HUIWillllllltIll(HillHllHI!(mINIIHIIUIIIUHIUW 31293 10406 6208 OCCUPATIONAL AS: OF Cathe This dissertatio: 2:31 and occupational i 3?::tance individuals a‘. 3.”; b; ' - 491 r corresponding I as": “H After reviewing 1 «’1‘ ABSTRACT OCCUPATIONAL AND ORGANIZATIONAL DETERMINANTS OF BEHAVIOR AT WORK BY Catherine Begnoche Smith This dissertation examines the role that organiza- tional and occupational variables play in determining the importance individuals attribute to their work involvement and their corresponding level of behavioral dedication to work. After reviewing the literature concerning work motivations, work settings and organizational structures, several propositions were presented for empirical assessment. These propositions centered around the thesis that perceived potential for reward at work directly affects individuals' levels of self-investment in work, which in turn determines extent of dedication to work. Furthermore, the greater the perceived potential for reward at work, the greater the probability that workers will quit only to enter more rewarding jobs. Those lacking self-investment in work were exPected to quit for random, non-career reasons. Tying these ideas into research on organizational and occupational factors, the general propositions were modified to suggest Lutsone aspects of is , meme upon workers The study used Zileleuentary and se. for occupational categ: settings could be compa: :rgnizational records is ;:;:sition-testing. In adapting the gilic education employe wary personnel face 5. iA ‘L . fl “5 Moan did teachers 1' ‘. u‘ in ' or». According to 2:14 ' - uEXhlbit greater < list a; «n those categ. user self-invest: Catherine Begnoche Smith that some aspects of formal structure should have great influence upon workers' perceptions of reward potential. The study used a sample of teachers and principals in 53 elementary and secondary schools in Michigan so that four occupational categories in a range of organizational settings could be compared and contrasted. Survey data and organizational records were used as sources of data for proposition-testing. In adapting the general propositions to the case of public education employees, it appeared that principals and secondary personnel faced a more advantageous reward struc- ture than did teachers in general or elementary personnel of any type. According to the prOpositions, then, principals should exhibit greater self-investment in work than teachers and, within those categories, secondary personnel should have higher self-investment in work than elementary per- sonnel, with related differences in level of dedication to work. Intra-occupationally, the propositions suggested that schools with higher scores on hierarchy, specializa- tion, use of universalistic criteria of evaluation and lower scores on centralization of authority should encourage gTeater self-investment and dedication in their employees. The propositions relating perceived potential for reward at work to both self-investment in work and dedica- tion received considerable support from the data comparing ‘the four occupational groups. However, intra-occupational analysis discriminating very finely among organizational cattems produced wea tie propositions. W5: axes were simply too be answered by furthex tainfer bases for jo‘: :tility patterns coul :rscositions. The prc from the fact that se' hiatior were tested tn". aster sex of worker satisfaction eXplaine Catherine Begnoche Smith patterns produced weaker and less consistent support for the propositions. Whether the intra—occupational differ— ences were simply too small to show much effect can only be answered by further research. Since it proved impossible to infer bases for job-leaving, the apparent differences in mobility patterns could not adequately test the relevant propositions. The propositions received indirect support from the fact that several competing explanations of work behavior were tested simultaneously and found lacking. Neither sex of worker, age of worker nor level of job satisfaction explained patterns of variation in self- investment or dedication to work as well as the structural explanation, although age was related to participation in union activity. This research makes a contribution to the subject of work motivation by showing that bureaucratic organization does not necessarily discourage involvement in work, even for professional workers. This study tested a new perspec- tive on the bases of variations in work behavior, rather than relying on the problematic and unconvincing theories cmfered by researchers concerned with the impact of job satisfaction or variations in human relations techniques. Instead, this dissertation offers a more complete explica- tion of the process of involvement in work and presents evidence that the structure of opportunity at work, as Shaped by occupational and organizational factors, may have 1“«31230: consequences for worker perceptions and behavior. OCCUPATIONAL .12" Y" in h (D Cathe . M. In partial :3: OCCUPATIONAL AND ORGANIZATIONAL DETERMINANTS OF BEHAVIOR AT WORK BY Catherine Begnoche Smith A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1976 There are sever - assistance helped t Dr. Philip “arc tardify the original it‘sting pool of data. literial, as well as hi collection of audit extend ACKNOWLEDGMENTS There are several individuals whose encouragement and assistance helped to bring this project to completion. Dr. Philip Marcus, dissertation chairman, helped me to modify the original research plan to take advantage of an existing pool of data. I greatly appreciate the use of this material, as well as his counsel and encouragement during the collection of additional data and during the tribulations of extended analysis. Dr. William Faunce was of great assistance in the formulation of the research propositions, while Dr. Harry Perlstadt and Dr. John Gullahorn evaluated early drafts of the proposal and research reports. The suggestions and criticisms of these three committee members were invaluable. For assistance in technical areas, I thank Tom Nichol, who wrote the necessary computer programs, Gary ShePherd, who gave me access to the Michigan School Census data and Mark Rideout, who helped in the coding of union contracts. Also, since I lacked outside funds for analysis, the Computer time provided by the Sociology Department was 0 - . f major ass1stance. ii Finally, my hu; ably at each step of t‘: :terful assistance in tuning programs, check references and, in gen Finally, my husband, Bruce Smith, helped immeasur- ably at each step of the dissertation. I appreciate his cheerful assistance in such tasks as keypunching cards, Immning programs, checking calculations and bibliographic references and, in general, encouraging me to finish. iii IKXNOWLEDGEMENTS . LIST OF TABLES . . IJST OF FIGURES . . Chapter TABLE OF CONTENTS Page . . . . . . . . . . . . 11 . . . . . . . . . . . . vii o o . o . o o o o o . . x1 I. INTRODUCTION TO THE RESEARCH PROBLEM . . . . 1 Outline of the Chapters . . . . . . . 5 II. REVIEW OF THE LITERATURE AND DEVELOPMENT OF PROPOSITIONS . . . . . . . . . . . 7 Introduction . . . . . . . . . The Job Satisfaction--Morale Hypotheses Exchange Theory and Dedication to Work General Propositions . .. . . . . Definitions . . . . . . . . Propositions . . . . . . . O O O O O O \l O O O O O O O O O N on Organizational Settings . . . . . . 29 The Organization of Public Education . . 35 Occupational Characteristics . . . . 38 Occupations in Public Education . . . 45 Effects of Organizational Variations Upon Occupations in Education . . . . 50 Adaptation of the General Propositions to the Case of Public Education . . . . 58 III. METHODS OF DATA COLLECTION AND ANALYSIS . . . 61 Introduction . . . . . . . . . . . 61 Research Design . . . . . . . . . . 62 The Sample . . . . . . . . . . 66 Measurement of the Variables . . . . . 69 Occupation . . . . . . . . . . 7O Organizational Structure . . . . . 7O Perceptions and Behavior . . . . . . . 84 Statistical Methods . . . . . . . . 89 iv Thu IV. SCHOOL STRUCTC‘ AND AGGREGATE Introducti: School Lev" Internal P Structure Interactior CompoSiti Conclusion V- VARIATIONS on . Occupation DemOg r Perce; Behavi ConClusio: FACTORS INFLL'E DIFFERENCES 1: OFTTACHERS IntIOduct: ffects o UPOn De EffeCtS 0 t0 Work EffeCts O Bedicat Effects C Upon Se Conelusic N. l;i. CEARACI'ERI ST] ORGHHZATION Patterns StruCtL Effects C Upon I Effects C UPOn TE CQn Chapter Page IV. SCHOOL STRUCTURE: PATTERNS IN ORGANIZATIONAL AND AGGREGATE VARIABLES . . . . . . . . 92 Introduction . . . . . . . . . 92 School Level Patterns in Bureaucratization . 93 Internal Patterns in Organizational Structure, by School Level . . . . . 97 Interaction of Organizational and Compositional Variables . . Conclusion . . . . . . . 104 109 V. ‘VARIATIONS ON OCCUPATIONAL DIMENSIONS . . . . 111 Occupational Differences . . . . . . . 111 Demographic . . . . . . . . . . 112 Perceptions and Self-Definitions . . . 118 Behavioral Differences by Occupation . 135 Conclusion . . . . . . . . . . . 142 VI. FACTORS INFLUENCING INTRA-OCCUPATIONAL DIFFERENCES IN PERCEPTIONS AND BEHAVIOR OF TEACIERS . O C O O O I O O O O O 14 3 Introduction . . . . . . . . . . 143 Effects of Personal and Family Variables Upon Dedication to Work . . . . . . 145 Effects of Satisfaction on Dedication to Work . . . . . . . . . . . 151 Effects of Self- Investment Upon Dedication to Work . . . . . . . 155 Effects of Organizational Variables Upon Self- -Investment and Dedication . . 162 Conclusion . . . . . . . . . . . 182 VLL. CHARACTERISTICS AND EFFECTS OF DISTRICT ORGANI ZATION O O O O O O O O O O O O 1 8 7 Patterns in District Organization . . . . 188 Interrelationships in District Structural Variables . . . . . . . 197 Effects of District Organization Upon Principals . . . . . . . . . 200 Effects of District Organization Upon Teachers . . . . . . . . . . 212 Conclusion . . . . . . . . . . . 226 ‘\ ‘o 1‘. 6" VII d Chapter Page VIII. SUMMARY AND CONCLUSIONS . . . . . . . . 230 Introduction . . . . . . . . . . . 230 Theoretical Background of the Research . . 231 General PrOpositions . . . . . . . . 232 Definitions . . . . . . . . . . 232 Propositions . . . . . . . . . 233 Adaptation of the General Propositions to the Case of Public Education . . . . 235 Research Evidence Bearing Upon the Propositions . . . . . . . . . 237 Differences in Organizational Patterns (Chapters IV-VII) . . . . 242 Occupational Differences (Chapter V) . 243 Relative Effects of Organizational Factors, Personal and Job Satis- faction Factors in Intra- Occupational Variations in Self— Investment and Dedication to Work . . 246 District Effects Upon Teachers and Principals . . . . . . . . 249 General Conclusions . . . . . . . . 251 Summary . . . . . . . . . . . . 260 BIBLIOGRAPHY . . . . . . . . . . . . . . . 261 APPENDICES . . A. Teacher Questionnaire . . . . . . . . . 273 B. IPrincipal Questionnaire . . . . . . . . 295 C- Clustered Variables: Teacher Data . . . . . 315 IL Clustered Variables: Principal Data . . . . 319 E. Contract Provisions Limiting Autonomy or Authority I O O C O I O O O O O 3 2 3 vi III, . OIQanizational Education 100 11. 12. LIST OF TABLES Organizational Dimensions: American Public Education . . . . . . . . . . Characteristics of Occupations in Public Education . . . . . . . . . . Social and Economic Characteristics of Cities Comprising Five School Districts Used in this Study 0 O O O O O O O O 0 Economic and Student Performance Character- istics of Schools, by School Districts . Total Number of Usable Administrative Units in Each Grade Category and Their Range of Faculty Size by District . . . . . Distribution and Mean Scores on School Hierarchy, by District . . . . . . Principal Perceptions of Decision-Making by Locus of Decision Competency, for Fourteen Decisions . . . . . . . Teacher Perceptions of Decision-Making by Locus of Decision Competency, for Fourteen Decisions . . . . . . . Independent Variables: Sources and Descriptions . . . . . . . . . Intervening and Dependent Variables: Teacher and Principal Perceptions and Behavior . Mean Scores on Organizational Variables, by Level of School . . . . . . . . Means for Aggregate Data by Level of School vii Page 37 47 64 66 67 73 79 79 82 85 94 96 Correlations . by School L Correlations .» by School L».- of SES Correlations ; Composition School Principals ' an SChOOl Leve:L Table Page 13. Correlations Among Organizational Variables, by School Level . . . . . . . . . . 98 14. Correlations Among Organizational Dimensions, by School Level, Controlled for Effects of SES . . . . . . . . . . . . . 101 15. Correlations Between Organizational and Compositional Variables, by Level of School . . . . . . . . . . . . . 105 16. Principals' and Teachers' Education, by School Level . . . . . . . . . . . 115 17. .Mean Professional Experience of Principals and Teachers, by Level . . . . . . . . 115 18. Perceptions of Power Relations, by Occupation . 120 19. Mean Reported Cooperation and Support, by Occupation and School Level . . . . . . 123 20. Importance of Work, by Occupation . . . . . 126 EL. Mean Importance and Idea Exchange with Other Personnel, by Occupation and School Level . . 130 22. Job Satisfaction by Occupation . . . . . . 133 23. Mean Behavioral Evidence of Dedication, by Occupation . . . . . . . . . . . 137 24- Summary of Occupational Differences in Level of Self-Investment and Dedication . . . . 141 25. Personal and Family Characteristics by Self- Investment and Dedication, Correlations by School Level . . . . . . . . . . . 146 26. Effects of Controlling for Age Upon Correla- tions Between Personal/Family Characteristics and Dedication to Work . . . . . . . . 149 27- Correlations Between Job Satisfaction and Dedication, by Occupation . . . . . . . 153 28° Correlations Between Job Satisfaction and Self-Investment, by Occupation . . . . . 154 viii “— (A) 3 Correlations ' Investment , Correlations Declication Correlation 0. Teachers‘ p. Work Settin “01k, by DC Eii'ects of St Upon Teachn Dedica Correlat 10“ “Wk Sett: work: by ( orgénizatio with Teac DiStributic ierarch} Table Page 29. Correlations Among Indicators of Self— Investment, by Occupation . . . . . . . 157 30. Correlations Between Self-Investment and Dedication to Work, by Occupation . . . . 160 31. Correlation of Organizational Structure and Teachers' Perceptions of Work Setting . . . 164 32. Correlations Between Perceived Quality of Work Setting and Self-Investment in Work, by Occupation . . . . . . . . ,. 167 33. Effects of School Organization and Composition Upon Teachers' Self-Investment in WOrk . . . 172 34. Correlations Between Structural Variables and Dedication to Work, by Occupation . . . 173 35. Correlations Between Teacher Perceptions of Work Setting and Level of Dedication to Work, by Occupation . . . . . . . . . 177 36. Organizational Characteristics Correlated with Teacher Turnover Rates . . . . . . 180 37. Distribution and Mean Scores on System Hierarchy . . . . . . . . . . . . 189 38. Formalization of Authority Relations, by School District . . . . . . . . . . 191 39. Constraints on Autonomy of Administration and Teachers, by School District . . . . . . 193 40. Areas of Emphasis in District Contracts . . . 194 41. Number of Departments and Size of Auxiliary Professional Staff, by District . . . . . 196 42. Spearman Rho Correlations Among Organizational and Contractual District Variables . . . . 198 43. District Variations in Principal Perceptions and Responses . . . . . . . . . . . 201 44. Correlations Among District Ranks on Selected Principal Variables . . . . . . 204 ix Correlations and Princip Effects of [)1 Variables Correlations 1 and Signi fi c Correlations 1 and Teacher: Status of p 1.0] ChaPters Table Page 45. Correlations Between Organizational Factors and Principal Variables . . . . . . . 207 46. Effects of District Upon Teacher Variables . . . . . . . . . . . . 213 47. Correlations Between District Characteristics and Significant Teacher Variables . . . . 216 48. Correlations Between District Characteristics and Teachers' Dedication to Work . . . . . 223 49. Status of Propositions Tested in Previous Chapters . . . . . . . . . . . . 238 ‘l I“ ‘q-u’e ‘O‘l‘. a Wives of SChoc LIST OF FIGURES Figure Page 1. Types of School Hierarchy . . . . . . . . 72 xi INTRODUCT 1' Work has been troughout the history entry, however, has iitepted systematic e 111"? In recent decac 5.3., t on the part of CHAPTER I INTRODUCTION TO THE RESEARCH PROBLEM Work has been an inescapable and central activity throughout the history of mankind. Only during the last century, however, has there been much interest in or attempted systematic explanation of the social aspects of work. In recent decades, we have witnessed an increasing effort on the part of social scientists concerned with classifications of types of work and explanations of individual differences in work behavior. For instance, the recent concern over the problems of professional workers in bureaucratic settings provides clear evidence that the sociology of work, which attempts to integrate occupational, «organizational and individual variables, has so far failed ‘torprovide consistent and logical explanations of worker behavior. In looking at the progression of dominant research ideas and assumptions in the area of work, we see two major assumptions underlying most modern attempts to explain worker behavior. The first assumption is that certain structural conditions and managerial principles maximize pesatisfactiona nee Awarding to the SECC tier levels explain ‘ gerz’orrance, absentee he further assumpti< serrations are char istication levels an irpact on female beh Staring directly fr E‘VrmL-v mural and manac it: ‘~ . .vr male worker: T‘id "‘3 any 0f the $2033 .aIlSOIl of m job satisfaction, need satisfaction or morale of workers. According to the second assumption, differences in satisfac- tion levels explain variations in such things as worker performance, absenteeism, quitting and interest in work. One further assumption has been that females and female occupations are characterized by lower performance and dedication levels and that structural differences have little impact on female behavior. The major research concern, stemming directly from these assumptions, focused upon those structural and managerial conditions that are most satisfy- ing for male workers. There has been little attempt to validate any of the assumptions and thus very little interest in comparison of males and females in the same occupations and even less interest in the possibility of establishing any non-attitudinal explanation of work behavior. This dissertation lies within the general tradition of the sociology of work in its emphasis upon structural 'variables affecting workers. However, this study diverges from the traditional perspective in two major ways: by directly focusing upon behavioral dependent variables, irrespective of satisfaction levels and by considering ‘variations in structural conditions in more than one occupa- tional category and for both males and females. The questions whose answers will be attempted in tflris dissertation are: "Are members of one occupation similar in behavior even in very different organizational environments?" and, even more importantly, "Do structural variables predict b tive of satisfactic The theoret- troad areas of soci‘ the dissertation deg the importance of pzl for continued invol'. area. The symbolic :rl c .3». rererence gIOUFS assertation. From occup at tier ...s of occupations the 1“firs of work rol Um" . we variables . T" 55:51 . Organizations rin variables predict behavioral patterns of workers irrespec- tive of satisfaction levels, sex or age of workers?" The theoretical basis for this study lies in three broad areas of sociolOgical research. From social psychology, the dissertation depends upon exchange theories regarding the importance of possible and actual rewards as a basis for continued involvement of individuals in an activity area. The symbolic interactionist concepts of self-esteem and reference groups are also important elements of this dissertation. From occupational research, this study uses descrip- tions of occupational differences in rewards, mobility and arbiters of work roles and some recent attempts to relate these variables. The third area of theoretical concern, formal organizations, is the source of several ideas that suggest a framework for occupational differences in oppor- tunities and for intra-occupational differences based upon characteristics of focal organizations. Theories of bureau- cracy, involving variables such as hierarchy, centralization and specialization suggest variations in the types of constraints upon workers and point to the importance of systematic variation of both occupation and organizational setting. While using these elements of conventional sociolo- gical research, this dissertation proposes very different types of relationships among the variables than those found .in typical research in the sociology of work. The most Spartant area of di bureaucratic organi: research emphasizes i‘irea'icracy and att negative effects (5 Earthauser, 1965). This disser elezevt .. S in the bur .;:n worker behavic “Critter . . ~51 by Permit: 1&1: c °‘ energy in “319:5 to Concert trim? .‘ng. I S‘Jgfie 1:31 I to OffEr t5 s’VER“ a: ‘ ture and “01" it“s' h’ill ' 11 1:9. P tc :1 :‘1,e 1 or. drg e 1.0- : “lC .‘gih‘. . k at' :S‘V-i 101‘s ‘ “"Qt \ ‘ and t1, 3;; Q ‘ Q Q CC}: “ patiohs ‘i. h ‘ ‘Vns ;\ four CO Vita,» ‘ Ce important area of difference centers around the effects of bureaucratic organization upon employees. Most contemporary research emphasizes the existence of negative effects of bureaucracy and attempts to eXplicate the reasons for such negative effects (see Blauner, 1964; Scott, 1965; or Kornhauser, 1965). This dissertation, however, proposes that several elements in the bureaucratic model may have positive effects upon worker behavior and self-esteem, even for professional workers, by permitting greater potential rewards for invest- ment of energy in work-role performance and by permitting workers to concentrate energy in areas appropriate to their training. I suggest that nonbureaucratic situations are less likely to offer these advantages to workers. I do not expect, however, that the relationship between bureaucratic structure and worker dedication will be linear and positive; rather, it seems likely that some elements of bureaucrati- zation will, up to some maximum point, encourage dedication on the part of workers. In the interest of simplicity, propositions will be stated in linear terms. Fifty-three public elementary and secondary schools in five large Michigan school districts comprise the sample of organizations studied. Bureaucratization at both the district and the school level will be analyzed. From the many occupations whose members are employed in these organ- izations four focal occupations were chosen for study: elementary and secondary teachers and principals. Data in: organizational r< lith‘gan records are ‘ this dissertation . This study off ratrihutions to the e attrition. First. I e Tuious organizational UV Eeianor and the perce hproduct of this part 9' .- taut: ty of the occupe . Elp‘gndx "Mary and second. he to identify 131' ascnool structural ' Br". :1 .es upon teache 5‘.‘ ~15 4“: ‘ ' ussertation wil litre ' -~ 0f lndiVidual v '5 ———-—v— "‘ " from organizational records, employee surveys and state of Michigan records are used to study the questions raised in this dissertation. This study offers both theoretical and practical contributions to the existing literature on work and on education. First, I expect to clarify the impact that various organizational conditions make upon the professional behavior and the perceptions of teachers and principals. A byproduct of this particular focus will be a test of the validity of the occupational distinctions made between the elementary and secondary levels in education. Furthermore, I hope to identify the relative importance of district and of school structural variables and of student compositional variables upon teacher and principal responses. Finally, this dissertation will attempt to identify the actual impor- tance of individual variables such as age and sex as factors underlying worker behavior and adherence to occupational norms 0 Outline of the Chapters The research shall be presented as follows: Chapter II contains the literature relevant to the research qu“33111011, the derivation of the propositions and the con- crete research hypotheses. Chapter III contains a description of the logic of the research design. the sites and sample characteristics. The . . . . - rationale for measuring the variables, the statist1cs in used and the analys. 11’ through VII pres: concludes the dis ser future research. used and the analysis procedure are also discussed. Chapters IV through VII present data and analysis, while Chapter VIII concludes the dissertation with a summary and ideas for future research. RE 71’?" \i2« 0? CHAPTER II REVIEW OF THE LITERATURE AND DEVELOPMENT OF PROPOSITIONS Introduction The organizational structure of public education has been undergoing radical change in recent years with the implications of new laws and new organizational tasks finally being felt at the local level. Population shifts and district consolidation have interacted with increasing specialization of school personnel and with recently success- ful attempts at collective bargaining to create very dif- ferent organizational patterns and problems than previously seen in the United States. One effect of these recent changes is the creation of a more heterogeneous pOpulation of school structures and of school personnel, even within school districts of similar size and locale. These facts present a prime opportunity for the study of occupational members across organizational settings. Most research concerning the effects of structure ‘upon workers has considered either occupational variation alone or else organiz defined classes like heater, takes advant ..ication to examine tional and organizati perceptions of worker accumu . :Ezzk 09.39, at al 1,. \. ' 1 alone or else organizational variation within broadly defined classes like "professional." This dissertation, however, takes advantage of cumulative changes in public education to examine systematically both specific occupa- tional and organizational determinants of behavior and perceptions of workers. The Job Satisfaction-—Morale Hypotheses During the last few decades, social scientists have devoted much time and effort to the study of workers and their jobs in an attempt to discover the determinants of effective performance. There has been no abatement of interest on the topic in recent years, largely because the net accumulation of hundreds of studies (1500 by 1957-- Herzberg, gg_§1., 1957) has failed to produce any consistent, logical explanation for phenomena such as job satisfaction, performance levels and efforts, turnover rates or intensity of interest in work. Any review of the literature on the subject of job satisfaction (for example, see Herzberg, 1957 or Vroom, 1964) shows that the most common working assumption in such studies has been the idea that workers' job satisfaction or morale directly leads to similar levels of productivity and dedication to work. For instance, in one specific test of the idea that satisfaction promotes greater productivity, Katz, Maccoby and Morse studied employees of an insurance company. “sharing highly 1310Ci they found no differe tion of workers (1950 311351 using employees results very similar li‘rert, however, in a as an important varia issed results like t .kil at t he level of ski 'j‘AyL “f-M Ears. t intervenim resonance will be 1 (1961) . Vroom‘ uat :Jhran‘ I ‘ We hls re :31. ‘h‘:‘act - Comparing highly productive and less productive sections, they found no difference in type or level of mean satisfac- tion of workers (1950). Katz, g2_§1., then replicated the study using employees of a railroad company and establishing results very similar to those of the first study (1951). Likert, however, in attempting to rescue job satisfaction as an important variable underlying worker behavior, dis- missed results like those of Katz eE_al, with the suggestion that the level of skill required to perform the job is an important intervening variable, such that satisfaction and performance will be more strongly related in highly skilled jobs (1961). Vroom's review of the empirical evidence for this proposition, however, showed that the differences observed across skill levels usually are very small, although in the direction predicted by Likert (1964). Further evidence that seriously damaged the credi- bility of the satisfaction-morale explanation of worker behavior was evaluated in a recent Department of Labor monograph. This review of research results showed that job satisfaction appears to affect turnover, absenteeism and sabotage but not productivity level (1974). Both Parnes (1971) and Mangione (1972) have offered considerable evidence regarding the negative relationship between satis- faction levels and turnover rates. The relationship of satisfaction and absenteeism is somewhat less clear (see Katz and Kahn, 1965 and Ingham, 1970). It appears, however, that the naive assumption that satisfied workers will be better in every “Y Y analysis. Even a (3111’sC ‘19 work behavior rai- :‘W ‘ i-psychologi ca 1 Frat makes workers ha: :aybe: Why have soc effects behavior at w A great nunbe 35 data c b ' an e Cited ai‘ent‘r "mares of these In answerinc actors underlying maziished researc‘ Lg" - ~ H~‘a: Chical sens 1‘31 wit h work (see I s; \ A L .‘y. ‘~.'~_~i« V S .‘~ 10 better in every way may require further specification and analysis. Even a cursory review of research findings regard- ing work behavior raises very basic questions about the social-psychological and structural bases of worker behavior. What makes workers happy? An even more critical question may be: Why have social scientists assumed that happiness affects behavior at work? A great number of contradictory ideas and fragments of data can be cited in attempting to explain the widespread acceptance of these assumptions. In answering the first question, concerning the factors underlying job satisfaction, one of the most firmly established research findings is that job level, in a hierarchical sense, has a great effect upon stated satisfac- tion with work (see Gurin §£_al., 1960, as an example). Attempting to explain that fact, some researchers have sug- gested that differences in work roles and situses (such as differences in complexity, status, pay, type of work, pro- motion possibilities, autonomy, etc.) are basically hier- archical in distribution and that these are critical deter- minants of job satisfaction. Supporting this argument, Katz cited the dissatisfying effect of repetitiveness of work (1954), which is more often a characteristic of jobs low in a hierarchy. walker and Guest studied the effects upon workers of mechanical pacing of work and control of only a few basic operations and found that both variables produced 11 dissatisfaction and both occurred mainly at low hierarchical levels (1962). Attempts to make low-level jobs more like higher level positions, by changing some aspects of job structure, have appeared to make employees somewhat happier (see Guest, 1957, and Lawler, 1969). Evidence from other sources suggests that use of workers' abilities, opportunity for self-expression and overall mental health appear to be hierarchically related and to affect satisfaction levels. Promotional opportuni- ties and history also affect satisfaction but are less clearly linked to position in the hierarchy (Brophy, 1959; Vroom, 1962; Kornhauser, 1965; Morse, 1953; and Spector, 1956). Whether one explains satisfaction levels in terms of the fulfillment of psychological needs, the opportunity for status-improvement or purely extrinsic factors like pay and conditions, each of these appears related to the hier- archical level of the job. But while this statement shows some progress has been made in identifying factors respon- sible for satisfaction levels, much less progress has been made in establishing the importance of satisfaction as a predictor of behavioral phenomena. Vroom, for instance, summarized an extensive body of literature describing bases of job satisfaction but found that this literature had identified no general effects attributable to satisfaction (1964). In particular, the assumption that job satisfaction 'would promote worker performance efforts has failed to receive general empirical support. One might question the ‘sases for this assure? basis for expecting s fcrzance? One might ju5‘ tier. is only a hairbr said not stimulate e ‘:”p A ,..s,-ective, that sat sttalating producti\ researchers, notably :esr . that workers re: :eefc‘u mg personal nee rated to eXpend much “1&9! Status, p981 fill ' .1 considered impc I5+ .' h. 85 yr :52er ““5 may even be sue “see: ed that Such 35.13- 12 bases for this assumption, though—~13 there any logical basis for expecting satisfaction to be reflected in per— formance? One might just as reasonably propose that satisfac- tion is only a.hairbreadth.from satiation, which probably would not stimulate efforts to improve performance. This perspective, that satisfaction is an ineffective basis for stimulating productivity, is a basic concern of several researchers, notably Argyris, Vroom and Faunce. They sug— gest that workers require stimulation and the potential for meeting personal needs through work, in order to be moti- vated to expend much energy at work. From this perspective, income, status, peer approval and sensitive supervisors (all considered important by mainstream management theorists), not provide adequate incentive for worker dedication. Some rewards may even be a source of further alienation by divert- ing attention away from the characteristics of the job itself, producing what Seeman has called "self-estrangement"-- a lack of involvement or interest in work itself. Israel suggested that such alienation produces in workers a concen- tration on time--how much time it takes to do a job, how 6 consrdere 33 careful to separate the various elements of bureaucracy in their attempt to isolate the critical variable causing some type of worker malaise (from non-professionalism to sabo- tage). Worse, there has been little attempt to separate organizational and occupational aspects of work. Integrating pieces of evidence from a wide variety of research efforts, I came to conclusions quite different from the typical assumption that bureaucracy and favorable worker behavior and attitudes are always inversely related. Instead, I decided that each of the elements of bureaucracy should be considered separately, unless there is overwhelming evidence that a few dominant patterns of bureaucratization can accomodate most formal organizations without loss of critical information. Looking at the elements of bureaucratization separately, there is reason to believe that not all of these elements contribute to worker alienation and lack of commitment. For instance, Scott and Mitchell suggested that both standardization and specialization of work may be related in a curvilinear manner to worker satisfaction, reasoning that lack of training and lack of knowledge about requirements of work are as frustrating to workers as are conditions that make work repetitive and boring, with satis- faction (and hypothetical dedication) being maximized at some point between these two extremes (1972). One might expect that a similar relationship might hold for the effects of formalization upon attitudes or involvement of work Procedures may elm: of workers, an oft-c toprofessional work Greater tech the ideal bureaucrac "93kg! Performance a: alcgical organizati <4 The unwieldll Ereaucracy to most 1 fiestrictive element C scar-3e of conflicts authority (Scott , l 9 6 l....otes close superv ‘57.!A 24‘ V suit. 34 involvement of workers. The existence of formal rules and procedures may eliminate the necessity for close supervision *of workers, an oft-cited irritant, especially frustrating to professional workers (Scott, 1966 summarizes this problem). Greater technical expertise required of workers in the ideal bureaucracy might also have positive effects on worker performance and attitudes, making autonomy of workers a logical organizational strategy. The unwieldly hierarchy of authority that typifies bureaucracy to most laymen is probably considered the most destructive element of bureaucratic structure, either because of conflicts between expert and administrative authority (Scott, 1966) or because hierarchy supposedly promotes close supervision and thus ritual conformity of workers (refuted by Blau,l968 and 1955 and by Kohn, 1971). If one looks at the hierarchy of authority from another perspective, it seems likely that at least a moderate number of higher positions open for worker career advancement would probably have positive effects upon worker commitment and dedication. Only if the ratio of upper to lower level posi- tions in a career chain is so low that no significant propor- tion of lower level employees are likely to move up would one expect the height of the hierarchy to have little effect. Some evidence for this idea was presented in a paper compar- ing quit rates of several occupations in a large organiza- tion. Both number of hierarchical levels and ratio of upper to lower positions exit rates (Smith, Blau prOPOS height of the organ worker dedication. roe-administrative 1 airinistrative compc sou-9h to handle the ii351 (see Blau, 1966 Thus, for a Level of bureaucrati related to ali enati C "'41 the fit and dedic :“C'I'St'er’ t , hat this 1- “Stead, expect th .ee‘ice worker dEdi 35 to lower positions were inversely related to occupational quit rates (Smith, 1974). Blau proposed one further reason why the size and height of the organizational hierarchy may be important to worker dedication. He suggested that the effectiveness of non-administrative personnel may be severely limited if the administrative component is not large enough and integrated enough to handle the non-technical details of the organiza- tion (see Blau, 1966; 1968). Thus, for a number of reasons, I propose that the level of bureaucratization of organizations is not directly related to alienation and job-leaving, but rather encourages commitment and dedication to work. It is not suggested, however, that this relationship will increase monotonically. Instead, I expect that excessive bureaucratization might reduce worker dedication--but so might insufficient levels of bureaucratization. This research project will include organizations varying in level of bureaucratization. The Organization of Public Education The public school system of this country would seem to be an ideal setting for the study of both organizational and occupational variations upon worker behavior since the range of such structural variation is fairly broad and yet there are basic similarities in the nature of work, back- ground of workers and typical organizational form. Unfor- tunately, descriptions of the typical features of school structure are large I limits for such str educators. Only re treat public school anal”itical classifi Eiéaell points out 1 organizational attri 1 :flj‘n «we tional researc.L 1n inter-persona lpal--teacher) , up' u n p.11. «Tin relations per: ”t2; Gutzels and 6‘ Recently thé Q \‘fifi ‘n’f‘ 3 ~ It s{mots of the 36 structure are largely based upon understanding of the legal limits for such structure or upon descriptive work by educators. Only recently has there been any attempt to treat public schools as formal organizations amenable to analytical classification and explanation. In one summary, Bidwell points out that educators have failed to consider organizational attributes of schools (1965). Instead, educational researchers have been more interested in morale and in inter-personal relationships (pupil--teacher or principal--teacher), generally looking at schools from a human relations perspective (for examples, see Griffiths, 1962; Gutzels and Cuba, 1957). Recently there has been some attempt to delineate some aspects of the structure of public schools (such as Anderson, 1967; Carlson, 1958 and Gross, 1962, among Others). One of the most comprehensive of the recent attempts to explicate school structure and show the effects attributable to structural variation is Corwin's study of 28 public high schools (1969). He found that the size, Specialization, hierarchy, complexity and heterogeneity of staff character- istic of the schools were positively related to incidence of organizational conflict. This type of approach repre- sents a significant advance from previous studies, typically centered around attitudes of personnel alone. Despite the amount of inter-school variation in structure noted by Corwin (1969), Given (1969), Moeller and.Charters (1966) and a few others, there are a number of similarities in school structures to be noted. :7ABLE l.-0rganizat Education ‘__ Dimension _ *- lenure of office Segaration of owner :rou adudnistratior ioztracts specify 5 money. ikitten records .emical relured f or hiring Kara rrchy of office al-t-acated 9 :1: ‘39 r S 1 . 37 TABLE l.