STRUCTURAL AND ATTITUDINAL EFFECTS ON ORGANIZATIONAL PRIORITIES . Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY HARLOWE G. HATLE 1975 ”MW“? , I/IIIIIIIIIIII/IIIIIIIIIIII III/IIIIIIII/II A ,m, v 3 1293 10463 2264 Michigan (:31: Universif}? . . I .. .. , M") "QR This is to certify that the thesis entitled STRUCTURAL AND ATTITUDINAL EFFECTS ON ORGANIZATIONAL PRIORITIES presented by Harlowe G. Hatle has been accepted towards fulfillment of the requirements for Ph . D. My” in Sociology (7/ Major professor } V mi: 0—7639 ./ ” 7 ABSTRACT STRUCTURAL AND ATTITUDINAL EFFECTS ON ORGANIZATIONAL PRIORITIES by Harlowe G. Hatle The problem under investigation in this study was the effect of (a) organizational variables and (b) perceived .need for increased supervision on salary priorities given selected activities in an institution of higher education. Specifically, the investigation attempted to determine (a) if the organizational variables, affluence, size, faculty salaries, complexity and faculty qualifications influenced perceived need for closer hierarchical control and/or closer collegial control of bureaucratic and/or professional activi- ties; and (b) what influence the above organizational and attitudinal variables had on the activities of academic advising of students, job counseling of students, personal values and ethics, popularity with students, publications, research, service activities in the community and service activities in the university. The data forming the organizational variables were gleaned from official university records, while the data on supervision and salary priorities were obtained from a mailed questionnaire given on a medium sized north central state Harlowe G. Hatle university. The 151 respondents were divided into five col- leges/schools based on their departmental affiliation. The five colleges/schools were then ranked on each of the orga- nizational variables. The study utilized percentage analy- sis, Yule's Q and multiple regression. The data indicate that faculty do not see an overwhelm- ing need for closer supervision from any source, but if there is to be closer supervision, strong preference is given to collegial versus hierarchical supervision and to closer supervision of bureaucratic versus professional activities. Increased size and complexity raised the probability that greater need for closer hierarchical versus closer col- legial supervision would be perceived, while increased fac- ulty qualifications and salary raised the probability that greater need for closer collegial versus closer hierarchical supervision would be perceived. Increased affluence reduced the probability that greater need for closer supervision from any source would be perceived. Regarding salary priorities, neither size nor complex- ity were as strong as influence as anticipated, but both increased size and complexity indicated a slight raise in the probability that greater priority would be given to the discipline oriented activities versus local oriented activ- ities. Increased faculty salaries affected salary priori- ties in a manner similar to size and complexity. Harlowe G. Hatle Faculty qualifications were positively correlated to increased priority for local oriented activities while afflu- ence had a positive correlation with both local oriented and discipline oriented activities. General statements are offered in regards to (1) source of supervision, (2) type of activity supervised, (3) the effect of organizational resources and organizational struc- ture on supervision, and (4) goals and reward priorities. Some suggestions for future research are offered. STRUCTURAL AND ATTITUDINAL EFFECTS ON ORGANIZATIONAL PRIORITIES BY mOU Harlowe Gf Hatle A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1975 ACKNOWLEDGEMENTS This dissertation owes its completion to the help of several individuals and organizations. On a technical level, particular appreciation is given to Philip Marcus, my major professor, and Professors Jay Artis, James McKee and Donald Olmstead, my committee members. The people of the Computer Center at the University of South Dakota were invaluable during the data analysis, as was the environment which Donald Habbe of the College of Arts and Sciences was able to create for this project. On a less technical level, the succession of people who have assisted in typing and coding the questionnaires, often for no compensation, are priceless. My friends on the faculty who have joked with me on delayed completion have provided help, both technical and social. On a totally nontechnical level, my wife, Lynne, my sons, Michael and Aaron, and my parents, Carl and Esther (who can all new wave at the camera) have been a constant and essential support. Maybe a vacation this year without the sheaf of papers. TABLE OF CONTENTS Chapter Page I. INTRODUCTION . . . . . . . . . . . . . . . . . . . l The Problem . . . . . . . . . . . . . . . . . . 1 Review of Literature . . . . . . . . . . . . . . 2 Structural Factors . . . . . . . . . . . . . . 2 Attitudinal Sets (Professional and Bureau- cratiC) O O O O O O O O C I O I O O O 0 O O 11 Hypotheses . . . . . . . . . . . . . . . . . . 18 II. DATA AND METHODOLOGY . . . . . . . . . . . . . . . 26 Collection of Data . . . . . . . . . . . . . . . 26 Operationalization of Variables . . . . . . . . 27 Complexity . . . . . . . . . . . . . . . . . . 29 Size . .1. . . . . . . . . . . . . . . . . . . 29 Resources . . . . . . . . . . . . . . . . . . 30 Professional and Bureaucratic Attributes . . . 30 Salary Priorities . . . . . . . . . . . . . . 34 Data Analysis . . . . . . . . . . . . . . . . . 39 III. FINDINGS . . . . . . . . . . . . . . . . . . . . . 45 Correlation Matrices . . . . . . . . . . . . . . 45 Structural Variables . . . . . . . . . . . . . ‘45 Supervision Variables . . . . . . . . . . . . 46 Supervision Indices . . . . . . . . . . . . . . 46 Type of Activity and Source of Supervision . . 46 Chapter Structural Influences on the Supervision Indices . . . . . . . . . . . . . . . . . Complexity . . . . . . . . . . . . . Size . . . . . . . . . . . . . . . . . . Resources . . . . . . . . . . . . . . . . Faculty qualifications . . . . . . . . . Faculty salaries . . . . . . . . . . . . Affluence . . . . . . . . . . . . . . . Summary of Structural Variables to Supervision Indices Relationships . . . . . . . . . . . Multiple Regression of the Supervision Indices on the Structural Variables- . . . . . . . . Multiple Regression of the Salary Priorities on the Structural Variables . . . . . . . . Multiple Regression of the Salary Priorities on the Supervision Indices . . . . . . . . . Influence of the Structural Variables on Salary Priorities, Controlling for the Supervision Indices . . . . . . . . . . . Affluence . . . . . . . . . . . . . . . . . Zero Order Correlations: Affluence . . . First Order Correlations: Affluence . . . Summary of Supervision Indices as Test Variables on Affluence and Priority Activities . . . . . . . . . . . . . Size 0 O O O O O O O O O O O O O O O O C O 0 Zero Order Correlations: Size . . . . . First Order Correlations: Size . . . . . Summary of Supervision Indices as Test Variables on Size and Priority Activities . . . . . . . . . . . . . . . Page 52 52 53 . 54 54 55 56 S7 59 62 68 69 7O 70 74 76 78 79 79 82 Chapter Page Faculty Salaries . . . . . . . . . . . . . . . 83 Zero Order Correlations: Faculty Salaries . . . . . . . . . . . . . 86 First Order Correlations: Faculty Salaries . . . . . . . . . . . . . 87 Summary of Supervision Indices as Test Variables on Faculty Salaries and Priority Activities . . . . . . . . . . . 88 Complexity . . . . . . . . . . . . . . . . . . 90 Zero Order Correlations: Complexity . . . . 90 First Order Correlations: Complexity . . . 92 Summary of Supervision Indices as Test Variables on Complexity and Priority Activities . . . . . . . . . . . . . . . . 94 Faculty Qualifications . . . . . . . . . . . . 94 Zero Order Correlations: . Faculty Qualifications . . . . . . . . . . 96 First Order Correlations: Faculty Qualifications . . . . . . . . . . 96 Summary of Supervision Indices as Test Variables on Faculty Qualifications and Priority Activities . . . . . . . . . 99 IV. SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . 103 Introduction . . . . . . . . . . . . . . . . . . 103 Supervision and Structural Variables . . . . . . 103 Affluence . . . . . . . . . . . . . . . . . . 106 Size . . . . . . . . . . . . . . . . . . . . . 111 Faculty Salaries . . . . . . . . . . . . . . . 116 Complexity . . . . . . . . . . . . . . . . . . 118 Faculty Qualifications . . . . . . . . . . . . 120 Chapter Salary Priority Activities Conclusions . General Conclusions . . . . . . . . . . Future Research . . . . . . . . . . . . Table 1.1 11.1 11.2 11.3 II.4 111.1 111.2 111.3 111.4 111.5 111.6 111.7 III.8 LIST OF TABLES Page Hypotheses . . . . . . . . . . . . . . . . . . l9 Correlation matrix of items used to form supervision indices . . . . . . . . . . . . 35 Intercorrelation averages for all activity/ supervision combinations . . . . . . . . . . 36 Faculty activities in order of the priority which should be given when considering faculty salaries . . . . . . . . . . ... . . 38 Study variables, variable Operationaliza- tion, and variable source . . . . . . . . . 39 Correlation matrix of the structural variables 0 O O O I O I O O O O O O O O O O 45 Correlation matrix of the supervision indices 0 O I O O O O I O O O O O O I O O O 46 Faculty response to opinion of need for closer supervision of the following activities by the legislature . . . . . . . 48 Faculty response to opinion of need for closer supervision of the following activities by their faculty peers . . . . . 48 Mean percentage of the faculty not opposed to closer supervision of the faculty by source of supervision . . . . . . 49 _Mean percentage of the faculty not opposed to closer legislative supervision of the faculty by activity supervised . . . . . . . 50 Mean percentage of the faculty not opposed to closer faculty peer supervision of the faculty by activity supervised . . . . . . . 50 Percentage differences between intra-source and inter-source supervision relative to activity supervised . . . . . . . . . . . . 51 Table 111.9 111.10 111.11 111.12 111.13 111.14 111.15 111.16 111.17 111.18 ’III.19 111.20 111.21 111.22 111.23 111.24 111.25 111.26 111.27 111.28 Correlation of complexity of the college/ school with the supervision indices . . Correlation of size with the supervision indices . . . . . . . .2. . . . . . . Correlation of faculty qualifications with the supervision indices . . . . . . . . Correlation of faculty salaries with the supervision indices . . . . . . . . . . Correlation of affluence with the super- vision indices . . . . . . . . . . . . . Diagrams of structural variables to legpro and perpro . . . . . . . . . . . . . . . Legpro . . . . . . . . . . . . . . . . Legbur . . . . . . . . . . . . . . . . . . Perpro . . . . . . . . . . . . . . . . . . Perbur . . . . . . . . . . . . . . . . . . Student advising . . . . . . . . . . . . . Job counseling . . . . . . . . . . . . . . Personal values . . . . . . . . . . . . . Popularity with students . . . . . . . . . Publications . . . . . . . . . . . . . . . Research . . . . . . . . . . . . . . . . Community service . . . . . . . . . . . . University service . . . . . . . . . . . . Correlation matrix of salary priority activities . . . . . . . . . . . . . . . Zero order correlations of the structural variables and salary priorities . . . Page 52 53 54 55 56 59 6o ' 6o 61 61 63 64 65 65 66 66 67 68 71 72 Table Page 111.29 Correlation of affluence and salary priorities by supervision indices . . . . . 73 111.30 Diagram of the effect of high supervision scores on the significant correlations between salary priority activities and affluence . . . . . . . . . . . . . . . . . 77 111.31 Correlation of size and salary priorities by supervision indices . . . . . . ,,, . . . 80 111.32 Diagram of the effect of high supervision scores on the significant correlations between salary priority activities and size 0 O O O O O O O O O O O O I O O O O O O 84 111.33 Correlation of faculty salary and salary priorities by supervision indices . . . . . 85 111.34 Diagram of the effect of high supervision scores on the significant correlations between salary priority activities and faculty salary . . . . . . . . . . . . . . . 89 111.35 Correlation of complexity and salary pri- orities by supervision indices . . . . . . . 91 111.36 Diagram of the effect of high supervision scores on the significant correlations between salary priority activities and complexity . . . . . . . . . . . . . . . . . 95 111.37 Correlation of faculty qualifications and salary priorities by supervision indices . . 97 111.38 Diagram of the effect of high supervision scores on the significant relationships between salary priority activities and faculty qualifications . . . . . . . . . . . 100 CHAPTER I INTRODUCTION The Problem On a theoretical level, this study attempts to link (a) the organizational theories which advance the proposi- tion that structural factors determine the type of control system in the organization and (b) the social psychological theories which advance the proposition that perception deter- mines activity. This research explores the relationships among (a) organizational structure, (b) perceived need for supervision relative to professional and bureaucratic orientations, and (c) perceived roles, goals and functioning of the university. At a time when higher education faces increasing pressure, both from inside and outside the system, we require addi- tional understanding of the relations between organizational structure and employee perceptions. The primary purpose of this study will be to determine if organizational structure affects the identification of certain activities as profes— sional, indicating collegial control, or bureaucratic, indi- cating hierarchical control. A secondary purpose will be to determine if the difference between a professional orienta- tion and a bureaucratic orientation results in a different perception of the university. 2 Review 9; Literature Structural Factors One distinguishing aspect of organizations is that they are social inventions which, as Etzioni states, specialize in "getting things done."1 The organization's effectiveness is based on its ability (1) to effectively coordinate its members and the subunits which comprise the organization,2 and (2) to adjust to a dynamic task environment and general milieu.3 A major assumption underlying this research is that this coordination necessarily presumes such a degree of interdependence that changes in any structural variable, or set of structural variables, has ramification throughout the organization. The higher education system is certainly not immune to the dual necessity of coordination and adjustment despite the-commonly held fallacy that academicians operate in an "ivory tower" and need not concern themselves with mundane efforts directed toward preserving and improving the organi- zations, i.e., colleges and universities, which provide the settings for intellectual work. ~The ivory tower myth is presently more likely recog- nized as a myth due to a number of developments which have impacted on higher education: (1) a rise in expenditures for higher education from 1.1 to 2.5 percent of GNP during the past decade; (2) recent increases in overt dissent and disruption on campuses; 3 (3) public attention focused on the contribution higher education may give in helping solve social problems of all types; (4) the increase of public policy control as exempli- fied by increased regulatory control in all aspects of society;4 (5) the propagation of research centers and institutes which actively solicit outside recognition and funding. The crux of these developments has been increased visi- bility for the university. Unlike the old saw attributed to Hollywood aspirants during the early days of motion pictures-- "Any publicity is good publicity"--higher education must operate in a system where not all publicity is good. The Carnegie Commission on Higher Education5 predicts a slowing down of the rates of enrollment growth for the 1970's and a situation in which the rates "will temporarily cease entirely in the 1980's." The Carnegie Commission predicts that at this stage expanding fields will not be "add-ons" but rather "replacements" in higher education.6 Facing this austere future, the only apparent substan- tive and unanimous agreement among researchers is that higher education institutions have increased in size and complexity in recent times.7 Even in this agreement, Ben-David points out that univerSities were large even in the middle ages; therefore, large size is not a totally recent development.8 If there is agreement on the existence of both increas- ing size and increasing complexity in higher educational institutions, this agreement disappears when the topics of their interrelations and consequences are broached. Good- 9 10 11 man, Hutchins, and Brown each decry the deterioration of academics due to the intrusion of bureaucratic mechanisms. The title of Presthus' article, "University Bosses: The Executive Conquest of Academe," is perhaps the best succinct description of the position taken by the above scholars of 12 education. Dykes writes, "the ability of faculties to play a meaningful role in decision making is increasingly challenged as institutions grow larger and more complex and as the decision-making processes become more bureaucratized and formalized."l3 14 15 Cowley and Blau, for different reasons, disagree with the previous writers regarding size and complexity. Cowley's argument appears a variant of Blau and Scott's cui bono classification system of organization.16 Cowley believes that professors have lost sight of those for whom the education system was created, i.e., students and the general society, and consequently have developed a collegial system to control higher education for the benefit of pro- fessors.17 The development of hierarchical mechanisms is perceived