) 33 ( 4.74%) YOungest of 1-3 109 (19.40%) 20 (12.42%) 124 (17.79%) Youngest of 4 or 32 ( 5.97%) 14 ( 8.70%) 46 ( 6.60%) more Middle of 2-3 82 (15.30%) 29 (18.01%) 111 (15.93%) Middle of 4 or 105 (19.59%) 32 (19.88%) 137 (19.66%) more Tbtal 533 (76.47%) 159 (22.81%) 692 (99.28%) APPENDIX F Frequency Distributions of Intervening Characteristics Table 41: Marital Status Status IMale Female Total n=536 n=161 Né697 Married w/ Spouse 494 (92.16%) 76 (47.20%) 570 (81.78%) iMarried w/o Spouse 2 ( 0.37%) 1 ( 0.62%) 3 ( 0.43%) ‘Widowed 5 ( 0.93%) 10 ( 6.21%) 15 ( 2.15%) Divorced 10 ( 1.87%) 16 ( 9.94%) 26 ( 3.73%) Never Married 22 ( 4.10%) 56 (34.78%) 78 (11.19%) Tbtal 533 (76.47%) 159 (22.81%) 692 (99.28%) Table 42: Do You Have Children Children Nhle Female Total n=536 n=161 Né697 Yes 474 (88.43%) 85 (52.80%) 559 (80.20%) No 57 (10.63%) 74 (45.96%) 131 (18.80%) Tbtal 531 (76.18%) 159 (22.81%) 690 (99.00%) Table 43: If You Have Children, HOW Many Number Children Male Female Total n=536fv n=161 Na697 l - 3 383 (71.46%) 71 (44.10%) 454 (65.14%) 4 - 6 79 (14.74%) 14 ( 8.79%) 93 (13.34%) More Than 6 11 ( 2.05%) O 11 ( 1.58%) Total 473 (67.86%) 85 (12.20%) 558 (80.06%) Table 44: Within Which Age Range Does Your Youngest Child Fall Age Youngest Male Female Total n=536 n=161 N=697 under 6 86 (16.50%) 4 ( 2.48%) 90 (12.91%) 6 - 18 223 (41.60%) 24 (14.91%) 247 (35.44%) Over 18 95 (17.72%) 47 (29.19%) 142 (20.37%) Total 404 (57.96%) 75 (10.76%) 479 (68.72%) Table 45: HOW Many Years Did You Teach NUmber Years iMale Female Total n=536 n=161 N=697 0 - 1 6 ( 1.12%) 1 ( 0.62%) 7 ( 1.00%) 2 - 5 188 (35.08%) 21 (13.04%) 209 (30.00%) 6 - 10 237 (44.22%) 46 (28.57%) 283 (40.60%) 11 - 15 71 (13.25%) 44 (27.33%) 115 (16.59%) 16 or more 21 ( 3.92%) 48 (29.81%) 79 (11.33%) Total 523 (75.04%) 160 (22.96%) 693 (99.43%) Table 46: Highest Earned College Degree at First Principalship Degree mle Female Total n=536 n=161 N=697 No College 3 ( 0.56%) 7 ( 4.35%) 10 ( 1.44%) Associate 4 ( 0.75%) 0 4 ( 0.57%) Bachelor 87 (16.23%) 32 (19.88%) 119 (17.07%) IMaster 421 (78.54%) 111 (68.94%) 532 (76.33%) Educ Spec 16 ( 2.99%) 9 ( 5.59%) 25 ( 3.59%) Doctorate 5 ( 0.93%) 1 ( 0.62%) 6 ( 0.86%) Total 536 (76.90% 160 (22.96%) 696 (99.86%) Table 47: Size of City/town Of Current Employment Size Male Female Total n=536 n=161 N=697 Rural NOnfarm. 20 ( 3.73%) 5 ( 3.11%) 25 ( 3.59%) Rural Farm. 71 (13.25%) 13 ( 8.08%) 84 (12.05%) Small wan 158 (29.48%) 30 (18.63%) 188 (26.97%) Small City 105 (19.59%) 24 (14.91%) 129 (18.51%) Medium City 84 (15.67%) 35 (21.74%) 119 (17.07%) Large City 42 ( 7.84%) 27 (16.77%) 69 ( 9.90%) suburb 53 ( 9.89%) 26 (16.15%) 79 (11.33%) Total 533 (76.47%) 160 (22.96%) 693 (99.43%) APPENDIX G Definition of Categories of Elementary School Principals' Descripter Characteristics 203 Table 48: Definition of Categories of Elementary School Principals' Descripter Characteristics Variable Original Recoded Definition of Categories Categories Recoded Categories N Age at First 25 or less 1 35 or under 475 Principalship 26 - 35 36 - 45 2 36 or over 222 46 - 55 56 or over Total Enrollment Less than 300 1 Less than 3000 248 (school system) 300 - 2,999 3,000 - 24,999 2 IMore than 3000 443 25,000 or more Total Enrollment Less than 200 1 Less than 400 287 (under direction) 200 - 399 400 — 599 2 IMore than 400 406 Over 600~ Salary Less than $8,000 1 Less than $16,000 81“"I $8,000 - 11,999 $12,000 ~ 15,000 $16,000 - 19,999 2 $16,000 - 19,999 207 $20,000 - 23,999 3 $20,000 - 23,999' 241 $24,000 - 27,999 4 $24,000 or more 168 $28,000 