--Organizational Dimensions: American Public Education. Dimension Extent of Variation Tenure of office Universal for teachers, after probation. Short—term con- tracts for principals and administrators. Separation of ownership Universal. from administration. Contracts Specify salary Universal. in money. . Written records Universal (many records required by law). Technical competence Universal (college courses required for hiring certify expertise). Hierarchy of offices Some hierarchy in all dis- explicated tricts but some schools and districts have taller struc- tures than others. Universalistic criteria Legally required everywhere, for hiring, promotion however, no uniform standards for measuring achievement. Division of work into True in most high schools and specialities. 40 percent junior highs (Corwin 1965) but not true in rural and elementary schools usually. Rules define procedures, Every district has some rules conditions and rewards but there is great variation in the content, complexity and application of rules. Orderly career movement True in districts with elab- possible for staff. orate hierarchies and special- ist departments (Lortie, 1969). Authority vested in Administrators usually have technically-qualified teaching backgrounds; wide officeholders. variation in the levels at which various decisions are made. nun l.--Continuec Euension Suture of work 38 TABLE l.--Continued. Dimension Extent of Variation Nature of work All schools take custody of minors in certain age ranges. Some districts and schools provide expensive auxiliary services, socialization, technical training or adult education beyond basic legal requirements. Occupational Characteristics Beyond differences in bureaucratization, organiza- tions also differ in the number and type of occupations whose members are required to carry out various aspects of organizational tasks. Since the kind of technology used in the organization shapes the nature of the work to be done as well as the kind and extent of training required of various categories of employees, it seems possible that the charac- teristics of many occupations may be affected by the struc- tures of organizations typically employing members of these occupations. Thompson, Avery and Carlson have thrown some light on the relationship between occupations and work settings in their explication of the bases along which occupations vary (1962). They prOposed that the main dimensions along ‘which occupations vary are career ceiling (early or late) and major source of occupational definition, which could be collegial or organizational. Combining these variables to form four basic predicted that tYPi type would be quiteI . I patterns-~occupat101 rut, both or neitha static picture of oc To compreher understand the re cil'n. mugs, we can tur areposed that the vi .w Incumbents is cr *luization ( 1967). well (1970). No 4908611: first, tr 39 to form four basic occupational types, Thompson §E_al. predicted that typical career patterns for each structural type would be quite different, revolving around four basic patterns--occupational commitment, organizational commit- ment, both or neither. This approach offers a useful static picture of occupational differences. To comprehend the dynamics of career movement and to understand the reasons why some occupations have late ceilings, we can turn to another idea of Thompson's. He proposed that the visibility of decision-making ability of job incumbents is crucial to career movement within an organization (1967). That idea has been advanced by Perrow as well (1970). Two elements may be extracted from that proposal: first, that some type of performance must be visible to superiors or colleagues and must be evaluated and, secondly, that the area of performance must be critical and valuable to the evaluators. For instance, decision- making may be a highly-valued, scarce resource to an organi- zation or originality of ideas may be important to colleagues but typing ability, for example, may be less valuable and less scarce, even though it is observable to superiors. It seems likely, then, that some occupations might impart to members scarce and needed skills but that lack of observation and evaluation of performance might produce an early career ceiling for typical occupational members. Other occupations may be concentrated in work settings where observation and evaluation of skill are routine, even though the skills observe colleagues. OrgaJ typical career ce; arose skills can 1 audiences. Lack < the inprobability lack of possible 1 dedication. Thes¢ tho showed that m< a. are ~€IS in the 0C1 “IQi-‘llzational ta' +: so: is I hOWEVer ' L . . he desrred ev IT. a: U. m: " 40 the skills observed are less important to superiors or colleagues. Organizational structure obviously can affect typical career ceiling, since the formal structure determines whose skills can be observed and evaluated by relevant audiences. Lack of supervisory personnel suggests not only the improbability of evaluation of some workers but also the lack of possible positions to be offered as incentives for dedication. These ideas receive some support from Woodward, who showed that most top management personnel began their careers in the occupation whose duties were most central to organizational task success (1965, p. 220). The structure of the work settings for some occupa- tions, however, makes it difficult for work to be observed by the desired evaluators. For example, much of the work of MDs is observed only by patients with little knowledge of medical standards just as much of the activity of profes- sors is observed only by students who have little basis for judging adequate performance and whose judgments may not be very important to colleague-oriented professors. Medicine and academic specialities offer members the possibility of reaching an audience of colleagues and receiving rewards by publishing in specialized journals. Many occupations struggling to achieve professional status, with the impli- cations of colleague-definition and specialized knowledge not available to lay people (Goode, 1960), lack such forums for assessing colleague performance and lack inter- organizational measures of performance. (For example, public school teach journals but the co llsgyecialists, not inectation that go: roles includes publ; my ) sob. 0 While genera pattern of occupatic taut to consider the tional structure may ~. various occupati Of {Each ob ers in small 41 public school teachers and nurses may read professional journals but the contents are typically written by PhD or MD specialists, not by other teachers or nurses; there is no expectation that good performance of these occupational: roles includes publications that advance knowledge in the field.) While generalizations about the range and typical pattern of occupational variables is valuable, it is impor- tant to consider the effects that variations in organiza- tional structure may have upon the place of employee members of various occupations. Obviously, the role and visibility of teachers in small rural schools may be very different from those of their counterparts in huge urban schools. The sc0pe of activities required may vary, the number and positions of evaluators may vary and the bases and rewards for good per- formance may also vary. In order to talk about either occu- pations or organizational structures as variables affecting work behavior, dedication, attitudes, interaction patterns or a great number of such dependent variables there must be systematic variation of both structural variables. It cannot be assumed that highly bureaucratic structures, for example, will always affect employees of a given occupation in one way (or vice versa). Instead, the nature and universality of such relationships must be established empirically. Probably, both structural elements will be important determinants of dedication to work. The types of activity that constitute dedication may also vary according to both structural bases si behavior may vary a lines. As an appliu| tion and structure, activity pattern cal ‘eristic of late-cei :nvrty that consti source of occupatior a lor to importar ill b lerobably occur 3:7; . euent possible. There is so 6'!- V “1:31 * e, Vroom and Fit 1 tions of future IEV Eation a1 Career st ‘rdnities Out s. fluid that m 42 structural bases since opportunities for various types of behavior may vary along occupational and organizational lines. As an application of specific principles of motiva- tion and structure, for example, one might expect that the activity pattern called dedication to work is more charac- teristic of late-ceiling occupations and that the type of activity that constitutes dedication will depend upon the source of occupational definition and the visibility of behavior to important referents. Simply phrased, dedication will probably occur in the form most likely to make a career movement possible. There is some support for this prOposition--for example, Vroom and MacCrimmon showed that the decision to quit one organization was a negative function of expecta- tions of future rewards (1968) and suggested that the organi- zational career strategy depends upon perceptions about Opportunities outside the organization as well. Sofer (1970) found that managers (organization-defined, late-ceiling occupation) were preoccupied with promotion chances and very concerned about trying to use more of their skills and improve their performances (pp. 300 and 330). Smith found that quit-rates were inversely related to height of job ceiling within one organization (1974). Also see Grusky, 1966; Buchanan, 1974 and Tannenbaum, 1974. Systematic variation of structural conditions of ‘work1may well eliminate some of the emphasis upon sex and age as variables en'gect that the C be a better Iii-"9di generality, alt} 3f structure and fat much of the betavior can be i :f'nale' and "ff it one organizati ale workers in 4 Upon exar 3f late ceiling . Zions most likel 33§efnit late c sine «ll occupatio Primt' ing late c Mrations. in it.» . Professione ‘wfiflpatible 43 age as variables supposed to explain worker behavior. I expect that the opportunity structure faced by workers will be a better predictor of behavior than age, sex, and personality, although these may certainly affect perceptions of structure and of alternatives. I expect, for example, that much of the difference between male and female workers' behavior can be attributed to differences in the structure of "male" and "female" jobs. Thus, older workers or females in one organization or occupation may behave like young or male workers in a different job setting. Upon examination of the characteristics and settings of late ceiling occupations, one realizes that those occupa- tions most likely to require complex or rare skills and thus to permit late career ceilings are managerial and profes- sional occupations. While the organizational structure promoting late ceilings seems to fit well with managerial occupations, in recent years there has been great concern that professionals and bureaucratic settings are logically incompatible. Many types of work bearing most hallmarks of profes— sionalism, however, can only be carried out in settings characterized by some elements of bureaucracy (see Friedson, 1974). Schools and school districts, for instance, appear to be somewhat bureaucratized, yet several types of expert work can be carried out only in schools. Teachers, princi- pals and school superintendents may vary in the nature and extent of expert training but each occupation has some claim to profess of such or cosmopolit traits as 44 to professional status. Is it then impossible for members of such organizationally situated occupations to develop cosmopolitan orientations, with such ideal professional traits as flexibility, innovativeness and independence? A variety of research evidence led me to believe that the affirmative answer suggested by Scott and others needs qualification. For instance, in Carlson's study of school superintendents, variations in cosmopolitanism were shown to occur even in an organizationally-linked occupation (1962). Furthermore, Blau showed that bureaucratic officials are not always ritualists determined to avoid change, and suggested that government agents discriminated between several types of change, accepting some. Blau found that those most resistant to change were workers lacking job security. This led Blau to suggest that, since tenure of office is a major criterion of bureaucracy, bureaucratiza- tion may produce less ritualism on the part of employees than would other work situations (Blau, 1955). It appears that only when insecurity pervades the work situation and when employees are completely dependent on superiors' evaluations that risks must be avoided, promoting rigid adherence to established procedures. Further evidence on this question was provided by Kohn. In a study of 3000 male workers in the U.S., Kohn found that those employed in bureaucratic settings were more intellectually flexible, self-directed in values, and open to new experiences, even when respondents' education, level in the hier trolled. Kohn sul ad requirements settings may have fear of reprisals Thus it a} fiesirable professi flexibility may nc 3.. «In -aucrati zati on of 45 level in the hierarchy of work and nature of work were con- trolled. Kohn suggested that job security, heavy work load and requirements for speed found in the more bureaucratic settings may have encouraged workers to try new ideas without fear of reprisals from superiors (1971). Thus it appears that the presence or strength of desirable professional traits like cosmOpolitanism or flexibility may not necessarily be negatively related to bureaucratization of work setting. This is an important point, since there is fragmentary evidence that such traits may be related to behavior I have called dedication (Glaser, 1964). Occupations in Public Education Several of the distinctions suggested regarding occupations in general may be critical in predicting the effects of variations in the structure of public education upon the behavior of members of occupations connected with education. This dissertation will consider two occupations where administrative and managerial skills are the main areas of expertise (elementary and secondary principals) and two occupations where non-administrative, technical training provides the basic skills and where the work itself is more technical than administrative in nature (elementary and secondary school teachers are in this category). Such a fourfold categorization offers many advantages since one L L . ‘- - ll ag: ‘aria ffec i A b 5 can sir gals in in tha‘ .l n n .. Riv as... V § 1" “C s n - 4 1 . grin sex as 46 can simultaneously compare behavior of teachers and princi- pals in any given school, elementary and secondary teachers in a given school system, elementary and secondary principals in that system, etc. Studying both elementary and secondary schools offers the advantage of slightly expanded range of variation in structure and also allows one to examine the effects of school system variables upon the structure of schools in a system. The inclusion of elementary school principals should permit the separation of occupation and sex as independent variables affecting behavior, just as the inclusion of secondary teachers allows for the comparison of male and female responses in the same occupation and organi- zation. The four occupations chosen, while similar in set- ting, are different enough to permit testing of some of the ideas already proposed, especially those concerning the effects of career ceiling, source of occupational defini- tion, visibility of performance to colleagues and superiors and criteria of performance. The following chart summarizes data supporting the treatment of these occupations as meaningful and distinct entities. While the distinctions between elementary and secondary teaching are well documented and accepted, it may be necessary to elaborate upon the rationale for treat- ing elementary and secondary principals as members of distinct occupations. One obvious reason for this decision is that individuals cannot move between the two types of HQQfiDCfiMQ EQCUEGE QCC fl U €13.0th o ageing EUZUHK nu fi HEN...— .~ .L 1 Rumble filo flvqn~zvhv1v nurse Pk»; fifu at hl~ MU U?» PIN Fecaull s~\ ..~.-.~..—.Ms“fi 47 .Hnom .omm .Hnom .Emam maflsomou mswaomou mcflnomou osflnomou 00cm mnmom 3mm whom» 3mm ucoosum ucoonum Iwnomxm Essflcwz Amuowuumwo oEOm ..oomm .omv +«z +us>wuom mo nmflm oumuoooz oumuoooz 304 cowumNHHmwoomm Ammuuxo oHnHmmom swam zoom pans 30A son sumamm msumum umonoflm oumuoooz 30A umosoq ameoom o>wuoaom A.EoHo soap mswawoo mung mama nonmav maumm >Hnmm umoumo Hmowmma anoumumfln noamu.ooz soamu.ooz .oozuzoq son .um ca woman mumocooom mumucosoam mnmocooom mumucoEon moaumwuouomumno Homeosflum Hosanna mGOwummsooo .cOfiumosom Ufiansm cw msowummsooo mo mowumfiuouomumnonu.~ mqmHGUCOO®m thuCOEOHB Xhmficooom kLmDCDEQHN HoQflUCeum Rezonofi $20 a. U. 613000 ll tl‘lsrbllllll I |.| I 1.x: .r? mOHumflhquQHm£U I.“.bf..l 1.4. ..I ‘l c~vmmv~u .hfixd phat-hull- navi. .afixfi -.\(\.3 48 .hmma .Humumw can Gmmcmm “mmma .wxmm .HmmH .mmfimm aroma .sxmcsemq lemma .oaoooq rooms .umoo romafl .cmeumnoaq “moms .Hcosuum "moonsom .ouncou m>ma uoa 0o mammwosHHmT2.um:m on madam I500 mucoummv noflmlmumuoooz oumuoooz ocoz ooflono mo Eooooum Amuconm masomv nmwmlooz oumnoooz 30a 30a muo> musmflao mo mdumum HMGOHumz Hmcowpmz mono .mumum mono .oumpm umxumfi Don memo» whoa no .Qoum nouns .Qoum Houmm N >Ho>o oanmsocou muomuusoo .ucwcmfinom .ucocmauom ousaoa sOHHMHuomos Hmsofl>flocH Hmsofl>flocH o>HuooHHoo o>HuooHHoo mumamm Mom mammm Amosomoaaoo . Eoumv xuoz 3oqn.ooz .ooz .oozlnowm swam um :ofluMHOmH Hoccomuom .moum Hooumo mcwuflHoHOOm .>m5m .cwsod .>m:m .GH864 mswsomoe mafinomoa usmusoo xnoz mhmocooom humucoeoam mumosooom mumucofioam moflumwuouomumso Homeosflum Honomme msowummsooo .owscwDGOUIl.m Manse 49 principalship without further training and experience, since districts usually require teaching experience at the elementary level as a prerequisite for principals at that level and require secondary teaching experience for prospec- tive secondary principal. A further reason for distinguishing these two occu- pations is that the content of these jobs may be very different, partly because of the differences in the age, needs and behavior of students and partly because school organization is very distinctive at each level. While there are typically a few large, diversified secondary schools in a district, there are usually many more small, internally homogeneous elementary schools. The kind of decisions the principal must make, the typical relationship between principal and teaching staff, the number of levels between teachers and principal, and the observability of principal decisions to teachers, parents and other principals and superiors may vary by school level. Although both prin- cipal occupations may have potentially later career ceilings than do the two teaching occupations and although the prin- cipals are likely to have a more cosmopolitan outlook, there are some bases for expecting elementary and secondary principals to differ attitudinally and behaviorally. Unfor— tunately, principals have not been subjects of research to the same extent that teachers have been and the existing research tends to concentrate on the relationships between Principal and teachers in a given school, rather than cog-paring and seek and Hemp” high school P E In th the structure fine, the pos less bureaucr teristics Web schools with 55.: .uered more 50 comparing and contrasting principals by level, district, experience, etc. One example of the latter approach is Rock and Hemphill's report on characteristics of junior high school principals (1966). Effects of Organizational Variations Upon Occupations in Education In the last section, several types of variation in the structures of public schools were described. For each type, the possible variations could be considered more or less bureaucratic, according to their fit with the charac- teristics Weber used to describe bureaucracy. For instance, schools with formal, written rules and records would be con- sidered more bureaucratic, along that dimension of bureau- cratization, than schools without such written rules. The possible effects of greater or lesser bureaucratization of organizations were considered, in a general way and also with specific reference to "professional" occupations (that is, those with some claims to that status, not just the traditional list). As a basis for making predictions about the behavior of those employed in public education, it is important to explicate the mechanism relating organizational structure to actiVity of occupational members. I propose that.various aspects of the organizational structure of schools‘will be consistently related to variations in the ‘behavior of employees, through changes in career ceiling and.visibility. As c tionshiPI CC or school 5} pmgl‘ESSr he limiting res rational cri workers (i.e permits deci for each dec bureaucratic cratization . iQh er caree Status ranki a. “- PIOgres s EPIiSte nce of lertical mos ‘thermore' P217 “‘1 H u:5~ rOHSibilj ‘NL‘Si (31155 The 51 As one clear example of the mechanics of this rela- tionship, consider the implications for teachers if a school or school system emphasizes standard evaluation of pupil progress, has well-defined rules for communication and for limiting responsibility of workers to specific tasks, uses rational criteria for hiring, assigning and evaluating workers (i.e., based upon expertise in field of specialty), permits decision-making at the level of greatest expertise for each decision area and, in general, approaches the more bureaucratic end of the various contina making up bureau- cratization. In such a situation, teachers may have a higher career ceiling, since specialization allows informal status ranking by expertise (and if formal specialization has progressed to the point of creating departments, the existence of the position of department chairman allows a vertical move without leaving the area of teaching expertise). Furthermore, individual performance and use of discretion may be more visible to superiors as well as to peers, since responsibilities are well-defined, records kept and some decisions made by teachers. In less bureaucratic schools, particularism, dif- fusion of responsibility and lack of records and hierarchi- cal distinctions might make it less possible to observe, evaluate or reward good teaching performance. Such struc- tures might also prevent the formation of effective colleague control of performance, since less Specialization means coworkers either would not be experts at all or would have 4‘. no basis for j ization would the basis of requirements lC'bS are much At fi 3:3 elementar 05 Variation that Secondar :Iatic in a n in this direc in their orga find 0‘19th slots 0f bure :" . «.23 t. ‘ 30"5 have ! 43!“. :‘\L EL .re pregame ‘\ 52 no basis for judging others' expertise. Lack of special- ization would make it very difficult to hire teachers on the basis of their knowledge or their records, since the requirements and standards of performance of very diffuse jobs are much more difficult to Specify. At first glance, it seems likely that secondary and elementary schools will not even overlap in their range of variation on the above dimensions. While it is likely that secondary schools have been forced to be more bureau- cratic in a number of ways, elementary schools have moved in this direction as well and districts have varied enough in their organizational change rates that one could hope to find overlap between levels on at least some of the dimen- sions of bureaucratization. For example, elementary teach- ing jobs have been much more explicitly defined recently, partly because of the collective—bargaining for contracts. The presence of auxiliary specialists--medical, social, psychological and testing experts, for instance--as well as the hiring of luncheon and playground supervisors and teachers aides relieve teachers of many of the non-teaching requirements that used to be part of their jobs. This process has gone on at both the elementary and secondary level. For both levels, then, one could classify schools on the basis of two aspects of specialization: are there auxiliary non-teaching specialists, so that teachers at least specialize in teaching?, and are there divisions Within the school on the basis of subject-matter expertise? lfurther type 0f asking how many CC teacher was requi school district ce certain amount of uncovered and that elementary and sec tion and hierarchy so me . terns of teache Some of t' 53 A further type of specialization could be measured by asking how many courses outside ones major in college each teacher was required to teach. As long as more than one school district can be studied, it is very likely that a certain amount of variation on these dimensions can be uncovered and that there might be some overlap between elementary and secondary schools. And, if the specializa- tion and hierarchy elements truly affect career ceiling and visibility, some concomitant variation in the behavior patterns of teachers might be expected. Some of the research on education and the profes- sional occupations in general may be useful in explicating the relationship between school structure and work behavior. I do not suggest that school structure differences can explain all the difference in behavior patterns of teachers and principals but the analyses of schools and teachers found in the literature tend not to consider the element of struc- ture very seriously. Many variables that seem to be occupational in nature may be greatly affected by the structural setting of work. For example, in one consideration of the character- istics of teaching, Geer pointed out the importance to scholars of an expert audience composed of colleagues com- petent to judge performance. Since pupils are in the process of learning, they have little basis for judging teachers' mastery of a subject or pedagogic ability and hence their approval is not as meaningful as is colleague recognition in 0t perspective may Ci of public school ‘ the impact of dif: highly bureaucrati develop systems of and reward for ex; i'nen departmental sequence of new cc subjects or to main Lnteraction provic' "~393ble of evalua1 urinal standards «8 an audience of steeple care more Fruit ' °ssmnS. part 54 recognition in other professions (Geer, 1966). While this perspective may correctly assess the historical situation of public school teachers, it does not take into account the impact of different organizational structures. In more highly bureaucratized schools, specialized departments may develop systems of consultation, observation, evaluation and reward for expertise in subject matter or technique. When departmental colleagues collaborate to design a sequence of new courses, to work as a team in teaching some subjects or to make suggestions on texts, such focused interaction provides each teacher with a trained audience capable of evaluating performance according to at least minimal standards held in common. Presumably, when there is an audience of specialists to appraise performance, people care more about performance levels. In traditional professions, part of the basis for such concern about per- formance lies in the fact that the critical audience of colleagues has some sanction powers. While teachers need not worry about losing a license, or if tenured, clients because of poor professional performance, schools organized into departments and allowing decisions to be handled by the level with needed expertise offer some sanction possi- bilities to teachers. Movement to the position of depart- ment chairman is probably somewhat affected by colleague evaluations of expertise and since this movement is prac- tically the only hierarchical change possible for a career teacher, this may . ubitious teachers In anotherJ teachig occupaticl. solitary training and the isolation £1969), produce an oriented approach . “’31 Support for < .; 0. teachers to deve “I". Paradoxical. the teaChing task . SiaQ¥ed ly‘ an anti‘pr. '9:" ‘U H Q‘H‘ 55 teacher, this may make colleague evaluation critical to ambitious teachers. In another explanation of the peculiarities of the teaching occupations, Lortie pointed out that both the solitary training and socialization of teachers (1968), and the isolation of teachers from each other during work (1969), produce an individualistic, rather than colleague- oriented approach to work problems. The lack of any struc- tural support for collegiality seems critical to the failure of teachers to develop strong ties in an occupational net- work. Paradoxically, it may be possible that breaking up the teaching task into less complete tasks (normally con- sidered an anti-professional trend) might help to provide some support for collegiality and thus produce work behavior geared toward colleague approval. In supporting the idea of specialization as a basis for collegiality, it must be remembered that the definition of colleagues would shift-- not all secondary teachers but only French teachers or math teachers would be classified as colleagues capable of evaluating performance. Another possible advantage of departmentalization ‘was suggested by Lortie. He pointed out that, through formal and informal meetings within a Specialty department, there might be developed at least a first step toward a body of expert knowledge to be applied by all the teachers during the course of work. This possibility would be strengthened by team teaching practices since teachers sorking together i the established P These ide of labor within SC study of bureaucré the two phenomenor that, specificallj negatively relate Eowesrer, he admit seen specializati 3. v- ' ' J: o.gan12ational 56 working together may begin to codify their experiences as the established professions have done (Lortie, 1969, p. 43). These ideas about the advantages of greater division of labor within schools appeared to be incorrect in Hall's study of bureaucracy and professional. He suggested that the two phenomenon were, in general, inversely related and that, specifically, the extent of division of labor was negatively related to professional attitudes of workers. However, he admitted that this relationship need not hold when specialization occurs as an occupational rather than an organizational phenomenon (Hall, 1966). Obviously, how- ever, Specialized teachers can be forced to teach non- specialized subjects if the school structure does not permit more than rudimentary division of labor. The other aspect of specialization within the organization (the creation of chairmanship positions) would probably have no ill effects upon professional attitudes, since Hall found that hierarchi- cal authority seen as legitimate by the professionals did not have such effect upon professionalism (1966). Since technical expertise is essential to beliefs of subordinates regarding legitimacy of superiors, the legitimacy of depart- ment heads who are trained in the departments' speciality probably is not questioned (and their legitimacy probably relieves the principal of problems caused by his lack of training in some subject areas). The extensive literature on bureaucratization and ‘Professionals need not be applied too rigorously to the occupations of te occupations can 0-- tional setting, w i‘inctions implied out, too, that prc mrelated to obje sional status (196 atti tudes in that he most popular 8 he best kind of d mites of profess & 57 occupations of teacher and principal, partly because these occupations can occur only within some type of organiza- tional setting, with at least rudimentary hierarchy dis- tinctions implied by principalships. Since Hall pointed out, too, that professional attitudes appeared to be unrelated to objective indicators of occupations' profes- sional status (1966» the great emphasis on feelings and attitudes in that body of investigation may be unwarranted. The most p0pular stereotype of a professional may suggest the best kind of data to collect to use in constructing indices of professionalism. The popular stereotype concerns the activity pattern of a real professional--someone who is totally involved in work. Such involvement probably requires that the work be interesting and challenging (more likely if Specialization in a field of expertise is required), that there be some way of measuring success or failure (more likely, for teachers, if standard tests are routinely given to students and if colleagues can evaluate teaching efforts), that there be some ways in which workers can make extra efforts (perhaps not possible in some occupations) and that there be some payoff for extra effort (such as a late career ceiling, visibility of good performance to others, etc.). I suggest that, for occupations in_general and for teachers and principals in particular, several aspects of bureaucratization encourage dedication to work. I expect to find sufficient variation among dis- tricts to permit some overlap of elementary and secondary school structures “d occupational As for th cation for teache of a professional trlily dedicated t. efforts to keep up reading or attends stay in education of dedication, su student problems , Euities in meanin tl‘cse activities that several a Spe iii-cg 58 school structures, so that the effects of organizational and occupational variables can be separated. AS for the kinds of behaviors that comprise dedi- cation for teachers and principals, the popular stereotype of a professional offers the appropriate indicators. A truly dedicated teacher or principal would make great efforts to keep up in his or her field, with coursework, reading or attendance at conferences, seldom miss work and stay in education indefinitely. Some alternative indicators of dedication, such as time-budgeting and concentration on student problems, will not be considered because of ambi- guities in meaning or probable biases in measurement. For those activities that will be measured, though, I expect that several aspects of bureaucratization of schools will encourage teachers and principals to exhibit high levels of dedication to work. Adaptation of the General Propositions to The Case of Public Education Given the evidence regarding differences in oppor- tunity structure among the occupations in public education, the general propositions lead us to expect that: 1. In general, the occupation of school principal offers more rewards than does the occupation of school teacher. (Principals have more autonomy, more scope for decision-making, less routinized work, earn higher salaries, receive greater statu' ceili org ar ln gs Offe: Elem the ener inte mOre org; POSS PEe: Sch. hig‘ SEC 59 status in the community, have later career ceilings and are more visible to collegial and organizational evaluators than are teachers). In general, occupations at the secondary level offer more rewards than occupations at the elementary level. (Greater specialization at the secondary level permits concentration of energy in areas of expertise, making work more interesting and allowing the recognized expert more freedom from parental interference or organizational rules on course content; the possibility of later career ceilings and expert peer audiences become important in departmentalized schools. The prestige of secondary personnel is higher than that of elementary, the "clients" of secondary personnel have higher social status, and the activities of secondary schools are usually the subject of community, not just neighborhood interest, making the performance of secondary teachers and principals more visible. a. Thus, ceteris paribus, principals should have .higher levels of self-investment in work and behave in a more dedicated manner than teachers. Secondary personnel should have higher self-investment in work than elemen- tary personnel and should behave in a more dedicated manner than elementary personnel. l Within ea organizat effects 0 to work. Cal level expertise Salistic : decision—r tise Shoui and great. their con Sex and a IEWard St against C importanc struCtUrE ing evide achieVe < Principe; mobility PErsOnne 60 Within each occupational category, variations in organizational structure should have noticeable effects on self-investment levels and dedication to work. Personnel in schools with more hierarchi- cal levels, greater division of labor, greater expertise of employees, greater emphasis on univer- salistic standards of evaluation and permitting decision-making at the level of appropriate exper- tise should exhibit greater self-investment in work and greater dedication, behaviorally, than do their counterparts in other schools. Sex and age of worker may affect perceptions of reward structure because of societal discrimination against older and female workers. The relative importance of sex and age of worker, compared to structural variables, will decline in school offer- ing evidence that older and female workers can achieve career success. Principals should exhibit a more purposive mobility orientation than teachers; secondary personnel should have stronger mobility orienta- tion than elementary staff. CHAPTER III METHODS OF DATA COLLECTION AND ANALYSIS Introduction This chapter outlines the design of the research, describes the collection of data and defines in operational terms the concepts used in the propositions developed in the last chapter. The research project weighs the impact of several structural variables--hierarchy, specialization, centrali- zation, formalization, universalism in evaluations, among others--upon teachers' and principals' dedication to work. Furthermore, this project attempts to clarify the relation- ship between dedication behavior and those attitudes generally considered characteristics of professionalism, such as idealism and cosmopolitanism. Finally, this project com- pares the relative importance of structural variables versus individual variables like age and sex in order to ascertain the role each plays in explaining dedication to work, pro- fessionalism and job-leaving of teachers and principals. As a basic step in accomplishing these objectives, the project also investigates the interrelations between 61 bureaucratic elem zations. providin tern to students . of organization '1. hegreeted area of Testing t? chapter requires a structure, within Toueet these bro Edistricts in Mi Of these schools a: t’a‘lsofi of four c ‘EfinL b “\v‘ue rs at both 62 bureaucratic elements and rational elements in formal organi- zations, providing new information in an area of major con- cern to students of bureaucracy. Explication of the patterns of organization in public schools will also contribute to a neglected area of educational research. Research Design Testing the propositions developed in the previous chapter requires a sample of schools that vary in internal structure, within districts that vary in structure as well. To meet these broad requirements, 53 public schools from 5 districts in Michigan were selected for study. Fifteen of these schools were secondary (three from each district) and 38 were elementary schools. This sample permits com- parison of four district occupations, principals and teachers at both levels, and also permits a greater range of structural variation than would be likely if only elemen- tary or only secondary schools were studied. The sampling strategy permits movement between several levels of analysis, since district patterns, school variables, occupational patterns and social-psychological variables can be isolated and analyzed separately. The inclusion of district organi- zational variables and community factors is of critical importance since many of the items usually cited in explana- tions of the satisfaction levels of professionals, such as autonomy, recognition and adequate working conditions, may be the results of policies. Original and assistants in that research tea tation requires t school and distri attitudinal and 1: Earlier analyses ”“3 never cons ide concurrent 1y . 63 be the results of community resources or of district policies. Original data were collected by Dr. Philip Marcus and assistants in 1968 and were partially analyzed by that research team (Betz, 1968; Given, 1969). This disser- tation requires the use of all data collected regarding school and district organization and most perceptual, attitudinal and behavioral data on teachers and principals. Earlier analyses did not consider the principal data at all and never considered both elementary and secondary teachers concurrently. For a full description of the logic of the sampling, see Betz, 1968, and Given, 1969. The research sites were selected so as to control for environmental variables that might contaminate the analysis. The comparability of the research sites is of major importance because of the tradi- tion of "grass roots" control characteristic of American education. When educational policy is formulated at the local level, the wealth and size of the community, its industrial base, its occupational and ethnic composition all become important factors shaping the educational policies of the district. Depending upon importance of education to local residents, one would expect to find variations in physical plant, salary range and success of bond issues, as well as in less concrete evidences of concern (such as prestige accorded educators, delegation of authority to trained personnel, insistence upon adequate training of ill QEoo mhmox Noocom Gusto? . . m.o~ NH cocoa . m.oa m as m 0H ooo.moH ooo.nn~ ooo~nm~ ooosmaa TNHW QWQND OOO~mm C> >H HHH HH H I..F.' “Nu-Irwin‘s ”fill null}- I .szum Och CH DEED mubflhumHQ H0050” flu>fi2~ HuCfiQfiHAkEOU aofld «NU «who QUflUmufihflydUmwhanhu Ufi—EnVCCUufl ~u:pv urefihuensfll .H. «QQ:—\.h In“.