as one means of returning control and benefit to their rightful place. Blau disagrees with the critics of increased size and complexity on a number of grounds. Blau states that 5 "universities and colleges have administrative structures that are similar to those of other bureaucracies," but that the same structural features do not have the same signifi- 18 Blau also illustrates that larger univer- cance for each. sities have disproportionately smaller administrative units and instead of being rigid, are more likely than smaller institutions to innovate by the development of new fields of study. Larger size also tends to reduce the "paternal- istic" centralization of authority possible in a smaller institution. “Blau believes that the increases in size and complexity actually serve to give professors increased autonomy and research possibilities rather than binding them by greater bureaucratic restraints.19 It should be noted that some of the bureaucratic features may have "deleterious consequences for educational performance," particularly for undergraduates.20 Much of the debate regarding the consequences of struc- tural changes either explicitly or implicitly revolves around professional-bureaucratic poles, with the profes- sional pole usually regarded as faculty supportive and the bureaucratic pole as administrative supportive. The diffi- culty with this conception is the zero sum dimension; any increase in administrative control or influence must be at the expense of the faculty and vice versa.21 This posture appears erroneous. Dyke's position that "faculty power and administrative power are, in a sense, 22 fused," with each depending on the other, appears more \/ 6 nearly to approximate the real situation. Even a cursory examination of the origin of prestigious (hopefully corre- lated with quality) institutions indicates the presence of "powerful” administrators during their development.23 The preceding portion of this literature review limited itself to studies of organizations engaged in higher educa- “tion, but universities have much in common with other types of organizations. Dykes, for example, writes, "as universi- ties grow larger and more complex, they tend to take on characteristics of other large organizations; structural superordination and subordination are accentuated, rules and regulations become more important, hierarchical authority increases, and universities move away from the characteris— tics of community and collegiality. In a word, they become bureaucratic."24 In reference to administrative structure, Blau states the "structure of universities and colleges is amazingly homologous to that of other types of organiza- tions."25 With at least a degree of commonality established between organizations engaged in higher education and orga- nizations engaged in other functions, the review will expand into the both more abundant and more abstract methodological and theoretical literature on organizations. The remainder of this review draws heavily from the work of Blau, both singly and in collaboration, direct and indirect, with his students and colleagues. Heydebrand compiled an inventory of structural variables for the "systematic, quantitative-comparative analysis of 7 large-scale, complex, formal organizations" regardless of whether the analysis be interorganizational or intraorgani- zational in nature.26 Using his review, we will stress those studies connected to particular structural variables most relevant to our work, i.e., (a) complexity, (b) size, and (c) resources. Complexity refers to the degree of differentiation within an organization, and is usually defined along two dimensions, horizontal and vertical. The latter refers to levels in the authority system, and the former refers to either (a) the number of distinct occupations in a unit or (b) the number of distinct subunits in an organization.27 Since this study is not concerned with authority levels as such, attention focuses on horizontal complexity, as meas- ured by the number of subunits, i.e., departments in a college/school. Complexity as utilized here approaches Rage and Aiken's indicators of occupational specialties,28 but is more nearly analogous to Blau and Schoenberr's "major subunits" of em- ployment security agencies which were termed "the first order of horizontal structural differentiation in an orga- 29 nization." Blau and Schoenherr operationalized a major‘ subunit as one in which the "head . . . always reports to the top management of an organization."30 The organiza- tional chart of the university investigated places the dean of each college/school directly responsible to the univer- sity president. 8 Size can be defined along a number of dimensions. Hall, Haas and Johnson operationalize size in terms of the number 31 while Hage and Aiken32 and Blau33 and the of members, majority of his colleagues operationalize size in terms of the number of employees. Other variables have also been ‘used to measure size. Price34 cites Melman as using five: (1) number of production personnel, (2) total assets, (3) average number of wage earners per establishment, (4) aver- age value added by manufacture per establishment, and (5) 35 Despite defensible alternatives, organiza- ‘net sales. tional literature generally defines size in terms of per- sonnel, either as members or employees. This research will define size in terms of employees. Size, one of the first structural variables used in organizational analysis, was initially considered the causal variable affecting the degree of complexity in organizations or societies. Simmel contended: “The size immediately de- 36 termines the form.” Durkheim thought that increasing popu- lation density led to increasingly complex patterns of orga- nization.37 Weber saw size as one of the major factors which lead to bureaucracy and the development of the ration- legal authority so necessary to effectively coordinate emerg- 38 ing complex systems. Michels saw size as one of the fac- tors which led to the "Iron Law of Oligarchy."39 According to Michels, increased size leads to increased complexity leads to increased specialization and finally increased centralization of information for coordination purposes. , 9 Utilization of the latter allows those occupying the higher positions to perpetuate their control with eventual develop- ment of an oligarchy. Recent research more likely posits a relationship be- tween a development in unit size and development in the administrative component of an organization rather than retention of specific personnel in administrative positions. Terrien and Mills in their study of elementary school dis- tricts substantiated the hypothesis, "The relationship be- tween the size of an administrative component and the total size of its containing organization is such that the larger the size of the containing organization the greater will be the proportion given over to its administrative component."40 Anderson and Warkov investigated the same phenomenon using data gathered from Veterans AdminiStration hospitals in 41 Anderson order to reassess Terrien and Mills' hypothesis. and Warkov did not refute Terrien and Mills' conclusions, but added important qualifying conditions. Sheer growth in organizational size does not always result in disproportional increase in the administrative component; "the explanatory variable is organizational complexity rather than organiza- tional size."42 The administrative component increases directly as (a) the number of tasks performed at the same place increases, and (b) the number of places at which the tasks are performed increases. The size of the administra- tive component decreases as the number of people performing the same task at the same place increases.43 lO Boland in a study of publicly supported colleges and universities mentions "degree of organizational complexity" as of "particular importance" in the distribution of power in the university system,44 yet the strong impact of size is evident when he summarizes the relationship as highlighting "the considerable importance of numbers in understanding "45 these matters. Blau's position with respect to the importance of size has been mentioned earlier.46 Many researchers have viewed resources, particularly resource surpluses, as the property of a system or organiza- tion which allows it to protect or expand its present domain.47 Surplus resources are necessary to increase in size and to develop a high degree of division of labor yet each of the latter two contribute to the accumulation of resources. Resources may be either concrete or abstract in nature. The abstract element of prestige has received less attention in the literature of organizations that the concrete element of finance, but of the major organizational assets, Thompson believes prestige is one of the most valued assets and yet one of the "cheapest."48 Lenski states that "power will determine the distribution of nearly all of the surplus," and power leads to prestige, but to be "complete” Lenski states this conception needs to indicate "some feedback from prestige to power."49 Blau states that "the antecedent conditions that extend most influence . . . can all be conceptualized as resources ll . . . human and institutional as well as financial assets."50 Blau believes the last makes the first two possible; affluent institutions pay better salaries, have better qualified and larger faculties and "tend to be decentralized with respect to educational affairs."51 Blau sees resources as a major factor in effectively allowing structural innovations in higher education. Attitudinal Sets (Profes- sional and Bureaucratic) Thus far we have emphasized variables which will func- tion as the original independent variables in this study. We now turn to dependent variables, professional and bureau- cratic, which function as dependent variables in the first portion of the findings, particularly as they refer to hierarchical and collegial control. Weber believed "bureaucracy and rationality necessarily 52 (Constas is one who disagreesS3) while went hand in hand" his more recent successors have concentrated on demonstrating an inverse relationship between the presence of "rationality" (now more likely termed professional) and "bureaucratic char- acteristics."54 Udy found support for the hypothesis that "mutual positive associations tend to exist between bureau- cratic elements and also between rational elements, but that rational elements tend to be negatively associated with bureaucratic elements."55 This study will investigate the relation of activities operationalized as professional or bureaucratic activities 12 specific to the organization studied, without an assumption of rational or nonrational distinctions. Much of the research to this time has operationalized professional and bureaucratic acCording to the proportion of administrative staff relative to "productive staff," i.e., those directly in the workflow pattern. The Terriens and Mills and Anderson and Warkov studies are two examples of this approach.56 A second common way of operationalizing professional and bureaucratic is according to entrance qualification and routinization of work activities. Hage and Aiken's work utilizes this second approach.57 In regard to the first approach, the conceptualization is overly simplistic, particularly since earlier studies did not properly account for the degree of labor-intensiveness in production. Blau, looking at the social structure rather than technology, writes that rather than viewing few man- agers as indicative of wider spans of control, hence less supervision and more autonomy for employees, it is possible to view few managers as implying a more "centralized author- ity structure, which encourages . . . one-sided directives with little feedback . . . thus reducing the autonomy of subordinates."58 Concerning the increasing internal divi- sion of labor and the resultant smaller supervision ratio, Heydebrand indicates the relation holds "only at the low 59 Reliance on the ratio of levels of professionalization." administrative to actual production employees does not ap- pear a true measure of professionalization/bureaucratization 13 for all situations. The second approach also has drawbacks as an absolute indicator of professional standing since qualifications can vary quite widely. For example, Stinch- combe refers to the efficiency of the "highly professional- ized manual labor force" in the construction industry,60 while Blau, Heydebrand and Stauffer operationalize profes- sionalization as "the proportion of the operating staff . . . who are required to have, at least, a college degree with a specified major."61 Skill levels, peg s3, do not equate to professionalization. A third approach, similar to the one which will be advanced later in the methodology portion of this paper, incorporates each of the above dimensions plus the type of work involved in fulfilling work expectations. Rage and Aiken refer to "highly non-routine" positions as an indication of a professional position.62 Not having a source to cite, perhaps because "what everyone believes no one be- lieves with conviction," I would advance the assumption that all work positions contain some elements of both routine and non-routine activities, albeit the mixture is different.. Reaching further, the assumption advanced is that there is a degree of isomorphism between professional organizations and the professional positions in them. Higher education faculty have been both included and excluded as professionals.63 This paper will follow the lead of Blau and his colleagues in considering higher educa- tion faculty as professionals.64 This paper will also refer to units, i.e., colleges/schools, according to degree of l4 professionalism and/or bureaucracy, again analogous to a procedure used by Blau, Heydebrand and Stauffer.65 The university will be considered a professional orga- nization since it meets the criteria advanced by Montagna: ”(1) profesSional groups define and achieve the primary organization goals, (2) the majority of the people in the organization are professionals . . ." (true, if students are excluded as "in the organization"), "(3) administrative hierarchy of authority lies within the firm, whereas author- ity in professional matters is placed in the hands of the professional association, and (4) the profession promotes norms of personal autonomy and altruistic action in all mat- ters relating to use of the body of knowledge."66 A caution, suggested by Price, is mentioned. Profes~ sionalization is a quality of an occupation, not of an orga— 67 The nature of the nization nor any other social system. occupations in the organization, not the organization Reg 52, is what designates an organization as a professional organization. The position taken in this study regarding the nature of professional and bureaucratic will differ from that of Weber, who considered the two phenomena not only compatible but a logical outgrowth of each other.68 The position taken here will also differ from that which views the two phenomena as independently or inversely related. V. Thompson has come closest to expressing the latter position in lay terms when he writes of "the growing gap between the right to decide' 15 which is authority, and the power to do, which is specialized ability."69 Kornhauser also reflects this orientation when he insists that examining professionals in bureaucracies is examining "the relation between two institutions, not merely 70 II between organizations and individuals. Dispensing with negatives, the position taken is simi- lar to Hall who sees "the assumption of inherent conflict between the professional or the professional group and the employing organization . . . to be unwarranted."71 I society in which many professions are becoming bureaucra- In a tized and many bureaucracies are becoming professionalized,72 insistence on their incompatibility appears forced. Hall's study of professionalization and bureaucratiza— tion deals with both structural and attitudinal attributes.73 The former consists of such items as formal education, en- trance exams, etc., while the latter consists of such items as a sense of calling to the field, colleagues as the major work reference group, etc. By the criteria advanced earlier, the structural attri- butes of professionalization can be taken as a given for this study. The respondents, due to their relationship with the university, are labeled as professionals. The attitudinal attributes, by virtue of their nature, cannot be created by fiat; therefore, their existence cannot be assumed. Perceived need for professional autonomy, autonomy being one of the major attributes of the professional ideal type,74 is an attitudinal attribute and a valid area for further research. 