or more *Category contains fewer than 20% of the total sample Table 48 (cont'd) variable Original Categories 204 Recoded Categories Definition of Recoded Categories N Contract 9 months 10 months 1 10 months or less 371 11 months 12‘months Other*** 11 months or more, and Other 326 State Illinois Indiana Michigan Ohio East of Mississippi River 395 Iowa Kansas Nfinnesota Mfissouri Nebraska North Dakota South Dakota Wisconsin West of Mississippi River 295 Level/levels of Elementary Elementary only 274 Teaching Junior High/ JMiddle Senior High College Other iMore than one level 423 ***Although the meaning of "other" as a category was unclear, it was determined preferable to retain the 37 individuals in that category fer analysis. Table 48 (cont'd) 205 variable Original Recoded Definition of Categories Categories Recoded Categories N Years a Principal Less than 5 1 Less than 5 154 5 - 14 2 5 - 14 361 15 or more 3 15 or more 182 Highest College No College 1 Master or less 549**f Degree Associate Bachelor ster Educ Spec 2 Over a master‘s 148* Doctorate Area of Elementary Educ 1 Other 222 Specialization Secondary Educ Counseling/ Guidance Content Area Other Supervision/Ed 2 Supervision/Ed 475 Admin Admin Number Schools 1 1 1 583** Under Direction 2 2 More than 1 111* 3 or more *Category contains fewer than 20% of the total sample **Category contains more than 80% of the total sample LIST OF REFERENCES LIST OF REFERENCES Acker, Joan, "WOmen and Social Stratification: .A Case of Intellectual Sexism”, American Journal of Sociology, V01. 78, No. 4, p. 937, pp. 936-945. Betz, Michael, "Inter-Generational Mobility Rates of Public School Teachers", The Journal of Educational Research, Vol. 67, No. 1, September, 1973. Blalock, H. M. et al (ed.,) Quantitative SociologL International Perchtives on Mathemat1ca1 and Statistical Mofiing, New York: Academic Press Inc. , 1975. Blau, Peter M. and Otis Dudley Duncan, The American Occupational Structure, New York: John Wiley 5 Sons, Inc., 1967. Bonelli, G., "Tree - Analysis-*The Method by Sonquist and Morgan", pp. 465-472, in H.M. Blalock et al (ed.), giantitative Sociology: International Perspectives on Mathematical and Statistical lvbdeling, New York: AcadEmic Press, Inc., 1975. Boudon, Raymond, Mathematical Structures of Social Mobi1i_ty, San Francisco: Jossey- -,Bass Inc., Publishers, 1973. Chase, Ivan D. , "A Comparison of Men's and Women's Intergenerational Mobility in the United States", American Sociological Review, Vol. 40, August, 1975, pp. 483-505. Counts, George S. , "The Social Status of Occupations: A Problem in Vocational Guidance", The Social Review, Vol. 33, January, 1927. DeJong, Peter Y., Milton J. Brawer and Stanley S. Robin, "Patterns of Female Intergenerational Occupational Mobility: A Comparison With Male Patterns of Intergenerational Occupational Mobility", Amirican Sociological Review, Vol. 36, December, 1971, pp. 1033- 10 2. Duncan, Otis Dudley, Introduction to Structural Equation Models, New York: Academic Press, 1975. Duncan, Otis Dudley, "Methodological Issues in the Analysis of Social Mobility", in Neil J. Smelser and Seymour Martin Lipset (ed.), Social Structure and Mobilgy in Economic Development, Chicago: Aldine Publishing Company, 1966, pp. 51-97. Duncan, Otis Dudley, "Path Analysis: Sociological Examples", The American Journal of Sociology, Vol. 72, No. 1, July, 1966, pp. 1- 16. 