-. q.lL..|‘.l“llHHfi cullnhlnfi I 64 .mamumEdc cmEom an on pmnnmumu on Haas muowuumwp nuuommocmm« .mmmH .cmmflsoflz ~mcwmcmq ummm lcowumeOmmd cowumosom cmmw£OHz “mmmalbmma .mumn uoauu Imam Hoonom oaansm cmmwnoflz ammma .cowumwoomm< GOHuMOSUM savanna: .hnuuoz .B can umxoom .m an .mwlnmma ~hocum mama uowuumwa cmmflsowz «mamcmo cmmwsowz coma unmounom he mm mm am am nonhuman an muHCd o>wumuumHGHEpm Hones: Hmuoa mo.oH o~.m oa.oH mm.m om.m was“: assumnmmo omumooaaa oo~.m ooo.m ooo.m omo.m comic mmmausmma How moaspmnom mumamm ESEHGHS 5mm.ma mm~.ma vsm.ms hom.sa mom.sm Human magmnmnEme mom coaumsHm> pmuwamswm mumum s.v~ m.o~ H.sm m.mm m.~m oflumm umsommauawmsm smo.a oms.a ooo.a omo.m smm.a mumnommu omamfluumo «0 gonads Hmuoa Hmmrm ssv.m mmo.m osm.m maa.m macocH masons o.H¢ o.mv m.ms «.mm «.mm mcoflummsooo HmHHOU muwnz ca Guacamfim usmoumm m.oa m.HH m.oH m.oa NH cmumameoo mamas Hoosom canvas ooo.mm ooo.moa ooo.ss~ ooo.sma ooo.~aa muam cans: «> >H HHH HH H Hoonom m>Hm mcwmwnmeou mowuwo .snsum may as com: muofluumno no mowumwuwuomnmso owfiocoom can Hmwoomll.m mqmfla teachers, etc.) -' cant effect upon organizational alI principals (Cole; Betz descr. 'To eliminate ex‘ urban areas were 11 are a diversifi corposition, simi values toward ed: éistricts corres; 3e Undertaken . l 65 teachers, etc.). These factors may in time have a signifi- cant effect upon the quality of the schools and upon the organizational and occupational commitment of teachers and principals (Coleman, et_al., 1966; Jencks, 1972). Betz described the rationale for district selection: "To eliminate extreme variations, larger districts serving urban areas were selected because they would most likely have a diversified industrial base, a broad occupational composition, similar tax bases and a comparable set of values toward education. Since most large urban school districts correspond roughly to city boundaries, the Michi- gan Census was used to select cities where the study could be undertaken. Michigan, in 1960, had six urban places where 100,000 or more people live, with five of these places containing districts of about the same size" (1968). Since the sixth, Detroit, was very much larger than the others, it was eliminated and the five comparable areas were studied. (See Table 3.) As the table shows, the five districts are similar in pupil-teacher ratio and in starting salary for teachers. Some differences in socio-economic status of district resi- dents is suggested (from median school years completed, income and percent in white-collar jobs) and verified in Table 4, but no district is different enough to introduce great bias. The most important difference among the dis- tricts is the state equalized valuation for District I, reflecting the high property values of that district and of Sch 1 Characteristic k SE5 mean score" 838 variance lean student ability score ability variance State Aid/ school \ scarce: Michigai 1970. . used he he .ae mean for 'I’as 45.0 th to 54. he ' b large indust atile this diffe ea “e SChools , 0e . raised during 66 TABLE 4.--Economic and Student Performance Characteristics of Schools, by School Districts. School District Characteristic l 2 3 4 5 SES mean score* 52.8 45.9 47.8 48.4 45.4 SES variance 9.4 10.8 11.1 11.0 11.0 Mean student ability score 53.7 47.1 48.1 48.7 46.2 Ability variance 7.8 9.2 9.5 9.0 9.4 State Aid/school $12,539 $29,366 $30,687 $29,890 $31,056 Source: Michigan Department of Education School Census of 1970. Composite scores of 4th and 7th graders are used here. *The mean for the State of Michigan was 49.6 and the range was 45.0 to 54.3 for 612 school districts. the large industrial concern located in that district. While this difference does not prohibit comparison among the schools, the effects of district-level variables may be raised during the analysis. The Sample Once districts were selected, individual schools in each district were categorized into three sampling levels: kindergarten through the sixth grade, seventh through ninth grades and tenth through twelfth grades. To insure com- parability in drawing the sample, several schools which did not conform to this gradation were eliminated. TABLE 5.--Total I Each C Size fl ‘ Total number of usable adminis— trative units: K-6 7-9 10-12 Page Of faculty size; filamentary Secondary Sources; 1967 I: 67 TABLE 5.--Tota1 Number of Usable Administrative Unites in Each Grade Category and Their Range of Faculty Size by District. District I II III IV V Total number of usable adminis- trative units: K-6 17 40 48 48 30 779 3 9 6 5 5 10-12 3 4 5 3 3 Range of faculty size: Elementary 7-26 14-49 4-23 8-30 7-40 Secondary 40-103 68-78 17—86 71-101 51-89 Sources: 1967 District Handbooks. The districts seem comparable, with a few exceptions. In District 1, there are only three 7-9 and only seventeen K-6 units from which to sample, since several schools in that district had grades K-9 and therefore were not included in the sampling population. Similarly, two primary schools (grades K-3) did not correspond to the stratifying rules and were eliminated from the sampling population. Despite such minor differences, there is considerable similarity in range of faculty size when districts are com- pared and the elementary and secondary schools seem to have fairly distinct size characteristics with little overlap. From this sampling frame, the researchers randomly selected one school from each of the 10-12 levels, two from each of the 7-9 levels and about 20 percent (but no fewer than 6) from tt were selected (I and V) , eigl‘ moderate numberlI the larger dist schools agreed teacher reSpons‘ tion behind the' 901ml in Given, distributed to schools, Yieldi tUre. F01- this 68 than 6) from the K-6 levels. Thus six elementary schools were selected from the districts with relatively few schools (I and V), eight were chosen from the district with a moderate number of schools (IV) and nine were chosen from the larger districts (II and III). All of the selected schools agreed to participate in the study. The average teacher response rate was 82.8 percent. (Further informa- tion behind the decision to use all sampled schools can be gound in Given, 1969; Betz, 1968). Questionnaires were distributed to teachers and principals in the sampled schools, yielding data on personal and career history, attitudes and activities and perceptions of school struc- ture. For this dissertation, however, improved measures of authority relations and of the schools' environmental setting were developed. To obtain better measures of Socio-Economic Status (SES) than overall census indicators for cities, data from the Michigan School Census of 1970, collected by the Michi- gan Department of Education, were used. Since SES and student body composition do not change radically from year to year, we assume that the 1970 data would provide a good estimate of conditions at the time of the original survey two years before. SES and skill levels are determined in the school census using the responses of 4th and 7th grade students to a battery of skill tests (e.g., verbal, mathematical and social cate- style, parental components of ST elementary schoc Standard deviati calculated , prov Variations in le Of School compos Survey questions 69 and social categories) and a survey regarding family life- style, parental occupation, education and other presumed components of SES. For every Michigan School district, elementary school and junior high school, the means and standard deviations for SES and student ability have been calculated, providing data on inter- and intra—district variations in level of SES and ability and on homogeneity of school composition. (Detailed information on the sample, survey questions and tests can be found in the Michigan Department of Education's Technical Report on the 1970 School Census.) Documentary sources of data included employment contracts between each district and its teachers' associa- tion for the 1967-8 school year (which were used to ascer- tain the level of formalization of each school district and to assess the legal limits and guarantees on principal and teacher autonomy), the district directories for 1967-8 (from which inferences about specialization and hierarchy were made), the Michigan Education Directory for 1966-7 and for 1967-8 provided rosters of principals in each dis- trict, thus permitting the calculation of principal turnover rates during the focal year. Measurement of the Variables The independent variables discussed in the previous chapter concerned the structure of work and some personal characteristics of individuals. While the personal characteristics the survey by as the structural i. sets of dimensic Occupation , org a CCCU ation 70 characteristics are simple and easily Operationalized in the survey by asking respondents' age, sex and work history, the structural variables are more complex. Basically, three sets of dimensions underly a job's opportunity structure-- occupation, organizational structure and environment. Occupation In the survey, respondents were identified by their school and position, yielding four categories--elementary and secondary principals and teachers. Teachers described their current positions so that student or substitute teachers could be eliminated and so that tenure status could be controlled, if desired. Data are available for 53 principals--15 secondary and 38 elementary--and for 1,413 teachers--850 secondary and 563 elementary. Data concerning different types of classroom teachers will be presented in the section on specialization. Organizational Structure Since schools are part of districts having dis- tinct organizational patterns, both types of organization may be relevant to the propositions. Within each organi- zational component, I will describe both types. The specific components of school structure measured, the source of data for each component and the range of variation for each item will be described next. School pattern of hie: possible forms were possible 5 Principal at tt Other Personnej Considered crit 8'98 3 are Cons sense, althougl and Span of cor The er ate position We Pals (See Apper possible to ass representing t} Sal and teache] Pcsitions leve l S 71 School Hierarchy.——For any particular school, the pattern of hierarchy could be described as one of only four possible forms (see Figure 1). Only two position levels were possible steps in a professional hierarchy with the principal at the top and classroom teachers at the bottom. Other personnel involved in auxiliary hierarchies were not considered critical and were not examined. Type 2 and type 3 are considered equal in meaning in a hierarchical sense, although they suggest differences in specialization and span of control. The existence or absence of each type of intermedi- ate position was determined from questions asked of princi- pals (see Appendix B). Using principal responses, it was possible to assign a number (0, l or 2) to each school, representing the number of position levels between princi- pal and teacher, excluding clerical and auxiliary specialist positions. Table 6 presents the mean number of intermediary levels in schools in each district, as well as the distribu- tion pattern. This table suggests fairly large inter- district variations in hierarchy, with district 1 schools appearing to have the most hierarchical arrangements. It is essential, however, to note that the number of elemen- tary schools studied in each district is not equal, and that there are different patterns in elementary and secondary schools. When similar ratios are calculated just for schools of each level, combining them across districts, the mean number of intermediary levels in secondary schools Teach ASSista Teacher 72 Principal l. Teacherl, T2, Tn Principal 2. Assistant Principal Teacherl, T2, T3, Tn Principal 3. Dept. Chair.1 Dept. Chair.2 T1 T2 T3 T4 T5 T6 Tn Principal Asst. P. Asst. P. 4. D.C.l D.C.2 D.C.3 D.C.4 T1 T2 T3 T4 T5 T6 T7 T8 T9 TX T Least hierarchical arrangement.(only two levels). Most likely found in elementary schools. Moderately Hier- archical (3 levels). Moderately Hier- archical (3 levels). (More Specialties) Most Hierarchical (4 levels). Rarely found in Elementary Schools. Figure l.--Types of School Hierarchy. TABLE 6.--Distr by Di ¥ I: __I District 0 TX; 2 8 3 9 4 7 5 6 73 TABLE 6.--Distribution and Mean Scores on School Hierarchy, by District. Intermediate Levels in Hierarchy District 0 l 2 Mean* Rank 1 5 l 3 .78 1 (High) 2 8 l 3 .58 3 3 9 3 0 .25 5 4 7 l 3 .64 2 5 6 l 2 .56 4 *[2(# schools with 2 levels) + 1(# with 1 level) + 0] t N schools in district]. is 1.67 while for elementary schools it is only .105. When means for each type of school in each district are calcu- lated and ranked, none of the elementary school components of any district rank higher. One interesting effect of this separation before ranking is that secondary schools in districts 1, 2 and 4 have the same mean number of inter— mediary levels--all the differences appearing in Table 6 were due to hierarchical differences of elementary schools in those districts. It appears that District 3 schools have more primitive hierarchy than schools in the other districts, while District 1 schools, in both elementary and secondary categories, showed the most complex hierarchical structure. However, the great intra-district variations in number of levels between district differ POlicy or envir While t archical arrang. the above table SECondary teach 52* $330 01 in the SE teacher and Prir schools had Suc‘r 74 levels between teacher and principal suggests that the district differences are not solely the result of district policy or environmental constraints. While the nature of district variations in hier- archical arrangements in schools cannot be uncovered from the above tables, these data do suggest that elementary and secondary teachers face different advancement opportunity structures within each school system. All but one secondary school in the sample offered at least one position between teacher and principal while only 4 of the 38 elementary schools had such a position that could permit intra- organizational career advancement for teachers. Although the differences between elementary and secondary schools in terms of hierarchy are huge, there are cases--where the con- nection between school level and hierarchy does not hold-- most districts have at least one atypical case, a secondary school in category 0 or 1 or an elementary school in cate- gory l and these few cases may help us to separate the effects of school level from those of hierarchy within the school. Hierarchy in the District Organization.—-When the district school system itself is the focal point, the hier- archy variable is much more difficult to measure. When compared to schools, districts use less universal position titles for administrators below the superintendent level. The most common title, "assistant or "associate" super- intendent is expected to denote an administrative area of control: rather nadministrative the hieral’Chical school PrinCipa supervise an ai‘ districts have ‘ type of district level programs directors} . Informat was gathered fro obtained from th titles and occup Position and any SKY Programs w 75 control, rather than subject-matter specialization. "Administrative assistant" is another common position in the hierarchical network extending from superintendent to school principal. Each of the above positions may also supervise an aide or assistant-to position. Finally, most districts have divisional directors responsible for some type of district-level work or for coordination of school- level programs (e.g., personnel, financial or business directors). Information about the existence of these positions was gathered from directories for the school districts obtained from the superintendents' offices. Position titles and occupants, clerical staffs connected to each position and any connection with federal,state or univer- sity programs were listed, allowing us to use reasonable criteria to distinguish staff from line positions. No survey data were used for this variable. Since assistant-to positions are not necessarily part of the main line of authority and since many divisional director positions appeared to function as staff specialists advising the superintendent and teachers, outside the main line of authority, the final hierarchy measure does not include these positions. Specialist positions and supervisors of course-work consultants were also eliminated from the hierarchy scale. These staff positions appear to be only peripherally related to the superintendent-principal line of control and are more likely to be seen by specialized teachers as op; whose expertise related. Centra- tic of bureaucr worker behavio r centralization s , “Clever. the co 76 teachers as opportunity for advancement than by principals, whose expertise is administrative, rather that course- related. Centralization of Decision-Making.--One characteris- tic of bureaucracies that researchers believe may affect worker behavior and perceptions of work is the extent of centralization of authority. In some research projects, however, the concept of centralization is either subsumed under the hierarchy dimension (such as Hall, 1969) or is subject to biased measurement, being based solely on percep- tions of managers at upper organizational levels (such as Reiman, 1973). Each of these problems requires attention. The first type of problem, the neglect of any independent measurement of centralization, stems from an assumption that any organization of a given configuration would have a decision-making pattern characteristic of that hierarchical structure. While this notion fits with Weber's analysis of the meaning of hierarchy, researchers have observed variations in the extent to which positions requir- ing ability of a given type have been entrusted with deci— sions appropriate to that position level and ability type (for instance, see Child, 1972). Thus it appears desirable to assess the decision-making structure of an organization empirically, rather than inferring it from the organization charts. At the same time, one must maintain the distinction between decision-making authority and personal influence or power. This st}. structure, alth included. The sec an over-relianc individuals at ¢ centraliZed. CC 1 I Mal as a Souro zation, a Practi Rpm—ts Cannot 1: research On Cent lower-level reSp 77 power. This study will concentrate on the formal authority structure, although a few questions about influence will be included. The second typical problem in research comes from an over-reliance by investigators upon the reports of individuals at one level of the organization. Thus, mana- gers may feel that the decision-making structure is not centralized. Complicating the problem of reliance upon one level as a source of information is the problem of using a very small number of informants for data on any one organi- zation, a practice which means that the reliability of the reports cannot be checked. Mansfield suggests that the most research on centralization suffers this weakness (1973). Yet lower-level respondents may not have realistic perceptions of the decision-making structure. In an attempt to surmount the problems mentioned here, both conceptual and methodological, centralization is measured independently of hierarchical structure and a variety of sources of information about decision-making are tapped. For this study, objective limits and subjective perceptions of the decision-making structure are analyzed. Types of decisions often required in schools were listed in a survey of teachers and principals who were asked to choose the usual actual decision-maker from a list of £6: positions. See Appendix A for a list of the decisions and positions. To simplify the analysis, the.l§7positions can be described as Teachers (acting individually or in concert) I PrinC System PerSOnne and state offi because no one decision-making The deC categorized , US. required by eacl limits on ce rtai opinions regardi Sion. The three ing teacher com; local administre System level km The ove; Principals of t 78 concert), Principal (and appointees or assistants) and System Personnel (including superintendents, school boards and state officials). One category, parents, was drOpped because no one believed parents played any regular role in decision-making. The decisions listed in the survey have also been categorized, using estimations of the kind of competence required by each type (see Goslin, 1965; pp. 229-30), legal limits on certain decisions and also using respondents' Opinions regarding the ideal decision-maker for each deci- sion. The three resulting categories are decisions requir- ing teacher competency, decisions requiring principal (or local administrative) authority, and decisions requiring system level knowledge or authority. The overall patterns of perception of teachers and principals of the levels at which the three decision cate- gories were handled confirms my original separation of the fourteen decision areas according to the type of expertise or authority probably required to make each decision. As the data in Tables 7 and 8 suggest, both teachers and principals thought that system administrators were the most powerful of the three levels. Both groups also saw teachers making more decisions than principals in these areas, although the teachers perceived a somewhat greater advantage than did the principals. ABLE 7.--Princ T of De Locus of Competency Teacher [5 decisions) Principal (3 decisions) System (6 decisions) let A uthOI‘i ty 79 TABLE 7.--Principal Perceptions of Decision—Making by Locus of Decision Competency, for Fourteen Decisions. Perceived Decision-Maker Locus of Competency Teacher Principal System NA Teacher 158 48 45 14 (5 decisions) Principal 16 87 52 4 (3 decisions) System 35 64 205 14 (6 decisions) Net Authority 209/742 199/742 302/742 32 53 principals x 14 decisions = 742 responses . TABLE 8.--Teacher Perceptions of Decision-Making by Locus of Decision Competency, for Fourteen Decisions. Perceived Decision-Maker Locus of Competency Teacher Principal System NA Teacher 3926 966 1208 920 (5 decisions) Principal 243 2120 1307 532 (3 decisions) System 1143 1391 4735 1155 (6 decisions) Net Authority 5312/19656 4477/19656 7250/19656 ’2607 1404 teachers x 14 decisions = 19656 responses The cen could only be 1“ teachers and Pr did not always often saw diffe... sane school. R] the other sourcrl list of componerl tions of extent tions of central In givin Perception of ce teaChfirs choosin 80 The centralization of authority within each school could only be measured from these subjective reports of teachers and principals.* As might be expected, teachers did not always agree internally and teachers and principals often saw different decision-making structures within the same school. Rather than judging the correctness of one or the other source of information, I put two factors into the list of components of bureaucratization--principal percep- tions of extent of centralization and teacher mean percep- tions of centralization. In giving each school one overall score for teacher perception of centralization, I calculated the number of teachers choosing each decision level for each decision in each school. If a choice of Teacher is called a Level 1 choice, a choice of Principal is called Level 2 and a choice of System is called Level 3 choice, the ideal decision-making pattern would be described as: 3 2 l, l > 2 but (1+2) > 3. Considering the distribution of choices in each school, a school will be labeled: a. Highly centralized if 3 > (2+1). (26 schools) b. Less centralized if 3 < (2+1). (27 schools) These variables can then be fit into a picture of the over- all bureaucratization of each school, using Child's point (adapted from Weber) that delegation of authority to the level whose positions involve adequate knowledge and authority to handle a given decision type is .one of the halhmarks of bureaucracy (1974). Low centralization thus should be characteristic of more bureaucratic schools. Distri more or less c Pals and teach Erovisions of w extent of deleg While t levels of distr. Principals and 1 Possibly bECauSi ferent Set of d! SQPQ ¥ rintendent ’ 81 District organizations can also be described as more or less centralized, using both net scores of princi— pals and teachers in each district and also using formal provisions of wage and benefit contracts that specify the extent of delegation of authority for certain decisions. While the contract provisions suggest only low levels of district differentiation on this dimension, principals and teachers see somewhat greater differences, possibly because they answered questions regarding a dif- ferent set of decision areas than those covered in the contracts. When the contract provisions are considered, it is assumed that, in a natural state, all authority rests with officials of the school system (i.e., the board and superintendent, and that the collective bargaining process represents a formal acknowledgement of the delegation of certain types of authority and autonomy to teachers (and in a few cases, indirectly, to principals). Thus the number of provisions restricting system authority and the ratio of such restrictions to restrictions on teacher and principal autonomy can be used as an indication of the extent of decentralization of authority in the district organiza- tions. (See Appendix E for the complete list of contract provisions limiting autonomy or authority.) The remaining independent variables are much easier to conceptualize and measure than the hierarchy and central- ization dimensions. Table 9 sets out the variable name, the source of the data used to measure the variable and, if ..-¢14:: :0 mcoflmw>oum HonEdz muocuucoo :OwuchNMEHOR “UHHDWHQ .mflsm HON .QUHQ meDOU Ouflswmh .n xOOQ OHDH MO mocwumfixm .6 GHHQCCOHDWODO HEQNOCHHQ COHDUNNHUEUON NOOCUW COHUQHHUNSQ If! QOHBCW thXJN-m. H,1.Naw\w twiddldln. 1'11. . ‘JI? ll. [nul'll' l4 Illull |.u.ll. I- o as :A v H. (a 0— a. ti .rJ Pu. rend BU FA «as aw. .ww yd. .5. u . AI..- .3. a psav~tf~na u 15> ,s .7.v~;..~.i«..~ewnu.~ ~ II. n \v. ...~.N:<.b 82 Hoosom sumo How muoom coo: memxo pump Icmum mean: muonomou w .n mmeo pudendum mo ucoEommnsooco HMfloflmmo .m coflumospo Hmmwocfium .0 «2 “4m usonuw3 w .Q HOGHE \HOnmE mowmuso scoop w .m mmmuw umHHMHoomm .cwEUMIcoc mo oNHm mumflawwoomm mcflnommulcoc HmQEDz ucmsmu mucosa mo menace .n unmsmu mmmusoo mo muoflum> .m uwuoE can muflnoflcom so mcowmfl>oum subsumed Incofluoeonm mo mwuoufiuo mason .huwuonusm so mCOMmH>0HQ Honasz .mnsm How .mmnm omHsoo muflsvmu .D goon mach mo oocmumwxm .m msmcoo Hoosom cmmwnowz ouflmccoflumoso Hmmflocwum moHHmQCOADmoSO Hmmwocwnm can Honomma mxoonpcmm uoauumfln ouwmcaoflummso Hmsomme mouwmccoflumoso Honomoa mucouucoo meHmccOHummso Honomoa new Hmmwocflum muomnucoo ouwmccowumono Hmmwocwum HOOQOmllmmm coflnmuueumeamum Hooaom omwuuomxm coonmEm Hoonom coflumNaamwommm nonhuman mc0flmmomoum mo cofiumanMHoomm Hoonom chADwmom mo cowumeHmwoomm Hoozom Emflamwuo>aco uoauumwo Emflammuo>HCD Hoonom GOADMNRHmEuom possumwa coflumuflHMEHom Hoonom cofiumwuomoo mouaom unassum> .mcowumfluomoo 0cm moouaom "mmflnmwum> ucmpcmmmocHll.m wands . ll. .. t ET}: noncou Hoocov. costumwmllxuwocooozto: mum HOOSUm £060 NOW UUCUHHM> MSWCUU HOO£UW HOOCUWIINUHQCQOOEOE WWW UUHHUWHO 3060 HOW @HOUQ C002 MSmCQU HOOSUW CQOHQUHZ uUwHuthllmmm COHUQHHUQOQ @UHSOMH znvul mmvN.fiN.mu H.1Nma\v (.il‘ol 4 ill i .Ill .Illdl.h|.l...ll.'| y ohtwre~nfl~iuu nhfiuviwvsl' Qkfx ...\)‘$.\<.Ns 83 sunshade some CH muflaflpm co mocmwum> msmcmo Hoosom uoanumHQIIMUHocmmOEom ucmpsum mouoow muflaflpm one so oocmanm> mamcoo Hoonom Hoonomllhuwmcomoeom ucmpsum uoanuwap zoom MOM ouoom you coo: mumcou Hoonom uowuumflnllmuwafind unopsum Hoonow some now ouoom no: coo: msmcou Hoonom Hoonomlumuflaflnm ucowsum uowuumwp sumo How consanm> mumcoo Hoonow uoauumflallhpflocmmOEom mum Hoosom comm new oocmaum> msmcoo Hoonom Hoonomllmuflmcomoeom mum uofluumwp comm How whoom coo: mamcmo Hoosom cmmwnowz uowuumwallmmm cowumfiuommo mousom mananua> .pmscwusooll.m mamas the source was tion/s used. E of each questic The int 0; V v' ‘ aLlOUS aSpe attitudes and b1 ErlnCipal turno mm were sathe #6,. b .s of PGrs< 'h) rESPORSES to gut me CluSter ana of responses to 84 the source was a survey item, a description of the ques- tion/s used. See Appendix A and B for the actual wording of each question. Perceptions and Behavior The intervening and dependent variables are composed of various aspects of teacher and principal perceptions, attitudes and behavior. Except for data on teacher and principal turnover rates in each school and district, which were gathered from district handbooks and district rosters of personnel, the intervening and dependent varia- bles were Operationally defined as responses or clusters of responses to questions on the teacher and principal surveys. The cluster analysis evaluated statistically the patterns of responses to each set of questions that were expected to tap one or another dimension of behavior or perception, to see if there was indeed great internal correlation between the elements of each theoretical cluster. This procedure will be described more thoroughly in the section on Statistical Analysis. The questions used to measure each variable and the content and reliability coefficient for each cluster obtained can be found in Appendix C and D. Table 10 displays the final list of intervening and dependent variables, the questionnaire or document from which data were drawn and a description of the character of the variable as it would be used for analysis (i.e., a rate, a cluster, a school mean, etc.). :..>..:.U. Tu: To: Hht .m >O>H5m HOSUMOE .€ HUDWSHUi WCOWMOK UNWQHHHNQ «HON Q05 CW CHMEQZ HfiQHUCflHQ om .HMWFNMUHWAUL... odu‘ uflUH—Jom EHOLH .....|..V=~PU2 NVNIQHVHNWHRV> |lll|lo .1 Idol vII'.EIiI Had: ‘I "Ilfllh'nll'q if I If.”|l|l| I - V- c .. ..J|kutlllul...fl.|llrlllu.ll‘| IIHO N.\r ~J~Nnuwu~ hens-u,» mflhoiflufiflooucnn I—lefitNiMUFNJ—WVHFH Hue-[ha “wnuhU-Mxnflvph u Bflfiflfl.fi~!¥ “(fl-Hy) .16 ~.~mu0‘=v-°fiholvfi~ “Uni-lo \v-iu fi!aU\r?~..U t L ~ ll I3 N. .\.H~.h..N<.~u 85 >o>usm Hmmwocwum .m mm>usm Honomoe .d Houmsau muwmuoucH mumpflmuso mw>nsm HMQAUCHHm .m mw>nsm nonomoa .< Hopwsao mosmmoaaou omoau >o>usm Hmmwocflum .m >w>u5m Honomwe .¢ Hmumsau mumpwmcH mo>usm Hmmflocfium .m >o>usm Honomoa .< noumsau Hoccomumm opflmuso "nuHB mmopH mmcmnoxm >o>udm Hmmflocflum .m >o>usm Honomoa .4 noumsao Hoonom opflmcH mo>usm Hmmflocflum .m mm>usm nonomoa .4 Houmsao Hoosom mpwwuso "Scum Duommsm can cowumuomoou uommxm ho>usm Hmmflocflum .m >o>usm uozomoe .< noumsau oaumflamopH\oflmcHuucH mo>unw HMQHUCfiHm .m >o>nsm Honomoa .¢ . Houmsao mcommom msumum mo>n5m Hmmflocflum .m >o>nsm Hmaomoa .d Houmsaoe mcommmm camcwuuxm ”How non cw chEom Hmanocnnm .m HOSOMOB .¢ "mousow Ehom $502 OHQMHH . m> .How>mzom 0cm mcoflumoouom Hmmfiocflum cam Honomoa “mudguaum> ucopcomoa new mcflco>uoucHll.oH wands it .§-< 7 . LdSL If» Kindrgwmll "Cuwk COqumwaumm >0>Hsm HcmflonHm .m x0>usm chume .< hmumDHU THOMHOuCH WEONQDHQ HQQHUCHHQ .m .HmwSU-wmwph .< "@UHSOW Shauna Uzzwz my figs» HUM. 6.3 do . Ill Ii .l‘lu'l‘ IHSfiUH- RN V vs -h\VpUI' «AV ~ 86 >o>nsm Hmmwocflum .m mo>nsm Hosomms .4 HoumsHo muopflmpso mo>u5m HMQAUCme .m >o>nsm Hosomma .4 noumsHU mumpmeH ”mo coacflmo @000 How :Hmocoo “Hoonom mpflmuso mocoDHmcH macov Hmmwocwum .m mo>nsm genomes .4 Houmcau muonuo Ho>o mucosamcH :30 wo>nsm Hmmwocflum .m >o>usm Honomma .4 Rwandan mnozomoa mumpcflm OB >o>usm Honomoe .4 nounsao muasomm muuommsm Hmmwocwum Auxoc paw mflau How Hmumsao ocov >o>u5m Hmmflocflum .m mm>nsm Honomoa .4 Houmsau coflumuuchHEU4 uoom ”usooo muco>m >m>usm umnomoe .4 noumsHo meowumaom wwwuonus4 >o>usm Honomoe .4 HovmsHU mowumwuouomumnu Hoowmmsmlcoz AmEouH cofluommmflumm co Houmsao can mco umsmv >m>nsm Hmmflosflum .m mowumwuouomumnu >o>nsm Honowoe .4 Houmsao moflumflnouownmno non Annamwnm unpw3 cowuommmwumm >m>usm Hmmfiocwum .m >o>u5m genomes .4 Hmumsau mummnmucH mEoHnoum Hmmflocwum .m Honomoe .4 «mousom Euom ofimz manmwnm> .possflucoool.oa mqm4a :t:s:U LCCOCOE .< HOquHU >0>H5m MOSUMOB .4 Mwumzsu IL Q05 0>wuwhumHCNEU< HON MSOHXCW DOD OQHSUTQE 302 MOM WSONKQT HUQHUCHHQ .2 MOEUQOE .< ”@UHDOW .. Illa-it} 1 ( «92:02 mVN. an. Hubs-“(w .( it‘lliillfllnrrtl ill“! I ll]. . - (I .y I Ill lull) . (till: Imvudiufl~fil¥a~fiyvV|ll ION: .MNWNQ~<.N. 87 mo>usm Honomma .4 Hmuwsau meapmmm mo>nsm seconds .4 HoumSHU 09 CH uoxnoz mm>usm Honomoe .4 “mumsao 09 CH mflnmuopmoq >o>nsm Honomma .4 Hospfl>w©cH Emflooucomn4 whom» mllmpuooom.uownumfla .m munch mllmpnooom uoaupmflo .4 spam Ho>ocuse >o>usm Hmmflocflnm .m >o>usm Honomma .4 HMSUH>MUGH humamm wm>nsm Hmmflocflum .m mo>msm Honomoe .4 Hmopfl>fich moonmoa mo>usm Hmmflocwnm .m mm>usm Honomoa .4 Hm5©H>HUGH xom >o>usm HMQHUCHHm .m >m>usm Honomoe .4 HMDUH>HGCH om4 mo>usm Hmmflocwum .m >o>usm Honomoe .4 Houmsao mufluonuz4 Hmmflocflum m0 momsvo©4 >m>usm Hmmwocflum .m >o>usm Honomoe .4 Hopmsau non m>fluouwmeHEU4 Mom macaxc4 >o>usm Honomoa .4 Hmumsao non mafinomma 3oz Ham msowxc4 Hmmflocflum .m Hmsomma .4 "condom Euom mama mflnmnum> .UODCHHWWWMH.OH mammm. :C:&:U LQrTL CL. «AN LQ‘U:FL - - . -\ .- >0>Hsm HMQHUCHHQ .m >0>H3m HOSUMOB .< HmumSHU MQOHOQO>EDU UQOQQW Homeocflum .m “@206.an -4 uQUHSOmw. 5.8054 .0532. @NQQ unholy; .Iv'I l. I; s no; u. . 2 ~.. 4 Cache -1 I. v C V. .3; . ~ IL. ill-ll.|II n I.I.l(l.l..1 88 .mump Honomou Ham How woumanoamo on HHfl3 memos Hoonomss .0 can 0 xwpcomm4 mom .muoumsao mo mucoucoo Home >o>usm Hmmwocflum .m mm>usm nonomma .4 Hm5©w>H©GH coupaflno mo om4 one Hmnazz >m>usm Hmmflocflnm .m Houmsau Hmmwocflnm mm mocoflummxm ambusm Homflocwum .m >o>nsm Honomoa .4 Hmumsao mcowumuoomxm Hmoumo >o>usm Hmmwocwum .m >0>Hsm nonomoe .4 noumsau nmmnsoo oxma mo>u5m Hmmflocwum .m >o>usm Hogomoa .4 nonmSHU mc0flucm>cou pcmuu4 Hmmwocwum .m Honomme .4 "condom Show mEmz mandaum> .ooscaucoouu.oa mamas For mar project adequat questions. Whi questions would consider the pc rect. Furtherrn tions are each the multitude 0 both of these p; all the mul ti -m. tation. The Sp 6 devised by Dr . od- " these h 9e teachErs cc 8 For inSt cunt . ence . in L*Oss 89 Statistical Methods For many of the variables considered in this research project adequate measurement required two or more survey questions. While I assumed initially that each such set of questions would be unidimensional, it seemed necessary to consider the possibility that my assumption might be incor- rect. Furthermore, even if my hypothesized sets of ques- tions are each unidimensional, some technique for handling the multitude of related measures is required. To answer both of these problems, cluster analysis was performed on all the multi-measure survey questions used in this disser- tation. The specific cluster analysis technique used was devised by Dr. John Hunter of Michigan State University. This technique permits one to see whether supposedly similar measures of a variable are actually similar in interaction with other variables and simplifies the further analysis of each group of similar measures of a variable. Once the statistically derived clusters were estab- lished, I compared them to my original list of variables. With only minor changes, the original variables were almost perfectly reproduced in the statistically-derived clusters. Of these changes, differences between principals and teachers accounted for most cluster content differences. For instance, in principal clusters, other principals are “close colleagues" while for teachers, others in the same specialty are "close colleagues." The only unexpected statistical cl cluster for pr idealistic, ex 30b aspects as actually did. EVidence regal—c tion as a Varig ing Was not tOC reason fol. aVO' a variable. 90 statistical cluster was the overall "Job Satisfaction" cluster for principals. Principals did not differentiate idealistic, extrinsic and status categories of satisfying job aspects as I had expected they would and as teachers actually did. However, since (in Chapter II) I presented evidence regarding the general invalidity of job satisfac- tion as a variable, this divergence from expected cluster- ing was not too disturbing. It simply gives an added reason for avoiding relying too heavily on satisfaction as a variable. The reliability of each of these statistically con- firmed and theoretically valid clusters is denoted by its alpha score in Appendix C or D. All principal clusters _ had reliability coefficients above .60 and most were above .70. Of the teacher clusters, only three had reliability coefficients below .60 and these were in the .57 to .59 range. Since these three were theoretically valid and seemed unidimensional and because of the large N for teachers (1403) these three clusters were kept as well. The teacher clusters were then subjected to a non- parametric two-way analysis of variance, to assess the independent and interaction effects of school level and school district upon teacher perceptions and beliefs. This analysis used a nested design with an unequal number of schools in the cells, since the number of schools varies by level and district. The .05 level was used to decide whether the di expected by ch To ass Won principal lid by school USGd F tests tr tur6:11 variablegl The he) is a multiple c the impact Up On 91 whether the distribution differs more than could be expected by chance. To assess the effects of district and school level upon principal responses, I calculated means by district and by school level for each cluster and each behavior and used F tests to determine the significance of these struc- tural variables. The next step in the analysis of the teacher data is a multiple correlation matrix, attempting to identify the impact upon teacher behavior of each of the structural variables subsumed under district and level differences. The theoretical framework suggests that several organiza- tional and occupational variables underly the overall dis- trict and level differences and that some of these may be more critical than others in influencing behavior patterns. Because only five districts are studied, organiza- tional variables presumably affecting principal behavior are harder to separate statistically. I used Spearman rank correlation coefficients to indicate approximately the importance of each of the district organizational variables upon principal perceptions and behavior. However, since so many structural variables differentiate the five districts, one cannot place too much reliance upon the results obtained here. SCHOOL CHAPTER IV SCHOOL STRUCTURE: PATTERNS IN ORGANIZATIONAL AND AGGREGATE VARIABLES Introduction In this chapter, I will consider several theoretical problems concerning the structural variables that were brought up in the first and second chapters of this dis- sertation. First, the rationale for treating elementary and secondary school structures as distinctively different will be tested by a comparison of actual patterns of organization at the two levels. If there are clearcut differences between the two levels on organizational dimensions, the next step is to see whether either level can be charac- terized as closer to the ideal bureaucratic type. In pursuing that question, it will be possible to see if the data provided in this sample present sufficient evidence to indicate that the bureaucratic elements are unidimen- sional. Turning to the interaction of organizational and other school level variables, I will check whether any 92 apparent patte buted to under ifferences in Lust ask wheth: ables are inde; If they are he‘- trolled before attitudinal an‘ 93 apparent patterns in organizational structure can be attri- buted to underlying variables (which will be inferred from differences in SES and staff composition). Finally, one must ask whether the aggregate school compositional vari- ables are independent of the formal structural variables. If they are not, the organizational patterns must be con- trolled before the effects of compositional variables on attitudinal and behavioral variables can be assessed. School Level Patterns in Bureaucratization To check the possibility that elementary and secondary schools had distinctive patterns of organization, mean scores for each level on each organizational variable can be compared and evaluated. Table 11 sets out the means by level as well as the results of t-tests comparing the values for the two sets of schools. It appears that secondary and elementary school structures are quite different, since secondary schools scored significantly higher in hierarchy, size of special- ist staff, use of universalistic criteria and standardiza- tion than did elementary schools. But elementary schools had significantly higher scores on both teacher and principal perceptions of centralization and on specializa- tion in major areas. For the other organizational variables, specialization in courses, expertise of teachers and formalization, secondary schools had higher scores, but the differences were not statistically significant. TABLE 11. --Me Lex Organizational Variable I Hierarchy Standardizatio: Fomalization Specializatiom Courses sIv‘ecialization. major :niversaliSml EXpel‘tise Centralization 94 TABLE ll.