16 Hall states that "an equilibrium may exist between the levels of professionalization and bureaucratization in the sense that a particular level of professionalization may require a certain level of bureaucratization to maintain social control."75 If Hall is correct, on both the existence of an atti- tudinal attribute and some type of equilibrium, a logical extrapolation is that professionals in organizations develop attitudinal stances regarding their work activities as pro- fessional, hence subject to collegial control, or bureau- cratic, hence subject to hierarchical control. I would interpret the statement in the Carnegie Commission report, Governance 9f Higher Education, "Selective independence--not autonomy—-is the issue as we see it,"76 to reflect a call for closer alignment of professional and bureaucratic spheres. Respondents in Dyke's study of faculty involvement in decision making showed a "strong tendency to dichotomize decisions into 'educational' and 'noneducational' categories," a dichotomy Dykes termed "arbitrary and impractical" because of the interrelatedness of decisions. "If the faculty's in- fluence is to be truly effective, surely it must be mani- fested in all areas."77 Faculty have proven an ability to deliberate, but not as great an ability to decide. Faculty should have input to the decision-making process, but to the present researcher, the distinction between educational and noneducational is neither arbitrary nor impractical; it is a matching of the 17 valid type of control system to the proper jurisdiction, and if faculty recognize this distinction, they should more likely be praised than blamed. "An institution cannot be well governed unless each of its components clearly recog- nized its obligations as well as its rights in the promotion of the common end."78 Defining the boundaries, if existent, for collegial and hierarchical control appears a logical starting place if obligations and rights are to be clearly recognized. This review has surveyed the literature on structural variables for (a) size, (b) complexity, and (c) resources, particularly as they occur in institutions of higher educa- tion. The review has also attempted to briefly trace the development of the study of bureaucracy and professionalism from the unity of existence Weber ascribed to them, to the inverse relation noted by Udy, to Hall who views them as phenomena which may need their influence balanced to some optimal equilibrium point if the organization is to be effective. The introduction of Hall's division of structural and attitudinal attributes in conjunction with professional and bureaucratic work activities provides the stimulus to ex- plore the relationships between the structural variables of an organization and attitudinal orientations relative to selected work activities. 18 Hypotheses Blau's work with the structural analysis of organiza- tions and Hall's work differentiating professional and bureaucratic characteristics by both their structural and attitudinal attributes were the major sources for the devel- opment of the hypotheses to be tested in this study. Com- bining the two, the following hypotheses are proposed. Table 1.1, on pages 19 and 20, shows the hypotheses. For the sake of parsimony the following shorthand will be utilized in stating the hypotheses: t = greater, in- crease, etc.; + = lesser, decrease, etc.; and + = affects, causes, etc. This same shorthand method will be used later in the study. 19 muwmuwbdas a“ moflufl>wuom ouw>uom How >uwuofium+ Suflcsaaoo ca moflufi>wuom oofl>uom How xufluofium+ mmfiufl>wuom coupomou Mow wuwuoflum+ mace» Imowansm new hufluofium+ mucopsum saws >uaudasmom new hufluoflum+ monopcmum Hoownuo new mosao> Ansonusm HON >uwu0wum+ nucopsun mo meadowsm nooumo can mafiaomssoo Issoo non How Suwuoflum+ mucousun Ho usoaomw>pm Dan—030M How huwuofluma. any Amv Amy Ame Apv on Any Any , +Acowumuwumuosmmusn+v sowumnwamcowmwomoum+ mmmHHoo 0:» tea mudsomH mo Honabc+ AH uuam muofiflmm Hmeomumm pom Hmeacauauua muouomm Housuosuum II“ monocuom»= H.H sands 20 H mammnuommc ca co>flq mm mowufluoflum How neocosqomcoo meow A mammnuomxn cw co>wm mm mofiufluoHum wow neocosqwmcoo 08mm H memonuomxn Ce ce>fim mm mmHuAHOwum mom moososqmmcoo meow H mflmmnuomxn ea cm>wo mm wofluwuoflum u0u neocosqomcoo meow +A20flumwwumuosmousn+v coflumuflamcoflmmowoum+ tacowumuflumuosmousn+v :oflumuflamcoflmmououm+ +Ac0HumNflumuosmousn+v cowumuflaocofimwmwoum+ +Acoflumuwumuosmousn+v COADmNAHmconmowoum+ +mocosammm+ Am mmoaaoo on» CA taumHmm auasomw oomuo>m+ Av +mcowuoowwflaosq muasoow+ Am mmoHDOmmm wmoaaoo on» :H +mucoauuomop mo nonasc+ Am auflxmaafioo mwmwaom assemuom umm Hmcacnuauua muouomm Housuosuum lemscuucooc H.H manna 10. 11. 12. 13. 14. 21 Footnotes Amitai Etzioni. A Sociological Reader pp Complex Orga- nizations, 2nd edition. New York: Holt, Rinehart and Winston, Inc. 1969. p. 197. James Thompson. Organizations ip Action. New York: McGraw-Hill. 1967. pp. 39—50. Ibido ' pp. 66-82. Carnegie Commission on Higher Education. Governance of Higher Education. New York: McGraw-Hill. 1973. pp:- 20-21. Ibid., p. 8. Ibid., p. 8. Ibid., plus see Walter R. Boland. "Size, External Rela- tions, and the Distribution of Power: A Study of Col- leges and Universities" in Comparative Organizations, edited by Wolf V. Heydebrand. Englewood Cliffs, New Jersey: Prentice-Hall, 1973; Peter M. Blau. 222 Orga- nization pf Academic Work. New York: John Wiley and Sons. 1973. Joseph Ben-David. "The Scientific Role," Minerva (4), 1965-1966, p. 19. - Paul Goodman. The Communipy pf Scholars. New York: Random House. 1962. Robert M. Hutchins. The Learning Society. New York: Praeger. 1968. Douglas J. Brown. The Liberal University. New York: McGraw-Hill. 1969. Robert Presthus. "University Bosses: The Executive Conquest of Academe" in The New Republic (152), Febru- ary, 1965. Archie R. Dykes. Faculty Participation £3 Academic Decision Making. Washington, D.C.: American Council on Education. 1968. p. v. W. H. Cowley. "Professors, Presidents, and Trustees," AGB Reports, Volume 9, No. 5 (February, 1967), pp. 14- 15. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 22 Peter M. Blau and W. Richard Scott. Formal Organipp- tions. San Francisco: Chandler. 1962. pp. 42-57. Blau, pp. pip., p. 279. Cowley, pp. pip., pp. 14—15. Blau, pp. pip., p. 279. £pi§., p. 279. Ipip., p. 280. Dykes, pp. pip., p. 40. lp;g., p. 41. Frederick Rudolph. The American College and Universipy. New York: Alfred A. Knopf. 1962; Brown, pp. pip.; The Carnegie Commission on Higher Education, pp. pip. Dykes, pp. pip., p. 13. Blau, pp. p1p., p. 269. Heydebrand, pp. pip., p. 1. James L. Price. Handbook pg Organizational Measurement. Lexington, Massachusetts: D. C. Heath. 1972. p. 70. Jerald Hage and Michael Aiken. Social Change ip Complex Organizations. New York: Random House. 1970. pp. 15- 18. Peter M. Blau and Richard A. Schoenberr. The Structure pf Organizations. New York: Basic Books. 1971. p. 401. Ibid., p. 401. Richard J. Hall, J. Eugene Haas and Norman J. Johnson. "Organizational Size, Complexity, and Formalization" in American Sociological Review (32), December, 1967, p. 905. Jerald Hage and Michael Aiken. "Program Change and Organizational Properties: A Comparative Analysis" in American Journal pf Sociology (72), March, 1967. Peter M. Blau. "The Hierarchy of Authority in Organi- zations" in American Journal p: Sociology (73), January, 1968. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 23 Price, pp. cit., p. 174. Seymour Melman. ”The Rise of Administrative Overhead in the Manufacturing in the United States, 1889-1947" in Oxford Economic Papers (3), January, 1951, p. 75, quoted in Price, _p. pi£., p. 174. Georg Simmel. "The Number of Members as Determining the Sociological Form of Groups" in The American Journal pf Sociology (8), 1902-3, quoted in Frederic W. Terrien and Donald L. Mills, "The Effect of Changing Size Upon the Internal Structure of Organizations" in American Sociological Review (20), February, 1955. Emile Durkheim. On the Social Division of Labor in Society, translated by George Gimpson. New York: MacMillan. 1933. Max Weber. “The Presuppositions and Causes of Bureau- cracy" in From Max Weber: Essays in Sociologyp edited by Hans H. Gerth and C. Wright Mills. New York: Oxford Press. 1946. Robert Michels. "The Conservative Basis for Organiza- tion" in Reader in Bureaucrapy, edited by Robert K. Merton, Ailsa P. Gray, Barbara Hockey and Hanan C. Selvin. New York: Free Press. 1952. Terrien and Mills, op. cit., p. 11. Theodore R. Anderson and Seymour Warkov. "Organiza- tional Size and Functional Complexity: A Study of Administration in Hospitals" in American Sociological Review (26), February, 1961. Ibid., p. 28. Ibid., PP. 27-28. Boland, pp. cit., p. 433. Ibid., p. 439. Blau, pp. cit., pp. 278-279. Karl Marx and Friedrich Engels. "Bourgeois and Prole- tarians" in Marx and Engels: Basic Writings pp Politics and Philospphy, edited by Lewis S. Feurer. Garden City, New York: Anchor Books. 1959; Gerhard Lenski. Power and Privilege. New York: McGraw-Hill. 1966; Thompson, _p. cit. Thompson, pp. cit., p. 33. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 24 Lenski, pp. cit., pp. 44-49. Blau, Organization p: Academic Work, op. cit., p. 249. Ibid., p. 251. Max Weber. "The Essentials of Bureaucratic Organiza- tion: An Ideal-Type Construction" in Merton, pp 31., _p. cit. Helen Constas. "Max Weber's Two Conceptions of Bureau- cracy" in American Journal pf Sociology (63), 1958, pp. 400-409. Talcott Parsons. The Structure pg Social Action. New York: McGraw-Hill. 1937. pp. 500-578; Stanley J. Udy. “'Bureaucracy' and 'Rationality' in Weber's Organiza- tional Theory: An Empirical Study" in American Socio- logical Review (24), December, 1959; Arthur F. Stinch- comb. "Bureaucratic and Craft Administration of Pro- duction: A Comparative Study" in Administrative Sci- ence Quarterly, September, 1959, p. 169; Richard H. Hall. "Professionalization and Bureaucratization" in American Sociological Review (33), February, 1968; Paul D. Montagna. "Professionalization and Bureau- cratization in Large Professional Organizations” in American Journal pg Sociology (74), September, 1968. Udy, pp. cit., p. 793. Terrien and Mills, pp. cit.; Anderson and Warkov, pp. cit. Jerald Hage and Michael Aiken. "Routine Technology, Social Structure and Organization Goals" in Adminis- trative Science Quarterly (14). Peter M. Blau. "The Hierarchy of Authority in Organi- zations" in American Journal p: Sociology (73), Janu- ary, 1968, p. 458. Heydebrand, Comparative Organizations, pp. cit., p. 473. Stinchcomb, pp. cit., p. 169. Peter M. Blau, Wold V. Heydebrand, and Robert E. Stauffer. "The Structure of Small Bureaucracies" in American Sociolpgical Review (31), April, 1966, p. 182. Hage and Aiken, "Routine Technology, Social Structure and Organization Goals," _p. cit., p. 368. 25 63. Burton R. Clark. "College Teaching" in Professionali- zation, edited by Howard M. Vollmer and Donald L. Mills. Englewood Cliffs, New Jersey: Prentice Hall. 1966. pp. 285-286; Everett C. Hughes. Men and Their Work. Glencoe, Illinois: Free Press. pp. 139-144. 64. Blau, Organization pg Academic Work, op. cit., pp. 11-13. 65. Blau, Heydebrand and Stauffer, pp. cit., pp. 181-182. 66. Paul D. Montagna. "Professionalization and Bureaucrati- zation in Large Professional Organizations" in American Journal pg Sociology (74), September, 1968. 67.. Price, pp. cit., p. 50. 68. Weber, "The Essentials of Bureaucratic Organization," _p. Cit. 69. Victor A. Thompson. Modern Organization. New York: Alfred A. Knopf. 1969. p. 6. 70. William Kornhauser. Scientists ip Industry: Conflict and Accommodation. Berkeley: University of California Press. 1962. p. 8. 71. Richard H. Hall. ”Professionalization and Bureaucrati- zation" in American Sociological Review (33), February, 1968, p. 104. 72. Amitai Etzioni. "Authority Structure and Organiza- tional Effectiveness" in Administrative Science gpar- terly (4), June, 1959. 73. Hall, "Professionalization and Bureaucratization," _p. cit., pp. 92-93. 74. Epig., p. 93. 75. 1229-: p. 104. 76. Carnegie Commission on Higher Education, pp. pip., p. 18. 77. Dykes, pp. pip., p. 40. 78. Robert M. MacIver. Academic Freedom ip_0ur Time. New York: Columbia University Press. 1955. p. 73. CHAPTER II DATA AND METHODOLOGY Collection pg Data The data were drawn from a project conducted on a medium-sized campus located in the North Central Plains. Two major sources of information were used to gather the data analyzed in this research: (a) a mailed questionnaire and (b) university records. The mailed questionnaire was distributed May 6, 1972. The mailing of the questionnaires was preceded by publicity in the student newspaper and an introductory letter mailed three days before the mailing of the questionnaire. The time difference between first return and last return was forty—six days. This does present a time lapse with poten- tial for responding to a changed environment, but nothing of great moment appeared to have happened during that span which would greatly alter responses on the campus. Other data were provided by the following three univer- sity documents: (1) the University's Qperating Budget 1971-72 (fiscal year 1972), (2) the University's Salarerist 1971-72 (fiscal year 1972), and (3) the registrar's corrected Spring computer class enrollment printout. 26 27 The Operating Budget 1971-72 publication provided (1) the number of departments in each college or school, and (2) the total budget allocated to each college or school.1 The Salary List 1971-72 provided (1) the number of faculty in each college or school, (2) the number of faculty with and without professional degrees, and (3) the average salary in each college or school.2 The registrar's spring enrollment printout provided the total number of students enrolled in each college or school in the spring of 1972. Since these documents are important to the operation of the University and are subject to scrutiny by the Board of Regents, the state legislature and the general public, it is assumed they are reliable and accurate. (See Appendix A.) Operationalization p£_Variab1es The organizational setting for this study is a medium— sized university, and much of the data is drawn from indi- vidual questionnaires, but the unit of analysis will be neither the University nor its personnel, but the subunits designated as colleges or schools on the University's main campus. The investigation of organizations using comparative structural attributes of the organization as variables allows for analysis using the organization as a whole, not as characteristics of particular individuals. This approach means the organization, ppp pp, becomes the unit of investi- gation rather than merely the environment in which the study 28 occurs. It is by the study of the organization, ppp pp, not as setting, that "the special needs of the field-—needs gen- erated by the emergent properties of this type of social unit," will be answered. Prior to specifically operationalizing each variable, the general class of variable as determined by their proper- ties will be advanced following the guides developed by Lazarsfeld and Menzel.4 It should be noted that each of the three types of properties--analytical, structural and global--describe collectives; organizations in this instance, not individuals. According to Lazarsfeld and Menzel, global properties are those characterized "by properties which are not based on information about the properties of individual members."5 The complexity variable in this study is developed from the global properties of the university. The analytical "properties of collectives . . . are obtained by performing some mathematical operation on some "6 Size and the average property of each single member. faculty salary variables in this study are developed from the analytical properties of the university. The structural "properties of collectives . . . are obtained by performing some operation on data about the relations of each member to some or all of the others."7 The resources variables (with the exception of "average faculty salary" which is analytical) are developed from what Lazarsfeld and Menzel would term the structural 29 properties of the university. The terminology used by Lazarsfeld and Menzel for this property may be somewhat confusing since the term is the same as that used through the rest of this study to refer to the whole set of vari- ables which are neither attitudinal nor phenomenological in nature. The supervision indices are developed from analytical properties of the organization as are the priorities for faculty salary. The Operationalization of the structural variables will draw heavily from Blau and his associates. Complexity Price notes that complexity is discussed, among other terms, as division of labor, specialization, functional dif- ferentiation, departmentalization, etc.8 Blau writes, "The index of the formally instituted academic division of labor is the number of departments."9 The indication of complexity for this study will be the number of departments in the college/school listed in the Operating Budget 1971-72.10 Size Operationalization of size follows Blau's approach of using employees rather than members. Blau operationalizes size as "Total number of faculty, both full-time and part- 11 time." This research will modify Blau's Operationaliza- tion along the lines suggested by Hall, Haas, and Johnson12 30 and will determine full-time equivalent faculty according to the percent employment listed in the university Salary List 1971-72.l3 Resources Resources will be operationalized using definitions advanced by Blau. One indicator of resources is affluence, the ratio of total budget to the total number of students in the college/school. This Operationalization is analogous to Blau's designation which is operationalized as "the total revenue in dollars divided by total enrollment of both under- graduate and graduate students."14 A second indication of resources, average faculty sal- ary, is operationalized as average academic year salary for 1971-1972. This Operationalization is limited to an aca- demic year and full-time equivalent faculty, but it is analogous to Blau‘s generationalization which defines sal— ary as "average annual salary of full-time faculty . . ."15 A third indicator of resources, faculty qualifications, is operationalized as the ratio of faculty with a profes- sional or Ph.D. degree to faculty members without a profes- sional or Ph.D. degree. Blau defines faculty qualifications as "The percentage of the total faculty with Ph.D.