206 207 Duncan, Otis Dudley, David L. Featherman and Beverly Duncan, Socioeconomic Background and.Achievement, New York: Seminar Press, 1972. Duncan, Otis Dudley and Robert W. Hodge, "Education and Occupational Mobility: A Regression Analysis", The American Journal of Sociology, Vol. 68, No. 6, May, 1963. Falk, William W. and Arthur B. Cosby, "Women and the Status Attain- ment Process: A Working Paper", A paper presented at the annual meeting of the Rural Sociological Society, Montreal, August, 1974, ERIC Abstract ED097237. Featherman, David L. and Robert M. Hauser, "Design for a Replicate Study of Social Mobility in the United States", in Kenneth Land and Spilerman (ed.), Social Indicator Models, New York: Russell Sage Foundation, 1975. Featherman, David L. and Robert M. Hauser, "Prestige or Socioeconomic Scales in the Study of Occupational Achievement?", Sociological Methods and Research, Vol. 4, No. 4, Why, 1976, pp. 403-422. Featherman, David L. and Robert M. Hauser, "Sexual Inequalities and Socioeconomic Achievement in the U.S., 1962-1973", American Sociological Review, Vol. 41, June, 1976, pp. 462-483. Glass, Gene V. and Julian C. Stanley, Statistical Methods in Education and P5 cholo , Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1970. Grasmick, Harold G. , "The Occupational Prestige Structure: A Milti— dimensional Scaling Approac ", The Sociological Quarter_ly, Vol. 17, Winter, 1976, pp. 90-108. Gross, Neal and Anne E. Trask, Sex Factor and the Management of Schools, New York: John Wiley 6 Sons, 1976. Haug, Marie R. and Harold A Widdison, "Dimensions of Occupational Prestige", Sociology of Work and Occgmtions, Vol. 2, No. 1, February, 1975. Hauser, Robert M. and David L. Featherman, The Process of Stratification: Trends and Analysis, New York: Academic Press, 1977. Health, Education, and Welfare, The Condition of Education, Washington, D.C.: National Center for Education Statistics, Education Division, 1975. Jencks, Christopher et a1, Inequality: A Reassessment of the Effect of Family and Schooliig in America, New York: Harper 8 Row Publisl'Ters ,’ 1972 . Littlejohn, James, Social Stratification, London: George Allen 8 Unwin, Ltd., 1972. 208 Lorenz, Jerome R. , "A Methodology for the Development of Empirically Based Differential Service Patterns for Clients in Rehabilitation Facilities", Doctor's dissertation, University of Wisconsin- Madison, 1973. Dissertation Abstracts International, 1974, 34 (10), 5171B (Imiversity—MicroTilms No. 74-3533). Matras, Judah, Social Inequality, Stratification, and Mobility, Englewood Cfiffs, New Jersey: Wentice-Hall, Inc., 1975. McClendon, McKee J. , "Structural and Exchange Components of Vertical Mobility", American Sociological Review, Vol. 42, February, 1977, pp. 56-74. McClendon, McKee J. , "The Occupational Status Attainment Processes of Males and Females", American Sociological Review, Vol. 41, February, 1976. McIntosh, Janet, "Differences Between Women Teachers Who Do and Who Do Not Seek Promotion", The Journal of Educational Administration, Vol. 12, No. 2, October, 1974. McNeil, Keith A. , Francis J. Kelly and Judy T. McNeil, Testing Research Hypgtheses Using Multiple Linear Rgression, Carbondale: Southérn Illinois University Press, 1975. Meskin, Joan D. , "The Performance of Women School Administrators - A Review of the Literature, Administrator's Notebook, Midwest Administration Center, TheTIniversity of ChicagoTVol. 