--Mean Scores on Organizational Variables, by Level of School. School Level Organizational Variable Elementary Secondary value of t Hierarchy 1.607 4.293 -6.07* Standardization 2.333 3.351 -l.76** Formalization ' 3.013 3.767 -1.26 Specialization-- ‘ courses 2.485 3.241 -l.28 Specialization-- major 3.660 2.464 2.65* Universalisml 2.584 3.184 -1.9o** Expertise 2.729 3.421 -1.15 Centralization Pr.l 4.234 2.386 4.04* Centralization T. 4.287 1.827 8.2* Specialist Staff 2.162 4.499 -4.59* 1 F tests were used to check the differences in variances of the two samples. Only universalism and centralization Pr. showed statistically significant evidence of differences in pOpulation variances, in 2- tailed tests with a : .10, with elementary schools showing greater variance. *Significant if (t) Z 1.96 for a = .05 in a 2-tailed test with > 30 d.f. **Significant for a =.10 in a 2-tailed test with > 30 d.f. I On the I variables des c secondary schc tion. More in approaches to some elements do not suggest burt‘laucrati c t other recent r. midimens ion a 1 B 1 la” e‘t\al.. zational Struc to consider Wh. reflect differ. To check this ‘ 56c Orldayy SChO Vallies for the SEt fOrth in T 95 On the basis of the two clusters of bureaucratic variables described above, it appears that elementary and secondary schools do have different patterns of organiza- tion. More importantly, those differences suggest different approaches to bureaucratization, since each pattern contains some elements usually described as bureaucratic. The data do not suggest that either type of school is clearly more bureaucratic than the other. Furthermore the data confirm other recent research showing that bureaucracy is not unidimensional (see Hall, 1967; Pugh, §£;gl., 1969; and Blau, gt_al., 1966). Before proceeding to further analysis of the organi- zational structure of the sampled schools, it may be useful to consider whether the apparent differences in structure reflect differences in ecological or school level variables. To check this possibility, I compared elementary and secondary schools on several different dimensions. Mean values for these variables and calculated values of t are set forth in Table 12. The data reported in Table 12 do not explain the large differences in organizational structure of elementary and secondary schools already observed. The only signifi- cant differences in the school level data concern the sex of the professional staff. Recalling conclusions from Census data noted in Chapter II, it is apparent that these reflect typical occupational differences, with a higher prOportion of males both in teaching and in administration TABLE 12 . “Meg Variable3 fl 855 Level 538 Homoqeneit Student AbilitI Student HomOgel Consensus Teac Principal Sex PrlnCipal EXpe lence 3' Female Teach l 2N 2 1| 3 ~ . 1 Principal SexL‘ 96 TABLE 12.--Means for Aggregate Data by Level of School. School Level Variable3 Elementary1 Secondary2 Value of t SES Level 3.042 2.920 .20 SES Homogeneity 3.070 2.614 .75 Student Ability 2.887 2.584 .49 Student Homogeneity 3.162‘ 2.499 1.12 Consensus Teacher 3.296 3.970 .95 Principal Sex 3.158 1.00 4.88* Principal Exper- ience 3.739 3.481 .49 % Female Teachers 4.461 3.264 3.11* 1N = 38. 2N = 15. 3 l = Low, 5 = High for all variables except Principal Sex, where 1 = Male. *Significant if (t) Z .196 for a =.05 in a 2-tailed test with > 30 d.f. at the second: that these va]I although this the effects of possibility th section, check 'w'hen sex Of st may be there Clc student abilitl leVel . Sectmn On the 5219.5 g The he 97 at the secondary school level. Although it appears unlikely that these variables determine school structure, and although this could only be established by observation of the effects of personnel changes, I will return to the‘ possibility that such a relationship exists in a later section, checking for patterns of organizational structure when sex of staff is controlled. One final observation that may be derived from Table 12 is that school district may be more closely related to such variables as SES, student ability and principal experience than is school level. This possibility will be considered later in a section on the bases and effects of school district vari- ables. Internal Patterns in Organizational StructureL by_School Level The next step in the analysis of organizational structure is to consider the interrelationships of the bureaucratic variables when school level is controlled. Udy (1957) and Stinchcombe (1957) proposed that bureaucratization should be separated into two multi- dimensional components, rational and legal components of organization. _I considered the pattern of internal corre- lation of elements and, although most correlations did not reach statistical significance, found some support for the separation of organizational items into two distinct groups. -Table 13 presents the actual correlations among Organizational variables . ‘ 1 ‘ II-|- 23~£C .«L C> ~1CO~¢2Ni C63h3 I! :0:.:‘ III: ._:>04 Hecztm x: .mcuntwsc> HeccaucNacszO 3COE< m:0wuc~ckh00|l.ma NQ=umucwfimam 9E3 mm. \Na. mm.-\mo.u ma.-\ma.u mm.-\mo.u Hm. \NH. ma. \va. ma. \Ho.- mm.-\mH.s NH. \FH. omfiuuodxm .OH NH.-\HN.- -.u\mv.u mo. \mH.u mm. \mo. an. \mo.u m~.-\~o. mm.-\no.u HH. \ o ummum\omdm .m as. \oa. mm.-\afl. hm.-\aa. mo.u\~o. so. \mm.- on. \ma. mm.-\oa.u gummauuucmo .w mm.-\nH. vm.n\mo. em.-\~a. mm. \Ho.- me. \No.- ss.-\~H.u cflumuuucmo .5 no. \no.- ma. \mH.n mo.-\v~.u mm.-\am.u 5H. \oa. .Hmmum>flaa .m as. \mo. va.-\mo.u am.-\mo. mm. \NH. Hoflmz\omdm .m ma. \NH. om.-\po.u mm. \mo. mmusoo\ommm .v Hmcofiumm hm. \mm. Hm.-\ao.u unameuom .m mm.-\¢a. numucmum .m swam .H mmmmm m m n m m s m m H HMCOHumm HMOQH mmHQMflum> Hmcoflum~flcmmuo .Hm>0q Hoocom we .mmanwfium> HmcoflumNficmmuo mCOE< mCOwumHmquUII.ma mqmda The a groups were n I used the pr correlations 1 tionships of . groups) to ad. reality, The Gm»; we 99 The anticipated intercorrelations within the two groups were neither as strongly nor as clearly demonstrated. I used the principles of cluster analysis (basing groups on correlations between each pair of items and parallel rela— tionships of each item in a group with the items in other groups) to adapt the theoretical groups to empirical reality. The two groups thus separated were: Group A: Hierarchy Specialization by major or minor Universalism Specialization in courses taught Expertise Specialization of support staff (weakly related to rest of group) Group B: Standardization Formalization Centralization-~Principal (weakly related) in secondary schools only Centralization--Teacher (weakly related) in secondary schools only Except for the inclusion of hierarchy in Group A, and cen- tralization in Group B, these groups correspond to the theoretical separation of rational and bureaucratic modes of organization that Udy and Stinchcombe suggested. Thus any reliance on one overall summary measure of bureaucrati- zation is likely to lead to confusing and contradictory results. Since the internal correlations within the groups are not exceptionally strong, I will continue to consider the effects of each dimension separately, rather than considering the effects of rational versus bureaucratic groups as a whole. Furthermore, in Group B, the weak relationships and both font. us that, even was to be dell be a mistake ‘ element of or: nOrlbureaucrat: well. The ne ti‘mShip of fc Positional var teachers and p available in t to tax level i. to see whether ables were Strr preSe L . Sch 1 00‘s Similar] 100 relationships between the two measures of centralization and both formalization and standardization should remind us that, even in Weber's idealized bureaucracy, authority was to be delegated to apprOpriate levels. Hence, it may be a mistake to consider centralization a bureaucratic element of organization--perhaps a third type, irrational, nonbureaucratic but authoritarian must be considered as well. The next step in analysis is to consider the rela- tionship of formal organizational structure to school com- positional variables so that their eventual effects upon teachers and principals may be assessed. One might expect 'that school structure would be affected by resources available in the district. Since SES level is related both ‘bo tax level in the district and human resources in the community (such as education level of parents), I checked ‘tc>see whether correlations among the organizational vari- ables were stronger for schools with similar SES. Data Irresented in Table 14 offer some support for the idea that SChools similar in SE8 share some similar organizational features and patterns. Partialing out the effects of school SES produced nc> significant changes in the correlations at the elemen- tary level. At the secondary level, however, it appears t"halt SES had masked the strength of the positive correla- tions between hierarchy and three variables: specializa- tJJDTl by subject, Specialization in majors and minor area . l j 1 _ .r: 3» i RPvK/nvnh Havfiyino 7.. 3 . $.23“ "375:0 HGCCflJTNwCCCLC CECE< UZ.L..LE~QLLCCII-VN WfiQQHF 2101 humucoemam mm .Mumpcoomm ma anaconom mm H z humpcooom\>nmucmfimam ma.-\ha.- -.u\mv.r om. \mm. mo. \VH.: Hm. \mo. em. \mo.: mm.-\mo. ~m.-\mo.u mo. \ o (mum comm ms. \mo. ~m.-\mo.u hm.-\ma. mm.-\mm. mo.r\so. Ho. \am.: an. \GH. am.-\~o.u anamuucoo mm.-\n~.- sm.u\ma. mm.u\mo. m~.u\ma. as. \ o as. \ o ms.-\aa.n muflmuucmo ms.n\mo.u mm. \mo. no. \NH. mo.-\ao. mH.u\mH.u mm. \ma. mmauuomxm no. \Ho.| ma. \ma.r no.u\v~.l mm.|\mH.l om. \ma. .muo>ch mm. \mo. sm.:\vo.u mm.:\so. Hm. \oa. none: .oomm mo.|\~H. mm.|\>o.l vm. \mo.l vownnsm .owmm. mv. \mm. o \ o .HMEHOh mv.|\ma. .pcmum . .Hmw: enucou mlucou omfluummxm .mum>fl:D Hoflmz .nnsm .HmEuOh .Ucoum .uwwm .oomm .oomw .mmm mo muoowmm mom pmaaouucou .Ho>wq Hoonom an .mcoflmcmsfla HocowumNflcmmuo mcoad mcowumamuuourl.vd manta and expertise stronger nega had appeared zation and fo tion with 8135 and Specializ with formaliz minOl' areas m : . . romaliZathn tional Cluste the Effects 0 organizatiOna Consist of hi of Ce . ntrallze 102 and expertise of teachers. Furthermore, hierarchy has a stronger negative correlation with standardization than had appeared when SES was not controlled, while standardi- zation and formalization show a stronger positive correla- tion with SES effects partialed out. Since both expertise and specialization by subject now show weaker correlations with formalization and since specialization in major and minor areas now shows a stronger negative correlation with formalization, the evidence that two distinct organiza- tional clusters exist appears quite a bit stronger now that the effects of SES have been separated from the direct organizational intercorrelations. One cluster appears to consist of hierarchy, universalism and all aspects of specialization and expertise, while the other is composed of centralization, formalization and standardization. Out of key with this interpretation is the fact that controlling SES also reduced the correlation between expertise and specialization by subject by .11. In every other case, however, the idea that two distinct clusters can be used to draw the broad outline of a school's structure received further confirmation from the purge of SES effects. One must ask, though, why no similar pattern is found at the elementary level. One possible answer, which might also help to explain the generally different patterns of correlation observed at the two levels, is that the range of variation in students and in tasks is narrower at the elementary level than at values on 50”" certainlY tqu zeasures 0f ‘3‘ these variabli research on t" than the prESE the number of zational varia research on SC enough meaning type of organi to get a reali It is tie very diffe scnools at dil 103 level than at the secondary, with not much overlap in values on some organizational characteristics. This is certainly true for hierarchy, teacher expertise and the two measures of centralization (see Table 11 for level means on these variables). If this is the problem, then future research on this question should discriminate more finely than the present data allowed me to do, so as to maximize the number of intervals of measurement of several organi- zational variables. However, it is also possible that research on school organizations alone cannot produce enough meaningful intervals on many variables and that the type of organization cannot be so narrowly defined if one is to get a realistic picture of organizational patterns. It is also possible that part of the answer lies in the very different tasks and problems faced by secondary schools at different SES levels. All elementary schools must inculate basic skills--the traditional three r's, at least. However, some secondary schools must prepare students for college entrance, a situation which might encourage the hiring and evaluation of very expert and specialized teachers who can stimulate creativity and who can work with highly motivated students. Lower SES schools may face a different set of tasks: keeping students in school, improving basic skills and controlling antisocial behavior. The next step in analysis is to consider the rela- tionship of formal organizational structure to school compositional variables so that their overall effects upon teachers and principals can eventually be assessed. Measul variables (pe, student variai homogeneity) a sex) . The Cox Organizational The Inc are StatistiCa few are signi f 'l',’ :0; OSitions b lily StrOng rel reassuring Us structure Carin cospoSitiOn’ s organizatiOn sch0°18. Perh llaimiZes the 104 Interaction of Organizational and 99mpositional Variables Measures of school composition inCluded two teacher variables (percentage female and level of consensus), four student variables (SE8, SE8 homogeneity, ability and ability homogeneity) and two principal variables (experience and sex). The correlations between these variables and the organizational variables appear in Table 15. The most important contribution of these data is that none of the organizational-compositional relationships are statistically significant for secondary schools and very few are significant for elementary schools. Since the propositions being examined in this research do not predict any strong relationships, this observation is useful in reassuring us that any eventual effects of organizational structure cannot be attributed to hidden effects of school composition, since there are few consistent patterns between organization and aggregate compositional data for the schools. Perhaps the uniformity of state legal requirements minimizes the potential effects of differences in district composition. Turning to Specific compositional variables, one sees that organizational patterns are only weakly related to percentage of female teachers and to sex of principal.. Elementary principal experience is only weakly related to hierarchy, formalization and specialization B. Of these three, formalization is the strongest correlate of experience. Since female principals are usually more ms. - : 3 ma .2 (. mU~QEwLQ> ~ECCwuNmOLECU lfll‘llLI-Il" III' I I ' .I\ b. d. [1.“ Ann 'I‘: [.III..I.".I- In nil". .1 II“! I.“|ol. I. l}.-1h I i in. .a IUHII. .-'Iv. Ilrl lain!- DQ‘N,I Ill J. .1...z.. w..‘1 u .Ilh .,t| tad-r I|| .III..|-d-[ I: l .. .dOOZUU. [a0 HU>OJ >3 .mQHQQwLnQ» H$EOnumm0QECU VS... utCOfiuENHC-wthc CQCRQCQ WECHdro~.v.N.~Crvll.mN .nsQQSa .mHoE mammoocflum whopcooom Ham “m u m .2 u H« %Hmpcoomm MOM ma i105 “mumucmfimaw How mm H z humpcoomm\>umuco8mam mm. \Hv. om. \mm. mm. \mm. om.|\mo.| Il\ov.| |n\mo. .moeom .psum .mH om. \mm. oo. \oo. oo.u\oa. u-\Hm.n ua\mm. soaaon< .osum .oo oH. \mm. m~.u\om.u au\oo.u su\mo.u .mosom mum .oH mo.u\ma. uu\oa.u u.\ o Hm>ou mmm .ma ul\mv. I|\vm. .uomxm .cflum .va Il\mo. «xmm :flum .ma .SOMOB m w .NH oH. \oo. om. \sm. mm.-\sm. oo. \om. so. \mo. --\om. uuxmo. mmaunmmxm .HH o~.u\mm. ms.-\oo. om.u\oo. oo.u\oa. oo. \oa.- -u\mm.- --\m~. .wum .ommm .oH o~.u\Ho.u mH.I\so.a Hm.u\~o.u mo. \H~.r so. \NH.- -u\mH. -u\~o. a .uucmo .o o \oo.u Ho. \oH.- oo.-\mm.r as. \oo.- mo.u\os. --\mm. --\H~.- o .oucmo .o oo.-\-.u om. \sm.n mm. \oo.- Ho. \H~.- mo. \mo.- -u\mo. u-\mo. .mnm>oco .o om. \oa. mm. \oH. oa.-\om. om. \om. ma..\om. uu\oo.u -u\oo. noon: .omdm .o om. \HH. mo. \Ho. mo.u\oo. om. \oH. so. \oo.- u-\so. u-\om.- .mnom .omdm .m oH. \oa. om. \oH. oo. \m~.- so. \mo.u sm.-\mm. uuxam. --\oH. .Hmsuom .o ~H.-\Ho.- mm.-\~a. oa.-\mo.u oH.u\HH. -.u\oa. -u\HH. --\oH. .eceum .o mo. \oo.u Na.u\oa.- oo. \oa. oo.-\HH. oo.-\o~. un\ao.u uuxoo. Hon: .m oH no oH ma «a ma NH mmfinmoum> sac mmHQMHHm> Hmcowuwmomeoo .Hoonom mo Hm>oq >9 .mmHDMHHm> HMCOwufimomEou new HmCOwumuwcmmno cmmsumm mcofluoamuuoonl.ma mqmda experienced principal se> t0 SECOndary . the ' . “ luteBrat: Years of ser Stho OlS . p r my Weakly Except that 106 experienced and since there is a weak correlation between principal sex and formalization, it is likely that experi- ence alone is not critical to greater formalization, but rather greater experience of female principals. Male principals may use different administrative tactics as they gain experience. Only size of Specialist staff is positively related to secondary principal experience, suggesting that seasoned principals may be entrusted with larger schools offering more services and requiring greater administrative Skill in the integration of this variety of services. Alternatively, assignment to more complex schools may be a reward for years of service in less prestigious, inadequately staffed schools. Principal experience at the secondary level was only weakly related to the other organizational variables, except that experienced secondary principals see themselves as part of a more decentralized authority structure than do elementary principals. This difference suggests to me that experience brings principals into closer agreement with teacher perceptions of the decision-making structure (see Tables 7 and 8, Chapter III). When the student compositional patterns are con- sidered, school level makes a great difference in the nature and strength of relationships. It is important to recall here that, on the whole, elementary schools scored at the extremes of the range for SES and ability and had great internal homogeneity, due to the location of elementary schools in di while seconda internal vari ofdistinctix' htrm turn t student compo: At the toaOgeneity we tosize of spe menay infer fiagmse diffi schools and th trained teache dent Problems Personnel reso Schools ’ Wi th < llkEly to Part: ms' a“thougl 107 schools in distinctive neighborhoods in each district, while secondary schools had less extreme values and greater internal variance, since they often covered a multitude of distinctive neighborhoods. Bearing these facts in mind, let us turn to a consideration of the interactions of student composition and other variables. At the elementary level, both types of student homogeneity were strongly related to teacher expertise and to size of Specialist staff. Based on these correlations, one may infer that administrators are better able to diagnose difficulties and weaknesses in the more homogeneous schools and that a concentration of specialists and well- trained teachers in schools with clearly identifiable stu- dent problems might appear to be a wise use of limited personnel resources. Possibly the more heterogeneous schools, with a wider variety of student problems, are more likely to participate in visiting teacher programs of various sorts, although such data were not available for this study. Although none of these correlations involving stu- dent compositional factors is statistically Significant, the data do suggest some differences from the patterns uncovered at the elementary level. Summing up these differences, secondary schools Showed a weak positive relationship between both formalization and teacher expertise and the compositional variable, student SES. Furthermore, SES homogeneity was weakly but inversely related to expertise and size of specialist staff while ability homogeneity was 108 inversely related to Specialist staff as well, unlike the strong positive correlations found at the elementary level. While these relationships are not critical to the preposi- tions at issue in this research, it is important to estab- lish that the organizational patterns cannot be considered reflections of these inconsistent relationships of compo- sitional and ecological variables. Furthermore, the weak and inconsistent relationships among organizational, staff and student variables suggest new areas for future research and confirm once again the apparent primacy of SES in predicting student ability levels. In Table 15, ability is also related, though not to a significant extent, to such rational elements of organization as specialization and teacher expertise, as well as formalization in secondary schools. SES is strongly related to each of the student variables but weakly or not at all related to the teacher and principal variables. Further research on the effects of bureaucratization upon student learning, independent of SES, might be a fruitful area of endeavor. Similarly, a longitudinal study of the ecological sources of school organizational change might help to assess the real importance of SES in explaining student ability. When staff and student correlations are considered, it appears that sex of staff is somewhat relevant to student compositional patterns. Male principals are found in schools with less homogeneity and lower student ability and male teachers predominate in schools with lower student ability. Th. district off; male and fem; sex and admir Cons: Positional Va tistically si ties concerni relationships tions link SE! allertise and tional feflture fitter-ms are features that f-. sndent hOmOge 109 ability. These surprising correlations may simply reflect district officials' beliefs about appropriate places for rnale and female staff, rather than any causal link between sex and administrative or teaching ability. Considering the whole range of data concerning com- puasitional variables, very few of the correlations are sta- tistically significant and yet a few interesting possibili- 'tjues concerning the observed trends and contradictions in reelationships have been detected. The strongest correla- ‘thDnS link SES, ability, student homogeneity, teacher encpertise and size of specialist staff. The only organiza- ‘tiJonal features that seem to be determined by ecological guitterns are the expertise and Specialist staff variables, features that a district official clearly might link to Student homogeneity of problems. Conclusion This chapter presented data testing the rationale for treating elementary and secondary schools as struc- 'turally different types. There were a few clearcut dif- ferences between the two school levels on organizational (iimensions, however, neither level was significantly closer ti) the ideal bureaucratic type. In fact, the data presented hei‘re offer evidence for the treatment of bureaucracy as a InLll‘ti-dimensional phenomenon. Turning to the interaction or organizational and Sch001 aggregate variables, I found that the observed patterns 0 whole, be compositio. tional pat variables a structural u trends rela tional vari eXPEI'tise a;' gate Variabl effects of c is assessed, varicibles . 110 patterns of organizational structure could not, on the whole, be attributed to differences in SES and staff composition, although SES interrelated with some organiza— tional patterns. Furthermore, the school compositional variables appeared to be largely independent of the formal structural variables, although there were some inconclusive trends relating principal experience to certain organiza- 'tional variables in elementary schools and relating teacher o um3om :30 emm.v v ma.v huwuongsfi Homeosflum mo momsvmpd ««~«mm.H A mN.v Hm>mq mfiflxmz IGOHmHomQ Hmowmwa "mumaomme mm H z «m>.H v mm.H mcflpafldm Hoonom mpflmuso Hm3om czo lace u as «mm.e v Hm.m wueuosusm :30 mo momsvmpm A.sus¢ uoeuumea u m .HMQHUGHHm u m .ngomme u as .mm.~ A mm.e Hm>ma omega: leewmflomo HMUHQMB "mammwocwnm unaccoomm humucmEmHm I.” Elna!!! .moaummsooo ha emnoeumamm umaom mo meowummonmmuu.ma mamas teacher-pri Although te their prime. considerabl:I ERCES. Furt references t their buildi heads‘Whilg interaction Superintend, 0f most of 4 if prinCipa fact, the m. ms do not Possibly in it is impo: between We: 31:01 . 121 teacmer-principal differences is less clear, however. Although teachers reported slightly more personal power than theuir principals, the elementary-secondary differences are considerably stronger than the principal-teacher differ- ences. Furthermore, the score for teachers included refEarences to their relationships with several peOple within fluaigr building--other teachers, assistants and department heads--while the score for principals refers only to their interaction with other principals, the school board and the superintendent. Since the district officials are the source of nuost.of the principals' authority, it would be surprising if purincipals all felt powerful vis-a-vis that group. In faCtJ' the most we can infer from this data is that princi- pals do not report more power or autonomy than do teachers. Employee Relations.--Turning to another factor POSSibly influencing the behavior of teachers and principals, it is important to assess the character of the relationship between worker and supervisor and between worker and other 9r°ups making demands on the organization. Several clusters of Variables measured these relationships: administration poor, principal supports teachers, teacher organization hinders, cooperation expected of outsiders and cooperation e"ipected inside school. The original propositions suggested that schools with more hierarchical levels, greater SPecialization, formalization, standardization and reliance on universalistic criteria would be characterized by better supervisor-v work roles 3 likelihood c apparent as autonomy (in less Strain Schools fit Personnel f i secondary pe; Smoother rel. than element. analYSis Of . IESponSeS co: Surp Ha though the 1 122 supervisor-worker relations because the more clearly defined worflc roles and the lack of particularism would reduce the likralihood of conflict. Occupational differences should be apparent as well, since workers with greater power and autrnnomy (in this case, secondary personnel) should report less; strain in the authority relations. Since secondary schcxols fit the organizational requirements and secondary personnel fit the autonomy requirements, I expected that seccuudary personnel would report more c00peration and smoother relations with school and district administration than (elementary personnel. Table 19 presents the results of analysis of variance for the effects of school level on respcnnses concerning these variables. Surprisingly, four of the five means indicate that the Galementary teachers perceive their environments as more COOPEErative and supportive than do secondary teachers, even thOugh the latter report more autonomy, more influence and cleazner, though more complex, organizational settings. In exPlaining this unexpected reversal, one might propose that“, in order to accomplish their jobs, elementary personnel muSt; develop channels of cooperation with others because 0f tflneir individual lack of autonomy and power. The more autonomous secondary teachers may be forced to bargain, rather than cooperate freely, because of a more delicate balance of power and autonomy in secondary schools. Also, Since elementary principals are considered to have little authority, it may be more difficult for teachers to blame TABLE 19.“! ( H“ c°°P~ Pare C00p. Dist COOP- In S- POOI‘ Admin tll'ation (Distric Level) TO HindErs Teachers Teah : fig COOP. Outs; COOp. Insi‘ POOI‘ Admin tration SchOol L Prin, Supp Hinders TeaChers tSignifiCanC erg. SEE AP Pendix Hard]: “9 ofq 123 TABLE l9.--Mean Reported Cooperation and Support, by Occupation and School Level. Significance Elementary Secondary Level* Principals: COOp. Parents 4.03 > 3.67 .09 Coop. District 3.68 < 3.83 >.10 COOp. In School 4.92 > 4.73 >.10 Poor Adminis— tration (District Level) 1.84 < 1.99 >.10 TO Hinders Teachers 1.69 > 1.51 >.10 Teachers: C00p. Outsiders 4.14 > 3.99 .0002 COOp. Insiders 2.99 > 2.65 .0001 Poor Adminis- tration (at School Level) 2.27 < 2.53 .0001 Prin. Supports T. 3.58 > 3.13 .0001 T0 Hinders Teachers 1.90 < 1.99 >.10 *Significance in F test with N = 52 principals, 1404 teachers. See Appendix C and D for contents of clusters and exact Wording of questions. problems on these principals than for secondary teachers to fault their more independent, more powerful principals. Since teachers were not asked to evaluate the performance of district officials, it is impossible to see whether elementary teachers might be more critical of the adminis- trators they see as most powerful. Whe: level makes cooperation tary than 56 be attribute other non-st Propositions 'i'l'liCh result is intereSti] more CO0E>.10 Importance Status 3.69 3.42 >.10 Importance Students 4.43 3.97 >.lO Importance Intellect 3.81 3.57 >.10 Re-enter Education 4.37 4.00 Teachers: Importance Extrinsic 3.28 3.17 >-10 Importance Status 3.44 3.21 .0001 Importance Intrinsic 3.67 3.58 >.lO Re-enter Education 4.11 3.76 .0001 h *Significance of F-test, 2-tailed, with N = N = 1404 teachers. (of work to self is justified. 52 principals, However, since elementary ‘teachers' means on extrinsic and status reasons for remain- :ing are significantly higher than secondary teachers' means mUQ WOOHNMUMH NU-3 COfiUfiwnwjnUUO \flad \Hflwccomyhflvm «.HHWH~J0 H~Uufi3 fiUHu-1v-pJufiv~ EflwfivH nah-n. MUPVF~5=~dtHnUn~Z~fi H~Hflufiwzlluqmw .fisfififif‘.“ .mumnommu «ova z .mammwocflum mm .mumnomou How =muamflommm as: .mammwocfium Mom =Hoosom cw: mance monmmmaaoo mmoHU«« z sues ammuum cmaamunm now mnemoamacmflm« 130 mo. mo.m A o~.m mocmwhm oocmuHomEH Hooo. ¢H.m A mm.m mHoHHmmom mocmunomEH Hoo. mm.m v om.m «*.HHOU mmoao mocmunomEH Hooo. mo.~ A mm.m mosmmoaaoo omoao omnmnoxw oa.A mm.a mm.H Hoonom mnflmcH mmcmsoxm moo. nm.m A mm.m Hoosom moflmuso chmnoxm “muonomoe oa.A mm.m mm.m mnowummsm mocmuuomeH 0H.A mm.m an.m ««.Haoo mmoHo mocmuuomsH oa.A oo.w mn.m mammoaaoo mmoao omcmnoxm mo. mm.m on.m Hoosom ca omcmaoxm oa.A v¢.m mv.m Hoonom moflmuso omcmnoxm “mammwocflum u c humocooom humucoEon «Ho>oq mocmoamwcmwm .Hm>mq Hoosow 6cm coaummsooo >9 .HOGGOmHom Hmnuo nuw3 omcmnoxm mooH paw mocmuHomEH amszI.HN mamfia pare pro: coli deli wor! 131 parents, friends and district officials) represent less professional idea sources but exchanges with close colleagues suggest a strong professional orientation as well, perhaps elementary teachers simply talk about their work more with everyone and have more opportunity to talk to everyone, given the smaller size of staff. Since their clients are very young, perhaps considered incapable of relaying accurate information to parents and since these children are put into programs rather than allowed to select classes, elementary teachers may be forced to confer more frequently with parents and other close teachers and may need parental cooperation in order to carry out their tasks of socializing incompetent clients. At the secondary level, where students sign up for classes and are acknowledged as semi—rational, almost fully socialized individuals, such comprehensive efforts on the part of teachers are not necessary. Data on the tOpics usually covered in idea- exchange might illuminate this problem. If elementary teachers continually search for ideas about how to work with specific children or handle their parents, these problems would require solution every year, as each teacher meets a new group of students. If secondary teachers con- sult on less socio-emotional problems, the need for con- tinual interaction with colleagues might be less, the observed differences between elementary and secondary exchange networks would be understandable, and the refer- ence group identification would not be problematic. Lac tea col ans ide the col 1'? a r SEC thy fer: 59cc tha; 132 Lacking this kind of data, we can only say that secondary teachers are more concerned with the good Opinion of close colleagues and that elementary teachers have stronger exchange networks with close colleagues. Perhaps the real answer is that secondary teachers are unwilling to solicit ideas from the people whose opinions are most important to them (see Blau's suggestion that consultation with a colleague shows deference to that colleague, 1955). When the principal data are considered, only one variable shows statistically significant differences: secondary principals report more idea exchange with their teachers and assistants than do elementary principals, who appeared to engage in fewer exchanges of any type. Support- ing further, though not at a significant level, the idea that secondary principals have a stronger professional orientation than elementary principals are the level dif- ferences in importance of superiors and colleagues. Secondary principals are more concerned about the opinions of close colleagues and less about the opinions of superiors than are elementary principals. On the whole, there seem to be somewhat stronger reliance on professional, rather than organizational or non-work reference groups among secondary personnel than among elementary, confirming the original propositions. g has been scrker be be consi< preposit on the b Sean sco Vith Wor ”JG D; f And B. r 'l ’1 u- l /: 133 Job Satisfaction.--Turning to another variable that has been assigned an important role in most research on worker behavior, variations in job satisfaction levels will be considered, so that one of the main alternatives to my propositions about worker behavior can be compared to mine on the basis of explanatory power. Table 22 presents the mean scores for the four occupational types on satisfaction with work. For contents of each cluster, see Appendix C and D; for specific questions on the survey, see Appendix A and B. TABLE 22.--Job Satisfaction by Occupation. Significance Elementary Secondary Level* Principals: General satisfaction 3.94 3.81 >.lO Teachers: Satisfaction Physical 3.54 > 3.22 .0001 Satisfaction Non-Physical 3.16 > 3.08 .02 Satisfaction Superiors 3.96 > 3.83 .0002 *Significance of F-test, 2-tailed, with N = 52 principals, N = 1404 teachers. pt? 88 a1 si sa ti 134 The data summed up in Table 22 show that elementary personnel expressed higher levels of satisfaction than secondary personnel for all categories of satisfaction, although the principal differences are not statistically significant. Since a score of 3.00 means "somewhat satisfied," it appears that no major areas of dissatisfac- tion were isolated for either level. Both sets of teachers found greatest satisfaction in relations with superiors and least in the non—physical aspects of their work. Comparing principals and teachers, it is likely that a single summary score for teachers would be lower than the very high princi— pal scores on general satisfaction but both groups reported the more than moderate satisfaction with their jobs that countless studies have led us to expect (Department of Labor, 1973). In summing up the occupational differences in perceptions of work situations and definitions of self, there seem to be very great differences between elementary and secondary teachers and between teachers and principals. Level of school exerts fairly small effects on principal variables, except in the areas of power and autonomy. On the whole, compared to elementary personnel, secondary personnel appear to feel more powerful, more autonomous, and more concerned with poor administrative practices; to get less c00peration from parents, superiors and colleagues; to be more concerned about the opinions of close colleagues but less concerned about superiors' and friends ' remaining between e1 these vari independen which is t occupation 135 friends' Opinions; to be somewhat less satisfied with work conditions, less willing to re—enter education and, finally, to be less enthusiastic about any of the reasons for remaining in their jobs. Since the demographic differences between elementary and secondary personnel may influence these variables, further analysis of the effects of the independent variables will control for the sex of respondent, which is the clearest demographic variable related to occupation. When teachers and principals are compared, principals report greater willingness to re-enter education, greater importance for each type of reason for remaining in educa- tion, more emphasis on professional reference groups, stronger collegial exchange networks, less awareness of union or administrative impediments to job performance, greater expectations Of OOOperation within the school but less outside the schools, greater autonomy and authority but slightly less personal influence over others at work than did teachers at the same level. The next step in analysis is to see whether there are occupational differences in behavior, as the documented differences in demographic and especially in perceptual and self-definitional variables suggest. Behavioral Differences by Occupation In this section I will examine overall occupational differences in the behavioral dependent variables. This prelimina importanc demographl possible I 1013 turnO‘I to dedicaI in this saI Should be 5€Condary Personnel, tions. Tb. trends. E fiCantly 5. report 3 i g 136 preliminary step allows the eventual comparison Of the- importance of specific occupational variables (such as demographic or perceptual) with organizational factors as possible determinants of behavior at work. Absenteeism, job turnover rates and several behaviors typically related to dedication of professional workers will be considered in this section. The propositions suggest that principals should be more dedicated to work than are teachers and that secondary personnel should be more dedicated than elementary personnel. Table 23 presents data concerning these proposi- tions. The data presented in Table 23 suggest four overall trends. Elementary and secondary principals are not signi- ficantly different in their behavior; secondary teachers report significantly more evidence of dedication than do elementary teachers; secondary personnel in general, have higher mean scores on dedication variables than do elemen- tary staff; and, finally, for the items where comparable data is available, principals generally exhibit more dedi- cated behavior than do teachers. Considering first the principal data, there is no statistically significant evidence that secondary principals Surpass elementary in dedication. However, the data do n0t refute that possibility, since the mean scores for Principals suggest that the trend is in the predicted dtrection. Since elementary principals reported lower hOpes of moving up the administrative hierarchy in any Hr n MC COLOkCCU CCODDW u MHMAWHHUCfluHAN «Tami/01H QUCflOflMflCDflW \nhmmufloomm \nHMUCUEmvHW ”fli‘ l, 1.!!! 1 “Hi oCOfiUQANJUUO «an \CCfiC-wujfiadnva N0 GUCSMUHKVWN .H-‘Ln.vH>-M-.J: n~f£uvzliuMJN ...~.Nm~s\.fl 137 .moumu Ho>ocusu moonlm “mammHOCflHm How muomuucoo umomlwaa .mnosomou vova u z .mammflocwum mm H 2 now m cmasmuum mo mocmoHMflcmflme mo. mm.H v mm.a Hoonom Houuom mo. Hm.a v mm.a .cHEO4 “momom Hooumu mo. N¢.H v hm.H umbamz woe Honuo .02 Ho. om.~ v om.a smumoum mumum Ho. bH.H v No.a Hoowmmo oumum Ho. qo.m v mb.H Hoofimmo Hmooq Ho. ov.m A Hm.m .moumom mcwvawsm Ho. mv.m v mo.m oouquEOU Hmooq Ho. vm.H v 5H.H mouuHEEOU oumum oH.A Hm.v v mv.¢ umbeoz "ooflcD mo. wo.~ A >v.~ momusoo oouflnvom waco 0H.A mm.m v mm.~ wuHMfioomm a“ momusou oxma m.Hm A m.om «*0umm H0>OGHSB mo. mo.m A mm.v Hmo>\ucomb¢ mama No. wn.m A v>.v mooaonomcou Ocouud No. vH.v v om.m nucoa\mcflomon mafia oa.A mm.H n mm.a bu:05\mamcuson .02 mo. vH.m v ma.v sucoE\mxoom .oz “mnonomoa A.qm xv ma w m.ma A m m.m~ .om o>moq u.uoomxm Hooumu A.qm xv m: w m.m~ v m m.ma do ":oHumuoomxm umonmu OH.A ~>.H v mw.a momom Hoonmu 0H.A oo.m v Hm.~ >ua>fluo¢ GOHGD A.am xv m: w m.ma A w m.m~ ««.«mumm um>oause 0H.A ma.v v mn.m moocouomcou ccoupd "mammwocwum «Ho>oq oocmOHMflcmwm wnmpoooom mumucofioam .coaummsooo an .cofiuMOfloon mo mocoww>m Hmuow>msom :mozll.mm mange school di represent the labor L ficant efl elementary and read a teaChers, Scores for fewer Work- reading fOJI their COHtII I‘OPES and E Furthemore energy intcl are Imamhers than Second which refle pom of can (Secondary faCulty men] teachers Soc Ileth‘er the i“. . d unlOniSt etion aha 1y 138 school district, the higher elementary turnover rate may represent more lateral movement and more withdrawal from the labor force. Looking at the teacher data, statistically signi- ficant effects of teaching level are plentiful. Although elementary teachers attend more conventions and conferences and read as many specialist journals as do secondary teachers, every other indicator of dedication shows higher scores for secondary teachers. Elementary teachers read fewer work-related, non-course books, spend less time reading for work, take fewer courses, not required by their contracts, are absent more often, have lower career hopes and higher turnover rates than secondary teachers. Furthermore, elementary teachers do not simply channel their energy into union activities. While both groups of teachers are members of unions, elementary teachers score higher than secondary only on history as building representative, which reflects not dedication to work but the much smaller pool Of candidates for the job in the elementary schools. (Secondary schools average four or five times the number Of faculty members.) For all other indicators Of union activity, secondary teachers scored significantly higher than elementary teachers. Whether the same teachers are exhibiting both professional and unionist dedication will be considered later in corre- lation analysis of the dependent variables. 