‘s or pro- fessional degrees."16 Professional and Bureau- cratic Attributes Self-regulation and autonomy are two related concepts widely used when characterizing professions. Self-regulation 31 is to grow out of "the concept of community" which "presup- poses an organization in which functions are differentiated and in which specialization must be brought together, or coordination, if you will, is achieved not through a struc- ture of superordination and subordination of persons and groups but through a dynamic of consensus."17 Control in a profession is collegial control which develops out of technical competence rather than hierarchal position.18 Autonomy means the professional should not only be charged with the "supervision" of his peers, but should be free to select the manner and techniques by which he accom- plishes the goals defined by him and his professional peers. The professional "ought to be able to make his own decisions without external pressures from clients, those who are not members of his profession, or from his employing organiza- tion."19 The elements above are disavowed in a bureaucracy. Friedrich refers to one of his six elements of bureaucracy as "centralization of control and supervision (hierarchical aspects)."20 The division of source of supervision to (a) legisla- ture as bureaucratic (hierarchical control) and (b) peer as professional (collegial control) is along accepted dimen- sions differentiating the two concepts. The division of type of activity into professional or bureaucratic is done on the basis of (a) complex routiniza- tion of activity and (b) specialized knowledge required for 32 each activity. Higher degrees of each of the above are assumed to be present in the activities: (a) content of courses, (b) selection of books and (c) assignments given, and consequently these activities are termed professional activities. Lower degrees are assumed to be present in the activities: (a) faculty absence from the classroom, (b) office hours for student appointments and (c) student advising. The data for professional and bureaucratic designations were generated by the following portion of the questionnaire. 1. In your opinion should the faculty of USD be under closer supervision by the legislature in regards to the following? CHECK (/) ONLY ONE ALTERNATIVE FOR EACH ACTIVITY. Definitely Undecided, Undecided, Definitely yes probably probably no yes no 1) Content of courses taught I I I ] I ] I ] 2) Selection of books [ ] [ ] I ] [ l 3) Assignments given I ] I l I 1 I I 4) Faculty absence from the classroom [ ] [ ] [ ] [ ] 5) Office hours for student appointments [ l I I I l I l 6) Student advising [ ] I ] I 13 I I 2. In your opinion, should the faculty of USD be under Closer supervision by their faculty peers in regards to the following? CHECK (/) ONLY ONE ALTERNATIVE FOR EACH ACTIVITY. 33 Definitely Undecided, Undecided, Definitely yes probably probably no yes no 1) Content of courses taught I l I ] I l I l 2) Selection of books [ ] [ ] I ] I ] 3) Assignments given [ ] [ 1 [ ] [ ] 4) Faculty absence from the Classroom [ ] [ ] [ ] [ ] 5) Office hours for student pppointments [ ] [ ] [ ] [ ] 6) Student advising [ ] [ ] [ ] [ ] Four variables were formed from Legpro which reflects attitudes the above. toward closer supervi- sion by the legislature of activities which are professional in nature (items 1 through 3 of question 1). Legbur which reflects attitudes toward closer super- vision by the legislature of activities which are bureau- cratic in nature (items 4 through 6 of question 1). Perpro which reflects attitudes toward closer super- vision by faculty peers of activities which are professional in nature (items 1 through 3 of question 2). Perbur which reflects attitudes toward closer supervi- sion by faculty peers of activities which are bureaucratic in nature (items 4 through 6 of question 2). Each of the last four variables consists of two dimen- sions: (a) source of supervision (legislative-external to the organization or faculty peers-internal to the organiza- tion), and (b) activity supervised (professional or bureau- cratic). 34 Table II.l is a correlation matrix of the items used to form the supervision indices. Table 11.1 indicates that the scales have a high degree of intra-category correlation as shown in the "blocks" along the diagonal of the matrix. Reading to the right of the diagonal, the cOrrelations are patterned in conformity to the logic advanced in the theo- retical formulations. Table II.2 is developed from Table II.1. Table 11.2 gives the arithmetic average of the cor- relations. The matrix adds support for these operationali- zations since the elements appear to form both clusters of correlation and non-correlation as would be predicted from the theory. Respondents were placed in the categories Legpro, Legbur, Perpro and Perbur based on a dichotomized summation of their scores on the items which made up the respective categories. A value of four was assigned for "Definitely yes,‘ a value of three was assigned for "Undecided, probably yes," a value of two was assigned for "Undecided, probably no," and a value of one was assigned for "Definitely no." The scores of each respondent were summed within each cate- gory, and the study population was dichotomized in the man- ner prescribed by Davis.21 Salary Priorities The priority responses were operationalized solely on the basis of faculty responses to the following set of state- ments concerning how much priority should be given selected activities when considering faculty salaries. 35 oc.H mma>6< ucmtsum mo mo. oo.H musom moamuo am we. mm. oo.H monomna suasomm we mm. as. ea. co-” coamma mmmau an we. mm. as. on. oo.H uomaom seem an as. am. as. mm. am. oo.a mmmuzoo unmucoo ma ov. «m. cm. ma. Hm. mm. oo.H oma>6¢ ucmcaum so am. am. am. mm. Hm. mm. «A. oo.H muses moammo Am mm. Am. mm. mm. EH. ma. mm. mm. oo.H mucomnm suaaoum a« HH. mo. OH. am. on. «H. mm. me. mm. oo.H :mamma annau gm oa. mo. no. mm. cm. us. On. em. om. cm. oo.H uooaom xoom gm mm. ow. ma. mm. mm. mm. me. am. ov. am. no. oo.H mmmusoo unmuaoo AH mm mm m? mm mm WA Q0 Am AG Am AN RH UHUMHUDMOHd—m HMCOHmeMOHm 0.." UMHOQMOHSQ HMGOHmmOMOHm >uw>fiuo< >uw>fiuod wumom huasomm musumamflmmq Cowmweuomsm mwOHccw cosmfl>ummsm away 0» Comp mamas mo xwuuds coflumawuuoo H.HH manna °zna KQIAIQDV szaad thnoeg ';ozd oma Kntntnov 'JOJd sznqetsrbeq uoTsTAJadns mmflu 6 Ausnnmmv mm. nfi>fiuom Oaumuosmmusn mo Amy 3 conH>Hmmsm ummm muasomm moan mv. Aoumnmmv om.. IH>Huom Hmcowmmmmoud mo AOV coflmfl>ummsm Home wuasomm . mmau am. am. Ansnmmqv em. ufl>wuom Oflumuosmmusn mo Abe cowmfl>nmmsm musumHmflmmq mews NH. mm. mm. Aoummmqv hm. Ifl>fluom accoammmmoum mo Amv cosmw>umm9m Ousumamwwmq A3 A3 By 23 mGOHumcaQEoo cofimfl>umm5m\mufl>fluom Ham How mmmmum>m coflumHmnuooumucH m.HH sense 37 Below is a set of statements concerning salaries paid within the university. How much priority should each of the following be given when considering faculty salaries? CHECK (/) ONLY ONE ALTERNATIVE FOR EACH ACTIVITY. A high Some Little Np_priority priority priority priority at all (a) Academic advisement of students I ] I ] I ] I ] (b) Job counseling and career guidance of students I ] L I I I I I (c) Personal values and 'ethical standards I ] I ] I ] I ] (d) Popularity with students I l I ] I l I I (e) Publications I l I 1 I l I ] (f) Research activities [ ] I ] I l I ] (g) Service activities in the community I l I ] I l I I (h) Service activities in the university ~ (epg., committees) I ] I ] I ] I ] Table 11.3 gives the order the faculty ranks the above activities in terms of the priority which the activities should be given when considering faculty salaries. The order indicates the faculty's for the various activities. While the are close in value, the highest degree priorities is for the priorities given lications. Table 11.4 gives a summary of the Operationalization and their source. relative priorities standard deviations of consensus on rank to research and pub— variables used, their 38 Table 11.3 Faculty activities in order of the priority which should be given when considering faculty salaries Rank Activity Standard Deviation 1 Research .38 2 Publications .40 3 Student Advising .48 4 Job Counseling .49 5 University Service .49 6 Community Service .50 7 Personal Values .50 8 Popularity with Students .48 Data Analysis The unit of analysis has already been presented in an earlier portion of the thesis. Thus this section will limit itself to the proposed statistical analysis. Buhler's P-Stat computer package is used for the analy— sis of the data collected for this research.22 For the per— centage analysis and Yule's 0 analysis, the data is dichoto- mized as closely as possible to the 50:50 split recommended by Davis.23 Dichotomization which falls outside the 30:70 range is avoided unless substantive logic dictates otherwise in which case the degree of skew is specifically mentioned. In this area of statistical analysis Alfred N. Whitehead's dictum is appreciated--"Seek simplicity and distrust it."24 More sophisticated analysis is used in the matrix and regression analyses. The matrices developed utilize Pear- son's r, while the major statistic from the multiple regres- sion is the standard partial regression coefficient. Table 11.4 39 and variable source Study variables, variable Operationalization, Variable Name Operationalization Source Complexity number of departments in the Operating college/school Budget, 1971-72 Size number of full-time equiva- Salary List, lent faculty in the college/ 1971—72 school Resources a) affluence ratio of the total budget to Operating the total number of students Budget, 1971-72 b) salary c) faculty Legpro- Legbur Perpro Perbur Salary pri- orities enrolled in each college/ school‘s spring 1972 classes average academic year salary for a full-time equivalent faculty member for the aca- demic year 1971-72 ratio of faculty with a pro- fessional or Ph.D. degree to faculty without a profes— sional or Ph.D. degree conjunction of legislative supervision and professional activities conjunction of legislative supervision and bureaucratic activities conjunction of peer super— vision and professional activities conjunction of peer super— vision and bureaucratic activities , self-responses to a set of questions concerning salary priorities for selected activities and registrar's spring 1972 en- rollment list Salary List, 1971-72 Salary List, 1971—72 questionnaire questionnaire questionnaire questionnaire questionnaire 40 One of the principal uses of standard partial regres- sion coefficients (beta weights) is as "measures of relative «25 importance. Blau used this type of regression analysis in The Opganization 9; Academic Work, stating that the "beta weight is indicative of the direct effect of each independent on the dependent variable, controlling the other variables 26 This study uses the same in the regression equation." technique. The beta weight is a measure of the direct effect of an. independent variable on a dependent variable while control- ling for the other independent variables being considered. The choice of independent variables to be included obviously affects the results of the regression. Constant sets of independent variables will be used throughout the regres- sions for this study in order to eliminate anyaccenting or suppressing effect based on exclusion or inclusion of selec— tive independent variables. In order to determine the direct or indirect influences by use of multiple regression techniques it is necessary to assign a sequence to the independent variables. The sequence for the structural variables is: (l) affluence, (2) size, (3) faculty salary, (4) complexity, and (5) faculty qualifi— cations. The sequence chosen is important since the vari- ables earlier in the sequence have a greater probability of being designated as having a direct effect on the independent variables. The variables would probably not be put in this same causal sequence by every researcher, and certainly it 41 must be admitted that reciprocal influences are exerted among the variables thereby making a claim of unilateral causality unrealistic. At the same time, the predominant causal sequence is believed to be reflected in the above order, and "no meaningful analysis of social structures is possible if the investigator always vacillates, attributes any concomitant variation of conditions to reciprocal influ- ences, and refuses to commit himself to a predominant causal direction."27 The sequence given to the independent vari- ables follows that given by Blau. It should be noted that the data do not meet the ideal statistical assumption of regression analysis. Much of the data are Ordinal, not interval, but this does not prohibit regression analysis. As Blau notes, "A number of methodolo- gists have noted in recent years that data that do not fully conform to the assumptions of regression analysis can be «28 used without distorting results. All of the data are at least ordinal and Boyle has illustrated the validity of data of this type in conjunction with regression techniques.29 The direct effect of the independent variable can be learned from the multiple regression results, while later' Yule's Q analysis can indicate the influence of the inde— pendent variable interactive with other independent varia- bles, designated and undesignated, which are implicitly in- cluded in the actual measurement of the independent variable under investigation. Although this procedure, i.e., the recognition that unknown variables may be influencing the result, is not as 42 “neat" as a.pure statistician would prefer, it is a reflec- tion of the state of measurement in the social sciences, particularly in the areas outside controlled experiments. An advantage of making this point explicit is that the inde- pendent variables can correctly be conceptualized as a first attempt at developing empirical parameters within which fur— ther refining can occur. Blau states, "Since there is little rigorous theory as yet in sociology, however, the typical role of theoretical conceptions and principles nowadays is to provide a guiding framework for largely, exploratory research."30 The nature of methodology utilized in soci- ology dictates its role as a guide to investigation rather than a fount of definitive proofs. The respondents do not represent a random sample, but rather a population; consequently, statistical techniques, findings and conclusions must be limited to case study parameters rather than the broader population of higher education pp pppp. One of the obvious statistical conse- quences of the case study method is that tests of signifi- cance related to the probability of study inclusion are ir— relevant. Tests of significance will not be reported. The units of analysis, i.e., colleges and schools in a university, are not independent of each other; consequently, the problems of multicollinearity and confounding effects may be a factor in portions of the analysis. Present sta- tistical techniques do not indicate a correction factor. 10. ll. 12. l3. 14. 15. 16. 17. 18. 43 Footnotes University of South Dakota. The University of South Dakota Operating Budget, 1971‘72'. 1972. University of South Dakota. The University pg South Dakota Salary List, 1971-72. 1972. Amitai Etzioni. A Sociological Reader pp Complex Orga- nizations, 2nd edition. New York: Holt, Rinehart and Winston, Inc. 1969. p. 495. Paul L. Lazarsfeld and Herbert Menzel. "On the Relation between Individual and Collective Properties" in Etzioni, Ibid. Ibid., p. 503. Ibid., p. 504. Ibid., p. 505. James L. Price. Handbook pf Organizational Measurement. Lexington, Massachusetts: D. C. Heath. 1972. p. 70. Peter M. Blau. The Organization 9; Academic Work. New York: John Wiley and Sons. 1973. p. 51. University of South Dakota, The University pf South Dakota Operating Budget, 1971-72, pp. cit. Blau, pp. cit., p. 286. Richard J. Hall, J. Eugene Haas and Norman J. Johnson. "Organizational Size, Complexity, and Formalization" in American Sociological Review (32), December, 1967, p. 905. University of South Dakota, The University pg South Dakota Salary List, 1971-72, pp. cit. Blau, pp. cit., p. 286. Ibid. Ibid., p. 287. John Millett. The Academic Community. New York: McGraw-Hill. 1962. pp. 234-235. Talcot Parsons. "Introduction to Max Weber" in The Theory pf Social and Economic Organization. New York: Oxford University Press. 1947. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 44 Richard H. Hall. "Professionalization and Bureaucrati- zation" in American Sociological Review (33), February, 1968, p. 94. . Carl J. Friedrich. "Some Observations on Weber's Analysis of Bureaucracy" in Reader pp Bureaucracy, edited by Robert K. Merton, Ailsa P. Gray, Barbara Hockey and Hanan C. Selvin. New York: Free Press. 1952. p. 29. James A. Davis. Elementary Survgy Analysis. Englewood Cliffs, New Jersey: Prentice-Hall, Inc. 1971. pp. 50-51. Roald Buhler. P-STAT: A Computing System for File Manipulation and Statistical Analysis pf Social Science Data. Princeton, New Jersey: Princeton University Computer Center. 1971. Davis, pp. cit., p. 59. Alfred North Whitehead. The Concept pf Nature. Cambridge: Cambridge University Press. 1920. George W. Snedecor and William G. Cochran. Statistical Methodp, 6th edition. Ames, Iowa: Iowa State Univer- sity Press. 1967. p. 398. Blau, pp. cit., p. 34. Ibid., p. 35. Ibid., p. 39. Richard P. Boyle. "Path Analysis and Ordinal Data" in American Journal pg Sociology (75), 1970, pp. 461-480. Blau, pp. pip., p. 45. b CHAPTER III FINDINGS Correlation Matrices Structural Variables Tables 111.1 and 111.2 are presented to give the inter- _Correlations of the variables which will function as inde— pendent variables. The data in Table 111.1 come from the University records. Table 111.1 Correlation Matrix of the Structural Variables _p. --.......-..._ 4.“..-a~1.— -4... mH.*maMb~-I MI-:-—_.—.cmo—l~.o«t 3‘91”»..C .a.‘ ..-....‘ -....‘ 3 .L--‘.. .. ,.- ._ a- _. —a_....,_,"_: Aff. Size Sal. Comp. Qual. Affluence 1.00 .35 .14 .43 .10 Size 1.00 -.82 .82 —.72 Salary 1.00 -.58 .88 Complexity 1.00 -.45 Faculty Qualifications 1.00 The strong associations do raise the question of multi- collinearity when the effects of the more highly correlated independent variables are considered relative to the depen- dent variable. The situation is not desired, but the rela— tion exists and cannot be wished away. Blau writes that "overriding substantive interests" resulted in his using size and number of departments (correlated .83), a correla- tion very similar to this study's correlation of .82 between 45 -7 .ID...’