23, No. 1, 1974. NEA Research Division, Elementag School Principalship in 1968, National EducationTssociat ion, Department of Elementary School Principals‘, Washington, D.C., 1968. Nie, Norman H. et a1, Statistical Package for the Social Sciences, New York: McGraw—Hill Book Company, 1970. Pavalko, Ronald M., Sociolo of Occmations and Professions, Itasca, Illinois: F. E. eacoc TuEliEhers, Inc., 1971. Pharis, William L. and Edward Keller, "Bucks, Benefits, and Bargaining: The Big Picture", The National Elementary Principal, Vol. 57, No. 3, March, 1978, pp. 22—27. Pohlmann, John T. and James F. Moore, "Interval Estimation of the Population Squared Multiple Correlation", Multiple Linear Regression Viewpoints, Vol. 8, No. 1, August 197 , pp. 1 -31. Ramsgfiy, Natalie Rogoff, "Patterns of Female Intergenerational Occupa- tional Mobility: A Comment", American Sociological Review, v61. 38, No. 6, December, 1973, pp. 806-807. 209 Reichardt, Robert and B. Schmeikal, "Theoretical Considerations and Simulation Models Related to the Method of Sonquist and Morgan", pp. 451-465, in H. M. Blalock et a1, Quantitative Sociology: International Perspectives on Mathematical andDStatistical IModeling, Nengork: .Academic Press, Inc., 1975. Reiss,.Albert J., Jr. et a1, Occupations and Social Status, New York: Free Press of Glencoe, 1961. Rogoff, Natalie, Recent Trends in Occupational Mobility, Glencoe, Illinois: The Free Press, 1953. Rosenfeld, Rachel, "W0men's Intergenerational Occupational Mobility", University of Wisconsin-Madison, Center for Demography and Ecology, CDE working Paper 75-28, 1975. Sewell, William H. and Robert M. Hauser, Education, Occupation and Earnings, New York: Academic Press, 1975. Siegel, Paul M., "Prestige in the American Occupational Structure", Unpublished doctoral dissertation, university of Chicago, 1971. Sonquist, John A., Elizabeth Lauh Baker and James N. Morgan, Searching f0r Structure, Ann Arbor, Michigan: Survey Research Center, 1974. Sonquist, John A., Mbltivariateledel Building: The Validation of a Search Strate , Ann Arbor, MiEhigan: Institute for SoCial Research, 1970. Sonquist, John A. and William C. Dunkelberg, Survey and Opinion Research: Procedures for Processing and Analysis, Englewood'Cliffs, New Jersey: Prentice-Hall, Inc., 1977. S¢rensen, Aage B. , "Growth in Occupational Achievement: Social Mobility or Investment in Human Capital", in Kenneth C. Land and.Seymour Spilerman (ed.), Social Indicator M0dels, New York: Russell Sage Foundation, 1975a. S¢rensen, Aage B¢ttger, "Models of Social Mobility", Social Science Research, V01. 4, 1975b, pp. 65-91. Sorokin, Pitirim, Social Mobility, New York: Harper E. Brothers, 1927. Treiman, Donald J., "Problems of Concept and Measurement of the Comparative Study of Occupational Mobility", Social Science Research, V01. 4, 1975, pp. 183-230. Treiman, Donald J. and Kermit Terrell, "Sex and the Process of Status .Attainment: .A Comparison of W0rking'W0men and Men", American Sociological Review, V01. 40, April, 1975, pp. 174-200. Tyree, Andrea and Judith Treas, "The Occupational and Marital Mobility of W0men", American Sociological Review, V01. 39, N0. 3, June, 1974, pp. 293-302. 210 Warwick, Donald P. and Charles A. Lininger, The Sample Survey: Theory and Practice, New York: McGraw-Hill Book Company, 1975. White, Harrison C. , Chains of fipgrtunity: System Models of Mobilig in Organizations, Caflaridge: Harvard University Press, 1970. Wolf, Wendy Carolyn, Occgmtional Attainments of Married Women: Do Career ContingenEies Matter7, Center for Demography and Ecology, University of Wisconsin-Madison, CDE Working Paper 76-3, 1976. Wright, Burton, II, John Weiss and Charles M. Unkovic, Perspective: An Introduction to Sociology, Hillsdale, Illinois: The Dryden Press, 1975.