139 Summarizing the teacher data, secondary teachers exhibit considerably more dedication to work than do elementary teachers. The only contradictory case is that of attending conferences, for which the elementary mean was significantly higher than the secondary. This activity does not appear to be such clear evidence of pure dedication to work as, for instance, time reading, so the contradictory evidence in this dimension cannot be considered a strong refutation of the prOpositions. A case could be made that going to conferences and conventions represents an escape from the job rather than an attempt to enhance job perfor- mance. In the next chapter, correlational evidence will be presented to further determine the relationship of conference- going to the other variables representing dedication to work. Finally, comparison of the few equivalent items for principals and teachers Offers some support for the prOposi- tion that principals are more dedicated than teachers. Unfortunately, there is no way to compare the most critical elements--reading and absenteeism-~although the ease of substitutability for teachers suggests that their rate of absenteeism would be considerably higher. Table 23 shows, though, that principals are more likely to plan hierarchi- cal career movement than are teachers, have lower turnover rates (not a clearcut evidence of dedication on its own) and are somewhat more likely to report being active in their professional unions (using teacher responses about local committees and local Officers as indicators of low-level principal them than B occupatio to work, my QEnera the level Table 24 E l"egarding t0rs of Se in dediCat Supper—t fc tions high means fOr Th only Stati relating F SuppOrt the UnfOrtunatE tes t the In 140 low-level activity). Once again, the smaller numer of principals in a district may require greater activity from them than is required of most teachers. Before proceeding to an examination of intra- occupational differences in self-investment and dedication to work, it is important to evaluate the current status of my general proposition that self—investment level determines the level of dedication to work displayed by an individual. Table 24 presents overall conclusions drawn from the evidence regarding occupational differences in the two major indica- tors Of self-investment and summary occupational differences in dedication (these are taken from Tables 20 and 21). Support for my proposition would require that the occupa— tions highest in level Of self-investment also have higher means for indicators of dedication to work. The summary information in Table 24 shows that the only statistically significant inter-occupational patterns relating mean self-investment level and dedication to work support the proposition that these are positively related. Unfortunately, the correlational data do not permit one to test the notion Of causation at all, only the strength and sign of a static relationship. Within these limits, it is encouraging to note that principals exhibit both greater self-investment and greater dedication than teachers. While none of the occupational level differences in impor- tance of intrinsic aspects of work is statistically signi- ficant, it is somewhat surprising that the Observed I COHUQUflUOQ WQDOGQHHOU QWOHU QOCQUHOQEH UflmCHHUCH Hm>mq "QUCQHQMQQ "UCUEUmmxyCHINHQW «COHUMQJUUO oCOHUerHfimUUQ mvnhm. UCOEUUOKICHimHHom «Ho Hm>oq :fl EQUCMVHQIHMJMQ Hfiucojfinuuwnwauv‘vo nNnU \fianwthrzumlioVN. 31:34.35 141 .Hm one om mOHbme oomllvGMOHMHcmHm >HHMOHumHumum mum mOOCOHOMMHU moose H Hsmwm Hsmwm 30H mumwcooom Hzoq H3OA anm humucoEOHm BO 30a 30 who OM09 H H H H A a cmHm anm 30H mumocooom 30A 30A anm humuGoEOHm Hausa Hesse flames mammnoanua GOHuMOHUoo mo5@MOHHOU omOHU oocmuuomEH OHmcHHucH Ho>oq "monouomom uucoaumo>cHIMHom «:OHummsooo .GOHHMOHOGQ can ucoaumo>cHIMHom mo Ho>oq GH moosOHOMMHQ HMQOHHMQDOOO mo kHMEESmII.vN mqmde w patterns from the investmen group. U Self-inve self-inve. the PrOposI Ir largely cc are more (5 bility of OCCupatiOn dieting be and distri 142 patterns on this dimension Of self-investment are different from the significant patterns on the other dimension of self- investment, importance Of close colleagues as reference group. Using only the latter, significant data concerning self-investment, secondary personnel do exhibit greater self-investment and greater dedication to work, supporting the proposition. In summary, the data presented in this section largely confirm the propositions that secondary personnel are more dedicated than elementary and that principals are more dedicated than teachers, although the lack of compara- bility of some data weakens these conclusions. Since these occupational variables and situations are critical in pre- dicting behavior, further analysis of the effects of school and district organization will be carried out separately for each occupational type. Conclusion In this chapter, I have presented evidence that elementary and secondary principals and teachers are dis- tinctive in demographic characteristics, perceptions of work situations and behavior. The next chapter will consider the effects of organizational variables within each occupational type, to answer the question whether occupational type subsumes most Of the variation in organizational structure and, if not, whether organizational structure has similar effects for across occupational types. CHAPTER VI FACTORS INFLUENCING INTRA-OCCUPATIONAL DIFFERENCES IN PERCEPTIONS AND BEHAVIOR OF TEACHERS Introduction This chapter will address the prOpositions that certain school organizational variables influence percep- tions, behavior and, less critical, attitudes of teachers in the school. The preceding two chapters have established some evidence supporting my predictions that occupation helps predict teacher and principal perceptions and behavior, and that the elementary-secondary occupational distinction is strongly related to several school organiza- tional variables. From these two sets of findings, one can see that it is important to maintain the elementary- secondary distinction in this chapter, in order to separate occupational from organizational influences. At the same time, controlling for occupational level will greatly reduce the range of organizational variation whose influ- ence is to be measured in this chapter. 143 '1 of the pr that seco less cent elementar secondary Professior teachers. PIOpositic occuPatio This Chapt Significan elements a Pation an d be confide zational t °°Cupation Sian the establishi2 PIOPOSitiO: Be: oonSidEr t}I in dedicati Ian able 144 This approach, then, provides a very stringent test Of the propositions at issue. We have already Observed that secondary schools are significantly more hierarchical, less centralized and have larger support staffs than do elementary schools. Furthermore, we have seen that secondary teachers exhibit more dedication to work (in both professional and union activity) than do elementary teachers. These Observations, while supporting the original propositions, are not conclusive, since the effects of occupational and organizational variables are not isolated. This chapter will attack that problem. Only if we observe significant correlations between individual organizational elements and the dependent variables, controlling for occu- pation and personal characteristics Of re3pondents, can we be confident that the Observed relationship between organi- zational type and teacher responses is not due merely to occupational or personal differences in responding teachers. Since the range of organizational variation is significantly narrowed by controls on school level, the difficulty Of establishing significant and meaningful support for the propositions is magnified. Before testing the critical propositions, I will consider the major alternative explanations Of variation in dedication to work: personal, family and satisfaction variable. in works sex, age Cal vari. in eXpla: elementar deciicatic are Press C tics Of 6 related t The Stl‘On Workers: participa. 145 Effects Of Personal and Family Variables 'Upon Dedication to Work Typical conventional explanations for differences in worker dedication levels allude to the importance of sex, age, experience, number and age of children as criti- cal variables. TO check the importance of these variables in explaining differences in dedication levels of both elementary and secondary teachers, correlations between dedication indicators and personal and family variables are presented in Table 25. On the whole, the personal and family characteris- tics Of elementary and secondary teachers are only weakly related to the indicators Of self-investment and dedication. The strongest relationships involve age and experience Of workers: older workers and those with more experience participated in union activities significantly more Often, particularly at the local and building level, are signifi- cantly less likely to take courses, whether required or not and are slightly more likely to read educational journals. Older secondary teachers are slightly less likely than their younger colleagues to be enthusiastic about re-entering education if they could make the choice again. Age and experience have almost no effect on career plans and have no significant effect on any of the indica- tors of level Of self-investment. Looking at the other individual variables, one notes that sex has almost no effect on self-investment or i (ii iii( . < 1C (iii. i..i...:~ 2.1—H «£0 3 _ a m H IN _ u N H I: t aLkm Q «Xi! ...l.< .-~CLX..,~ g 21:5 31>2~ . 13.5. .::... SC _ 2r... H _ ..> >~ TECH 1:... Hr...Z.:.....L . ~..>ru._ Haunt—71.. >£ :C:~ of. u :& .parv . :7..— LP..H..—Vm...: «I... d Artty 3...»): H IL ~§..n x»; .4.» H 5 1 H .w- I .5; 5:. v \n ~ ~._:..._~ ....... ~ ....::|~-- ZIMI . .. AV :4: (L. 146 m “on u H«. Heruht mmn\moo n z xumpcouomxxumucoEon oo. \Ho. oo. \No. oo. \Ho.- mH. \ o oo. \oo.: oH. \oo. mo. \mo. oo. \oH. ucmmnd mH.u\vo. oo.u\Ho. oH.u\mo. NH.u\mo. so. \Ho. NN. \oo. NH.u\No. mH.n\Ho.u umucmuom oo.-\Ho.u oo.u\No.n oo.-\No. oH.n\Ho. No. \Ho. o \NH. NN.-\NH. MN.-\oN.- a.emuHsamm momuoou oH.-\No.u oN.u\NH.- oo. \oo. Ho.-\No. oo. \No. Ho. \NH.- oo.-\mm.r mo.n\oo.u mmmusoo NH. \No.r Ho.u\mo. NH. \mo. mo.u\mo. oo. \mo.-. NH.-\vo.u mo. \NH. oo. \hH. oe m>Huo< NH.u\NH.n No.-\mo. HH.-\NN.- oH. \oN.- HH.u\Ho. mH. \Ho. oo. \oo.- oo..\ o cmHa ummnmo No.u\ o mo. \No. Ho.u\No.u HH.-\No.u mo.u\mo.- oo. \ o «H. \oo. NH. \oo. some .one. mo.-\oo. No. \NH. so. \No. oo.u\No.- oo.u\oo.- mo. \Ho.u oH. \oH. oH. \HH. mHmcnsooH oo.u\No. Ho.u\oo. mo.u\Ho.u No. \Ho.u oo.u\Ho. oH. \Ho. oo. \Ho. so. \Ho. mxoomo No. \oo.. oo. \mo. No. \No.- oo.u\vo.u mo.-\oH.u mo. \mo.: NH. \oo. NH. \oH. .ucoO¢ oo. \oo. oo. \oH. mo. \No.u No.-\No.u o \mo.- No.u\HH.r om. \om. Ho. \No. .one .mon oH. \oo. oo. \HH. HH. \Ho.- mo. \ o Ho. \mo.a oo.:\oH.u Hm. \oN. Nm. \om. .Esoo Hmooq oH. \mo. oo. \oH. oo. \ o No. \eo.u No. \oo.- mo.-\oo.- NH. \oH. oH. \oN. .eeoo mumum so. \oo. oo. \oH. so. \ o oo. \oo.- oo.-\oo.u so. \No. oH. \oN. NN. \Nm. mos . oo. \Ho.- No. \oo. No. \mo.u oo. \oo.. Ho.-\oo.u No.n\oo. oH. \NN. oH. \oN. .mouo onmnm so. \No. mo. \oo. No. \mo.r o \Ho.- Ho.u\mo. Ho. \Ho.- oH. \No. oH. \No. .ouo oumum HH. \Ho.- mo. \No. no. \mo.: oo. \mo. o \ o so.-\HN.- oN. \HN. Hm. \NN. .one HmooH mo. \oo. HH. \HH. mo. \Ho.. Ho. \oo.- mo.u\oo.n No.-\No.- NN. \oN. NN. \oN. OB noose: "COHDMOHOOQ Ho. \HH.- o \oo.: Ho.-\oo.- No.:\mo.u mo. \mo.- o \Ho.1 Ho. \oo.. o \oo.. cmuHmoo a: Hooumu Ho.-\oo.u o \No.- mo.a\NH.u mo.-\oo.- no. \Ho.- mo. \Ho. No.:\No.u No.-\ o .masm Home Ho. \Ho.. oo.u\Ho. o \oo.. oo. \No.n mo. \Ho. mo. \ o Ho.u\ o oo.u\mo. .mmHHoO "mom no. \vo.a mo. \No. No.u\oo.u o \oo.. Ho.-\Ho.n Ho. \mo.u mo.-\mo.- oo.a\No. mmoHonmmnH ,o \Ho. oo. \oH. Ho.-\oo.n No.u\ o Ho.a\No.u o \mo. No.n\oH. No.u\oH. .uucH .asH mo. \ o oo. \No. No.-\mo.u oo. \No.u o \mo. o \oo. No. \so. mo. \oo. museum .QEH coupHHnu mH* mHINHu NHImu mu>m a .xom omd .ummxm acoEumo>cH u Hmuoa uuHom moHanHm> AHHEmm new HMCOmHmo E I .Ho>OH Hoocom >2 mcoHumHouuou .coHuonomo can DCOEDmo>cHerom an moHumHuouomumcU >HHEmm can Hmcomuomnr.m~ mqm<9 dedicati< are 51191 absent ar and about activity level. 8 in work. T with chil less ambi attendancr Children 1 Effect 0n 147 dedication of workers. At the secondary level only, males are slightly more likely to read professional books, to be absent and to express enthusiasm about re-entering education and about career plans but females report slightly more activity in teacher unions, especially at the elementary level. Sex has almost no effect on level of self-investment in work. Turning to family compositional variables, teachers with children under age 5 are distinguished only by slightly less ambitious career plans at the secondary level, lower attendance at conferences and slightly fewer journals read. Children between ages 4 and 12 have the most depressing effect on time teachers spend reading (-.14 correlation) and willingness to re-enter education. The presence of children 5-12 is slightly correlated with rate of absenteeism but is also positively correlated with taking courses not required by contracts. The relationships between presence of children and both self-investment in work and union activities are very weak. However, the presence Of chil- dren aged 12 to 18, is related to lower career desires, especially among elementary teachers and, at the secondary level, less frequent enrollment in college courses. Chil- dren over 18 strengthen the trends Of the last category but also correlate with teacher participation in union activity. The total number Of children Of respondents has a slight negative effect on teachers' desire for upward career movement, on taking courses and on the eagerness of secondar positive clear ex: age and Since thi founded rI a Partial was run t children, A most diff and Seve r SECOndary 148 secondary teachers to re-enter education but a slight positive effect on union activity of secondary teachers. None Of the correlations suggest any strong or clear explanation of self-investment of teachers, although age and experience appear to encourage union activities. Since the data on effects Of children are probably con- founded by systematic variation in the ages of respondents, a partial correlation, controlling for the effects of age, was run to see the independent effects Of presence of children, sex and experience. As Table 26 shows us, controlling for age made the most difference in the relationships between experience and several aspects of dedication. Eleven Of these correla- tions were reduced by .10 or more for both elementary and secondary teachers, although experience does appear to play an independent role in a few Of these cases, notably hold- ing local Office, number TOs joined, working on state or local committee, acting as building representative and taking courses. The negative correlation between experi- ence and taking only required courses was explained as an effect of age for secondary teachers but not for elementary teachers. Table 25 sets out all the cases where controlling for age affected the original correlations by .10 or more. None of the correlations between aspects of self-investment and personal or family characteristics was affected by controlling for age. Age greatly affects the importance Of family composition, reducing several correlations between mmh\mow u z a . mhwvcooom\>umucoson 1149 nn nn nn nn nn nn oo.n\om.n NN.n\om.n nouHsaom Nn.n\No.n «H.n\mo.n oo.n\mo. oN.n\mH.n nu nn oH.n\HN.n m¢.n\oo.n mmmusoo nn . nu nn nn nn nu o \mo. NH. \oo. comm mum nn _ nn . nn nn nn nn No.n\HH. mH. \HN. .nu5\mHo nn nn HH.n\No.n oo. \oH. oo.n\HN.n oo.n\HH.n NH. \om. He. \No. .one .oon nn nn oH.u\oH. so. \HH. nn nn NH. \oN. NN. \om. .ssoo HmooH nn nn . nn nn nn nn mo. \mH. oH. \oN. .eeoo oumum nn nu No.n \oo. oo. \oH. nn nn HH. \NH. NN. \Nm. mos . nn nn nn nu nn nn so. \mH. oH. \NN. .ooum mumum nn nn oH.n\HH.n mo. \No. nn nun oH. \NH. Hm. \NN. .ouo Hmooq nn nn o \No.n HH. \HH. nn nn oo. \oo. NN. \oN. . nos noose: ooHHouucoo . Hmuoa ooHHOHucou HmuOB ooHHouucoo Hmuoe poHHOHucou Hmuoe mOHQMHuw> , coHumoHooo cmunHHsu . Hones oHA .nuo xwm mocmHnmaxm .xHOS 0» coHuMOHooo one mOHumHuouomumnu aHHEmm\Hmcomuom coo3uom mcoHuMHouuou com: mod MOM mcHHHOHucoo mo unadummnu.m~ mqm:HIMHmm mo mHOHMOHch UGOE¢ meHuMHOHHOUII.mN mqmHuoa Hooumo mHOHHomsm .HOU omoHo .cHuucH .QEH .umum .QEH Hooumv GMHm cOHHMOHUoo ouflmma i‘ .cOHammsooo an .xuoz Op cOHHMOHcoo com pcmEumm>eHnMHom coosuom mcoHHMHoHHOOnn.om names 161 number Of children each suppressed a stronger relationship between importance of close colleagues and participation as a building representative for elementary teachers only and a slightly stronger positive relationship between close colleagues and number Of job-related books read by both elementary and secondary teachers. Despite these controls, the correlations between self-investment and dedication remain very weak. It appears, though, that collegial reference groups have a slightly greater effect on the behavior of teachers who are younger, have smaller families and/or fewer young children. The net effect of controlling for personal and family characteristics is a very slight increase in support for the propositions. However, from the data at hand, the proposition that high self-investment in work produces dedication to work for both occupations cannot be confirmed. Nevertheless one piece of information may be a useful pro- duct of this limited data. If self-investment at work is only one kind Of self-investment and if work competes with other activities like family and recreational activities, then the discovery that one indicator of self-investment is most centrally affected by evidence Of other types of commitment suggests that this indicator, importance of close colleagues as a reference group, may be the best indicator Of self-investment in work. On the other hand, it is possible that the questions about whose opinion is 162 important to the respondents had different meanings for those committed to young families and to other teachers. In summing up the effects of self-investment in work upon dedication to work for elementary and secondary teachers, it is apparent that the prOposition linking the two has not received significant support from the data at hand, despite the introduction of controls on possible suppressor variables. Nevertheless, it remains possible that the organizational variables hypothesized to explain variations in both self-investment and dedication may exert some effects on either or both of these variables. Effects of Organizational Variables Upon Self-Investment and Dedication According to the propositions presented in Chapter III, even when occupation Of worker is controlled, several organizational variables should encourage high self— investment in work and, hence, greater dedication. While the connection between self-investment and dedication has not been supported by the intra-occupational data, it is still possible that some effects due to variation on organi- zational dimensions might be isolated, particularly if the problems encountered above stem from inadequate measurement of self-investment. The organizational variables expected to encourage self-investment and dedication are: greater hierarchy, specialization, standardization, formalization, universalism, expertise Of staff and lower levels of centralization. 163 The first step in the analysis of organizational effects is to see how variations in the organizational setting influence teachers' perception of work setting, since these perceptions are assumed to be critical to self- investment in work. Table 31 sets forth the correlations between organizational variables and perceptions of work by occupational group. Several school compositional variables are considered as possible independent variables as well. From Table 31, one sees that the relationship between organizational structure and perceptions of work environment are weak and, in many cases, inconsistent across occupa- tional lines. It is important to remember here that the range of variation in organizational structure is very narrow for both elementary and secondary schools. It is possible that the actual differences are too small to produce great differences in teachers' perceptions. Since there is more difference between elementary and secondary schools than within either category alone, the fact that effects Of several organizational variables appear to be inconsistent by occupation may indicate a curvilinear relationship between structure and perceived problems or advantages for some items. Looking at specific organizational influences, it is apparent that only four variables can claim any real effect at all upon teacher perceptions of organizational setting. The three types Of specialization each appear to 1654 .mHooaom mH\oo ch mumnommu oosxooo u z mumpcoomm\%umucoEon Ho.n\mo.u mo.n\mo.n mH.n\OH. so. \oH.n mo.n\mo.n No. \Ho.n no.n\Ho. Ho. \so.n mo.n\mo.n uwmum . .oomm ouHm no. \mo.. Ho. \mo. no. \so. Ho. \so.n 0 \H0.1 H0. \ o so.n\mo. no. \No. .men.u»cou eo.u\~o.n no. \Ho. No. \ o MH.n\mo. mo.n\mo. mo.n\vo. oH.n\mo. mo. \mo. .Hmnuucoo mo. \Ho.n 0 \mo.: no. \mo. mo..\mo.| mo.n\Ho.n Ho.n\~o. so. \vo. Ho. \ho.n Ho. \so.n omHuuomxm vo.n\so. mo.n\oo. so.n\oH.n mo.n\Ho.n mo. \mo.n mo. \mo.n mo.n\mo.n HH. \mo.n ao.n\Ho. .meuo>HcD o \wo.n mo.n\No.n MH. \Ho. HH.n\mo.n No.n\ o No. \mo.n mo. \Ho. mo. \mo. mo.n\mo. Hoflme :H .comoe ".oomm .vo.n\mo.n mo.n\vo.n MH. \NH. o \Ho. Ho. \ 0 mo. \ o o \mo.l HH.n\mo. Hm>OH uoc oHlo an .oomm so. \Ho. No. \so. mo. \so. oo.n\oo. Ho.n\Ho.n o \mo. oo.n\oo. mo.n\mo. .NHHmsuom Ho. \mo. No. \mo. o Ho.n\oo.n mo.n\so.u No.n\ o oo.n\so.n so. \so. .uchmeamum vo.n\ o voun\mo. so. \no. Ho.n\Ho. mo. \vo.u mo.n\Ho. No.n\~o.n mo.n\mo. anonmuon .Hmch3o anoe omOHUnOH :HmmpH unomoOH cHnOU ago @000 coHuMNHcmmuo mcoHudoouom .mcHuuom xuoz mo mcoHumoOHom .muocomoe can ousuosuum HmcoHuMNHcmmno wo coHucHouHoonn.Hm wands 165 make the administration appear more competent to teachers and also lead teachers to say that their principals support teachers (except for secondary teachers in schools with many specialists, a condition that appears to make principals seem less supportive of teachers). The only other charac- teristic of organizational structure that has even minimal effects on teachers' perceptions of work setting is the principals' assessment of centralization of decision-making in each school, which is negatively related to in-school OOOperation among secondary teachers and which inhibits idea exchange with secondary teachers' close colleagues. To check whether any organizational characteristics, structural or compositional, meaningfully affect teachers' perceptions of work situations, I examined correlations between these perceptions and several school compositional variables. Since these variables are not central to the propositions being tested, I will not consider them in detail. However, it is essential to note that none of the compositional variable, either--including student ability, SES, homogeneity, percentage of female teachers, sex and experience of principals--show any significant correlation with teacher perceptions of school features. Each of the correlations involved is S .14. Thus, it seems safe to conclude that teachers' perceptions of good and bad aspects of their schools are only tenuously related to identifi- able features Of school structure or composition. 166 Given this setback to the propositions, the next question to be considered is the determination Of the best predictor of teachers' self-investment and dedication to work--teacher perceptions Of organizational problems or structural features identified by the researcher? Turning to Table 32 one can Observe the correlations between teachers' perceptions of work setting and the indicators of level of self-investment in work. The data in Table 32 suggest a moderate positive correlation between favorable perceptions of the job setting and level of self—investment in work. Once again, the teachers who actually plan to move up to administrative posts within five years show different responses to percep- tions about school than do teachers exhibiting other evidence of self-investment in work. None of the perceived charac- teristics or settings appear to have much effect on planned upward movement, although there are some moderate effects upon desire for upward movement. These facts add to the evidence that planned career movement is quite different than the other five variables in meaning and implication. Whether it belongs as an indicator of self-investment or not, it appears quite conclusive that planned mobility is not produced by any Of the variables I had thought would affect it. Turning to the indicators of self-investment that do appear to be affected by teacher perceptions of setting, one sees that two types of perceived problems, TO hindrance 167 mmh\mom u z mnococoe mhmccooom\humucoEon mm. \mm. ms. \Nm. mm. \Nm. no. \ 0 NH. \oH. mo. \so. .nchH :30 mH. \mm. mH. \mH. mH. \om. No.n\mo.u no. \oH. No. \mo. Hooch OB oH. \sH. mm. \mm. mm. \om. NH. \ao. mo. \mo. Ho. \o . .mmnm .cHum mm. \mm. mH. \mH. NH. \NN. so.n\mo.h so.4\eo.n mo. \mo. .cHeo< Hoom No. \mo. mm. \mH. om. \mH. om. \mH. mH. \mH. mo.n\mo. omOHU mopH NN. \oH. oN. \NN. oH. \NH. NH. \HH. mN. \NN. mo. \oo. .cH mmeH No. \mo.n oN. \sH. oN. \mN. NN. \NH. mH. \oH. mo.n\mo.n .uso mmeH oH. \oH. NN. \NN. oN. \oH. NH. \HH. NN. \oN. o \Ho. .cH .mooO no. \Ho. mH. \mH. sH. \NH. us. \ms. os. \Hs. so.n\ o .uso .mooo Hooumu .HHOU .HU mHOHHomsm .cHHucHn.mEH msumuwn.mEH .HmUIQMHm umcHuuom oHHmoo . co>HoOHom ucoEumo>cHanom mo mHOHMOHOGH .cOHpmmnOOo an .xu OKAH/ udoeumo>cHIMHom pom mcHuuom xHoz mo muHHmso po>HoOHom coospom chHumHOHHoun . 168 Of teacher performance and poor administration, each moderately encourage teachers to express a desire for upward career movement. However, the strongest single correlate Of desired career movement is the teachers' perception Of extent of own influence over others at work. Thus, in explaining the bases of teachers' differential eagerness for career movement, the most significant influences appear to be the perception Of difficulties in the school environment plus the respondent's feeling of efficacy in the present school setting. It is possible that perceived efficacy, however, is not a determinant of career planning but that both stem from a common source, such as the teachers' confidence in own abilities and willingness to initiate action. Since perceptions regarding own influence were not found to correlate significantly with level of centraliza- tion in the schools, it is possible that perceived efficacy is based on each teacher's comparison Of self with others in the school, thereby holding constant actual centralization of authority. Turning back to Table 32 to look for other possible effects of teachers' perception of problems in their schools, the only other indicator of self-investment that appears to be related to such perceptions is teachers' reliance upon superiors as an important reference group. Surprisingly, teachers who saw poor administration also tend to use superiors as an important reference group. This suggests to me the possibility that teachers who see 169 administrators as a major source of problems find it necessary to try to avoid administrative action or attention concerning their own work, making it important that superiors hold good or neutral Opinions about the teacher. Teachers who use superiors as an important refer- ence group also see the Teacher Organization as a source of problems hindering the achievement Of educational objec- tives. It would be useful to know whether difficult unions drive teachers into alliance with administrators or whether concern for the Opinions of superiors leads teachers to agree with administrators about the extent of difficulties caused by the unions. This intriguing question cannot be answered with cross-sectional data. Turning to teachers' favorable perceptions regard- ing work environment, the propositions lead us to expect that more favorable perceptions will produce greater self- investment in work on the part of teachers. The correlations reported in Table 32 lend some support to these ideas; except for one indicator Of self-investment, planned upward movement, correlations between favorable perceptions and self-investment are positive and, in some cases, fairly strong. In general, it appears that cooperative work settings with frequent exchange of ideas among personnel, support of teachers by principals and feelings of efficacy on the part of teachers promote self-investment in work. Considering differences in the strength of these correlations, 170 one notes that perceived principal support and perceived efficacy affect career plans and importance of status or intrinsic aspects Of work very little but greatly encour- age the use of work-related reference groups and teachers' desire for upward mobility. Cooperation within the school moderately correlates with importance of status and intrinsic aspects of work but has slightly less effect on the importance Of Opinions of coworkers. Extent of idea exchange Of all sorts positively correlate with the impor- tance of status and intrinsic aspects of work and the importance of coworkers Opinions but these correlations are fairly weak, ranging from .10 to .29. Comparing elementary and secondary teacher data, one sees great consistency in the relationship between variables across school level. Except for a few pairs of correlations near zero, the elementary and secondary teacher coefficients for each relationship are the same in sign and, in most cases, very close in degree of association. This evidence supports my expectation that occupational and personal variables would not explain all the variation in level Of teachers' self-investment in work. Instead, the data suggest that similar perceptions Of work structure produce some similarities in teacher reactions across occu- pational lines and despite differences in personal and family characteristics by level. In summing up the evidence presented in Table 32, the proposition linking favorable perception of work climate 171 to extent of self-investment in work is moderately supported for both occupational groups. Whether teacher's perception of structure is superior to more objective indicators of structure in predicting teacher's self-investment remains to be seen, however. Although my propositions suggested that perceptual indicators would be superior, I had assumed that there would be significant correlation between percep- tual and documentary data about work setting. The lack of any strong correlation (already noted in Table 31 and con— sistent with other research evidence showing typical low correlations between perceptual and objective measures) invalidates that assumption and transforms this attempt at proposition-testing into a more exploratory endeavor. At this point, one must ask whether the non-perceptual data on school structure can exert any independent effect on level of self—investment in work. Data presented in Table 33 may suggest an answer to that question. Contrary to the propositions, the correlations reported in Table 33 do not Offer any support for the idea that structural or compositional school variables affect teachers' level of selfvinvestment in work. Not only are the correlations very low (all under .12) but also they are inconsistent by school level. Since it is conceivable that structural variables might have differential impact depend- ing upon certain characteristics of teachers, partial corre- lations were run to control for the effects Of age, sex, number Of children and teacher's level of education. Also, 172 .oHMSom unwound co =30H= pouoom mHoocOm humpcooom Hche .m u onEom NH u OHM: onE u .GHHQ mumccooom HHfis mHoocom mH\mm «muocomou mmh\mom n z mumpcooom\MHmucoEon Ho.o mo. 0 mo. Ho. mo. ««.comoa m w o mo. No.n Ho. Ho.n so. «xom .CHHm No.n\Ho.n no. \Ho.n ~o.|\mo.n No. \ho. so. \mo. mo.n\mo. .Hmmxm .cHum so.n\mo.n oo.n\so.n mo.n\so.n mo.|\oo. oo.n\ o Ho.n\Ho. >UHHHQ¢ so.n\mo.n mo. \mo.| so. \mo. Ho.n\Ho. mo.n\Ho.n mo. \so.n .moeom mo.n\no.n so.n\oo.n Ho. \mo.n so.n\so. ho.n\ o mo.n\ o mmm mo.n\oo.n mo.n\mo.n No. \mo. No. \mo. mo. \ o o \mo. mmmum .oomm oNHm so. \Ho. Ho.n\mo. mo. \No.n so.n\Ho.n mo.n\mo.n Ho. \mo. .ma .unamo so. \mo. oo.n\ o No. \mo.n mo.n\No.n oo.n\No. so. \oo.n .um .uuamo so.n\mo.n mo.n\no.n oo \Ho. Ho. \mo.n mo.n\mo.n mo.n\mo.n omHuHomxm o \mo. mo. \Ho. mo.u\oo.n mo. \so.| so. \mo. mo. \mo. .Hmmuo>HcD oo.n\mo.n so.u\mo. mo.n\oo. mo. \oo. No. \oo. mo.n\mo.n Home: cw .oomw oH.n\ o oH.n\so.n oo.n\ o so.n\Ho.n so.n\ o mo.n\ o nomnnsm GH .oomw Ho. \No. mo.n\so. ~o.n\oo. Ho.n\mo.n ~H.|\oo. Ho.n\mo. .Hmeuom mo. \No.n mo.n\mo. so. \NH. mo.|\mo.n oo.n\~o.n mo. \Ho. .ocmum mo.n\Ho. mo.n\~o.n oo.n\Ho. mo. \mo. mo. \mo. mo.I\Ho.n mauumuon Hooumu osmmoHHOU .uomsm .HucH QEH umum mEH .Hmo GMHm moHQMHHM> oHHmoo .mEH . .mEH .. Hoocom ucoEumo>cHIMHom .xnoz cH ucosumo>cHIMHom .muocomoa comb :OHuHmOQEOU can GOHMMNHccmHo Hoosom mo muoommmnn.mm wands 173 to eliminate the possible effects of teacher interaction with more or less able principals and students, effects of principal experience and mean student ability level were partialled out as well. Nevertheless, it appears that none of these variables affect the correlations between aSpects Of school structure and teachers' level of self—investment by more than .04. Thus, I conclude that, if there is a relationship between structure and self-investment within the fairly narrow range of variation found in elementary schools or in secondary schools, that relationship either is masked by the effects of some unknown variable or requires finer measurement of the relevant variables than was achieved in this study. Data from this study suggest that teacher perceptions Of setting are much more critical than other indicators of structure as predictors Of level of self- investment. The final check on the relative importance Of struc- tural factors and teachers' perceptions of setting in explaining teacher activity is to consider the effects of each variable upon teachers' dedication to work. Given the lack Of correlation between structural variables and both the expected sets Of intervening variables, perceptions of structure and level of self-investment, one cannot antici- pate much support for the prOpositions. Data presented in Table 34 confirm those gloomy predictions, since school structure and teacher behavior appear to be totally unrelated. None of the correlations exceed .18 and most fall between 1374 can mocowuomxo HmmHOcHHQ .xom HMQHOCHHQ .huHOCOUOEoc .mmm .huHHHnm ucocsum mo mu00mmo How mCOHumHouuoo HMHuumm .mHoocom mH\om nauseous» soN\sos u z .mcoHumHouuoo o>onm Eouu mcoHumHum> HsumchmoE on pmpHlo muonomou mHmEow w H mumpcooom\aumucoEon No.n\so. so. \No. so. \so.n Ho.n\sH. so.n\so.n No. \so. so. \Ho.n so. \so. o \mo.n so. \oH. seesaw No. \NH. No.n\No. Ho. \HH. .mo. \mo.n mo.n\so. Ho. \No. No. \mo.n o \HH. Ho.n\No. No. \Ho.n nonsense so.n\mo.n No. \No. so.n\so. HH.n\No.n No. x o oo.n\so.n mo.n\so. so. \oo.n so.n\so. oo. \Ho. omuHsamm Ho. \mo. mo.n\mo.n so.n\so.n so. \so. so. \ o No. \NH.n so. \mo. so.n\so.n so. \No. No. \No.n momusoo Ho. \Ho.n No. \No. No. \so.n Ho. \mo.n mo.n\ o so.n\so.n oo.n\so.n mo.n\mo.n so.u\No.n so.n\so.n swam meHa so. \so. so.n\No.n Ho. \so. No. \Ho. No. \mo. o \mo. mo.u\so.n Ho. \so.n so. \so.n mo.n\No. mHmcusoo so. \mo. o \so.n Ho. \so.n No. \No. so.n\so.n No.n\oo.n Ho.n\so. Ho. \so.n Ho. \so. Ho. \so.n mxoom so. \oH.n so.n\No.n oH. \so. mo. \so.u so. \so.n so. \Ho. No. \ o so.n\so. No.n\mo. Ho. \so. maoamxsoz o \Ho.n Ho.n\Ho. so.n\Ho.n NH.n\No.n so. \No.n sH.n\No. oo.n\Ho.n Ho. \so. so.n\so.n so.n\No. .amm .osHm No.n\oo. Ho. \No.n No.n\so.n Ho.n\so. so. \NH.n No.n\so. so.n\No.n o \Ho. mo.n\mo. Ho.n\Ho. .asoo HmooH Ho.n\so. o x o No.n\oo. so. \so. mo. \so.n so. \so. o \so. Ho.n\mo.n mo. \Ho. Ho. \Ho.n .ssoo muons HH. \oH.n oH.n\mo. NH.n\oo.n sH. \NH.n sH. \No. sH. \so. o \oH.n sH.n\so.n Ho. \so.n o \Ho. mos . so. \so. No.n\Ho.. so. \so.n Ho.n\ o No. \so.n so. \HH. so. \so.n so.n\mo. Ho.n\mo.n Ho. \ o .oono muons so. \No. so. \ o so.n\mo.n No.n\oo. so. \so.. Ho. \so. so. \so. Ho.n\so. No. \mo. No.n\so. .oso muons so.n\Ho. No. \ o Ho. \so.n so.n\Ho. Ho.n\so.n oo.n\mo. so.n\Ho.n No. \oo. Ho.n\No.n No.n\mo.n .oso HsooH o \Ho.n Ho. \sH.n No.n\mo.n Ho.n\No. oH. \HH.n so. \so. so.u\HH.n so.n\so.n so. \so. No. \No.n os m>Huo< so.n\sH.n mo.n\so. HH.n\No. so.n\HH.n HH. \so.n so. \ o so. \so.n so.n\so. so.n\No.n No. \No. 09 bongo: scans scmo acmu ysnmoxm .snm>Hco noon: .Hnss .scsum .Hmsnoa .umHm coHusuHsoa .oomw .oomw .oomw wOHbmHHm> Hausuosuum .cOHummsooo xn .xuoz ou coHumoHpmo can mpomHum> Hmusuosuum coo3uom mcoHumHmuuounn.sm mqmde 175 .05 and -.05. When partial correlations are run to separate out the effects of age, sex, number of children, teacher education, student ability, SES, principal sex and experi- ence and teacher career plans, only a few correlations are changed by more than .05, and each of these changes involves only the effects of universalistic criteria of evaluation for secondary teachers. At the secondary level, control- ling for teachers' education raises the correlation between UNIV and #TOs from .16 to .23; controlling for student ability raises the correlation between UNIV and Local Office holding to -.15; controlling for age reverses the weak correlation of UNIV with reading journals (from -.13 to .06) and of UNIV with time reading (from .06 to -.1S). None Of the other controls have any noticeable effect on the correlations between organizational aSpects and teacher dedication to work. Even those noticeable changes do not Offer significant support for the proposi- tions, since universalism was expected to encourage all types Of teacher dedication to work. Furthermore, one also wonders why those particular controls affected the influence Of use of universalistic criteria in schools. It makes some sense that universalism and age would have joint effects upon dedication, since age might easily be a basis for dis- crimination in schools relying upon particularistic criteria of evaluation. From that line of reasoning, one would also expect sex or family composition to interact with univer- salism, rather than extent Of teacher education or student 176 ability levels, which do not suggest any basis for discrimi- nation. Looking at the data from a different perspective, though, one might expect both teacher education and student ability to affect the appearance of teacher's success when universalistic standards of evaluation are used. From this perspective, it makes sense that education and student ability interact with universalism in affecting level of dedication, although the reason for a lack of interaction with sex or family composition as well as the failure of elementary school variables to show similar interactive effects remain unexplained. Since Table 34 and further refinements of these data show the proposition that school structure determines level Of dedication when occupation is controlled to be lacking in support, one must ask whether the revised proposition, that teachers' perception Of work setting determines intra-occupational differences in dedication to work, receives any more support from the data than did the original structural proposition. Table 35 presents correla- tions between the several aspects Of teacher perceptions Of structure and dedication to work for both elementary and secondary teachers. Since perception Of setting did prove to be a better predictor than structural variables when self-investment was the dependent variable, one might expect the same pattern to hold when dedication is the dependent variable. However, the data in Table 34 do not support that expectation, since there are only weak correlations between 1377 moo\ooo I z . unaccooom\>unu:oeme Ho.n\ o Ho. \No.n Ho. \No. so. \No.n so.n\No. mo.n\ o No. \so. o \Ho.n Ho.n\No. oceans o \Ho.n No. \Ho.n No. \so.n so.n\so.n so. \ o Ho. \mo. so. \ o no. \Ho.n No. \No. Nonsense Ho.n\No.n NH. \No.n No.n\No.n so. \oo.n No. \No.n so. \oo.n so. \NH. so. \so. o \so. censuses so. \No. so. \so. so. \so.n so. \so. o x o so. \Ho.n so. \Ho.n Ho.n\Ho.n mo.n\Ho. sosuaoo No.n\NH. so.n\mH. so.n\No. so. \so. so. \so. No. \ o no. \so. Ho.n\mo. so. \Ho.n some oeHa Ho.n\No.n so.n\so.n No.n\so. so.n\so.n Ho.n\No. Ho.n\oo.n Ho.n\so. Ho.n\ 0 so. \so.n mHsnuaoo No.n\so.n so.n\ o so.n\mo.n Ho. \No. so.n\No.n so. \No.n Ho. \Ho.n No.n\Ho. o \HH.n axoom so. \No.n mo.n\so. No.n\Ho.n o \so.n so. \so. so. \Ho. o \Ho. so. \No.n mo.n\so.n smoamxuoz No.n\so. No.n\so.n so.n\oo. mo.n\oo.n Ho.n\ o so.n\No.n so.n\oo. Ho.n\No. so. \so. .one .osHm mo.n\so.n so.n\so.n so.n\No. mo.u\so.n Ho. \mo.n Ho. \No.n No.n\No. Nod \ o Ho.n\Ho. .ssoo HmooH No.n\No.n Ho. \so.n o \No. Ho. \No.n Ho. \so. o \No.n Ho.n\oH. so. \Ho. Ho.n\ o .2500 mumps Ho. \so.n No.n\No. so.n\Ho.n so.n\No.n so. \so. o \Ho.n Ho. \so. HH. \oo. Ho.n\Ho.n mos s so.n\mo. so.n\No.n o \so. No.n\No.n Ho.n\ o Ho. \so.n Ho.n\No. o \Ho. No.n\so.n .ooua mumps No.n\No.n o \Ho.n No.n\ o so. \ o mo. \No.n Ho.n\so.n so. \ o No. \so.n No.n\so. .oso muons Ho.n\mo. so.n\oo.n so.n\No. so.n\so.n No. \No.n o \No. mo.n\so. so. \mo.n so. \so. .oso HmooH so. \so. Ho.n\No.n so. \so. so.n\mo.n No. \so. so. \No. Ho. \so. so. \oo. so. \so. as w>Huu< oocooncH Hmnoe mamasum .Uduoom .HoomooH chopH usomopH chooo usomooo ocHuuom mo mcoHumoouom coHusoHooo .coHucmsooo he .xHOK ou coHuMOHpoo mo Ho>oq pom ocHuuom xuoz mo mcoHumouuom Hocomoa cmo3uom mcoHumHouuoonl.mm manta 178 the two sets of variables. Furthermore, when partial correlations were run to separate the effects of age, sex, number of children, teacher's education level, career plans, student ability and SES, none of these controls affect any of the original correlations by .05 or more, for either elementary or secondary teachers. The unexpected data presented in Tables 34 and 35 do not offer any meaningful support for the proposition that school structure affects teacher dedication to work, either directly or through teacher perceptions of setting. Several interpretations can be made regarding this lack of support: it is possible that the propositions are wrong (although the initial support drawn from comparison of elementary and secondary schools and teacher behavior Opposes that inter- pretation); it is possible that the variables supposedly measuring dedication are not valid indicators of dedication; finally, it is possible that the range Of structural varia- tion within the Occupational categories is too narrow to be discriminated by teachers-~although that explanation does not appear to be the case for the range of differences in perceptions of work setting, since we did observe that per- ceptions of setting significantly affected teachers' level Of self-investment in work (Table 32). Finally, turning to another possible indicator Of dedication, in trying to predict teacher turnover rates, one might draw two different explanations of turnover from the original propositions. First, if it is true that 179 teachers have, on the average, lower self-investment in work and hence less concern for upward career movement than do principals, one might expect teacher turnover rates to depend largely on random or extraneous factors. However, one might also draw from the prOpositions the idea that some school organizational variables might play a role in determining turnover rates, since some teachers do evidence strong self-investment in work. If the latter perspective is at all true,«one would expect that higher scores on school hierarchy would predict lower teacher turnover, since those teachers interested in mobility would find hierarchical schools to Offer more potential for such mobility. Furthermore, other bureaucra- tic features Of schools, such as standardization, formaliza- tion, required expertise, specialization and the use of universalistic criteria of evaluation, might also be expected to depress teacher turnover rates if these features make the schools better places to carry out educational objec- tives. The most critical feature of school structure that I would expect to find linked to turnover rates is centrali- zation of decision—making, which determines the autonomy teachers can exercise at work and probably affects even teachers lacking strong self-investment in work. To test these ideas and, by implication, the original propositions, I checked the relationship between such struc- tural variable plus student ability, SES and homogeneity and total teacher turnover rates for 1965-68. Schools with 180 less than 50 percent turnover were considered "low" in turnover; the 20 schools with turnover ranging from 50 to more than 100 percent were considered "high" in turnover. Table 36 presents Yule's Q coefficients for each pair Of categories for the 53 schools in the sample. With onlyng secondary schools, it is not meaningful to calculate separate Q coefficients for elementary and secondary schools. TABLE 36.