{ l:...e 46 size and complexity (number of departments).1 Substantive interests also indicate pursue in the above areas despite the strong association between some variables. Supervision Variables Table 111.2 gives the inter-correlations of the Super- vision Indices. The data come from faculty responses to the questionnaire. Table.III.2 Correlation Matrix of the Supervision Indices Legpro Legbur Perpro Perbur Legpro 1.00 .51 .33 .18 Legbur 1.00 .28 .39 Perpro 1.00 .64 Perbur ' 1.00 The correlations indicate that the same source of supervision, either legislative (Legpro to Legbur .51) or faculty peer (Perpro to Perbur .64), tends to give the high- est correlations followed by the same type of activity, either professional (Legpro to Perpro .33) or bureaucratic (Legbur to Prebur .39), with mixed source of supervision and type of activity (Legpro to Perbur .18 and Legbur to Perpro .28) having the weakest association. Supervision Indices e pg Activity and Source pg Supervision The relations among the major dimensions which form the indices will be investigated in this section. The major 47 dimensions are (a) source of supervision and (b) activity supervised. Since each supervision index consists of a combination of each of the above, their effect on each other will be given prior to their combination. The most parsimonious presentation is to give the rela- tions as hypotheses and show the support for each hypothesis. Three hypotheses are presented in the following. The first hypothesis dealt with in this section is: 1. Faculty are less likely to oppose closer super- vision if the source of supervision is faculty peers versus the legislature. In order to investigate hypothesis one, Table 111.3, which gives the faculty response to closer supervision of selected activities by the legislature, and Table 111.4, which gives the faculty response to closer supervision of the same selected activities by faculty peers, are presented. The data are presented as percent ppp opposed to closer supervision with the other half of the dichotomy being the combined responses of faculty both opposed and undecided. The former category was chosen for reporting since it is a category more distinctly defined by the actual respondents rather than through Operationalization procedures. A cursory investigation of the table indicates that the faculty indicate less opposition for closer supervision by the faculty peers than by the legislature. In fact, the lpppp opposed of any activity suggested for closer supervi- sion by the legislature is still eleven percentage points 48 Table 111.3 Faculty response to opinion of need for closer supervision of the following activities by the legislature Percent not opposed to Activity closer supervision N a) Content of courses taught 16% (23) 148 b) Selection of books 9% (13) 148 c) Assignments given 9% (13) 148 d) Faculty absence from the classroom 37% (55) 147 e) Office hours for student appointments 21% (31) 146 f) Student advising 28% (41) 146 Table 111.4 Faculty response to opinion of need for closer supervision of the following activities by their faculty peers Percent not Opposed to Activity closer supervision N a) Content of courses taught 70% (103) 148 b) Selection of books 53% (78) 148 c) Assignments given 48% (69) 146 d) Faculty absence from the classroom 73% (108) 148 e) Office hours for student appointments 62% (97) 147 f) Student advising 58% (85) 147 49 less than the most opposed of any activity suggested for closer supervision by faculty peers. Table 111.5 Mean percentage of the faculty not Opposed to closer supervision of the faculty by source of supervision Percent not opposed to Source closer supervision Legislature , 20% Faculty Peers 61% The percentage difference of 41 percent between the means supports the first hypothesis. Supervision originat- ing from a source internal to the organization is less likely to be opposed than supervision originating from a source external to the organization. The second hypothesis states: 2. Faculty are less likely to oppose closer super- vision of bureaucratic activities versus profes- sional activities. Referring to Tables 111.3 and 111.4, where parts a), b), and c) are professional activities and parts d), e), and f) are bureaucratic activities, Table 111.6 and Table 111.7 can be generated. In each table, the percentage not opposed to closer supervision was higher for the bureaucratic activities than for the professional activities. Relative to legislative supervision, bureaucratic activities had a difference of seventeen percentage points over professional activities, 50 and relative to faculty peer supervision the difference was thirteen percentage points. Support is thus found for the second hypothesis. Supervision of bureaucratic activities is less likely to be opposed than supervision of profes- sional activities. Table 111.6 Mean percentage of the faculty not opposed to closer legislative supervision of the faculty by activity supervised Percent not opposed to‘ Activity closer supervision Bureaucratic 28% Professional 11% Table 111.7 Mean percentage of the faculty not Opposed to closer faculty peer supervision of the faculty by activity supervised Percent not opposed to Activity closer supervision Bureaucratic 64% Professional 57% It should be noted that hypothesis two is supported regardless of whether the source of supervision is the legislature or faculty peers. The third hypothesis states: 3. Faculty are more likely to give greater weight to the source of supervision than the activity being supervised. 51 'Referring to Tables 111.6 and 111.7, where parts a), b), and c) are professional activities and parts d), e), and f) are bureaucratic activities, Table 111.8 can be generated. Table 111.8 Percentage differences between intra-source and inter-source supervision relative to activity supervised Intra-source difference: Bureaucratic/Professional Percent difference Legislative 17% Faculty Peers 7% Inter-source difference: Faculty Peers/Legislature Percent difference Bureaucratic 36% Professional 46% Table 111.8 shows that the largest differences in opposing closer supervision are attributable more to the .source of supervision than to the activity being supervised. The percentage mean of the intra-source values is 12 percent whereas the percentage mean of the inter—source values is 41 percent. Support is thus found for the third hypothesis. I Faculty are more concerned with the source of supervision than with the activity being supervised. Summing up the findings, each hypothesis was found to be supported by the data. Faculty are less likely to oppose peer supervision than legislative supervision. Faculty are less likely to oppose supervision of bureaucratic than 52 professional activities, and the source of supervision is more influential in generating opposition to supervision than is the type of activity. Structural Influences pp the Supervision Indices Complexity The review of literature pointed to a complexity as one of the major qualifiers of the effect of size on administra- tive components. Perceptions of supervision were predicted to be influenced by the complexity of the unit in which the respondent was employed. Table 111.9, Correlation of com— plexity of the college/school with the Supervision indices, gives the correlations for the variables. Table 111.9 Correlation of complexity of the college/ school with the supervision indices Supervision Indices Yule's Q Legpro +.02 Legbur -.ll Perpro -.4l Perbur -.28 Legpro and legbur are not significant (+.02 and -.11, respectively) while perpro and perbur are significant (-.41 and -.28, respectively). The ordinal ranking for the super- vision indices is opposite that predicted from the hypoth- esis, perpro having the largest negative correlation (-.41) and legpro having a positive correlation (+.02). Support 53 is thus not found for the first two variable relationships of hypothesis 1 (see page 19). Size The review of literature showed that one of the conse- quences of increasing size was to alter the administrative unit of an organization. Perceptions of supervision were predicted to be influenced by the size of the unit in which the respondent was employed. Table 111.10, Correlation of size of the college/school with the supervision indices, gives the correlations, and they support the position that size does influence perceived need for selective types of supervision. Table 111.10 Correlation of size with the supervision indices Supervision Indices Yule's Q Legpro +.03 Legbur -.l6‘ Perpro -.34 Perbur ' -.39 Legpro is directly related to size (+.03), but the cor— relatiOn is not significant. Legbur is inversely related to size (-.l6), but again the correlation is not significant. Both perpro and perbur are inversely related to size and are significant (-.34 and —.39, respectively). Support is 54 thus not found for the first two variable relationships of hypothesis 2 (see page 20). Resources Faculty gpalifications. The review of literature indi- cates that as professional qualifications increase there is less need for, and less utility derived from, hierarchical supervision. Perceptions of supervision were predicted to be influenced by faculty qualifications such that the more highly professional colleges/schools would favor Closer col- legial supervision, while the less qualified would more likely favor closer hierarchical supervision. Table 111.11, Correlation of faculty qualifications with the supervision indices, gives the correlations. Table 111.11 Correlation of faculty qualifications with the supervision indices .... -1 —‘ ._ - -—.I-. - .-__.. -—-.-- .g.---—-.—»A--»..-—- 5.” -.-...— ”‘gg. ~_. Supervision Indices Yule's Q Legpro -.12 Legbur -.03 Perpro +.30 Perbur +.29 The first thing noted on Table 111.11 is that the ordinal ranking is exactly that predicted. Legpro is the highest negatively correlated (-.12) and perpro is the high— est positively correlated (+.30). Again, legpro and legbur were not significant (-.12 and -.O3, respectively) while 55 perpro and perbur were significant (+.30 and +.29, respec- tively).. Support is found for the first two variable rela- tionships Of hypothesis 3 (see page 20). I Faculty salaries. The review of literature indicates thatthe best paid,faculty tend to be the most productive and the most professional faculty. .Perceptions*of super- ~vision were predicted to be influenced by faculty salaries such that colleges/schools with the highest average salaries would be the colleges/schools most likely to favor closer collegial versus closer hierarchical supervision. Table 111.12 shows that another indicator of resources correlates as predicted. Table 111.12 Correlation of faculty salaries with the supervision indices ~Y— f Supervision Indices ’ ‘Yule's Q Legpro -.26 ‘Legbur ‘ +.04 Perpro +.27 Perbur +.28 Legpro, predicted to be the least likely to be posi- tively correlated with faculty salaries, is a negative. significant, -.26, whereas perpro and perbur, predicted to be most likely to be negatively correlated to salary, .are significant at +.27 and +.28, respectively. Legbur is not significant (+.04), but ranks with the predicted order. 56 Support is found for the first two variable relationships of hypothesis 4 (see page 20). Affluence. The first resource indicator is affluence of the college/school. The variable is the result of the ratio of finances to output; therefore, a clarification shouldbe stated. The variable is not, nor is it intended to be, a measure of efficiency. In fact, the variable is more likely a measure of the degree to which a college/ school can afford to be inefficient in the pursuit of being effective. Again, the predicted relation was the higher the ratio of budget/student, the more likely to favor collegial control. Table 111.13 Correlation of affluence with the supervision indices -.—- .1..- ,.....- -r... Supervision Indices Yule's Q Legpro -.ll Legbur -.10 Perpro -.36 Perbur -.39 Table 111.13 indicates legpro and legbur are not sig- nificant (-.11 and -.10, respectively), while both perpro and perbur are significant (-.36 and -.39, respectively). The total budget to student majors is the only resource connected variable in the study which does not support the hypothesis predicted (see page 20). 57 Summary of Structural Variables to §gperv sion Indices Relationships The following significant relationships were discovered. 1. Complexity is negatively related to perpro and perbur. 2. Size is negatively related to perpro and perbur. 3. Faculty qualification is positively related to perpro and perbur. 4. Faculty salary is negatively related to legpro, and positively related to perpro and perbur. 5. Affluence is negatively related to perpro and perbur. Support is thus found for the hypothesized relation- ships between the first two variables in hypotheses 3 and 4 on pages 19 and 20 of this study. The findings reject the original hypothesized relationships between the first two variables of hypotheses l, 2. and 5, and give support to conclusions opposite the originally proposed relations. Increased size and increased complexity both appear to have a dampering effect on perceived need for collegial supervision, while higher faculty qualifications and higher faculty salaries both appear to encourage closer collegial supervision. The resource variable, affluence, introduces a variant in the pattern developed by the other relations. As the ratio gets larger, need for closer supervision of any activity from any source is perceived as less necessary. This ”abberation" in the apparent pattern is not disruptive, but is expected. 58 1f alternate types of supervision are conceptualized as one means of (l) more successfully coordinating the internal organization and (2) more favorably influencing the task environment of the organization, then if the organization is receiving the greatest share of resources per unit produced, it desires closer supervision from neither a hierarchical source which may institute more austere practices, nor from a "less prosperous" peer whose potential jealousy might dis- rupt the present favorable situation. Legpro, hierarchical supervision of professional activ- ities, can be viewed as a mismatched supervision system which is least likely to be favored in a professional orga- nization. Perpro, peer supervision of professional activi- ties, can be viewed as a matched supervision system which is most likely to be favored in a professional organization. A juxtaposition of these two indices, which shows their rela- tive relationship regardless of their individual statistical significance, might be of interest. Table 111.14, Diagrams of structural variables to legpro and perpro, serves the dual function of clarifier and summarizer for this section. The same shorthand system is used as that in the original hypotheses section. Table 111.14 gives a more easily visible support to the original hypotheses 3 and 4, while showing hypotheses 1, 2, and 5 are not supported. The nature of the hypothesis modi— fication needed is also quite Clear in Table 111.14. 59 Table 111.14 Diagrams of structural variables to legpro and perpro l) +Size + +Legpro (+.03) .2 less likely col- ' +Perpro (-.34) legial super- vision 2) +Complexity+ +Legpro (+.02) .3 less likely col- +Perpro (-.41) legial super- vision 3) +Faculty +Legpro (-.12) .2 more likely col- Qualifications+ +Perpro (+.30) legial super- vision 4) +Faculty Salaries+ +Legpro (-.26) .2 more likely col- +Perpro (+.27) legial super- vision 5) +Aff1uence+ +Legpro (-.ll) .2 less likely +Perpro (-.36) either collegial or hierarchical supervision Multiple Regression of the Supervision Indices on the Structural Variables The first tables presented will utilize the structural variables as independent variables and the supervision indices as dependent variables. Beta weights will be given for all relations, but if the beta weight is less than twice its standard error, they are considered to have no appreci- able effect.2 I Only affluence and faculty salary are indicated as having significant direct effects on the belief that there should be closer supervision of the faculty's professional activities by the legislature. 1n the former instance a negative effect is registered, while in the latter a posi- tive effect appears. 60 Table 111.15 Legproa Simple Beta Weight Correlation l. Affluence .l4** -.06 2. Size -.40 .21 3. Salary -.61 -.09 4. Complexity .01 .01 5. Qualifications .20 -.04 R2=.O4 **more than three times its standard error 3The title in the table always refers to the dependent variable and the row stubs to the independent variables. Table 111.16 Legbur _— . ~——- Simple Beta Weight Correlation l. Affluence -.24** -.16 2. Size .00 -.08 3. Salary .31* .02 4. Complexity ~.l4 -.06 5. Qualifications -.20 -.Ol R2=.04 A V *More than twice its standard error **More than three times its standard error Again affluence, the ratio of the total college budget to the total number of college‘majors, and faculty salary, the average academic year salary, are indicated as the only two having a significant direct effect on the belief that there shquld be closer supervision of the faculty's bureau- cratic activities. The beta weights indicate a selective 61 factor operating with the more affluent colleges/schools being more likely to favor closer legislative supervision of bureaucratic activities. The higher salaried colleges/ schoOls indicate preferences exactly contrary to those of the more affluent colleges/schools. Table 111.17 Perpro Simple Beta Weight Correlation l. Affluence .02 -.07 2. Size . .03 -.17 3. Salary -.17 ' .13 4. Complexity -.24 -.20 5. Qualifications .20 .14 R2=.05 None of the relations between the structural variables and perpro indicate a significant direct effeCt. Table 111.18 Perbur Simple Beta Weight Correlation l. Affluence -.21* .16 2. Size -.07 -.18 3. Salary .17 .12 4. Complexity .12 -.12 S. Qualifications .00 .12 R’a.os *More than three times its standard error 62 Only affluence is indicated to have a significant ef- fect on the belief that faculty peers should exert closer supervision over the bureaucratic activities of faculty members. It is interesting to note that only the structural variables which are based on resources had significant direct impact on the belief sets concerning supervision. Size and complexity have received much more emphasis in organizational literature relative to appropriate styles of supervision, but in this instance_resources appear to exert a greater effect. As indicated earlier in the review of literature, most studies of size now explicitly consider the additional factor complexity in explanatory or predic- tive schema; perhaps both size and complexity should be con- ceptualized as intervening variables between resources and the phenomenon to be explained. Multiple Regression 9; the Salary Priorities pp the Structural Variables This section will present the tables generated by the multiple regression of salary priorities on the Structural variables. Table 111.19 presents the multiple regression on the priority student advising. Affluence, faculty salary and faculty qualifications are the three variables indicated to have a direct effect on beliefs regarding the priority which should be given the activity student advising when considering faculty salaries. The higher-salaried colleges/schools are indicated as putting 63 a lower priority on this activity than their less wealthy counterparts, but the more affluent and more highly quali- fied colleges/schools are indicated as placing a higher priority on this activity than their less highly qualified counterparts. Table 111.19 Student advising Simple Beta Weight Correlation l. Affluence .21* .00 2. Size p-014 -006 30 salary . “088* 003 4. Complexity -.26 -.08 5. Qualifications .68* .15 R’=.11 *More than three times its standard error In view of the high correlation between faculty salary and faculty qualifications, this finding is unexpected. One possible explanation, which cannot be investigated with the present data since the necessary questions were not asked, is that more junior faculty with doctoral degrees are being hired, but due to their lack of seniority in the colleges/ schools they are (a) less well paid and (b) relegated to the task of student advising. Consequently, the junior faculty feel that student advising, since advising is the work the junior faculty is doing, is the activity that should have a high priority when salaries are considered. 64 Table 111.20 presents the multiple regression on the' priority job counseling. Table 111.20 Job counseling Simple Beta Weight Correlation l. Affluence .02 -.05 2. Size .10 -.08 4. Complexity - -.24 -.ll 5. Qualifications .49* .14 R .006 *More than three times its standard error Faculty salary and faculty qualifications are indicated as having direct effect, negative and positive, respectively, regarding the salary priority which should be giVen to job counseling and career guidance of students. Again the ap— parent lack of common effect from two variables closely cor- related was not expected. Table 111.21 presents the multiple regression on the priority personal values. Complexity is the only variable indicated as having a direct effect on beliefs regarding salary priorities for personal values and ethical standards. The effect of com- plexity is negative. Table 111.22 presents the multiple regression on the priority popularity with students. 65 Table 111.21 Personal values Simple Beta Weight Correlation 2. Size -.15 . -.12 .3. Salary -.11 .03 4. Complexity -.33* _ -.21 5. Qualifications .08 -.02 R22.06 *More than two times its standard error Table 111.22 Popularity with students Simple Beta Weight Correlation l. Affluence ‘-.37* .20 2. Size .82* -.10 3. Salary .45* , .06 4. Complexity -.44* -.23 5. Qualifications .08 .05 R’=.12 *More than three Affluence, size, times its standard error salary and complexity are each indi- cated as having a direct effect, the first and last having negative effects while the middle two have positive effects. priority publications. Table 111.23 presents the multiple regression on the Faculty salary is indicated to have a direct positive effect on the belief that publications should be regarded when considering faculty salaries. 66 Table 111.23 Publications Y" Simple Beta Weight Correlation l. Affluence .03 .07 2. Size -.14 A -.01 3. Salary .32* .05 4. Complexity .20 .05 5. Qualifications -.31 -.02 R’=.03 *More than two times its standard error Table 111.24 presents the multiple regression on the priority research. Table 111.24 Research Simple Beta Weight Correlation 1. Affluence .10* .19 2. Size ‘ .00 .08 3. Salary .39** .03 4. COmplexity .16 .12 5. Qualifications -.28 -.03 R2-.06 *More than two times its standard errOr **MOre than three times its standard error Affluence (positive effect) and faculty salary (posi- tive effect) are the two variables that have a significant effect on beliefs regarding the priority which should be given to research when considering salaries. Colleges/ 67 schools that are more affluent are more likely to promote research than their less affluent counterparts, and the higher salaried colleges/schools are more likely to promote research relative to the lower salaried colleges/schools. Table 111.25 presents the multiple regression on the priority community service. Table 111.25 Community Service Simple Beta Weight Correlation l. Affluence .02 .10 2. Size .22 -.25 3. Salary .35* . .36 4. Complexity ’ -.13 -.20 5. Qualifications .12 .33 R2=.14 jfi. *More than two times its standard error Only faculty salary is indicated as having a signifi- ‘cant effect on salary priority fer community service, the higher-salaried colleges/schools being more likely to per- ceive service activities in the community as valid criteria regarding salary priorities. Table 111.26 presents the multiple regression on the priority university service. The more affluent colleges/schools are more likely to promote service in the university than are the less endowed colleges/schools. The higher salaried colleges/schools are more likely to reject university service as a priority. The 68 more affluent units are more nearly consistent in their ap- proach to service both in the university and the community, while the higher salaried units support community service as a priority while rejecting university service. It should be noted that no compensation is allocated for university service, while additional monetary rewards usually accompany community service. Table 111.26 University service ‘ Simple Beta Weight Correlation l. Affluence .26** .01 2. Size -.29 -.32 3. Salary -.33* .30 5. Qualifications .20 .28 R2=.15 *More than two times its standard error **More than three times its standard error Multiple Regression of the Salary Priorities on the §ppervision Indices The additional information gained about salary priori- ties by means of multiple regressions on the supervision indices was not extensive. 1n the interest of parsimony the tables will not be presented since the only independent vari- able which has a significant direct effect is perpro which has a positive effect on both publications and research. The more likely a perceived need for closer supervision of 69 professional activities by faculty peers, the more likely a favoring of publications and research as salary priorities. Influence of the Structural Variables pp Salapy Eriorities, ControlIIng for the Supervision Indices This section will present the findings regarding the effect of the structural variables upon the perception of how much priority should be given selected activities when considering faculty salaries. Since the supervision indices will be utilized as test variables, listwise deletion is used rather than pairwise deletion. Rosenberg states, "In a sense, all relationships may be considered conditional . . . ."3 At this stage, all the relationships will be termed conditional relationships with no attempt made to designate contingent associations as due to, for example, a suppressor variable versus a distorter variable. This strategy is employed with the intention of later attempting to develop an explanation which succeeds in simultaneously accommodating all, or most, of the_divergent contingent associations into an integrated interpretation rather than developing separate interpretations for each pair of contingent associations. Only correlations in which (1) the conditional Q's dif- fer by 10 or more units and (2) both conditional tables meet the standards for expected cell frequencies will be treated as being significantly influenced by the supervision indices. The activities to be considered in regard to faculty salary priorities are (a) academic advisement ofstudents, 70 (b) job counseling and career guidance of students, (c) per- sonal values and ethical standards, (d) popularity with stu- dents, (e) publications, (f) research activities, (9) ser- vice activities in the community, and (h) service activities in the university, e.g., committees. A correlation matrix of the activities is given in Table 111.27. Table 111.28 presents the zero order correlations of all the salary priority activities by all the structural variables. Affluence The findings of this section concern the effect of affluence on salary priority considerations controlling for the supervision indices. Table 111.29, Correlation of af- fluence and salary priorities by supervision indices, pre- sents the correlations. Zero Order Correlations: Affluence I Academic advising (+.38), job counseling (+.33), ser- vice in the community (+.44), service in the university (+.59), and research (-.34) are each significantly related to more affluent colleges/schools. With the exception of service in the community the relations are as hypothesized. The zero order correlations of affluence and salary priorities tend to indicate that affluent colleges/schools are more likely to promote salary priorities based on ser- vice to the local student, university or community rather than discipline advancement. 71 oo.a wuamuc>wsb mow>umm .m mm. oo.H muessEEoo cow>umm .h co. mc. oo.H noummmmm .o mo. mo. em. oo.H mcowumowansm .m mo. ma. ma.| mo.u oo.a musmosum muwucasmom .v ma. no. mo.1 «0.: mm. oo.H mesam> HecOmuom .m 5N. ea. om.u NH.L on. He. oo.H mswammssou non .N ma. Ha. NH.I mH.| H~. hm. up. oo.H msfimfi>p< basswood .H m h w m w m N A mmwuw>wuoc huwuowum humane mo Nausea sowucamuuou 5N.HHH OHQMB 72 He.+ em.u om.+ oe.u .mm.+ >uamum>aso moa>umm so.+ Hv.u sm.+ m¢.- se.+ suaaseeoo moa>umm mo.| Hm.+ no.+ H~.+ vm.n gossamer eo.u, NH.+ ~H.+ Hc.- 4a.- maosumoaaasa HH.+ mv.| ma.+ HN.I mH.+ mucmosum muwumasmom .mo.+ me.| mo.+ m~.I N~.+ mosam> HMCOmem om.+ mm.a oH.I mH.a mm.+ usflHmmssoo non vm.+ ma.n mo.+ ~H.n mm.+ mswmw>c< oascomom msowveu auaxmamaou ausacm exam cosmsammm -mmuum nowuwHOaum humane use amanmwuc> acusvosuum can no escaucacuuoo Heouo oucu m~.HHH,manB 73 .mOfl>uwm cam >uwnmasmom “mad “oaa u c xuwssEEoo mOfi>umm u s mosqm> HMCOmHmm “mad u s mcflammssoo non .ofiam> uoc mum mcofiumHsoamo 0 haucmswomcoo cam m>flw BOHOQ mp3 mHHOO OHOE so oco cw >Ocmsowuw dame cwuommxm one were moumowpsw >c umfluOmQSm onen>c “GHH H C SUHMQWGH «OHH .mHH n c muamuo>flcs u c mcoHUMOflansm Lona u c marmosum sues .>Hco mufluofium mcwmfl>pm Owempmom scum. mm mm mm mm as mm mm mm HNH u .c NE84+ me.+ Hm.+ am.+ mo.+ mv.+ oo.+ >com.+ mm.+ spamsm>sce mos>smm mm.+ ms.+ Hm.+ mm.+ om.+ om.+ oe.+ >amm.+ ve.+ suscsssou mus>smm >cem.+ me.- >cmm.- Hm.- se.- >co~.u Hm.u >cmm.+ em.- nusmmmmm >cma.+ mm.- >csa.- mo.+ mm.- >cmo.u em.- >cmv.+ ea.- mcoaumuaaasa vH.I hm.+ mo.l 5v.+ 0a.: hv.+ no.+ >com.+ mH.+ mucmosum auwumHsmom ~m.+ mo.+ mm.+ oH.- mm.+ mo.+ m~.+ >cma.+ m~.+ macaw» Hmaomsma os.+ ea.+ ms.+ >aam.+ mm.+ mo.+ se.+ >aom.y mm.+ masswmcsoo non mo.+ oH.+ EH.+ >ame.+ sm.+ HH.+ He.+ >:¢H.+ mm.+ mcsma>cm oaswcmua 0H as on an 0a a; 0H as Hanson OHQHOA Hsnmcq Oumwmq umpuo _ OHUN .L moowch sowmw>uomsm Away t. '5'! huwbwuod mufluowum moamsamma “was IIIIII.-- mmoflocw COHmH>Hmm5m >3 mmHuHHOMHm %HmHmm cam OOCOSmem mo COflumHmuuou cm.HHH cance 74 An examination of the affluent colleges/schools reveals that these colleges/schools have (1) the smallest average class enrollments, (2) the highest faculty/student ratio and (3) the least number of service courses to nonmajor students. These conditions encourage the development of primary relationships between faculty and student and conse- quently greater emphasis on these relations as salary pri- orities for the faculty. A possible reason why the more affluent colleges/schools are more active in university activities is that this par-' ticipation is one means of retaining a favorable position in the allocation of resources. The only hypothesis for which significant results are generated supportive to that hypothesized was service to the community. Although a surprise, the lack of support for the hypotheses in this area does logically follow from findings later in this study. The implications will be considered at a later point in the analysis. First Order Correlations: Affluence Legpro pp test variable. Utilizing the above criteria, none of the correlations formed when legpro was used as the test variable are valid correlations to designate as signifi- cant or non-significant. This result is statistically deter- mined, although the ultimate source of determinacy is sub- stantive. The faculty's rejection of legislative super- vision of professional activities, legpro, is so definitive 75 that the legpro cannot be utilized in any first order corre- lations in this study. There simply were so few faculty members who supported closer supervision by the legislature that the type of statistical analysis used in this study would not be valid. Since none of the correlations meet the criteria for significance advanced earlier, neither support nor nonsupport can be claimed for any of the hypotheses concerning size and priority activities by legpro. Legbur pp test variable. A high legbur score signifi- cantly decreases the positive relation between more affluent colleges/schools and beliefs concerning priorities for aca- demic advising, job counseling, personal values, service in the community and service in the university. In no instance is the effect of a high legbur score strong enough to change the relation from a positive to a negative value. ‘The significant correlations indicate a high legbur score for an individual tends to dampen the effect of a positive relation between the organizational variable afflu- ence and the priorities for academic advising, job counsel- ing, personal values, service in the community and service in the university. With the exception of popularity with students and service in the community, support is not found for the last two variable relations advanced in hypothesis 5. (Publications and research, while not significant, changed in the opposite direction hypothesized.) 76 Perpro pp test variable. Only three priority activi- ties, personal values, popularity with students and service to the community, meet the statistical requirements for sig- nificance. The effect of perpro was opposite that predicted on popularity with students and service to the community, but was as predicted for personal values. Perbur pp test variable. Four of the priority activi- ties, academic advising, job counseling, personal values. and service to the community, meet the statistical require- ments for significance. The effect of a high score on perbur is to dampen the positive effect of the organizational variable affluence on the priority activities academic advising, job counseling, personal values, and service to the community. Only for service to the community does the effect not support the hypothesized relations. Summary of Supervision Indices as Test Variables on_Affluence and Priority Activities Due to the relatively small n (22) of respondents scoring high on legpro and consequent lack of statistical significance, legpro will not be included in the summary. Utilizing the same shorthand system established earlier, Table 111.30, Diagram of the effect of high supervision scores on the significant relationships between salary pri- ority activities and size, is presented. Analysis of Table 111.30 shows that regardless Of the supervision index being investigated a high score on the 77 Table 111.30 Diagram of the effect of high supervision scores on the significant correlations between salary priority activities and affluence Supervision Direction of Index Correlation Change Activities + Legbur + + Academic Advising (.11)* + Job Counseling (.08) + Personal Values (.02) f Popularity Students (.48) + Service Community (.36) + Service University (.48) + Perpro + + Personal Values (-.10) + Popularity Students (.47) + Service Community (.28) + Perbur + + Academic Advising (.10) + Job Counseling (.14) + Personal Values (.02) + Popularity Students (.27) + Service Community (.19) *The correlation coefficient is the Yule's Q from the high supervision conditional table formed when the super- vision indices are used as test variables. The zero order correlations are given in Appendix A. ~ 78 indices decreases the positive correlations between personal values and affluence, and increases the positive correlation between popularity with students and affluence. Perceived need for supervision pp; pp appears to have a uniform, gen- eral effect on the ranking of these two salary priorities. Both of these variables can be conceptualized as resulting in large degree from the individual's personality factors, (and thus the general effect which develops, while not pre- dicted, certainly does not distract from the discriminating value of the indices regarding the organization's profes- sional and bureaucratic dimensions. Personal factors, in this instance the desire, or non-desire, for personal popu- larity and/or the desire, or non-desire, to proselytize per- sonal values, are excluded from the ideal type of both pro- fessional and bureaucratic models. - Looking at the remaining significant relationships, the direction of change for academic_advising and job counsel- ‘ing (the two activities most likely associated with under- graduate education) is the same for legbur and perbur and while not at a significant level, opposite that for perpro. The only other activity which is significantly affected by a high score on each of the supervision indices is service in the community. Size The findings of this section concern the effect of size on salary priority considerations, controlling for the 79 supervision indices. Table 111.31, Correlation of size and salary priorities by supervision indices, presents the cor- relations. Zero Order Correlations: Size Only two activities, (a) service activities in the com- munity (-.48) and (b) service activities in the university .