—-Organizational Characteristics Correlated with Teacher Turnover Rates. School Turnover Rates, School Characteristic 1965-68, N = 53 Hierarchy -.86 Standardization -.65 Formalization —.58 Specialization by Subject -.21 Specializaton--Major -.34 Universalism .39 Expertise -.05 Centralizaton—-Principal P. .36 Centralization-~Teacher Perc. .60 Specialized Auxiliary Staff -.12 Student SES .25 Student Ability .25 Student Homogeneity -.03 181 Data presented in Table 36 Offer fairly substantial support for the idea that school structure does affect teacher turnover rates. Particularly, school scores on hierarchy show a strong negative correlation with teacher turnover (although this relationship is complicated by the fact that secondary schools tended to have both high scores on hierarchy and low scores on turnover); a modern negative correlation is established between turnover and both standardization and formalization; teacher perceptions of centralization of schools are moderately and positively related to teacher turnover rates. Surprisingly, univer- salism in schools is positively related to teacher turnover, although the correlation is not strong. Yet another unexpected finding is that specialization (by subject, by major, presence of staff and expertise Of teachers) does not have the strong negative effects on turnover that I had anticipated. Since Specialization by Subject and Expertise of teachers were not significantly related to school level (see Chapter V), the low correlations for these two variables, at least, cannot be attributed to some occupa- tional variable. The best way to check the meaning of these low correlations would be to obtain turnover rates for teachers with high self-investment in work and for those with low self-investment and see whether Specialization of various types has different effects within each group. Unfortunately, the data on teacher turnover concerns mainly 182 teachers who left before the survey was made, so that the characteristics of teachers who left are unknown. Turning to the contextual variables, student ability and SES weakly encourage turnover while homogeneity of students appears unrelated to actual turnover rates. On the whole, it appears that certain characteris— tics Of school structure underly variations in school turn- over rates, although the exact specification of the roles of occupational and organizational variables cannot be determined from this data. Conclusion This chapter tested the proposition that certain school organizational variables influence perceptions, behavior and attitudes Of teachers. Confirmation of this proposition required that the data show significant and meaningful correlations between organizational elements and the dependent variables, when occupation and personal characteristics of the respondents are controlled. The explanatory power of two alternative explanations of worker behavior, involving personal/family characteristics and level of satisfaction with various aspects of work, was considered as well, so that the usefulness of the proposed explanation of work behavior could be compared with that Of the major competingexplanations. In testing the propositions and the alternative explanations on worker behavior, the following conclusions were established: 183 Personal and family variables do not signifi- cantly affect teachers' level of self- investment in work and have only weak effects on non-union indicators of dedication, except for likelihood of taking non-required courses. When union activities are considered, age and experience positively correlate with history of union participation, probably reflecting the differences in Opportunity of Old and young workers to have such a history. My expectations that young and female workers would be more militaristic, because of lack Of advancement Opportunities through regular channels, was not supported. Teachers' level of satisfaction with three aspects of work--physical, non-physical and authority relations--has no significant effects upon level of dedication to work but satisfaction with authority was mildly related to one indicator of self-investment. However the most plausible explanation of that phenomenon is that concern with intrinsic aspects Of work leaves teachers inattentive to power problems, yielding a high score on satisfaction with authority. The indicators of level of self-investment in work do not correlate internally enough to support the idea that each is a critical element of a 184 unidimensional variable, self-investment. Instead, three separate groups can be identified, although it is difficult to say if these dimensions are equally part of self-investment. Contrary to expectations, none of the aspects of self-investment strongly correlate with any of indicators of level of dedication to work.. When partial correlations are run to isolate the effects Of personal and family variables, we see that collegial reference group have slightly more impact on dedication for teachers who are younger, have smaller families and/or fewer young children. The original propositions suggested that teachers' perceptions of good and bad aspects of work setting would be determined by school structure. Correla- tions between each aspect of structure and each perceptual variable are very weak and offer no support for this idea. However, it is possible that the intra-level organizational variation is so narrow that teachers do not discriminate these fine differences. The proposition remains "not proven" rather than conclusively refuted because Of that possibility. The proposition linking favorable perception of work environment to level of self-investment in work is moderately supported for both occupational groups. However, school structure does not appear 185 to have any effects upon level of self-investment within occupational categories. 7. Neither school structure nor teachers' perceptions Of the work climate meaningfully explain variations in teachers' level of dedication to work. At the secondary level only, controls on level of teachers' education raise the correlation between universalism in schools and the number Of teacher unions a teacher joins to .23, the strongest relationship uncovered. Structure does influence turnover rates, although controls on level are lacking. While the data presented in this chapter support only a few aspects of my prOpositions, two critical facts must be kept in mind: of the major alternative explanations of dedication and attachment to work, neither personal/ family characteristics nor satisfaction with work better explained level Of self—investment in work, compared to structural variables, while teachers' perception of struc- ture proved to be the best predictor Of self-investment; furthermore, comparing the three explanations of dedication, only age and experience Of worker had significant effects and only upon historical aSpects of dedication to work, not upon current behavior. Thus, while my propositions are not fully supported, the independent variables I expected to be critical fared no worse as predictors than did the competing explanations of variations in worker behavior and, in a few 186 cases, my perceptions of structure proved to be better pre- dictors of teacher responses. Finally, it is important to bear in mind that this chapter considered a very stringent set of tests of the propositions at issue, since occupational and organizational influences were separated, thereby wiping out most of the variation along organizational dimensions. If this stringent test Of the propositions had yielded much support, one could have been quite certain that organizational dimensions are critical to worker behavior. As it stands, one can only say that the propositions were not confirmed. However, neither did we find any evidence that bureaucratization of schools discourages teachers from exhibiting self-investment in work or from behaving in a dedicated manner. Further research in which a broader range of organizations can be studied, holding occupation constant, is required in order to make any more conclusive statement about the status of the propositions. CHAPTER VII CHARACTERISTICS AND EFFECTS OF DISTRICT ORGANIZATION The hypotheses regarding behavioral and perceptual responses to organizational setting require specification of different organizational settings for principals and teachers. Since teachers carry out most of their activity and interaction in individual schools, rarely moving into district administration, the effects Of school organiza- tion upon teachers was expected to be more critical than the effects Of district organization, although broad corrOlaries of district variables will be considered later in this chapter. When principals' perceptions and responses are under consideration, however, only occupational level and district organization are expected to affect them. School organization is of less importance than district organi- zation in Opportunity for advancement, for reOOgnition and in interaction with close colleagues (other principals and assistant superintendents). Thus it is necessary to relate 187 188 the organization of the five school districts to the :principal data. Unfortunately, there are only five school districts in the sample and a great many ways in which these dis- tricts might be differentiated. In considering possible organizational differences at the district level, the following procedure was used: documentary sources of data were consulted and coded to yield a score for each district on each variable and then Spearman rho correlations were calculated between each pair of variables to ascertain patterns of district organization. Eventually, these patterns of organization will be related to those aspects of principals' perceptions and behavior that show statisti- cally significant variations in a one-way analysis of variance where district is the independent variable. Patterns in District Organization The dimension usually considered most critical in determining level of bureaucratization is hierarchy, involving the number of levels of authority in an organi- zation. From the propositions developed in the first chapters, I expect hierarchy to be of critical importance in explaining principal behavior, since the major element in the intra-district career possibilities Open to princi- pals is the number of higher positions in which vacancies might occur. Using criteria described in Chapter III, the number of "assistant superintendent" and "administrative 189 assistant" positions in each district organization were established. Table 37 presents the number of such positions in each district as well as the ratio of possible career moves to possible competitors (i.e., all principals) in each district. TABLE 37.--Distribution and Mean Scores on System Hierarchy. Number of Positions Ratio Assistant Administrative Prin. District Superintendent Assistant Positions Rank 1 5 3 .267 l 2 3 2 .093 3 3 3 2 .048 5 4 4 2 .111 O 5 3 2 .077 4 From Table 37 we see that District 1 has more administrative positions between the principal and the superintendent and that the ratio of such positions to principals is highest in that district. Districts 3 and 5, with the lowest number of such positions also have the lowest ratio Of administrators to principals. In terms of the original propositions, then, one would say that District 1 offers the best organizational advancement structure, while District 3 and 5 Offer fewer possibilities 190 for principal promotion inside the district. There is no evidence regarding differences in opportunity for secondary and elementary principals, except that secondary principals probably have an advantage based on greater administrative experience because of the generally more complex adminis- tration Of secondary schools. Formalization Of rules and procedures is another organizational variable that could be detected from documen- tary sources. Each of the 5 districts worked out teacher- district contracts for 1968. While each of the contracts included provisions specifying salary, school calendar, status of the teacher organizations and grievance procedures, the contracts differed along two main dimensions: the extent to which authority was explicitly granted or denied teachers in a number Of disputed areas and the total number of pro- visions explicitly setting the locus Of authority for various decisions and procedures in each district. Using these items as indicators Of the formalization of authority relations, I gave each district a score on formalization corresponding to the number of separate contract provisions dealing with authority and autonomy. Separate analyses of the contract provisions were carried out by two coders. For the few items where interpretations of contract pro- visions differed, mutually satisfactory compromises in coding were reached. From this coding procedure, 55 limits on administrative or teacher decision-making and autonomy were identified. None of the districts had each 191 of the 55 provisions in its contract. Table 38 sets out the actual incidence Of these items in each district. TABLE 38.--Formalization Of Authority Relations, by School District. Number of Provisions District on Authority Rank: Formalization l .618 3 2 .636 2 3 .382 5 4 .582 4 5 .800 1 Looking at Table 38, one Observes that District 5 has a much more formalized authority structure, Districts 1, 2 and 4 are similar and moderate in formalization while District 3 has the least concern with formalization, con- siderably lower than the others. Data from the district handbooks and contracts did not suggest any major variation by district in adherence to universalistic criteria of evaluation, expertise of administrative staff or in standardization Of procedures and content (all districts had standard texts and a set school calendar). While there may actually be variation on these dimensions, such variation could not be inferred from the district documents. 192 The provisions of teacher-district contracts did, however, suggest some differences in the centralization Of authority in the 5 districts, independent of level of formalization. In coding the contract provisions regarding locus of authority and retention of autonomy, coders kept two tallies for each district: the number of provisions in each contract that explicitly limit the authority or autonomy of district Officials and the number of provisions restricting the autonomy of teachers. In the contracts, principals appear as agents Of the school board and super- intendent, so the few cases of explicit restrictions on principals are considered equivalent to other restrictions on district authority. Table 39 sets out the number of restrictions of eacy type Observed in each district con- tract, as well as the ratio Of limits on administrators to limits on teachers that I use as an index of centralization. Using this ratio, rather than any simple measure of number of provisions, assures that level of formalization does not intrude on this measure of centralization. When the substance of the contract provisions is at issue, other district differences appear. Some district contracts emphasize teacher professional autonomy, some are largely concerned with benefits, hours and security, while others strongly emphasize procedures to be followed in resolving conflicts. Each of these types of emphasis involves some restrictions on teachers, some on district 193 .mGOHmH>OHm uomuucoo mm HHm mo umHH m MOM m xHocommm mom N mm. Hm HH m m mm. Hm h s H mm. sH m m m ms. mm HH N s ms. sN OH H cOHpmNHHmuucoU “xcmm _m OHumm coHumnpchHecd so muocomoa so muchuumcoo uOHHumHo .uOHHumHQ Hoonom an .muocomoa can GOHucHuchHEO< mo mfiosousm so mucHnHumcounn.mm mHmHumHoH co msmcomcoo OOOHOHm chHumoonom Honomoe cam HmmHocHHm« os.n o.H Hs.n sm. so. os.n os.n NN. NuHHHna os.n Nm. Ns.n sN. HN. Hs.n NN.n .osom Hs.n sm. so. ss.n os.n NN. sum os.n HN.n os.n HN. so.n scans HN. os.n sH. Hs. .muomo NN. os.n ms.n «muncmo sH. mo.n Onucmo so. Enos .Hon NuHHHna .oosom sum swans .muomo anncmo Onucoo .euom .umsm .moHamHHm> uOHHumHQ HmsuOMHucou can HmcoHumNHcmmHo mcoem mcoHHMHOHHOU ocm :mEHmommnl.Ns mHmms 199 of measurement of these variables. At least the data Show that hierarchy, formalization and specialization have stronger positive intercorrelations than their correlations with centralization. As expected, student ability and SES levels are significantly related, with districts having identical ranks on these variables. One interesting and unanticipated find- ing is the strong yet not significant correlation between contractual evidence of centralization and student homo- geneity. Perhaps the recurrence of Similar problems and the probable existence of community consensus about education encourage the standardization of problem-solving at the district level. In districts where students are very dif— ferent and problems in teaching and administration are very different depending upon neighborhood characteristics, it may be much more difficult to handle anticipated problems at the district level. The greater diversity of students may encourage greater autonomy of principals and teachers in those districts. Since the correlations among organizational elements approach statistical significance only in a few cases (although the trends that emerge are consistent with pre- dictions based on Weber's exclusion of centralization from the list of elements of ideal bureaucracy further analysis of the effects of these variables will consider each organizational dimension separately. 200 Effects of District Organization Upon Principals The first step in uncovering the relationships between district structure and principal reactions is to see whether or not there are any significant differences in principal perceptions and responses by school district. Accordingly, a one-way analysis of variance was run to check the importance of district identity as an independent variable. For those variables that Show a Significant district effect, I will then attempt to identify the Specific components of overall district identity that are responsible for these patterns of district differences. Table 43 presents two sets of data: the results of the analysis of variance for the Six principal variables that showed a significant district effect (5.10) and data on principal turnover in each district from 1967 to 1969, which were subjected to a Chi-Square test for significance. The data in Table 43 present several interesting items Of information. First, there are no Significant district differences for most of the principal perceptual and behavioral clusters. Despite the many differences in district structure already established, these structural differences had no effect upon extent of idea exchange, cooperation with others, job satisfaction, perceived own influence, reference groups, career plans or turnover among the sampled principals. Although the turnover rates varied 201 s m m m H mmo. .cus< :30 .qoo¢ oH.A message .aooo oH.A mucoumm .mooo oH.A .mumon .QOOU oH.A .one .B .moou oH.A mucosHmcH :30 s H m N s omo. .mnmsnoucH .one .9 oH.A .cHEcm Hoom oH.A cOHuommmHumm 0H.A HMUHmmnm .QEH s s m N H Noo. osmcsnuxm .msH s N m s HmuH soo. oncHuucH .osH oH.A namosum .unomsH oH.A .coHHoo monocoxm oH.A c30o omnmcuxm oH.A mo monocoxm oH.A mmocmxuoz 0H.A mocoHHomxm m s m m H «Ho>oH oemz mquOHchoHs oHamHum> mEouH ucoHoMMHo hHDGMOHMHcmHm co .HOHHumHQ comm mo xcmm .momcommom can mcoHumoOHom HmmHocHHm cH mCOHDMHHm> uOHHumHann.ms mHm Hocuo How >om NHo>ocusu How umou mumsvmano Eonm« oH.A opmm HO>OGHDB N s m s H soo. .ouo .anm m>Huoe oH.A mCMHm m H N s s soo. .OHHHHO NuHanssoo oH.A .Hom HObcH .boum oH.A .cHum Hooumu oH.A .msm .cho .QEH oH.A comma .cho .mEH m s m N H «Ho>oH oEmz oOSMOHMHcmHm OHQMHHm> mswuH uanmHMHo hHuGMOHchmHm co .uOHHumHa comm mo xcmm .cossHucooln.ms mqmda 203 from 8 percent to 45 percent, even these apparently large differences were not statistically significant. Second, for those variables that do Show significant district effects, distinctive patterns in the data emerge. In District 1, principals Show least concern for interfer- ence or criticism by the Teacher Organization or the com- munity and report the greatest importance of both intrinsic and extrinsic aSpectS of work, believe their authority is adequate and are most active in the Principal Organization. That pattern is almost exactly reversed in District 4, where TO and community problems are greatest and where principals are inactive in their organization andsee little importance in extrinsic aspects of work. Districts 3 and 2 have more moderate scores on all the variables, while Dis- trict 5 is like District 1 in some ways. Perhaps a better way of handling this data is to ask how much the principal variables interrelate, rather than labelling each district. Table 44 presents Spearman rho correlations for the district ranks on each of these variables, SO that one can check the evidence pertaining to the proposition that favorable perceptions Of the organi- zation (such as greater adequacy of own authority, less TO interference, less community criticism) should produce greater levels of self-investment in work, suggested by the importance of intrinsic rather than extrinsic aspects of work and greater dedication to work, suggested by activity in principal organization mobility orientation. 204 .oo.m can sH oH. us namoHsHaon mH o .s u 2 nos nn NN.n os. HN. NN. oo.n NN. um>ocnsa .nsnm nn NN.n ss. Hs.n sm. NN. .o .aHHN m>spo¢ nn so. Hs. Hs.n os.n .uHNO .ssoo nu os.n sm. so. .amse .ausa nn os.n os.n moussnmscH oa nn oN. osmasnuxm In OHmGHHucH noocmuuomEH Ho>ocusa O .uHHU .voo¢ .ucH OB OHmcHuuxm OHmcHHucH .cHua .ssoo .asoa .mpoMHHm> HmmHocHHm wouooHom so mxcmm uOHHumHQ wooem mGOHHMHOHHOUnI.ss MHmHB 205 The only statistically significant positive correla- tion presented in Table 44 is that between principal per- ception of adequacy Of own authority and importance Of intrinsic aspects of work, which provides some support for the proposition that favorable perceptions of the organization induce greater self-investment in work. Fur- thermore, importance of intrinsic aSpects of work and perceived adequacy of authority were also strongly, though not Significantly, related to principal turnover rates. For this sample, it appears that favorable organizational climate and high self-investment in work produce higher turnover rates among principals than do negative aspects Of the organizational setting. Unfavorable settings, here indicated by community criticism and interference by teacher organizations, do not lead to significant principal turnover (or escape) while favorable organizational features not only produce greater self-investment but also more mobility, suggesting that these principals are not escaping from a bad situation but more probably, maximizing important rewards through job mobility. Looking at the other predicted patterns, perceived adequacy of authority is the best single predictor of greater activity in principal organizations, followed by importance of extrinsic, then intrinsic, aspects of work. This mild confirmation of my propositions runs counter to another possible explanation Of principals' concern with their principal organizations, that unfavorable aspects of 206 work settings leads to such concern. On the contrary, we see that difficulties with teacher organizations and level of community criticism greatly decrease principal to involvement in principal in principal organizations. It is important to distinguish here between prin- cipal perceptions of organizational support or problems and other evidence regarding organizational structure, in assessing the true importance Of structure and Of percep- tions of structure upon principal responses. The original prOpositions suggest that self-investment in work, here shown by the importance of intrinsic aspects of work, and dedication to work, suggested by activity in principal organization and district turnover rate, should be higher in diStricts that score higher on hierarchy, formalization, and Specialization and that have lower scores on centrali- zation of authority. Student ability and SES might also exert an independent effect upon principal self-investment and dedication. Furthermore, perceived adequacy Of own authority Should be related to lower centralization in the decision-making structure of the district. The pattern of relationships between organizational variables and these intervening and dependent variables is set forth in Table 45, which Shows the Spearman rho correlations between each set of variables, based upon the district ranks on each variable. The data in Table 45 present some interesting sur- prises. First, although hierarchy, the best Single 207 Com. M och NH Hm>oH oH.w um quOHchmHm mH a .m n 2 Mom NN. oo. so. No. so.n Hs. so. smm NN. oo. so. No. so.n Hs. so. ouHHHn< NN. os.n NN. os.n No. os.n oN.n NHHmcmoosom Ns.n HN.n so. so.n NN. Ns.n os.n Humps muss oo. HN.n Ns.n so.n HN. Ns. sH. .mnoon s os. so. os.n sN. HN.n os. os.n .omonmmnNHHmuucmO oo. sN. so.n oo. os.n HN.n os.n .NncoonNHHmnucmO mo.n sN. so.n Ns.n os.n HN.n os.n coHHmNHHmsnom oH. oo. os.n oN. so.n os. Hs. NonsensHm Ho>ocuse .o.m .uHHU .cpsd .muoucH OchHHuxm OHmcHHucH .cHHm o>Huo¢ .EEOU .UOOd OB .mEH .QEH noncommom HmmHocHHm .mpomHHm> HmoHOcHHm com muouomm HchHumNHcmmHo coospmm mcoHpMHmHHOUnu.ms mqmda 208 indicator of level of bureaucratization, is moderately and positively related to degree of emphasis principals put on the intrinsic aspects of work, none of the other purely organizational variables correlates much with level of importance of intrinsic factors. Only one organiza- tional factor, number Of specialist departments, moderately correlates with principals' expressed emphasis on extrinsic aspects of works and this correlation neither supports nor challenges the propositions, except in that I had antici- pated that each aspect of bureaucratization would encourage importance of intrinsic rather than extrinsic factors of work. That does not appear to be the case, except for a slight difference according to district rank on hierarchy. The propositions are futher challenged by the fact that level of activity in principal organization and prin- cipal turnover rates do not Show the expected strong corre- lations with organizational features. Except for a strong negative correlation between district formalization and principal turnover, it appears that organizational variables have little impact on these two measures Of principal dedication. A final set of unexpected results concerns the effects Of Size of district auxiliary staff, which shows a negative correlation with adequacy of principal authority and a significant positive correlation with principal awareness Of community criticism. It is possible that the auxiliary professionals, mostly experts in some curricular 209 field, are felt by principals to have tOO much influence over the programs and procedures of the district and to compete with principals for the attention of teachers. It is also possible that the community may be critical of the expense such large staffs require, or the types of programs such staff members initiate. Using cross-sectional data, however, it is impossible to rule out the chance that community problems that produce criticism of schools also produce greater district efforts, in the form of larger staffs of consultants and advisors trying to correct these problems. Turning to the effects of contextual variables, student ability and SES are Significantly related to principal emphasis on intrinsic aspects of work and moder- ately related to perceptions that principal's own authority is adequate, suggesting that principals find student compo- sition to be an important part of their work climate. Furthermore SES and student ability are moderately related to principsls' turnover rate, further indicating that stu- dents may affect principals' dedication and career orienta— tion. An alternative explanation must also be conSidered-- Officials in higher SES districts may renew fewer principal contracts and may require more evidence of high self— investment in work as a condition of contract renewal. Since the data do not distinguish voluntary from involun- tary job turnover, it is impossible to confirm either explanation. 210 Finally, student homogeneity is moderately related to principal perceptions of community criticism, perhaps because more homogeneous communities presenting one con- sistent brand of criticism are more effective in presenting these items of criticism. However, the repercussions of student contextual variables will not be considered in more detail since they do not directly bear on the propositions at hand. It is possible, though, that contextual variables have affected the develOpment of district structure and patterns. Of the many non-organizational factors differenti— ating the five districts, two critical aspects of district history appear to be most relevant to a discussion of teacher reactions and behavior. When this survey was con- ducted, collective bargaining for teachers was a new pheno- menon and there was competition between two unions to see which would represent teachers in each district. Four of the districts studied were represented in collective bar- gaining by locals Of the Michigan Educational Association, which played the role of a professional association rather than a traditional union. Teachers in District 1, however, had voted to be represented by the Michigan Federation of Teachers, a branch of the AFL-CIO, which held a more tradi- tional, trade-unionist approach to bargaining. Along with these differences in representation, the five districts varied greatly in the amount Of conflict associated with contract negotiations completed just prior 211 to the survey. District 1 had experienced a lengthy strike, District 5 a short strike, District 3 had no strike but very difficult negotiations, while District 4 had fewer problems and negotiations were easiest in District 2. While the effects of such differences in the history of collective bargaining in the 5 districts are beyond the scope of the present research project, it is important to recognize that these and other district contextual differ- ences may be responsible for the Observed differences in contract emphases and may have effects on teacher and principal perceptions and behavior. In summing up the status of the original proposi- tions, taking into account this set of data on district organization and principal responses it is only possible to say that bureaucratization at the district level does not Significantly discourage principal self-investment and dedication but there is no significant evidence that bureau- cratization encourages these positive principal responses. However, student SES and ability levels do significantly correlate with indices of self-investment in work and both are strongly related to perceived adequacy of principal authority and to principal turnover rates. Thus, any fur- ther research on principal behavior Should consider pro- positions linking student contextual variables to principal behavior and should distinguish voluntary and involuntary turnover. 212 Effects of District Organization Upon Teachers Although the school organization's characteristics are presumed to be most critical to teacher behavior, it is conceivable that district organization may affect teachers both directly and through limits on school organization. For instance, the auxiliary professional staffs at the district level are composed of specialists in curricular areas, many of whom began their careers as teachers. Hence, the teachers in a given district may see these positions not only as potential aids for planning and evaluating programs but also as potentials for upward career movement. Other aspects Of district organization may affect teacher behavior as well, affecting possibilities for OOOperation, idea exchange and importance Of various reference groups. In fact, when the perceptual variables are subjected to analysis of variance to check the independent effects of district, thirteen of the 21 variables Show a strong dis- trict effect, as Table 46 shows. Table 46 Shows that, despite my expectation that school level and organizational characteristics would be most important in explaining variations in teacher responses, some dimensions of district organization must be responsible for the significant district effects shown for thirteen of the teacher perceptual and response variables. Those teacher variables that showed a significant district effect must be further analyzed, to see which 213 TABLE 46.--Effects Of District Upon Teacher Variables. Significance of F for District Variable Effects* Importance of Extrinsic ASpects .0001 Importance Of Status >.lO Importance of Intrinsic Aspects >.lO Cooperate Outside School .06 Cooperate Inside School .01 Exchange Ideas with Outsiders .06 Exchange Ideas with Insiders >.10 Exchange with Close Colleagues .Ol Satisfaction--Physical ASpects .0001 Satisfaction--Non-Physical .0001 Satisfaction--Relations with Superiors >.lO Poor Administration >.lO Principal Supports Teachers .0001 Teacher Org. Hinders Teachers .0001 Own Influence >.lO Reference Group: Close Teachers .06 Reference: Superiors .002 Reference: Friends Outside Education >.10 Adequacy of Principal Authority .0003 Career HOpes--Teaching Level .003 Career Hopes--Administration >.lO *Part of a 2-way Friedman Analysis of Variance checking the independent effects of district and school level. Data on effects of level is in Tables 18-23. 214 specific features of district organization can be isolated as probable reasons for these strong district differences. Recalling the original prOpositions, I expect that hier- archy, formalization and specialization will encourage self- investment in work, indicated here by reliance on colleagues rather than superiors as reference persons, by providing a more professional climate (one in which principals support teachers, OOOperation is high, teachers and administrators exchange ideas and principal authority is considered ade- quate). Furthermore, more bureaucratic districts Should, from the propositions, encourage stronger career orienta- tions on the part of teachers, with the typical low mobility teaching career characteristic only of the less bureaucra- tized districts. Since centralization of authority is a non-bureaucratic form Of organization, Offering few advan- tages for professional employees, low centralization should be related to each Of the anticipated effects of bureau- cratized structure. Emphases displayed in teacher contract provisions should bear out individual teachers' perceptions of prob- lems in their districts. Thus bureaucratic districts should permit more emphasis on protection Of professional autonomy while less bureaucratic districts should display more concern with benefits, safety and other extrinsic asPects Of work, since those districts do not encourage a Strong professional orientation on the part Of teachers, aCoording to the prOpositions. 215 To test these expectations with the available data, districts were ranked on the mean teacher responses for each of the thirteen significantly different teacher vari- ables. Each set Of rankings on teacher responses was then apired with each rank on district characteristics so that Spearman rho correlations could be calculated. These correlations between organizational characteristics, con- tract provisions and teacher responses are presented in Table 47. The correlations presented in Table 47 offer several interesting confirmations and contradictions to the prOposi- tions in question. First, there iS only moderate evidence that more bureaucratized districts encourage greater reliance on professional or collegial reference groups. Although the correlations of reference group use with both hierarchy and number of departments suggest some support for greater reliance on colleagues than superiors, staff size and formalization, as well as centralization, appear to discourage relative importance of colleagues as refer- ence group. These trends, though not statistically signi- ficant, suggest that district specialization does not greatly encourage teacher self-investment in work (indi- cated by choice of reference group). The correlations With size of staff might be interpreted in another fashion, though, as indication of the importance of a reference group composed of superiors who are experts in curricular fields, hence colleagues. 216 .oo.M onu NH .OH.w um MGMOHMHcmHn pH 8 .m u 2 Mom oN.n os.n o oo. o os. oo.n oo. Ns.n so. oo.n NN. NN. .Huaoo Hs.n NN.: oo.n oo.n oN.I NN. oo. om. NN. mo. oo.n mo. oo.n oo. oo. ouommm oo.n NH. oo.n oH. so.| so. oo.: oo.: HN.n Ns.n sH.n Hs.n so. so. .cuadom oo.n om. oo.n oH. HN. HN. oo. on. no. om.n oH. oo.n so. so. .mocom oo. Ho. oo. oo. Ho. oH.: os. oo. om. oo.n NN. NN. sH.n sH.n .woum “youuucoo NN. oo.n oo.: no.1 so. oo.n No. os. om. o oo. oo.: o. o .HHnt NN. so. om. so. sH.n NH. oo.n os.n oH. NH. Ho.n NN. Hs.n Ho.n mason NN. oo.n oo.n oo.n so. oo.n No. ms. om. n oo. oo.: o o mam us. oo. oo. os. oo.n o oo.: oo. NN. oo.n NH. NN. oo.: oo.: onmwum oo. oo.n oo. oo.n o oo.: oo. os. oH. Hs. om. oN. . oo. oo. mason. oH. os. sH.n oo.n oo.n om. oo.n oo.n Hs.n oo. hm.u oN.: oo. oo. mason oH.n so. oH.n HN. Hs.n NN. oo.n mo.n oo.n oo. mo.n oo.n oo. oo. oucou oo.n oo.n Ns.n oo. Hs.n NN. oo. o oH. Hs.n oo. oo.n oo. oo. .Euom sH.n oo.n oH.: No.1 oo. os.n o.H no. mo. mN.n oo. oo.: oo. on. .Hon Ho>ocuse Hmoe us< mobm Boom ucH now one new Broxm ocoxm Hmoou omooo .Huxm uOHHumHQ Hum 09 Hum :02 ham moHQMHHm> Hocomoe .mpoMHHm> Hocomoe ucmoHuHcmHm can mOHumHHouomumco uOHHumHo coosuom mcoHumH0HHOUII.hs mance 217 The most surprising data concern the strong rela— tionship between both SES and student ability and the relative importance Of colleagues rather than superiors. It is possible that districts with high SES, high ability students attract more expert teachers who bring with them a strong interest in colleague ideas and opinions but it is also possible that student composition exerts some independent effect upon teachers, whatever their backgrounds. These rival explanations cannot be tested with the available data Since the small N for districts makes it impossible to consider more than one possible source Of variation at a time. Leaving for the moment the effects of district organization upon teachers' choice of reference group, let us turn to a consideration of the effects of district variables upon dedication to work, and importance Of work even though the role of the hypothetical intervening vari- able cannot be completely determined. The original prOposi- tions suggested that the more bureaucratic the work struc- ture, the greater the employees' dedication to work. The rationale for this proposition was the belief that bureau- cratic structures Offer more Opportunity for advancement, more autonomy and support from superiors, more opportunity for Observation by close colleagues and as a further conse- quence, a different.less anti-professional role for the teacher union to play. The data in Table 46 Show that district hierarchy Significantly encourages exchanges outside 218 schools, satisfaction with non-physical aspects of work and greater principal support for teachers. Furthermore, in more hierarchical districts, teacher organizations are not seen by teachers as Significant hindrances to profes- sional objectives of educators. Not all the indicators of bureaucratization have such positive effects upon teachers, however. Contrary to the propositions, formalization does not significantly correlate with any of the variables, though it Shows a sur- prisingly strong relationship with teacher perceptions of their unions as hindering educatinal Objectives. Number of Specialized departments and size Of auxiliary professional staff Show only weak and insignificant correlations with perceptions Of work climate, level of self-investment in work and teacher dedication. Size of staff is strongly related to teacher perceptions that principals had adequate authority, however, supporting the original propositions. On the whole, the correlations between each aspect of district centralization and teacher perceptions Of the character of relationships and problems in the districts support the propositions I offered. Table 46 shows us? significant negative correlations between centralization and principal support Of teachers, and a moderate but insignificant correlation between centralization and teachers' perceptions that the teacher organization hinders teachers. Highly centralized districts are characterized by little idea exchange between teachers and others 219 outside their schools, little satisfaction with physical or non-physical aspects of work and little reliance upon collegial reference groups. These results were predicted by the propositions, where centralization appeared as an alternative to bureaucratization and centralization was expected to produce an unprofessional work setting. The strongest support for the propositions being tested comes from the data concerning teachers' career plans. Only centralization of authority is strongly and positively related to planned immobility (i.e., plans to remain in teaching with no upward movement). District scores on hier- archy and number Of specialized departments seem to encour- age teachers tO plan upward career movement, as the proposi- tions predicted. It is interesting to note here that the only other variable significantly encouraging immobility is student homogeneity. Presumably, when students are very similar, the problems teachers face are more predictable hence less threatening, perhaps encouraging teachers to remain in a stable and predictable environment. When data on actual teacher turnover are considered, though, one realizes that none of the district variables significantly affects actual turnover. It must be remem- bered that not all such turnover is voluntary, since probationary teachers are not necessarily granted tenure and Since mandatory retirement and maternity leaves might account for some Of the Observed turnover in each district. However, the teacher turnover rates probably have a higher 220 component of voluntary turnover than do the principal rates, since principals have only two year contracts. Turning to the relationships between observed areas of emphasis in teacher contracts and district ranks on teacher reSponse variables, Table 46 shows us, not unexpectedly, that teachers' expressed concern for extrinsic aspects Of work is strongly correlated with contract emphasis on benefits, on safety and on retention of board authority. Another unsurprising strong correlation pairs contract emphasis on professional aspects of work and expressed teacher satisfaction with non-physical aspects of work. Contrary to expectations, there are no Significant correla- tions between contract emphasis on professionalism and either use Of colleagues as important reference group, or extent of idea exchange with close teachers. Extent of contract emphasis on retention of rights of the school board shows a strong positive correlation with teacher perception that teacher organizations hinder educational concerns, a rather intriguing finding. It is possible that the struggle between conservative school boards and aggressive teacher organizations requires the teacher unions to ask more of teachers. Unfortunately, the cross-sectional data cannot tell us whether teacher organi- zations are forced to use more aggressive tactics by resistant boards or whether school boards take tougher positions when threatened by aggressive teacher organiza- tions. 221 Not all the correlations from Table 47 are so easily explained. For instance, teacher perception of cooperation with those outside own school building is Significantly correlated to three kinds Of contract emphasis: benefits, board rights and safety. Perhaps COOperation is possible between teachers and district Officials when these extrinsic aspects of school and district activity are at issue, pro- ducing consensus in the form of contract provisions on these items. When intrinsic aSpects Of work, such as limits on professional autonomy in conducting courses, rights to privacy and appeal for punishments, etc. are at issue, there may be less agreement that these constitute apprOpriate areas for bargaining. Thus, teachers concerned with profes- sional rights and autonomy may find OOOperation with district Officials much more difficult both for contract negotiations and for everyday problem—solving. Emphasis on typical union bargaining areas like benefits and safety is more in line with administrative perceptions of what issues are nego- tiable, hence contributing to COOperative interaction within the district. However, the absence Of strong collegial reference groups undermines this interpretation somewhat. Considering all the asPects of district organization and their effects upon teacher perceptions, the data provide mild support for the prOpositions that more bureaucratic structures provide favorable environments for professionals, hence encourage self-investment and probably dedication to work. Looking at the array of Significantly different 222 teacher responses, only 3 Of the 14 areas Showing strong district effects cannot be partially explained using one or more of the district characteristics identified in this study. Idea exchange with close teachers, satisfaction with physical aspects Of work and actual turnover rates are not significantly related to any of the district characteris- tics studied. Even if we treat the data regarding emphases in contract provisions with Special caution, Since these could most plausibly be considered either dependent or independent variables since contract contents change every year, these differences in contract provisios were the sole Significant correlates of only three teacher variables: importance of extrinsic aspects of work, and cooperation outside and inside school, none of which are central to the propositions being tested. The eight most important sets Of teacher variables can be at least parti- ally attributed to structural and compositional district characteristics, more enduring than contract provisions and more likely to independently affect teacher variables. But do district structural characteristics have Similar effects on teachers' dedication? Table 48 presents correlations between structure and those indicators Of dedication showing a strong overall district effect. The data presented in Table 48 Show that district structure appears to be more important in explaining teacher behavior than I had originally anticipated. I had expected bureaucratization at the district level to have 223 .oo. M can sH oH. w as unsoHchoHs as .moosuumss s u 2 nos .HMOHHEOQM OHM HHQ‘ HOW UGM mmm HON chHHMHOHHOUc oH. oo.n No.n oo.. Ns.n ss. os. os.n oo. oN.n oN.n os. Hammad oN. . No.n NN.n os.n oo.n os.n os. oH.n os.n oH. oo.n oN.n nonsmom o oo. oo. oN. NH. oo.n oo.n o oo.n oN. oN. os.n smNHsaomncoz ss.n Ns.n Ns.n so.n NN. oo.n so. os. os. sN.n sN.n so. mmmusoo oH. No. oo.n oo. NH.n ss. oo. oo.n oN. oo.n oo.n oo. some mass oN. so. ss. oN. No. ss. oo. oH.n o os.n os. oN. mHscusoo os. oo. oo. oH. NN. ss. os. o oN. oo.n oN. oo. mHoom oo.n ss. No.u oo.n so. so. oN.n oo. oH.n os.n os. oo. smocmnmscoo oH.n NH. oo. oo.n ss. Ns.n os.n oo. oo.n oN. oN. os.n mom osHm os.n NN.- NH.n oo. oo. No.n os.n oo.H oo.n oo.n oN. oH.n seco Hsooq ss. No. No.n oo.n ss. Ns.n o oN. oN. oN.n oN.n oH.n mos o oN. NN. oo.n os. NH. No. oN. oH.n oH.n os.n oH.- oN. .oouo mumps oo. ss. ss. os. oo. No.n oH.n oN.n oN.n oo. oN. oH.n .oso HsooH oo. .os. so. os. oo.n NH., oo.n os.n os.n os. os. oo.n oa o>Hso< soHHsaou Nsmsss oases swans «sous .oosom .sss Nssns noses some .euoo .umHs .ooHusoHsoo mOHumHuouomumcU uOHHumHo .xuoz ou coHuMOHUmQ .mHQSDMQB GEM mOflflmHHOUUMHMSU #UwuumflQ C003H0m mCOflDmHOHHOUn-I.mv BQANB 224 only weak effects on teacher dedication, though in the same direction as the anticipated effects of school organization. It is interesting that district organization appears to be related to the variation in teacher dedication more than school organization appeared to be and in a manner that largely supports the ideas, if not the form, of my proposi- tions. For instance, looking at Table 48 one notes that district ranks on hierarchy are strongly and positively correlated with district means on number Of books read, time spent reading for work, taking non—required courses and participating in state programs. Furthermore, number of Specialist departments also moderately correlates with time reading and size of auxiliary staff positively corre- lates with several indicators of dedication--committee work, building representation, attendance at conferences and membership in many teacher organizations--while negatively correlating with absenteeism. Further confirmation of my ideas about the effects Of structural variables upon workers is apparent in the very different effects of centralization of authority, which is negatively related to participation in state programs, attending conferences, reading books, time spent reading and enrollment in only required courses. The only Observations that strongly contradict my expectations are the negative correlations between district formalization and teacher willingness to re-enter education and the positive correlation between number of specialist 225 departments and days absent, as well as the overall depres- sing effect of number of departments upon teacher partici- pation in union activities. However, since the implications for teachers Of district specialization are not immediately apparent, these apparent contradictions are less interest— ing to me than the unexpected strong support for the notion that bureaucratization encourages teacher dedication detailed in the last paragraph. Turning to the other district characteristics, it is interesting to note that SES and student ability appear to encourage teachers to spend more time reading and to take courses other than those required by their contracts, although these two variables do not encourage any union activity except for participation in state programs. When the areas of contract emphasis are related to teacher dedication, several interesting patterns emerge. First, emphasis on professional autonomy is strongly related to participation in union activities and at conferences and to low absenteeism, but not to any evidence of truly pro- fessional dedication. On the contrary, one sees that the best contractual predictor of time reading is emphasis on benefits, while emphasis on safety and on conflict also positively correlate with reading books and journals and taking courses not required by contract. Emphases on safety and conflict resolution appear to predominate in districts where teachers are active in union activities other than local committee or representative work. Once 226 again, the correlations on teacher behavior and contract emphases cannot be interpreted causally, since both may undergo major changes from year to year and since correla- tional data alone cannot suggest direction of relationships. In summary, the proposition that bureaucratic forms of organization encourage dedication to work received sur- prisingly strong confirmation from data on district charac- teristics, although not all supportive evidence met the criteria of statistical significance for the small number Of districts. Since each district had approximately the same percentage of secondary teachers, compared to elemen- tary teachers, these effects cannot be attributed to occu- pational differences but appear to reflect a genuine reac- tion to structural differences. Conclusion Data on the structure and effects of other aSpectS of district organization mildly confirm the propositions of this study. As expected, the measures of centralization were moderately correlated but were only weakly related to other organizational dimensions. However, hierarchy, formalization and Specialization showed only weak and inconsistent intercorrelation, Offering no adequate basis for choosing between the models of bureaucratization other writers have propounded (see Chapter II for summary). As the data on schools led us to expect, district scores on SES and ability of students were significantly 227 correlated and showed little correlation with homogeneity of students. The Observed variations in district structure seem to have only a few important effects on principal percep- tions and responses. While bureaucratization at the dis- trict level does not appear to discourage principals' self-investment and dedication to work, there is no significant evidence that it encourages these responses, L. although the data suggest moderate tendencies in that direc- tion. Student composition variables seem to play a more important role in predicting principal responses than was anticipated, Since SES and ability levels of students correlated strongly with indicators of principals' mean levels of self-investment and dedication. Surprisingly, the district variables appear to be more critical in explaining Significant differences ' in teacher responses across districts than in explaining principal differences. Particularly, hierarchy at the dis- trict level affects teachers more than principals, contrary to my expectations. Although not all the evidence is sta- tistically Significant, data on district effects upon teachers Offer somewhat more support for the prOpositions relating bureaucratization to high self-investment and dedication than did the principal data. Considering the role various district characteris- tics appear to play in explaining variations in teacher responses, district hierarchy appears to be the most 228 significant Single element, strongly correlating with three sets of teacher responses that indicate that teachers in more hierarchical districts find more support fOr self- investment and dedication to work. Since hierarchy is Often considered the best single indicator of level of bureau- cratization (see Hall, 1967), this set of data must be considered to at least moderately support the propositions in question. For teachers as well as for principals, though, SES and ability of students,which are related to hierarchy, appear to encourage self-investment in work, evidenced by greater reliance on close colleagues as a reference group for teachers. Data concerning possible causes and effects of teacher contract emphasis showed that districts where teachers cooperated with outsiders and expressed concern with extrinsic aspects of work produced contracts emphasiz- ing extrinsic aspects of work. Contracts emphasizing pro- fessionalism occurred in districts where teachers derived great satisfaction from non-physical aspects of work. Less purposive career orientations of teachers predominated in districts higher in centralization of authority, as my propositions predicted, although actual turnover rates could not be attributed to any organizational variables. On the whole, the data on district structure and effects provide moderate support for the prOpositions that bureaucratization encourages self-investment and dedication 229 by providing a favorable environment for professional employees. Since several aspects Of organizational struc- ture could not be considered, due to limitations in the data making differentiation Of districts impossible, it is encour- aging tO find that the few variables that could be measured yielded some confirmation of the propositions. CHAPTER VIII SU MMARY AND CONCLUSIONS Introduction This cha pter s ' ummarizes the theoretical background Of the r eSearch - . describes the current status Of the tested PrOpOSiti on S and offers some conclusions about the Contribut' 1°38. lim' ' itations and implications Of this research Project. This di . organizational ::e1:tat10n weighed the impact 0f several tralizathn, merléTbleS-'hierarchy, specialization, cen- and prinCipalsn daszation and universalism-«upon teachers' out the effect edlcation to work, as well as separating 5 0f structural variables represented by' OCCupati Onal titl es Upon the reactions and behavior of this clarify the r racteristics self- ple. ‘1 th I - Ship be twee ° n dedication behavior and two cha Of pr ofess' iona ' 118m. here considered indicators of invest ment 1 n W’ - ork (relative importance Of intrinsic n collegial re relative imPor‘ asPect s of work and worker rel' iance o ference grOUPs) . . F lnally' this Project compared the ndividual variables tanc e of s tructural variables versus 1 230 231 like age and sex in order to distinguish the role each factor plays in explaining dedication to work, self- investment and job-leaving of teachers and principals. As a basic step in accomplishing these Objectives, the project also investigated the interrelationships between bureaucratic elements and rational elements in formal organizations, providing new information in an area of major concern to students of bureaucracy. This explication of the patterns of organization in public schools also makes an important contribution to a neglected area of educational research. Theoretical Background of the Research To reiterate the rationale for these research activities, I will summarize the basic arguments presented in Chapter I and II: 1. Previous studies of worker behavior have failed to explain bases for behavior variations of workers except as a function Of hierarchical differences. 2. Particularly, research on job satisfaction as a determinant of behavior has produced no ade- quate, consistent explanation of worker behavior. Not only that, but job satisfaction hypotheses do not clearly link behavior at work to any accepted social-psychological explanation of human behavior in general. 232 Faunce, Vroom and Dubin each contributed to a more promising approach to understanding variations in worker behavior. Basically, they suggested that job conditions will greatly affect behavior only if workers see their occupational roles as very critical to self-esteem or ego. Those workers to whom the job iS an unimportant necessity of life will be unwilling to devote more than minimum time or energy to work, regardless of extrinsic induce- ments or managerial warmth. Finally, my extension of these ideas focused on the importance of structural (occupational and organizational) determinants of probably worker self-investment and consequent dedication to work. I expected that structural factors would be more important than personal or family characteristics of workers and also more important than reported level of job satisfaction in predicting behavior at work. Definitions of the key concepts and specification of the general propositions plus adaptations to the case of public education follow. General Propositions Definitions A. Self—investment: a process through which the degree Of importance Of social encounters upon 233 self esteem becomes differentially distributed among social roles (see Faunce, 1972). Indicators Of area of self-investment include use of various standards of performance, use of various refer- ence groups. B. Rewards: intrinsic or extrinsically derived gains from activity, perceived by individual to be contingent on own behavior, within limits imposed by structure (such as rules). C. Dedication to work: efforts to perform the work role well. "Well" is determined subjectively, based upon reference group standards. Indicators include low absenteeism, preparation for work in free time, extra time and attention devoted to work (or improving work situation) beyond that formally required for employment. D. Purposive mobility orientation: Career planning strategy aimed at moves to more rewarding jobs; random, geographic and affiliative factors are unimportant in decision to quit or stay. Bropositions Given that an individual believes rewards crucial to self—esteem are contingent upon own efforts, then 1. +perceived potential for reward at work+ +self-investment in work. 234 2. +self-investment in work+ +P (dedication to work roles). a. +perceived potential for reward at work+ +P (dedication to work). 3. tdedication to work+ +P (decision to quit is based on mobility orientation). a. +self-investment to work+ +P (decision to . F‘“'-r-r-H=&1 quit based on mobility orientation). b. rperceived potential for reward at work+ +P (decision to quit is based on mobility orientation). 4. If occupational and/or organizational structure make purposive career orientation impossible to achieve (or to continue achieving), rself- investment in work+ tattempt to change structure of work (such as through unionization). Thus attempts to change structure may be equivalent to dedication activity. Furthermore, it appears that per- ceived potential for reward at work includes perceptions regarding likely success of attempts to change structure. Thus, a. iperceived success of change efforts+ +p (continued self-investment in work). 235 Adaptation Of the General Propositions to the Case of Public Education Given the evidence regarding references in oppor- tunity structure among the occupations in public education, the general propositions lead us to expect that: 1. In general, the occupation of school principal offers more rewards than does the occupation Of school teacher. (Principals have more auto- nomy, more scope for decision-making, less routinized work, earn higher salaries, receive greater status in the community, have later career ceilings and are more visible to collegial and organizational evaluators than are teachers.) In general, occupations at the secondary level Offer more rewards than occupations at the elementary level. (Greater specialization at the secondary level permits concentration Of energy in areas of expertise, making work more interesting and allowing the recognized expert more freedom from parental interference or organizational rules on course content; the possibility of later career ceilings and expert peer audiences become important in department- alized schools. The prestige of secondary personnel is higher than that Of elementary, the "clients" of secondary personnel have higher social status, and the activities Of secondary 236 schools are usually the subject Of community, not just neighborhood interest, making the per- formance Of secondary teachers and principals more visible.) a. Thus, ceteris paribus, principals Should have higher levels of self—investment in work and «1 "Col behave in a more dedicated manner than 1' '- teachers. Secondary personnel Should have (2' 1r #- higher self—investment in work than elementary personnel and should behave in a more dedicated manner than elementary personnel. Within each occupational category, variations in organizational structure should have noticeable effects on self-investment levels and dedication to work. Personnel in schools with more hierarchi- cal levels, greater division of labor, greater expertise of employees, greater emphasis on universalistic standards of evaluation and per- mitting decision-making at the level of appro- priate expertise should exhibit greater self- investment in work and greater dedication, behaviorally, than do their counterparts in other schools. Sex and age of worker may affect perceptions of reward structure because of societal discrimina- tion against older and female workers. The relative importance of sex and age of worker, 237 compared to structural variables, will decline in school Offering evidence that older and female workers can achieve career success. 5. Principals should exhibit a more purposive mobility orientation than teachers; secondary personnel should have stronger mobility orienta- tion than elementary staff. Research Evidence Bearing Upon the Propositions The next step in this chapter is to evaluate the propositions in the light of the data presented in Chapters IV, V, VI, and VII. Table 49 presents a summary of the evidence that was required to test each prOpositions and the status of each tested prOposition. In general, it appears that the prOpositions con- cerning the effects of occupational differences in reward structure were largely confirmed by the data on mean self- investment and dedication levels of elementary and secondary teachers and principals. These differences could not be attributed to differences in age, experience or educa- tional background of respondents and, although secondary personnel had a lower proposition of females than did elementary occupations, the predominantly female elementary principals exhibited higher mean self-investment and dedication than the predominantly male secondary teachers. Contradicting my expectations, though, these respondents did not perceive the organizational dimensions of reward RW‘I 238 .coHumEsmmc chu now unommom o: mzocm om oHnma usofiumo>cHIMHom mo ucmcHEHouop HMOHHHHO umOE mH wEocousm co>Hoonom umcu mzocm ss oHnma memHocHHm Mom .muocomou How Hm oHnme CH unommsm Hmuooow .cOHumNHHMHoon wo poommm oco co mH Hmmno>ou oHco msocm HH pomBnlpounommsm oHomHMH .oH pome .oH oHntIIOounommsm .mHmconum ouMHoHHOO OHsocm mGOHu Imoouom some can moHQMHHm> HmcoHumNHcmmHo co mouoom .ucofipmo>cHIMHmm mo mnouMOHUcH co Hoach ouoom cHsocm mmcHuuom Huos o>HuHommsm cam mcHoumsmH oHOE o>Hooumm 0:3 mnocomoa .cOHumNHHmuucoo co monoom Ho3OH .GOHHMNHOHMO ncmum can coHumNHHmEHOw .omHuHomxo .EmHHmmHo>H:s .coHumNHHMHoomm .thHm nHoHc co mouoom Hocch o>mc OHsocm mHoocom .oom .mnocomou omcu ucoEoOcm>om Hoouco How mono: can MHMHmm .GOHp nommmHumm meocousm .Hosom oHOE uuomou oHsocm .CHHQ use topmou hHuHOHmeo uoz HMHpcouom cumzou cH moocouom IMHO HchHumNHcmmHo com HccOHummsooo on» o>HoO nuom mnoxuoz ucOHumEDmmd xno3 CH unoEumo>cH IMHoms npumsos new Hchcouom po>Hoouom+ .H 0 .mumucoEoHo .mnmccooom NHo>oH Hoocom on HoMMHO OmHm xnos um moumsou How mHMHucouom .N.m .mnogomou can» woumsou whom o>HoOHom mHmmHocHHm .H d coHuHmomonm How unommsm Mo ucouxm cam monsom GOHumEHHmcoo now couHonm mama .coHuHmOQOHm ooummcm Ho Homoeow .muoummco mDOH>oum cH poumoa chHuHmomOHm mo msumumll.os mnm<9 9 3 2 .s. .MHCon .ueum HHe HOC «oneuCoEoHe cecu Hocch oneoCooom .muocoeou Ceca mCeoE HoCoHC e>eC mHemHOCHHm .MN poeanluuommsm oEOm .mHeoHOCHHm mCOEe uuommsm OC msocm ms poeB “mnocoeou mCOEe unommsm OC mSOCm oN poeB HemHOCHHm Co UOHHumHo mo wooemmo How unommsm uCeOHMHCon OC mBOCm ss poeB “muoaoeou CO powwow meow we: .mHO uOHHumHo msocm os poeB NCOHueNHceoHO Hoonom Ou omCommoH whosoeou How uuommsm OC mBOCm Nm poeB muoomme OHmCHHuCH Co mHmeCmEe Mom DOC .one moonm .mou How uuommom pCeOHm anon zoom HN one oN mmHnma .m oneuCoEoHo Ceca mmoH ucomne on oCe monsoon ooanooHCOC oxen .oHOE oeou oHsOCm HoCCOm [Hoe .oom oCe ..mmHo .monm CH o>Huoe whoa on .moOCoHom ICOO whoa oCoupe oHCOCm .m mueoCooom oCe mHemHOCHHm xnos ou COHueOHooo oce Hum Coosuob mCOHueHeHHOO mCOHum oCHm oHSOCm .hHHeCOHuemsooonenuCH .CoHueNHHeHquO oCe Hum coozuon COHueH noHHOO o>HuemoC NuCoEpmo>CH IMHom Op ooueHoH Cons COHueNHoHeoCeum oCe CoHu neNHHeEHOm .omHuuomxo .EmH IHemHo>HCC .COHueNHHeHoomm .mcoueHoHC Consume mCOHueH IoHHOO o>HuHmom oCHm oHCOCm HoCCOmHom mueuCoEoHo oCe whosoeou Ceca oHOE .xnos mo muoomme OHmCHHUCH oNHmecmEo .mmsoum .mmu HeHmoHHoo one oHCOCw .mnom mneoCooom oCe .CHHm HoCCOmHom oneuCoEoHo Ceca mueoCooom omnocoeou mom Ceca mHeeHOCHHm How Houeouo CoHueOHooo .eN m coHueoHemcs +Hnm+ .N o acoEHmo>CHIMHom mo Ho>oH moCHEHouoo onsuoouum HeCOHueNHCemHO .moHuomoueo HeCOHuemCOOO CHCqu .m m HoCCOmHom when nCoEoHo Cecu mneoCooom How .mnmnoeou Ceca mHemHOCHHm How noueouo “CoEumo>CHIMHom .eN m CoHpHmomonm How unoemsm mo uCouxm oCe monsom COHueEHHmCOU How oouHCoem eueo COHuHmomonm ooumeom.no Heuocow .ooCCHuCOUnI.os mqm<fi 240 COHueuHHeHoemm mHCO .uoeNNe OC mes NCOHeHoHC .mHemHOCHHm oCOE¢ .COHCC CH muH>Huoe No uCouxo ou mnosmCe mCHHoon ICOO .uuommsm uCeOHNHCon uOC was sHHs zoom ss was ss mmHnma .oN mHnma .smeNHscoo noz NCoEe>OE ooHHmeo on .one .Non HeHmoHHOO HON umooxo oN oHoeB .ooEHHNCOO uoz .sN oCe MN moHneB "CoHueEHHNCOU eueo HemHOCHHQ CH mCoHueHoHHOO uCeOHNHCmHm OC msoCm ms poeB “mumsoeop mueuCoEoHo mCOEe uuommsm OC mBOCm so poeB muH>Huoe COHCC oCHNHmeCmEo mHemHOCHHm oce whosoeou o>eC ON COHueNHHeHoomm .HCOHeHoHC 30H\3 muOHHumHo poomxo “HeseHoCDH3 one» NOCCeO .eueo OHueum CNHZ uCoEo>OE Hooneo onesms oonHmoo oce oocceHm CuOb oCe muoueOHoCH .ooo Coosuon mCOHueHoHHOO mCOHum quEo>OE onesms ooHHmoo oCe ooCCeHm oCe Hum NO mnoueOHoCH Consume COHueHoHHOO mConum mHeuCeEoHo Ceca oneoCooom mCOEe .wuocoeou Cecu mHeQHOCHHm mCOEe mCHCCeHQ Hooueo who: HHOB Op COHueOHooo oCe onsuonnum NO mCOHumoOHom poeHO>eN Coosuob COHueH nounoo mCOHum oCHN oHsocm .NHHeCOHuemoOOOIeHuCH Hmzmuesnss HO muHONNo COHCC+ Hnms .mHnHmmoosH NHHHHnos moxea onsuosnum NH .s m COHueucoHHO MNHHHQOE o>HmomHCm+ +aoHusoHsmc+ .no u .coHnmucoHno NHHHanm o>HmomHCm+ +Hlm+ .em w oneuCoEoHo Ceca oHOE oneoCooom “whosoeou Cece CoHueuCoHHo Hooueo o>HmomHCm oHOE o>eC mHeQHOCHum .mscu NouHHHQOe Hooneo ooCCeHms +xH03 ue mouezous .m fl oCe m m Huos Op CoHueOHooo moCHEHouoo HeHuCoupm oHeSoH oo>HoOHom .eN w COHuHmomOHm HON uuommsm No pCouxm oCe monsom CoHueEHHNCOU HON omHHCvom euea ICOHuHmomonm ooumeoe Ho HeHoCoU .emsaHuaoOnn.os names 241 ome No NCooComooCH HoeNNo oHHNHH o>e£ CoHoHHCO .EmHCOHCC mHCHeE mHOHomHm woe .xom No muooNNo HON NHommom OC 3ocm mN oCe sN mpoeB ooHHOHNCOO HoxHos NO xom Coos Co>o .NHommsm OC zoCm oN oCe oN mpoea He>oH Hoocom HoCNHo He Hocoeou No xom NO muooNNo HCNmCH nCeoE OC .COHCC CO oueHuCoO ICOO mHmcoeou HooHo mBOCm mN oCe sN moHneBnnooEHHNCOO Noz mmoHoHHCO oCe moHeE mCOEe .ooo mHOE ..ooo oCe ome Coozpon mCOHNeHoHHOO con .ooo No oCe .Nem NO mHOHeOHoCH oCOEe mCoHHeHoHHOO cmHm COHNeOHooo NO muoomme HeCOHmmoNOHm CecN COHCHH £NH3 Hum NO mCOHN neHoHHoo HomCOHum oCHN OH Hoomxe .NHCO mHecoeoN No mmCOHm omocu Hom COHN neOHooo NO Hm>oH oCHE nHouoo mOHumHHouoeHeCO oHHEeN oCe HeComHom .N COHHeOHooos +coHuostHums+ .H “mCOHueCemem oueCHouHe COHNeOHooo NO muoomme umHCoHCC CO mHmeCQEo moOCoOHm Hnms .HoHeoCooom oHeEoN oCe >HeuCoEeHo .mCsooo HCoEo>OE oHesmC No oOCeno mmoH CuHs mHoCOeou Hom .o m COHNHmOQOHm HON uHoomCm NO NCouxm oce oOHCOm COHNeEHHNCOU HON ooHHsoom eueo ICOHNHmOQOHm ooumeod HO HeHoCoo .eoscHucoonu.os enema 242 structure as I thought they would. Hence, the proposition that perceived setting influences behavior was confirmed, while the proposition that one can use observable aspects of organizational structure to predict intra—occupational variations in behavior across organizational settings was not confirmed. Since the summary presented in Table 49 requires some elaboration, I will recapitulate the major conclusions of each data chapter. Differences in Organizational Patterns (Chapters IV-VII) At the school level, two main clusters of organiza- tional variables were identified. The correlations among the variables supported somewhat the idea that hierarchy, universalism, expertise and all aspects Of specialization of task and worker covary; at the same time, however, the data also Show that formalization, standardization and centralization of authority covary. I had not expected centralization to correlate with any aspects of ideal type bureaucracy, but the data really only suggest that school organizations do not conform perfectly to the ideal type. The major contradictions to Weber's ideas of patterns of organization receive little support either: it would be difficult to label either cluster "bureaucratic" or "rational," as Udy (1959) suggested. Far less information was available for measurement of district organizational patterns. However, even the 243 existing data suggested somewhat different organizational patterns than we Observed at the school level. In parti- cular, formalization of decision—making was moderately and positively related to each of the other organizational variables, although the two measures Of centralization Showed only weak correlations with hierarchy and aspects of specialization, just as at the School level. It is possible that the different problems, audiences and responsibilities connected with schools and with school districts account for some of the differences in patterns. Since the proposi- tions could be tested for the effects of each dimension of organization, I did not attempt to classify either school or district organizations into only two or three overall categories. Moving from a consideration Of organizational patterns, the next step in this chapter is to summarize the evidence that occupational differences exist and permit ranking on the bases of reward Opportunity. Occupational Differences (Chapter V) In summing the data concerning the variety of possible occupational differences between elementary and secondary teachers, it is sufficient to note that the most important difference found was the proportion of female teachers--more elementary teachers are female, more secondary teachers are male. Differences in educational background, while not strong, correlated with school level, 244 so that fewer elementary than secondary teachers had com— pleted the Masters' degree. There were only minor differ- ences between the two groups when age and years of professional experience were considered. Among teachers, salary differences depend absolutely upon degrees completed and years of experience; neither group had a higher salary level when those two factors were controlled. Summing up the data on demographic differences at the principal level, the most Obvious difference here, too, concerned sex Of worker. All secondary principals are male, less than half the elementary principals sampled are male. Furthermore, secondary principals are somewhat younger, slightly better educated and report fewer years of teaching and administrative experience. Sex of worker interacted with the latter variables, Since females appeared to wait longer than males before making the move from teaching to administration. Salaries of principals are higher than those of teachers but Since almost all principals checked the highest category for salary, school level differences cannot be established. Turning to occupational differences in perceptions of favorable and rewarding work settings, we saw in Chapter V that, compared to elementary personnel, secondary per- sonnel feel more powerful, more autonomous and more con- cerned about poor administrative practices but report less cooperation from parents, superiors and colleagues, and report lower satisfaction with working conditions and 245 somewhat greater self-investment in work (that is, secondary personnel place greater importance on the judgment of colleagues than superiors and outsiders but they also express less enthusiasm about each of the reasons for remaining in education, including intrinsic aspects of work). When teachers and principals are compared, principals perceive greater autonomy and authority but slightly less personal influence than teachers, less awareness of union or administrative impediments to job performance and greater expectations of cooperation within the school but less cooperation with outsiders than teachers expected. On the whole, principals appear to see a more rewarding and more advantageous work environment than teachers. This difference in perception is reflected in differences in level Of self- investment in work, since principals report greater willing- ness to re-enter education, more emphasis on professional reference groups and stronger collegial exchange networks than teachers. The above differences in perceived setting and self— investment suggest corrolary differences in dedication to work. Data presented in Chapter V largely confirm the propositions that secondary personnel are more dedicated than elementary and that principals are more dedicated than teachers. The next step is to consider the effects of organi- zational variation for each occupational type, to see whether my structural eXplanation of self-investment and dedication can be confirmed intra—occupationally as well. 246 Relative Effects of Organizational Factors, Personal and Job I Satisfaction Factors on Intra- Occgpational Variations in Self— 'Investment and Dedication to Work In Chapter VI I ascertained the relative importance of organizational features, personal and family character- istics and level Of job satisfaction in predicting variations in teacher responses and behavior. I expected to find strong correlations between each characteristic of bureaucracy (including a strong negative correlation with centraliza- tion) and teacher responses. Furthermore, I expected that variations in perceived favorableness of the school environ- ment would also correlate with variations in responses of teachers. The latter expectation received some support but none Of the other independent variables significantly influenced teacher responses. The following conclusions were established in Chapter VI: 1. Personal and family variables do not significantly affect teachers' level Of self—investment in work and have only weak effects on non-union indicators of dedication, except for likelihood of taking non-required courses. When union activities are considered, age and experience positively corre- late with history of union participation, probably reflecting the differences in Opportunity Of Old and young workers to have such a history. My expectations that young and female workers would 247 be more militaristic, because of lack of advancement Opportunities through regular channels, was not supported. Teachers' level of satisfaction with three aspects Of work-~physical, non—physical and authority relations--had no Significant effects upon level of dedication to work but satisfaction with authority was mildly related to one indica- tor of self-investment. However, the most plausible explanation Of that phenomenon is that concern with intrinsic aspects of work leaves teachers inattentive to power problems, yielding a high score on satisfaction with authority. The indicators Of level of self—investment in work do not correlate internally enough to support the idea that each is a critical element Of a unidimen- sional variable, self-investment. Instead, three separate groups can be identified, although it is difficult to say if these dimensions are equally part of self-investment. Contrary to expectations, none Of the aspects of self-investment strongly correlated with any of indicators of level of dedication to work. When partial correlations are run to isolate the effects of personal and family variables, we see that collegial reference group have slightly more impact on dedication for teachers who are 243 younger, have smaller families and/or fewer young children. The original propositions suggested that teachers' perceptions of good and bad aspects Of work setting would be determined by school structure. Correla- tions between each aspect of structure and each perceptual variable are very weak and Offer no support for this idea. However, it is possible that the intra-level organizational variable is so narrow that teachers do not discriminate these fine differences. The prOposition remains "not proven" rather than conclusively refuted because Of that possibility. The prOposition linking favorable perception of work environment to level of self-investment in work is moderately supported for both occupa- tional groups. However, school structure does not appear to have any effects upon level of self-investment within occupational categories. Neither School structure nor teachers' percep- tions of the work climate meaningfully explain variations in teachers' level of dedication to work. At the secondary level only, controls on level of teachers' education raise the correlation between universalism in schools and the number Of teacher unions a teacher joins to .23, the strongest relationship uncovered. 