(-.60), are significantly related to size. Both of these activities would appear to be focused on issues of local concern. Of the relations not significantly supported, the two priority activities least negatively related to size were (a) publications (-.01) and (b) research (+.21). These relations are noted since they concern activities which appear most likely to be focused on issues of extra-local concern. If a pattern develops along these dimensions the . concepts of localite and cosmopolitan may have some explana- tory value for the conclusions chapter. First Order Correlations: Size Legpro pp test variable. Utilizing the criteria ad- vanced earlier, none of the correlations formed when legpro was used as the test variable are valid correlations to designate as significant or non-significant. Since none of the correlations meet the criteria for significance advanced earlier, neither support nor non-support can be claimed for any of the hypotheses Concerning size and priority activi- ties by legpro. 80 wOH>uOm cam >uHHMHsmom “mad “OHH n c huflCSEEoo moa>umw “mad n s consumes u : mmsHm> HMCOmHOm “mad u .CHHm> so: mum mcoflumH5OHsO 0 >HL:O:oOm:OO ocm w>flm 30Hwn mm3 mHHOO ones no use CH >Ocmoowuw HHOO pmuooaxc emu umzu mmumoapcfi >2 seasonesm scen>c : @sflammcsoo now .maa n c >uflmuo>asd “mad u c mcofluMOfiHndm Nona n : mucmpsum nufi3 .xaco >ufiuownm msflmfi>cm Owsoocom Bones mm mm mm mm mm mm mm mm HNH n as om.n mm.u nm.n $0.: me.- me.u mo.u >eem.u 00.: .suamsm>aca moa>umm Hm.u mm.n om.u me.u oo.u mm.. mm.: >cmm.: me.n suassesoo moa>umm mm.- mm. >cmo. >coa. em. No. «a. >cmm.u Hm. nosmmmmm em.. as. No. mm.u ms. sm.- mo. >cmv.u Ho.u mcoaumossnsa mo. em.n NH. 00.: No.1 ov.u mo.u >com.n Hm.s mucmtzum musssasdoa mm.u ma.u mm.n mo.u mm.u oo. mm.n >amH.u m~.a mossm> Hmcomuma No.1 OH. HO. mm.u mm.u «H. mm.: >com. ea.u masammcsoo non me.u mm. mo. sm.n Hm.- NH. mH.a >cva.u ma.- ocams>c< osswcmoa 0H as 0H as 0H «a 0H a: usauwm oumumm usumTH oumomq Hwouo OHTN moOfipsH soamw>ummsm Axnv muflm Axnv >uw>fiuom >uwuowum mOOflUCfl cowmw>ummsm an wmflueHOAs; >scacm pcc mNem no :oHucHOuuou HM.HHH OHQMB 81 Legbur pp pppp variable. According to the criteria advanced earlier, legbur influences the relationship between size and each of the priority activities. Among the high scores on legbur, the cOnditional corre- lations between size and academic advisement (+.12), job counseling (+.12), personal values (.00), and service com- munity (-.36) were significantly more positive than their zero correlations. Popularity with students (-.40), publi- cations (-.27), research (+.02), and service university (-.75) were significantly less positive than their zero order correlations. A high legbur score appears to cause a more positive increase in the correlation between size and academic advise- ment, job counseling, personality values, and service com- munity, while a high legbur score appears to cause a more negative increase in the correlation between popularity with students, publications, research, and service university. There does not, however, appear to be a discernible pattern at this time which can be pointed to in the data. Perpro pp test variable. According to the criteria advanced earlier, perpro influences the effect of size on salary priority activities in five out of eight activities. Only in personal values is the difference between con- ditional correlations more positive for the high scores on perpro. Among the high scores on perpro, the differences between the conditional correlations were significantly less positive for academic advisement, job counseling, popularity of students, and publications. 82 Perpro's influence appears to more constantly have an inverse effect on the correlations of size and priority activity than does legbur. With this exception, again there does not appear to be a discernible pattern of selective in- fluence. Perbur pp test variable. According to the criteria advanced earlier, legbur influences the relationship between size and salary priority activities in seven out of eight activities. Among the high scores on perbur, the differences be- tween the conditional correlations were significantly more positive for academic advisement, job counseling, personal- ity values, publications, service community, and service university. Of the significant differences, only pOpular- ity of students becomes more negative with a high score on perbur. Again, the only discernible pattern for perbur tends to be general in its influence, with the impact causing a more direct relationship between size and priority activities. Summary of Supervision Indices as Test Variables on Size and Priority Activities Due to the relatively small n (22) of respondents scor- ing high on legpro and consequent lack of statistical sig- nificance, legpro will not be included in the summary. Utilizing the same shorthand system established earlier, Table 111.32, Diagram of the effect of high supervision 83 scores on the significant relationships between salary priority activities and size, is presented. The effect of a high supervision score on personal values and popularity as explained earlier under affluence also holds for the relation between size and salary priori- ties. I Looking at the remaining significant relationships, the direction of change for academic advising and job counSeling' (the two activities most likely associated with undergraduate education) is the same for legbur and perbur, as noted earlier in the affluence section, but opposite for perpro. This apparent association is unexpected since it modifies the earlier position that similarity based on source of supervision would be more influential in affecting choices than similarity based on type of activity supervised. A similar effect on the correlations between size and salary priority activities appears more likely to be the result of indice commonality regarding type of activity rather than source of activity. Faculty Salaries The findings of this section concern the effect of faculty salary on salary priority considerations, control- ling for the supervision indices. Table 111.33, Correla- tion of faCulty salary and salary priorities by supervision indices, presents the correlations. 84 Table 111.32 Diagram of the effect of high supervision scores on the significant correlations be- tween salary priority activities and size 44‘ Supervision Direction of Index Correlation Change Activities ,+ Legbur + + Academic Advising (.12)* + Job Counseling (.12) + Personal Values (.00) + Popularity Students (-.40) + Publications (-.27) + Research - (.02) + Service Community (-.36) + Service University (-.75) + Perpro + + Academic Advising (-.27) + Job Counseling (-.38) + Personal Values (-.08) + Popularity Students (-.66) + Publications (-.25) + Perbur + + Academic Advising (.22) + Job Counseling (.10) + Personal Values (-.l3) + Popularity Students (-.27) + Publications (.11) + Service Community (-.29) + Service University {-.53) *The correlation coefficient is the Yule's Q from the high supervision conditional table formed when the super- vision indices are used as test variables. .The zero order correlations are given in Appendix A. .pHHm> uoc mum mcowumadoamo a saucwsvomcoo was m>Hm sonQ was mHHwo mnoe so 0:0 as >ocmsvmuw Hamo bwuoomxo msu umsu mmumofipcfl >s umwuomnsm ozan>c .maa u c xuflmum>flcs woa>uwm 6cm “baa u : >uflcsseoo wofl>som umHH n c :uummmos amHH u c mcofiumowandm Nona u c mucmcoum cuw3 >uwansmom “mad u : mwsHm> Hmcomumm “maa n c mafiaomcsoo non .maco wuwuowum wcfimfl>cm oafimomom Eon»a 85 mm mm mm mm no mm mm mm HNH n «a >ch. em. mm. mm. on. em. oo. >cmo. om. spamsm>aca mofl>smm mm. mm. mm. as. me. mm. vs. >cnm. so. suscsesoo ous>umm >cam. mo. >ch. mo. ms.- >cem. vo.- >coo.H so. nuummmmm >cNH. NH. >cvm. as. 00.: >cmm. 00.- >coo.a NH. maoaumofiansm mH.u mo. vH.u ow. mo. ma. Ho.u >cms. ma. mnemosum suflsmasmom as. Ho.- no. so. ms. ~H.- so. >cvs. mo. mmsam> Hmcomsmm >cmv. mo.- mo.- oe. em. ¢N.- om. >coo.u as. masammasoo non >com. sm.- mo.- mm. av. mm.u ms. >asm.- mo. mcflms>n¢ oasmcmoa 0H a: 0H a: 0H an 0H a: Hanumm oumsom Manama osmmmq noose Ohms mmowncH cOAmH>ummsm A>nv >HmHmm muHsomm Away >uw>fluo< hufluoflum mmofincw COHmfi>smmsm >n mmfiuHHOAnm mumamm cam >HMHMm wuasomw mo coauMHmuuoo mm.HHH magma 86 Zero Order Correlations: Faculty Salary Service in the university (.67) and service in the com- munity (.60) are both significantly related to higher sal- aried colleges/schools. The hypothesis of a direct relation between higher salaried colleges/schools and service activi— ties in the community is supported while the hypothesis of an inverse relation between higher salaried colleges/schools and service in the university is not supported. ‘ The zero order correlations for the higher salaried colleges/schools tend to indicate a relatively high local profile. Investigation of the colleges/schools finds.the schools of Medicine, Law, Business, and Education in the higher salaried category. Each of these schools has devel- oped a very pragmatic interpretation of their role as that of developing specialized vocational training. Each had operated, until the spring of 1974 University reorganiza- tion, its own placement bureau and/or program for their graduates. Each school has what might be termed a separate "Recognition Program“ in which community (state-wide) leaders and faculty members are honored for achievements in the ad- ‘vancement of the "profession" represented by the particular schools. Each is supported by a distinct cohesive constitu- ency in the state. The School of Education has been particularly active in promoting and encouraging their faculty to involve themselves in service to the University. A high priority by the Educa- tion and Business Schools is the probable reason for the 87 discrepancy between the priorities set by the higher sal- aried and more affluent colleges/schools. The zero order correlations tend to indicate that the higher salaried colleges/schools tend to place a greater emphasis on activities which are more easily visible to larger numbers of people in the local area. First Order Correlations: Faculty Salaries Legpro as test variable. Utilizing the above criteria, none of the correlations formed when legpro was used as the test variable are valid correlations to designate as signifi- cant or non-significant. Since none of the correlations meet the criteria for significance advanced earlier, neither support nor non-support can be claimed for any of the hypoth- eses concerning size and priority activities by legpro. Legbur as test variable. Six of the priority activi- ties, academic advising, job counseling, personal values, popularity with students, publications, research, service in the community and service in the university, meet the sta- tistical requirements for significance. A high score on legbur dampened the positive relations between high salaried colleges/schools and each of the above salary priorities. Support is thus found for the hypothe- , sized relations with the exception of service in the com- munity which was hypothesized to effect a more positive relation between higher salaried colleges/schools and belief in service in the community as a high salary priority. 88 Perpro as test variable. Only one priority activity, popularity with students, meets the statistical require- ments for significance. The effect of a high perpro score on the relation between higher salaried colleges/schools and the priority belief concerning pOpularity with students, was to accentuate the positive relationship between the latter two. This was not the predicted effect. Perbur as test variable. Only two of the priority activities, personal values and service in the community, meet the statistical requirements for significance. The effect of perbur is to dampen the apparent positive rela- tions between higher salaried colleges/schools and the pri- ority beliefs. The dampening effect is as predicted for personal values, but is opposite in effect to that predicted for service in the community. Summary of Supervision Indices as Test Variables on Faculty Salaries and Priority Activities Analysis of Table 111.34 does not indicate any clear pattern which follows from imposing the supervision indices as controls, although the earlier trend of type of activity commonality, i.e., legbur and perbur versus perpro, is noted if non-significant results are included. (The high perbur cells considered are valid cells.) Although the results of the discipline oriented activi- ties, publication and research, were non-valid, the direc- tion of change of the significant results of the undergradu- ate activities, academic advising and job counseling, are 89 Table 111.34 Diagram of the effect of high supervision scores on the significant correlations be- tween salary priority activities and faculty salary Supervision , Direction of . Index Correlation Change ’ Activity + Legbur + Academic Advising (-.39)* + Job Counseling (-.24) + Personal Values (-.12) + Service Community (.52) + Service University (.37) + Perpro + Academic Advising (.25) + Job Counseling (.40) + Popularity Students (.46) + Perbur + Personal Values (-.01) + Service University (.56) *The correlation coefficient is the Yule's Q from the high supervision conditional table formed when the super- vision indices are used as test variables. The zero order correlations are given in Appendix A. 9O interesting when compared with the structural variables investigated earlier. For the structural variables afflu- ence and size, introduction of legbur and perbur caused a positive change in the original relation and the introduc- tion of perpro caused a negative change. For the structural variable faculty salary, the direction of change is exactly opposite. At the present, a logical explanation of this finding escapes the present writer since the finding is so completely at odds with the original hypotheses advanced. Complexity The findings of this section concern the effect of com— plexity on salary priority considerations, controlling for the supervision indices. Table III.35, Correlation of com- plexity and salary priorities by supervision indices, pre- sents the correlations. Zero Order Correlations: Complexity Personal values (-.42), popularity with students (-.46), research (.31), service in the community (-.4l), and service in the university (-.64) are each significantly related to complexity. The only priorities with a positive relation to complexity are publications (.12, non-significant) and re- search (.31). The evidence tends to indicate that a higher degree of complexity is related to discipline oriented activ- ities while a low degree of complexity is more likely related to more locally oriented activities. 91 .pflam> so: mum mcoHumasuamo O >Husmsvxn:oo can o>Hm Bonn mos maamo once so wco cw wocmsvmsw Hamo bosommxm on» non» mmumowbca >s umfisomnsm mzau>z .mHH n : >uflmum>flcs oua>uwm was “mad n : xuflssEEoo mofl>umm “mad u : nosmmwmu «mad n c mCOHUMOAHnsQ “oma u c mucopsum cuss auflumasmom amHH u : mosam> assemsmm amHH u : mswaomssoo non .>Hco upwHOflum msflmw>cm owEmnmom Sosa. mm mm .mm mm so mm mm mm HNH u .: >sam.u oo.n oo.u Hp.| mm.u mv.n mo.u >:H>.u vo.u >usmsm>flco moe>umm «5.- mm.. om.- mv.a mm.- mm.- ov.- >cmv.u H¢.- susqsesoo mos>umm >csm.u me. am. mo.u ow. >cvm. mm. >:mo. Hm. noummmmm >c~H.- mm. mm. mm.. mm. «0.- ma. >cvm.u NH. mcosumossnsm mH.- em.- mm.- on.: mm.- nm.- Ne.n >ceo.u mq.- mucmnoum sueumssmoa H¢.| av.l @¢.n n~.n om.n o~.n mm.l >cma. mv.u mosam> Hmcomumm >cme.- mo.- mH.- >cmH.- mv.- oo.- m~.u >coo. mm.u massmmcsoo non >cs~.- mo.- mo.u >cso.- mH.- os.- «H.u >csv.n mH.