249 It is necessary to recall at this point that the inra— occupational tests Of the propositions were extremely stringent ones, Since most of the variations along organiza- tional lines was wiped out in the separation of elementary and secondary schools. The data do not permit one to decide whether the propositions could be supported if the range of organizational variation were wider. The last step in this recapitulation Of the evidence is a consideration Of the effects of district structure upon principals and teachers. District Effects upon Teachers and Principals Data on the structure and effects Of district organi- zation mildly confirm the prOpositions Of this study. AS expected, the measures of centralization were moderately (correlated but were only weakly related to other organiza- tional dimensions. However, hierarchy, formalization and Specialization Showed only weak and inconsistent moderate intercorrelation, offering no adequate basis for choosing between the models of bureaucratization other writers have propounded (see Chapter II for summary). AS the data on schools led us to expect, district scores on SES and ability of students were Significantly correlated and Showed little correlation with homogeneity Of students. The Observed variations in district structure appeared to have only a few important effects on principal 250 perceptions and responses. While bureaucratization at the district level does not appear to discourage principals' self-investment and dedication to work, there is no signifi- cant evidence that it encourages these responses, although the data suggest moderate tendencies in that direction. Student composition variables seem to play a more important role in predicting principal responses than was anticipated, since SES and ability levels Of students correlate strongly with indicators of principals' mean levels of self- investment and dedication. Surprisingly, the district-level variables appear to be more critical in explaining Significant differences in teacher responses across districts than in explaining principal differences. Particularly, hierarchy at the district level affects teachers more than principals, contrary to my expectations. Although not all the evidence is statistically significant, data on district effects upon teachers Offer somewhat more support for the propositions relating bureaucratization to high self-investment and dedication than do the principal data. For teachers as well as principals, though, SES and ability of students encourage self—investment in work, evidenced by greater reliance on close colleagues as a reference group for teachers in high SES schools. Data concerning possible causes and effects of teacher contract emphasis Show that districts where teachers cooperate with outsiders and express concern with extrinsic ‘ 251 aspects of work produce contracts emphasizing extrinsic aspects of work. Contracts emphasizing professionalism occur in districts where teachers derive great satisfaction from non-physical aspects of work. Less purposive career orientations Of teachers predominate in districts higher in centralization of authority, as my propositions predicted, although actual turnover rates can not be attributed to any organizational variable. On the whole, the data on district structure and effects provide moderate support for the propositions that bureaucratization encourages self-investment and dedication providing a favorable environment for professional employees. Since several aspects of organizational structure could not be considered, due to limitations in the data making dif- ferentiation of districts impossible, it is encouraging to find that the few variables that could be measured yielded some confirmation of the propositions. General Conclusions In drawing some overall conclusions regarding the contributions of this dissertation research to scientific explanation of worker behavior it is apparent to me that this study illuminated some aspects of the phenomenon while, at the same time, uncovering new difficulties that could not be handled with the available data and resources. 252 I will attempt now to list the major contributions and areas for future refinement: 1. Distinctive elementary-secondary school patterns of organization have been revealed; at the district level, organizational patterns appeared to be less coherent, perhaps reflecting the need for a larger sample size. The predicted relationship between favorable perceptions Of work environment and self-investment in work received substantiation. However, the indicators of self-investment either lack equal validity or suggest that self-investment is a multi-dimensional concept. Further research and consideration of this problem is indicated. Predicted occupational differences in both self- investment and dedication to work have been con- firmed; furthermore, the inter-occupational patterns support the expected positive relation- ship between these two variables. When occupation is controlled, no meaningful effects Of school structure upon teachers' self- investment and dedication were revealed. However, neither did bureaucratization of structure appear to depress levels of self-investment or dedica- tion of teachers. Thus one assumption running through social science, namely, that formalized work settings alienate workers, appears to be in 253 error. However, it would be premature to con- clude that organizational characteristics are irrelevant to worker behavior until a replication of this study, using a broader range of organiza— tional structures but still intra-occupationally focused. The difficulty with this approach will be the separation Of organizational from occupa- tional influences, requiring selection of an occupation whose members are distributed across a broad spectrum of organizational forms-~perhaps accounting or engineering specialists. None of the prOposed interrelations between career orientation and either self-investment or dedica- tion to work received confirmation from the data. When actual turnover rates were considered, principals and secondary personnel showed lower turnover, even when sex of worker was controlled. However, since it was impossible to distinguish quitting from involuntary termination or, among those who quit, establish bases for quitting, the prOpositions could not be tested completely. Further research on this question would have to establish more comprehensive information on career plans and desires than the present data yielded, as well as separating voluntary from involuntary turnover. 254 Two major weaknesses in the data on dedication to work were revealed in the analysis. First, the questions asked principals were not nearly as com— prehensive as those asked teachers. Not only was it difficult to compare principals and teachers, it was difficult to ignore the fact that possible areas Of principals dedication were insufficiently covered. Since the original survey was not designed with an eye to the propositions I tried to test, it naturally did not cover in detail all aSpectS of the problems Of interest to me. For this reason, my conclusions regarding relative strength of principals' and teachers' dedication to work must be considered tentative. The second weakness in the measures of dedica- tion stems from the phrasing of questions on the teacher survey. Although teachers are asked about current professional activities (reading, course work, etc.) they are asked about the history Of their union involvement, not just current parti- cipation. This difference in phrasing has two important effects: an inflation in scores on union activity relative to professional involve- ment and a bias in favor Of Older, more experi- enced teachers, who have had more years in which they could participate in union activity and more years of interaction with other teachers, 255 probably essential to achieving union Office. To be really comparable with other measures of dedica- tion, measures of union involvement Should refer only to present involvements. Despite the problems and inconsistencies in data supporting my prOpositions, this study strongly suggests that two alternative explanations of worker behavior are inadequate. Neither level of job satisfaction nor personal and family characteris- tics predict self-investment or dedication levels as well as occupational factors do. Furthermore, neither alternative explanation fared better than my propositions regarding the effects of intra- occupational organizational structure upon workers. It is at least as important to clear away mistaken explanations of phenomena as it is to establish a correct explanation, so I believe that this dissertation makes a valuable contribution from that perspective, apart from the status of my alternative explanation of worker behavior. Socio-economic status Of the community is critical in at least two ways: in its effects upon principals' level of self-investment in work and in its effects upon the pattern Of organization of secondary schools. The data showed that the original prOpositions Should be refined to include the idea that SES Of students somehow affects the 256 reward potential for principals~~perhaps community SES is a major determinant of the prestige asso- ciated with principalship of given schools. How- ever, SES did not play any such role for teachers. With respect to the second point about the role of SES, the data Showed that secondary school organi- zational patterns were considerably strengthened when SES was controlled. Perhaps different organi- zational practices are required to process teenage students of different SES or, perhaps different kinds of specialists, and procedures are demanded by high and low SES communities. The implications of SES for organizational structure require further investigation for adequate explanation. Detailed study of the effects of district variables upon teachers might be fruitful, Since my data suffered from a paucity of cases (only 5 districts) and an overabundance of variables, both structural and dependent. Even this problematic data, though, suggested that district variables may be more important determinants of teacher behavior than I had expected. Particularly, district Specializa- tion may affect potential career moves for teachers, while district hierarchy, formalization and cen- tralization may place limits on school organiza— tional patterns. The latter possibility Opens up a whole new area of potential importance in the 10. 257 explanation of school structure and is certainly worthy Of further attention. If the district variables actually are critical in determining teacher and principal behavior, one would expect to find considerably lower levels Of self- investment and dedication in smaller districts, although comparison across any broad range of districts probably would introduce variations on many other dimensions as well. The complex data and inconsistent results of this dissertation underscore the need for further research on the range of variables potentially affecting variation in self-investment in work. It is apparent that organizational and occupational factors alone cannot account for all the variation in teacher and principal responses. At the same time, the data demonstrate the failure of variables like sex and marital status as adequate predictors of level Of self-investment in work. It is possi- ble that some previously neglected factor, like family economic background, occupation of Spouse, parents or friends or individual differences in the importance of self-esteem (relative to other personal and interpersonal needs, such as needs for love, companionship, power, security, etc.) may play some role in determining variations in self—investment in work. Further research Should 258 focus upon the independent and joint effects of these variables, as well as the structural variables this dissertation considered. AS an example of how these variables might interact to influence self—investment in work, consider the many male teachers from working class backgrounds. It seems very possible that teachers (or principals) from such families may feel great self-esteem just from achieving entrance into a "profession"--that is, they may make only inter- occupational comparisons of achievement and decide they have proved their worth relative to others in their family (Faunce, 1972). In such a case there may be no motivational basis for continued striving for occupational success. Furthermore, one could argue that peOple who go into teaching may be more concerned about security than self-esteem, have low levels of need-achievement, like their position of authority relative to students, or in some other way do not require continued testing of own abili- ties and performance Of the sort that high self- investment in work involves. Quite clearly, Faunce's idea that need for self esteem is a variable, not an absolute, must be taken into account in more detailed future research. One would expect organizational and occupational structure to affect self—investment only for 11. 259 workers who have some minimum interest in self- esteem maintenance through intra—occupational comparison of performance. My assumption that teachers and principals would not completely lack interest in self-esteem could not be tested with the data at hand; furthermore, there may be critical differences in the degree of concern for self-esteem of various individuals in the sample, although one would expect such differences to be random across schools and hence not confound the effects of structural variables. Any further testing Of the bases of self-investment in work must consider these problems. This dissertation indicates the need for investi- gation in at least one more area--the presumed effects Of teacher dedication to work. How do school structure and teacher behavior affect student learning? It is possible, for instance, that some indicators of teacher self-investment in work may actually inhibit learning for some kinds of stu- dents. More generally, one might ask if "professionalism" always works for the good of the client. This line Of inquiry deserves further attention. In a similar vein, one might ask whether increasing self—investment and dedication of teachers and principals has any implications for the school district organizations. For 260 instance, what happens if teachers begin to emphasize purely professional concerns in collective bargain- ing? In fact, what happens to teachers as a special interest group when dedication to work leads teachers to identify only others in their specialty as colleagues? In this dissertation and in most research on worker behavior, investigators assume that dedication to work encourages goal-attainment for the organization, i.e., that dedicated workers are better workers. This basic assumption should not be left unchallenged. Summary In summarizing this concluding chapter, it is apparent to me that this research, while not consistently supportive Of my original propositions, at least suggests some Of the advantages to be gained from this new way Of looking at work behavior. The perspective of Faunce, Dubin and Vroom, modified in this dissertation to include structural determinants Of importance of work, is a reasonable synthesis of ideas borrowed from research on formal organizations, in occupations and in the field of symbolic interactionism. (This dissertation has not only Offered some confirmation of the idea that self-investment in work is critical to dedication and is dependent upon work structure, but it has also produced evidence seriously damaging the conventional explanations of variations in 261 worker behavior--that over-bureaucratization, low job satisfaction or certain personal and family characteristics lead workers to disengage themselves from work. Whether or not the structural and self-investment explanations of worker behavior are adequate on their own must be determined by further research. Despite the tentative state Of the conclusions in this dissertation and despite the clear need for continued investigation Of many Of these problems, I believe that this research has presented a useful and workable integration Of organizational theory and logically grounded social- psychological explanations of behavior, producing a synthesis that considers human behavior in structural context. Further- more, this dissertation goes far beyond typical research on the topic Of worker behavior with its over-reliance upon measures of group composition and supervisory style as measures of work context. In demonstrating that occupational and organizational factors can affect perceptions and behavior of workers, this dissertation helps to retrieve the study of worker behavior from the realm of predominantly psychological, individual problems and to Show that the common problems and restraints faced by individuals in similar structural settings may be an important factor in behavior. BI BLI OGRAP HY BI BLI OGRAP HY Adams, J. Stacy. "Wage Inequities, Productivity and Work Quality, Industrial Relations. 3(1963). . "Inequity in Social Exchange." In L. Berko- witz (ed.), Advances in Experimental Social Psychology. Vol. 2, N. Y. Academic Press, 1965. Anderson, James. "The Authority Structure of the School." Educational Administration Quarterly. III (1967), 130-148. Argyris, Chris. Personality and Organization. N.Y.: Harper, 1957. Armknecht, P. A. and J. F. Early. "Quits in Manufacturing." Monthly Labor Review. 95(1972). Backman and Secord. "The Self and Role Selection." In Gordon and Gergen (eds.), The Self in Social Interaction. N.Y.: Wiley, 1968, p. 295. Becker, H. "The Career of School Teachers." American Journal of Sociology. 58(1952), 470-7. ' . "The Teacher in the Authority System of the Public School." Journal Educational Sociology. 27(1953). Behrend, H. "Absence and Labour Turnover in a Changing Economic Climate." Occup. Psyghol. 27(1953), 69-79. Bell, Gerald D. "Determinants of Span of Control." American Journal of Sociology. 73(1967), 100-109. Betz, David. "Organizational Correlates Of Occupational Orientation and Support of Negotiating Organiza- tion." Unpublished Ph.D. dissertation, Michigan State University, 1968. Bidwell, C. E. "The School as a Formal Organization." In J. G. March (ed.), Handbook Of Organizations. Chicago: Rand McNally, 1965. 262 263 Blau, Peter. Theggynamics of Bureaucrac . Chicago: University Of Chicago Press, 1955. Blau, P. and W. Scott. Formal Organizations. San Francisco: Blauner, Robert. Alienation and Freedom. Chicago: U. of Chicago Press, 1964. . "Work Satisfaction and Industrial Trends in Modern Society." In W. Galenson and S. M. Lipset (eds.), Labor and Trade Unionism. N.Y.: Wiley, 1960. Blumer, Herbert. "Attitudes and the Social Act." Social Problems. 3(1955). Box, S. and S. Cotgrove. "Scientific Identity, Occupational Selection and Role Strain." Brit. Journal of Sociology. 17(1966). Brophy, A. L. "Self, Role and Satisfaction." Genet. Psych. Monograph. 59(1959). Buchanan, Bruce. "Building Organizational Commitment: The Socialization of Managers in Work Organizations." Administrative Science Quarterly. 19(1974). Caplow, Theodore. The Sociology Of Work. U. of Minnesota Press, 1954. Carlson, Richard 0. Executive Succession and Organiza- tional Change. Chicago: U. of Chicago Press, 1962. . "Research on the School System as an Organiza- tion." School Review. 66(1958), 473-475. Child, John. "Organizational Structure and Strategies of Control." Administrative Science Quarterly. 19(1974). Chinoy, Eli. The Automobile Worker and the American Dream. Garden City: Doubleday, 1955. Coleman, James S. Equality Of Educational Opportunity. Washington, D.C.: Government Printing Office, 1966. 264 COOpersmith, R. The Antecedents of SeleEsteem. San Francisco: Freeman, 1967, pp. 259-260. Corwin, Ronald. A Sociology of Education. N.Y.: Appleton, Century, Crofts, 1965, pp. 241-265. .__° "The Professional Employee: A Study of Con- flict in Nursing Roles." Americal Journal of Sociology. 66(1961). . "Patterns of Organizational Conflict." Administrative Science Quarterly. 14(1969). . "Education and the Sociology of Complex Organi- zations." In D. Hansen and J. Gerstl (eds.), On Education: A Sociological Perspective. N.Y.: Wiley, 1967. DeFleur, M. and F. Westie. "Attitudes as a Scientific Concept." Social Forces. 42(1963). Dubin, R. ”Industrial Workers' Worlds." Social Problems. 3(1956). Etzioni, A. The Semi-Professions and Their Organization. N.Y.: The Free Press, 1969, pp, 4.9.183. Faunce, W. "Job Satisfaction and the Meaning of Work." Paper for U.S. Dept. of Labor Conference on Job Satisfaction. Washington, D.C., December, 1968. . "Self Investment in the Occupational Role." Paper presented at meeting of Southern Sociological Society. New Orleans, April 6, 1972. Friedman, A. and P. Goodman. "Wage Inequity, Self- Qualifications and Productivity." Org. Behavior and Human Performance. 2(1967). Geer, Blanche. "Occupational Commitment and the Teaching Professions." School Review. 74(1966). Gerstl, Joel. "Education and the Sociology of Work." In D. Hansen and J. Gerstl (eds.), On Education: A Sociological Perspective. N.Y.: Wiley, 1962. Gerstl, J. W. and E. J. Guba. "Social Behavior and the Administrative Process." School Review. 65(1957). 265 Given, Charles. "Integration and Conflict in Educational Organizations." Unpublished Ph.D. dissertation. Michigan State University, 1969. Glaser, Barney. Organizational Scientists: Their Profes- sional Careers. Indianapolis: The Bobbs-Merrill Company, Inc., 1964. Goldner, F. and R. Ritti. "Professionalization as Career Immobility." American Journal of Sociology. 73(1967). Goldthorpe, J. "Attitudes and Behavior of Car Assembly Workers." Brit. Journal of Sociology. 17(1966). Goode, W. "A Theory of Role Strain." American Sociological Review. 25(1960). Goslin, David. The School in Contemporary Society. Scott Foresman and Co., 1965. Gouldner, Alvin. "Cosmopolitans and Locals." Administrative Science_guarter1y. 2(1957). Grass, Neal and Anne Trask. Men and Women as Elementary School Principals. Cambridge, Mass: Harvard, 1964. Griffiths, David. Human Relations in School Management. N.Y.: Appleton, Century, Crofts, 1962. Gross, Llewellyn. "A Theory of Power and Organizational Processes." School Review. 70(1962). Grusky, Oscar. "Career Mobility and Organizational Commit- ment." Administrative Scienceguarterly. 10(1966). Guest, R. H. "Job Enlargement: A Revolution in Job Design." Personnel Admin. 20(1957). Hage, Jerald and Michael Aiken. "Relationship of Centrali- zation to Other Structural PrOperties." Adminis- trative Scienceguarterly. 12(1967). Hall, Richard. "The Concept of Brueaucracy." American Journal of Sociology. 69(1963). . "Professionalization and Bureaucracy." Adminis- trative Science Review. 33(1968). 266 Hansen, Donald and Joel Gerstl (eds.). On Education-- "Sociological Perspectives. N.Y.:firwiley, 1967. Herzberg, Frederick; B. Mausner; and B. Snyderman. The Motivation to Work. N.Y.: John Wiley and Sons, 1959! pp. 80-82. Heydebrand, S. W. and R. Stauffer. "The Structure of Small Bureaucracies." Administrative Scienceguarterly. 31(1966). . "The Hierarchy of Authority in Organizations." American Journal of Sociology. 73(1968), 453-67. Hickson, David et a1. "Operations Technology and Organi- zation Structure." Administrative Science Quarterly. 14(1969). Homans, George. Social Behavior. N.Y.: Harcourt Brace and World. 1961. Hrebiniak, L. and J. Alutto. "Personal and Role-Related Factors in the Deve10pment of Organizational Commitment." Administrative Scienceguarterly. 17(1972). Hughes, Everett. "Work and the Self." In J. Rodner and M. Sherif (eds.), Social Psychologyiat the Cross- roads. N.Y.: Harper and Row, 1951. Hyman, H. "The Psychology of Status." Archives of Psychology. #269. 1942. Ingham, R. G. Size of Industrial Organization and Worker Behavior. Cambridge. England: Cambridge Univer- sity Press, 1970. Israel, Joachim. Alienation from Marx to Modern Sociology. Boston: Allyn and Bacon, 1971. Jencks, Christopher, et a1. Inequality: A Reassessment of the Effect of Family and Schooling in America. N.Y.: Basic Books, 1972. Kassalow, E. "White-Collar Unionization in the U.S." In White-Collar Trade Unions. A. Strumthal (ed.). Urbana, 111.: U. of Illinois Press, 1967, p. 337. 267 Katz, Daniel and R. Kahn. The Social Psychology‘of Organi— zations. N.Y.: Wiley, 1965. Katz, F. E. Autonomy and Organization. N.Y.: Random House, 1968. . "Explaining Informal Work Groups in Complex Organizations." Administrative Science Quarterly. 9(1965), 204-223. Kelly, S. "Two Functions of Reference Groups." In Readings in Social Psychology. Edited by Swanson, et al., 1952. Kohn, Melvin. "Bureaucratic Man: A Portrait and an Inter- pretation." American Sociologiéal Review. 36(1971). Kornhauser, Arthur. Mental Health of the Industrial Worker. N.Y.: Wiley, 1965. . Scientists in Industry. Berkeley: U. of Calif. Press, 1963. Ladinsky, J. "Occupational Determinants of Geographic Mobility." American Sociological Review. 32(1967), 48-53. LaPorte, T. "Conditions of Strain and Accomodation in Industrial Research Organizations." Administrative Science Quarterly, 10(1965), 32. Lawler, E. E. "Equity Theory as a Predictor of Productivity and Work Quality." Psych. Bulletin. 70(1968). Lawler, E. E., et a1. "Inequity Reduction Over Time in an Individual Overpayment Situation." Org. Beh. and Human Performance. 3(1968). Lawler, E. E. "Job Design and Employee Motivation." Personal Psych. 32(1969). Lieberman, Myron. Education as a Profession. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1956. Likert, Rensis. New Patterns of Management. N.Y.: McGraw-Hill, 1961. Litwak, E. "Models of Bureaucracy Which Permit Conflict." American Journal of Sociology. 67(1961). 268 Litwin, G. and R. Stringer. Motivation and Organizational Climate. Boston: Harvard University, 1968. Lodahl, T. and M. Kejner. "The Definition and Measurement of Job Involvement." Journal of Applied Psychology, 49(1965). '6 Lortie, Dan. "Control and Autonomy in Elementary Teaching." In A. Etzioni (ed.), The Semi-Professions and Their Organization. N.Y.: The Free Press, 1969. . "Shared Ordeal and Induction to Work." In H. S. Becker, et al. (eds.), Institutions and the Person. Chicago: Aldine, 1968. Luck, P. and J. Heiss. "Social Determinants of Self-Esteem in Adult Males." Soc. and Soc. Research. 56(1972). Mangione. Thomas. Turnover: A Model and a Review of the Literature. Ann Arbor: Survey Research Center, 1972. Manis, M. "Social Interaction and the Self-Concept." Journal of Abnormal and Social Psychology. 51-1 11951). Mansfield, Roger. "Bureaucracy and Centralization." Administrative Science Quarterly. 18(1973). March, James and Herbert Simon. Organizations. N.Y.: Wiley, 1963, pp. 94—101. Marcson, Simon. The Scientist in American Industry. Princeton: Industrial Relations Section, 1960. Marvick, D. Career Perspectives in a Bureaucratic Setting. Ann Arbor, Michigan: U. of Michigan Press, 1954. Maslow, A. Motivation and Personality. N.Y.: Harpers, 1954. Maurer, John. "Work Role Involvement." Unpublished Ph.D. thesis, Michigan State University, 1967. Mausner, B.; R. Peterson; and D. Copwell. Job Attitudes: Review of Research and Opinion. Pittsburgh: Psychological Service of Pittsburgh, 1957. Michigan Department of Education. Technical Report: Michigan School Census of 1970. 1970. Moeller, Gerald and W. W. Charters. "Relation of Bureau- cratization to Sense to Power Among Teachers." Administrative Science anrterly. 10(1966). 269 Montagna, P. "Professionalization and Bureaucratization in Large Professional Organizations." American Journal of Sociology. 74(1968). Morse, Nancy and Everett Riemer. "The Experimental Change of a Major Organizational Variable. Journal Abn. and Soc. Psych. 52(1956). Neter, John and William Wasserman. Applied Linear Statisti- cal Models. Homewood, Ill.: Irwin, 1974. Orzack, Louis. "Work as a Central Life Interest of Profes- sionals." Social Problems. 7(1969). Parnes, Herbert, et al. The Pre-Retirement Years. Washing- ton: U.S. Government Printing, 1970. _. Career Thresholds. Washington, D.C.: U.S. Government Printing Office, 1971. Patchen, Martin. "Alternative Questionnaire Approaches to the Measurement of Influence in Organizations." American Journal of Sociology. 69(1963). . Participationy Achievement and Involvement on the Job. Englewood Cliffs, N.J.: Prentice-Hall, 1970. Pellegrin, R. and C. Coates. "Executives and Supervisors: Contrasting Definitions and Career Success." Admin- istrative Science Quarterly. 1(1957), 506-17. Pelz, D. and F. Andrews. Scientists in Organizations. N.Y.: Wiley, 1966. Perrow, Charles. Organizational Analysis. Belmont, Calif.: Wadsworth, 1970, pp. 45-7, 77-85, 153. Perrucci, Robert and Joel Gerstl. The Engineers and the Social System. N.Y.: John Wiley, 1969, pp. 283, 358. Porter, L. "Job Attitudes in Management." Journal of Applied Psychology. 46(1962). Porter, L. and Lawler. "The Effects of Tall and Flat Organizational Structure on Managers' Job Satis- faction." Personnel Psych. 17(1964), 35-48. 270 Pugh, D. S., et a1. "Dimensions of Organizational Structure. Administrative Science_guarter1y. 13(1968). Reiman, Bernard. "On the Dimensions of Bureaucratic Struc- ture." Administrative Scienceyguarterly. 18(1973). Reiss, Albert. Occupations anngocial Status. N.Y.: The Free Press of Glencoe, 1961. Robinson, John; Robert Athanasian; and Kendra Head. Measures of Occupational Attitudes and Occupational Characteristics. Ann Arbor: Survey Research Center, 1969. Rock, Donald and John Hemphill. Report of the Junior High School Principalship. Vol. 2. National Association of Secondary School Principals, 1966. Rosenberg, Morris. Occupations and Values. Glencoe, Ill.: The Free Press, 1957. . "Psychological Selectivity in Self-Esteem Formation." In C. Gordon and Gergen (eds.), The Self in Social Interaction. New York: Wiley, 1968, pp. 1-32, 55-9. Saxe, Richard W. Perspectives on the Changinngole of the Principal. Springfield, Ill.: Charles C. Thomas Publishing, 1968. Schacter, Stanley, et a1. "Emotional Disruption and Indus- trial Productivity." Journal of Applied Psychology. 45(1961), 201-213. Scott, William R. "Professionals and Complex Organizations." In H. Vollmer and D. Mills (eds.), Professionaliza- tion. Englewood Cliffs: Prentice-Hall, 1966. Scott, William and Terence Mitchell. Organization Theory. Homewood, Ill.: Richard D. Irwin, Inc., 1972, p. 107. Seeman, Melvin. "On the Personal Consequences of Aliena- tion in Work." American Sociological Review. 32(1967). Sheldon, Mary. "Investments and Involvements as Mechanisms Producing Commitment to Organizations." Adminis- trative Science Quarterly. 16(1971), 142-7. 271 Shibutani, T. "Reference Groups and Social Control.” In Rose (ed.), Human Behavior and Social Proce3ses. Boston: Houghton-Mifflin, pp. 128-138. Smith, C. "Effects of Opportunity Structure and Sex of Worker: Job Leaving Patterns in the Michigan Civil Service." Paper presented at North Central Associa- tion Meeting, 1975. Sofer, Cyril. Men in Mid—Career. Cambridge U. Press, 1970. Spector, A. J. "Expectations, Fulfillment and Morale." Journal Abnorm. Soc. Psych. 50(1956. Stinchcombe, Arthur L. Creating Efficient Industrial Administration. N.Y.: Academic Press, 1974. Stinchcombe, Arthur. "Bureaucratic and Craft Administration of Production." Administrative Science Quarterly. 4(1959). Tannenbaum, Arnold S., et a1. Hierarchy in Organizations. San Francisco: Jossey, Bass Publishers, 1974. Telly, C. W. French and W. Scott. "The Relation of Inequity to Turnover Among Hourly Employees." Administrative Science Quarterly, 16(1971). Thibaut, John and Harold Kelley. The Social Psychology of Groups. N.Y.: Wiley, 1959. Thompson, James. Organizations in Action. N.Y.: McGraw- Hill, 1967, p. 107. Thompson, James and Richard Carlson, and Richard Avery. Occupations, Personnel and Careers. Pittsburgh: Administrative Science Center, U. of Pittsburgh, 1962. Udy, Stanley. "'Bureaucracy' and 'Rationality' in Weber's Organization Theory: An Empirical Study." American Sociological Review. 24(1959). United States Department of Labor. Job Satisfaction: Is There a Trend? Washington, D.C.: U.S. Government Printing Office, 1974. Vroom, Victor. "Industrial Social Psychology." In G. Lindzey and E. Aronson (eds.), The Handbook of Social Psychology. Vol. 5. Reading, Mass.: Addison-Wesley, 1969, pp. 199-204. 272 . Work and Motivation. N.Y.: Wiley, 1964. . "Ego-Involvement, Job Satisfaction and Job Performance." Personnel Psychology. 15(1962), Vroom, Victor and K. MacCrimmon. "Toward a Stochastic Model of Managerial Careers." Administrative Science Quarterly. 13—l(l968). Waisanen, Frederick. "Self—Evaluation and Performance Estimates." Soc. Quarterly. (1962). Walker, Charles and Robert Guest. "The Man on the Assembly Line." Harvard Business Review. 30(1962). Weber, Max. The Theory of Social and Economic Organization. A. M. Henderson and T. Parsons (trans.) and T. Parsons (ed.). Glencoe, Ill.: The Free Press, 1947. Winer, B. J. Statistical Principles in Experimental Design. N.Y.: Mc~Graw-Hill, 1962. Woodward, Joanne. Industrial Organization. London: Oxford University Press, 1965. Yuchtman, E. "Reward Distribution and Work-Role Attractive- ness in the Kibbutz." American Sociolpgical Review. 37(1962), 81-95. Zaleznick, A.; G. Dalton; and L. Barnes. Orientation and Qonflict in Career. Boston: Harvard U., 1970, pp. 54-60, 132-7. APPENDIX A TEACHER QUESTIONNAIRE EDUCATION ADMINISTRATION STUDY Department of Sociology Michigan State University TEACHER QUESTIONNAIRE In what grades are you teaching? (Include any pre— first grade under "K." Please circle all the grades in which you are teaching this year.) K 1 2 3 4 5 6 7 8 9 10 11 12 13 What is your employment status in the school system? CHECK ONE. a. I have a regular full—time appointment with tenure. b. I have a regular full-time appointment but not on tenure. c. I am a substitute teacher part-time. Counting the present year, what is the total number of ears of full—time teaching experience you have had? (Consider counseling as teaching experience.) (Write in number) I have had years of full-time teaching experience Counting the present year, what is the total number of years of full-time teaching experience you have had ig’ this school? (Consider counseling as teaching experi- ence.) (Write in number) I have had years of full-time teaching experience in this school. Are you a member of an affiliate of MBA or MFT? CHECK ONE NO .Go to Question 6) Yes w 273 lo. 11. 12. 13. 274 a. If yes, which one? CHECK ONE. MEA MFT b. How many years have you been a member? (Write in number) years c. Have you held an office in the local organization? CHECK ONE. Yes No d. Have you held an office in the state organization? CHECK ONE. Yes No e. Have you presented or contributed to one or more state program(s)? CHECK ONE. Yes No How many other teacher organizations do you belong to? (Write in number) teacher organizations Wm--Hir-' w‘ ‘ 4 Have you ever served on a committee, commission, council or held office in the MBA or MFT? a. At the state level? CHECK ONE. Yes No b. At the local level? CHECK ONE. Yes No Are you now, or have you ever been, a building representative? CHECK ONE. Yes No How many teacher conferences, conventions, and workshops have you attended since September, 1966? (Write in number) meetings How many books have you read in the last 6 months that are related to your teaching or teaching subject area? (Do 223 include reading required by college courses you may be taking. Write in the number of books.) books - Do you subscribe to any teaching or subject-matter journals other than MBA or MFT membership publications? CHECK ONE. Yes No How many teaching journals do you read regularly other than the MBA or MFT membership publication? (Write in number) journals About how much time do you spend reading teacher journals in a typical month? (Do 225 include reading by college courses you may be taking.) (Write in number.) hours per month 27S 14. People remain in an occupation for many different rea- sons. How important are each of the following for remaining in teaching? CHECK ONE ON EACH LINE. Very Great Some Slightly Not great impor- impor- impor- impor- impor- tance tance tant tant ’ tance at all a. The students with whom I have contact in this school b. The friend- ships I develop with the people at work c. The opportunity to be creative and original in the work I do d. My salary com- pared to others at my level of education e. My present job in the light of my career expec- tations f. The physical conditions under which I work 9. The number of duties I have besides actual teaching h. My teaching load i. Educating the future genera- tion j. The opportunity to be a profes- sional 276 Very Great Some Slightly Not great impor- impor- impor- impor- impor- tance tance tant tant tance at all k. My vacations and free time 1. The prestige and respect I receive ) from the community p,. m. Working with books and ideas N 15. To what extent can you expect cooperation and support for your ideas about doing your job from each of the follow- ing? CHECK ONE ON EACH LINE. To a To a To To a To no very great some slight extent great extent extent extent at all extent a. Member(s) of the school board b. Other teachers in your specialty in your school c. Your department head d. Superintendent of schools e. Officers of the local chapter of your teacher organization (not the building representatives) f. Principal of your school g. Officer(s) of the PTA in your school 277 To a To a To To a To no very great some slight extent great extent extent extent at all extent h. Parents of the children in your classroom i. The superin- tendent's staff j. Other teachers not in your specialty in your school k. The principal's assistants Locum“ m and unawam .cssaou ummH ocu :H .COmeuop umnu Lm>o wocwaaucw umoe ecu C>Mc vasocm on: usonm :OMCH .qua :uo oucuaawcw umoe unu mm: Madmauum 0:3 lull!) IE . ao>L=o mcwvwnmv .uuo «wwawuw .vommma xaamsno: an on mucousuw mo ommucoOLUA 0:» anaemEAOuwn .m cemummom >nomm>uvm=m hozuo ho cum: acuEunuaov ou Lonumou m meduOEOLm .h howaom anaemuuzpumcm msquASEhom .o mucwvsum ou xmonw on mmzopm woqmuno ou co«mmwsnoa mam>qo .m uponumUu :wc mcwhwx .3 mwmnzou 00 vanqsaun on» mcw=MELouoo .n HI quums mCMUNVn Ga AhmucuEuaaazm mewuooaom .u mxoonuxuu cwpwzvop wcqpoonw .H Sc C: E A: C: 23 r: E :3 35 2: 13 20388 .8 FE ucm -vcwuc« and Iguana quCwLa . we wnu acme ML: cow» Amvucmu Ah own upth no mu umummm ucocuuou -wwovam mo .comumu muuszquH vLmon .uaoo ucwuc« -mnumq 4mm w.ama mo you n mo xuazumw nflcmmho Hmo: mucwumu Hoonom mutum -hwdsm lawecm -Hocwta -wocwtd uuMEEoo >uHJUMM ohmucw thomwu hwnumou mxh uxdu denm 0:: mozuaqmzH Fmo: mxh m<= >qqwu -mfimmm. mnwcomvu -mauwam mo .co..3m~ Egon $de 1.13.: umnumm and mLma mo Emu m. we banana Iwcwmno 3:866“ 406.com Snow 1596 155.3. $05.33 2.35.3 LUZ—zoo >ua3umw upwpcw Lunomwu Locomwu 8sz4.15 .50: m5. mn emu Irma no (u: :53 on“ .3: lmwmmm 33:3 -cfioaam Imp ow floccum fm.» 5 37:: .3 to Dawn g: .33» 862063 323% mo .coUmu «muzmpqmzH venom Lawn .935 hp»? and nLém we mu» m up 51.6mm .1595 .50: mucwhma Hoocum mume £596 .556 Lucia Lucia uuMEEou 316mm mecca Cotonou szumwu uxu Han :. m 0:: ”.6554qu .50: ME. mqg<2b< 0:: .6363 :5 L66 mucosa—J umoe wnu w>mn 3:05.... 0;: Hana 5:76 .30» mcqumuwucw 3:3 .353 m and ammo: £5300 um: 92 5 dz: was? «5:033 05 Lo>o 35:35 umos ecu mm; wadmsuom on: :65 20 uzo 5me $8163. to )n x 302 vnwamu on cu am 39500 awn: wcmcwEouoo .1.— um: vasecm Locummfi m mvonuoe mcwcgfiouuo .2. 3.300 hmasuHVLma m cm 29.6» on on as 3:1...» can muauocou awn: magi—L88 .Na 3755me no 9.13. 2: 95:3 «>2 3:05 .595... 5163 m .3505: wcmcfifivuon .oa nzmnug wn gnozm pwzumwu Humwuw>onucou m w... wcqzufiouoo .m A: C: A: A: C3 A3 CV Amy :3 Rev A3 A3 22303 mo amt. ucw .9555 Han upwaam Joana mo .23 paws .E: :0: Amvuzmu .3 v3 Lawn ho ua=omw lqcmmpo Scana «ongun 3me £33 -555 $05.5 Lucia Lac—coo .3163 8:5 Lwcumwu Lwcumwu uuzuquzH .50: ME. m