- wasme>o< ossmomoa 0H a: 0H as 0H a; 0H as usnswm oumucm Hanmmq A - onmmmq umwuo OHwN .. mwowwcH scamfi>ummsm ~>nv muflxmamsoo A>nv >uw>auo¢ >uwuoflum mavens“ coamfl>hmmsm >n mmwuquflHm whoamm new >uflxmamaoo mo cowumHmuuou mm.HHH magma 92 Of the significant correlations, only the hypothesized direct relationship between complexity and service in the community was not supported. I First Order Correlations: Complexity Legpro as test variable. Utilizing the above criteria none of the correlations formed when legpro was used as the test variable are valid correlations to designate as signifi— cant or non-significant. Since none of the correlations meet the criteria for significance advanced earlier, neither support nor non-support can be claimed for any of the hypoth- eses concerning size and priority activities by legpro. Legbur as test variable. Six of the priority activi- ties, job counseling, personal values, popularity with stu- dents, publications, service to the community and service to the university, met the statistical requirements for signifi- cance. The effect of a high score on legbur was to dampen the negative effect of the organizational variable complexity on four of the priority activities. The exceptions were the activities popularity with students and publications, for which a high legbur score increased the apparent negative relation between them and complexity. The hypotheses advanced were supported, with the excep- tion of the hypothesis regarding service to the community. Perpro as test variable. Six of the priority activi- ties, personal values, popularity with students, research, 93 publications, service to the community and service to the university, met the statistical requirements for signifi- cance. I The effect of a high score on perpro was to accentuate the negative direction in five out of six of the relations. The lone exception was the relation between complexity and personal values, in which a high perpro score dampened the negative relation present in the non-controlled situation. Only two of the controlled situations support the hypotheses advanced. A high perpro score does accentuate the negative relation between complexity and popularity with students while it dampens the negative relation with service in the university. Perbur as test variable. Two of the priority activi- ties, popularity with students and service in the community, meet the statistical requirements for significance. The effect of a high score on perbur is to acCentuate the negative relation between complexity and popularity with students, while dampening the effect of the negative rela- tion between complexity and service in the community. Both of the controlled situations support the hypoth- eses advanced. A high perbur score accentuates the nega- tive relation between more complex colleges/schools and the belief that popularity with students should have a high sal— ary priority while a high perbur score dampens the negative relationship between more complex colleges/schools and the 94 belief that service to the community should have a high salary priority. Summary of Supervision Indices as Test Vari- ables on Complexity and Priority Activities Due to the relatively small n (22) of respondents scor- ing high on legpro and consequent lack of statistical sig- nificance, legpro will not be included in the summary. Utilizing the same shorthand system established earlier, Table 111.36, Diagram of the effect of high.supervision scores on the significant relationships between salary pri- ority activities and complexity, is presented. Analysis of Table 111.36 shows that regardless of the supervision index being investigated, a high score on the indices affects the relation between complexity and the pri- ority of popularity with students in a negative fashion. Service in the community is affected in a negative manner by legbur and perbur, but in a positive manner by perpro. Service in the community still continues to be a salary pri- .ority which is more likely affected by the distinction be4 tween type of activity supervised versus source of super— vision. Further patterns resulting from the imposition of the supervision indices as control variables are not apparent. Faculty Qualifications The findings of this section concern the effect of faculty qualifications on salary priority considerations, Table 111.36 95 Diagram of the effect of high supervision scores on the significant correlations be- tween salary priority activities and com- plexity Supervision Direction of Index Correlation Change Activity + Legbur + Job Counseling (-.OO)* + Personal Values (-.20) + Popularity Students (-.57) + Publications (-.04) + Service Community (-.28) + Service University (-.43) + Perpro + Personal Values (-.27) + Popularity Students (-.70) + Research (-.05) + Service Community (-.49) + Service University (-.71) + Perbur + Popularity Students {-.54) + Service Community (-.23) *The correlation coefficient is the Yule's Q from the high supervision conditional table formed when the super- vision indices are used as test variables. The zero order correlations are given in Appendix A. 96 controlling for the supervision indices. Table 111.37, Correlation of complexity and salary priorities by super- vision indices, presents the correlations. Zero Order Correlations: Faculty Qualifications Academic advising (.34), job counseling (.30), service in the community (.67) and service in the university (.61) are each significantly related to higher qualified colleges/ schools. Only for the relationship between higher qualified colleges/schools and the belief that service in the community should have high priority in salary considerations is the hypothesis supported. It is of interest that the only negative correlations, albeit on a non-significant level, are the discipline ori- ented activities of publication and research. The strong- hold of discipline research and publication does not appear to be in the more highly qualified schools/colleges but rather in the less qualified schools/colleges. ‘ This set of correlations could be interpreted as sup- porting the position that for the majority of university faculty the "peak" professional resocialization is mani- fested immediately after leaving the shelter of graduate school before the weight of routine faculty obligation dims the ideal. First Order Correlations: Faculty Qualifications Legpro as test variable. Utilizing the criteria ad- vanded earlier, none of the correlations formed when legpro 97 .cflam> so: our mCOwumHoono 0 maucwoommcoo new m>flw Scamp mmB mHHmo once so use ca >ocm9omum Hamo owuommxm on» was» moumoHUCfi >c umfiuomnsm mnBu>c .mHH u c wufimum>wcs o0fl>umm was “bad u c wuficsesoo moa>Hmm Nada u c noummmos “mad n c mcoflsmoflanom noma n c mucmcsum new: auwsmasmom “mad u c mwsHm> Hmcomsom “mad u c mcfiHmmcooo non .waco >ufiuoflum mcfima>om anwcmom Eouma mm mm mm mm so mm mm mm HNH u ..c >coo.a me. me. om. mo. he. so. me. so. susmuw>eca moa>umm >coo.H om. as. oo. mm. mm. vs. >cem. so. suscsesoo mow>umm >ams. HH.- >cms. >:HH.- >cs~.- >cmm. -.u >coo.H mo.- soummmmm >cmo.- mo.- >cmo. 00. «H.. ms. om.- >coo.H ao.- maoaumossnsa >cmo. os. 50.- mm. mo.- mm. mo.u >cms. Ha. mucmesum suaumasmom >cmv. ma.n ma. mm.. «a. HH.- so. >cva. mo. mosam> Hmcomumm >cmm. 6H. NH. >cmm. mm. «0.- we. >coo.- om. mcsflmmcsoo non >cqa. so. NH. >cms. mo. ms.- as. >csm.u um. mcamn>ea osEmemoa 0H a: 0H as 0H as 0H a: ssnuom oumsmm Manama .osmmmq aumcso . OHGN mwowvcH cmew>Hmmsm Axnv mcofiumowmwamsa mufl>fluo¢ seasons Ashe susuoanm moowocfi cofimfi>ummsm >9 mwfluflsowum >HMHMm ccm mGOHDMOHMfiHmsU >uasomw mo coHuMHmusou hm.HHH manna 98 is used as the test variable are valid correlations to designate as significant or non-significant. Since none of the correlations meet the criteria for significance advanced earlier, neither support nor non-support can be claimed for any of the hypotheses concerning size and priority activi- ties by legpro. Legbur as test variable. A high legbur score signifi- cantly dampens the positive relationship between higher qualified colleges/schools and individual beliefs about the priorities which should be given to academic advising, job counseling, personal values, popularity with students, pub- lications, service in the community and service in the uni- versity. The effect of legbur on the relationship between higher qualified colleges/schools and beliefs regarding priorities for academic advising, job counseling, personal values, pub- lications, and service in the community is opposite that predicted. The hypothesized effect of legbur on popularity with students and service in the university is supported. Perpro as test variable. Perpro has a significant ef- fect on the relationship between higher qualified colleges/ schools and beliefs regarding salary priorities. The nega- tive effect on the relationship involving the priorities personal values and service in the university supports the hypotheses advanced while the effects of perpro on the rela- tionships involving popularity with students and service in 99 the community, positive and negative effects respectively, were not as hypothesized. Perbur as test variable. Utilizing the above criteria none of the correlations formed when perbur was used as the test variable are valid correlations to designate as sig- nificant or non-significant. Since none of the correlations meet the criteria for significance advanced earlier, neither support nor non-support can be claimed for any of the hypoth- eses concerning size and priority activities by legpro. The non-valid statistical situation exists because so few faculty of the low faculty qualification colleges/schools placed low priority on any of the salary priority activities. Summary sf Supervision Indices as Test Variables as Faculty Qualifications and Priority Activities Due to the relatively small n (22) of respondents scor- ing high on legpro and consequent lack of statistical sig- nificance, legpro will not be included in the summary. Utilizing the same shorthand system established earlier, Table 111.38, Diagram of the effect of high supervision scores on the significant relationships between faculty qualifications and priority activities, is presented. Analysis of Table 111.38 shows that for the first time perbur joins legpro as a test variable whose application results in the developing of conditional cells which are non-valid for each priority activity. Legbur and perpro are thus the only two supervision indices which will be dealt with in this summary. 100 Table 111.38 Diagram of the effect of high supervision scores on the significant relationships between salary priority activities and faculty qualifications . ...—._—_—— -—.— -..._.- ____—..——_ ... ‘ Supervision Direction of Index Correlation Change Activity + Legbur + Academic Advising (-.18)* + Job Counseling (-.O4) + Personal Values (-.11) + Popularity Students (.22) + Publications (.13) + Service Community (.58) + Service University (.47) + Perpro + Personal Values (-.22) + Popularity Students (.26) + Service Community (.60) + Service University (.50) *The correlation coefficient is the Yule's Q from the high supervision conditional table formed when the super- vision indices are used as test variables. The zero order correlations are given in Appendix A. 101 The effect of legbur and perpro is the same for the significant relations they have in common. The imposition of either increases the positive relation between complexity and the priority popularity with students, while either ef- fects a negative change on the relation between complexity and the priorities personal values, service in the community and service in the university. The zero order correlations of complexity and salary priorities most nearly approaches the impressionistic reward system which four years of participatory observation have formed. The high priorities on student advising and service activities with low priority for research and publications certainly match the perceived system. The conclusions chapter will attempt to integrate what appears to be a set of diverse findings with few unifying relations into a more abstract, but more cohesive, set of ‘explanatory statements. 102 Footnotes Peter M. Blau. The Ogganization sf Academic Work. New York: John Wiley and Sons. 1973. p. 40. Blau uses this procedure for determining whether or not the independent variable has a significant effect on the dependent variable. Morris Rosenberg. The Logic sf Survey Analysis. New York: Basic Books. 1968. p. 156. CHAPTER IV SUMMARY AND CONCLUSIONS [I Introduction The main focus of this study is on the effects of selected organizational variables, supervision, and beliefs concerning the organization's salary priorities. The organizational variables dealt with are affluence, size, faculty salaries, complexity and faculty qualifica- tions. The supervision indices are developed around two dimensions: source of supervision, hierarchical or col— legial, and type of activity supervised, bureaucratic or professional. The supervision indices represent every pos- sible combination of source of supervision and type of activ- ity supervised. The salary priorities are academic advising of students, job counseling and career guidance of students, personal values and ethical standards, popularity with students, pub- lications, research, service activities in the community and service activities in the university. Sapcrvision and Structural Variables The data indicate that faculty do not see an overwhelm- ing need for closer supervision from any source, but if there is to be closer supervision, strong preference is given to collegial versus hierarchical supervision. The 103 104 faculty also are less opposed to supervision of bureaucratic activities than of professional activities. The data indicate that even in professional organiza- tions an increase in size or complexity results in greater perceived need for bureaucratic supervision and less per- ceived need for collegial control. These findings support Dyke's conclusions that increased size and complexity leads to increased need for bureaucratic supervision. The data do not appear to support Blau's conclusion that while "universities and colleges have administrative structures that are similar to those of other bureaucracies . . . the same features do not have the same significance."1 Larger schools/colleges appear more likely to generate hier- archical structures than collegial structures. The univer- sity system thus faces a dilemma in that larger units which help create greater opportunity for specialization also cre- ate greater pressure for bureaucratic controls which create pressures for routinization standards which reduce the faculty member's autonomy and allowable discretion in pursuit of his specialty. An increase in the structural dimensions of the organization which generate the opportunities for‘ specialization also generate barriers to effective special- ization. A professional organization which desires to retain a collegial atmosphere must weigh the advantages and disadvantages of growth and complexity since increased size and increased complexity appear a mixed blessing regarding these results. 105 Few organizations consciously resist expansion because of any or all of the excellent reasons Starbuck advances for growth;2 consequently, if the deleterious effects of expan- sion on professional dimensions is to be avoided, some moderating variables need to be considered simultaneously with size and complexity. The three possible moderating variables studied were (1) affluence, (2) faculty salaries and (3) faculty qualifications. Faculty qualifications and faculty salaries both in- crease the probability that collegial supervision will be favored over hierarchical supervision. The effect of expan- sion on supervision would appear capable of modification if the expansion which occurs involves highly qualified faculty who warrant high salaries. This study does not have hard data on the hiring policies in expanding colleges/schools, but this researcher’s nonsystematic observation of expansion practice in higher education indicates that the majority of expansion occurs on the junior faculty levels involving ‘ master's degree or A.B.D. personnel who are less expensive to acquire. If a collegial atmosphere is worth keeping, resistence to, or at least more caution regarding, expansion based on junior faculty appears warranted. High affluence lessens the probability that either increased collegial or hierarchical supervision would be emphasized. Expansion in an affluent unit also appears less likely to result in an increase in hierarchical pressure 106 since the affluent unit is more likely able to hire higher qualified and higher salaried faculty as additional staff. In sum, the university as an example of a collegial system appears likely to weaken its collegial supervision system if its expansion takes place within either the con~ fines of economic contingencies or high affluence. If a person feels compelled to find a silver lining, the effect of the present retrenchment in institutions of higher educa- tion may result in the revitalization of the collegial sys- tem among the less affluent colleges. The following sections summarize the findings regarding the structural variables. Affluence Blau states "the antecedent conditions that exert most influence . . . can all be conceptualized as resources . . . human and institutional as well as financial assets."3 Af- fluence is a measure of resources, and its range of effects on the colleges/schools of this study indicates its impor- tance. Given the causal sequence developed for the multiple regression procedure, affluence has a direct significant relation with the presence of legpro, legbur, and perbur. The direct effect on legbur and perbur is negative, while the direct effect on legpro and perpro (non-significant) is positive. The evidence tends to indicate that more affluent col- leges/schools do not perceive a need for closer supervision 107 regardless of whether the control be hierarchical or col- legial, but if closer supervision exists, it should be directed at professional activities. Looking at the effect on salary priorities, affluence has a direct effect on student advising (positive), popu- larity with students (negative), research (positive) and service in the university (positive). The effect of introducing attitudes toward supervision as a conditional variable on the structural variables and beliefs about salary priorities was mixed. The effect of using legpro as a conditional variable was not ascertainable for this section or any of the other sections of this study since the number of faculty who favored legislative supervision of professional activities was too few to be validly analyzed by the statistics utilized. This fact does, however, illustrate that the faculty still considers regulation of professional activities almost exclu- sively their domain. The effect of the faculty believing there should be . closer supervision of bureaucratic activity by the legisé lature is to significantly decrease the positive relation between more affluent colleges and beliefs about the salary priority which should be given to academic advising, job counseling, personal values, service in the community and service in the university while accentuating the positive relation between affluent colleges/schools and popularity with students. A low legbur Score increases to a significant 108 degree the probability that individuals in high affluence colleges/schools will oppose publication and research as salary priorities. An attitude which fosters rejection of hierarchical control over bureaucratic activities thus re- sults in a greater likelihood that local-oriented activities versus discipline-oriented activities will receive higher priorities in more affluent colleges/schools than in less affluence schools/colleges. It is a plausible interpreta- tion that those not opposing closer legislative supervision of bureaucratic activities are neglecting obligations in this area in order to concentrate more attention on reSearch and publication. It is also plausible that those not oppos- ing closer legislative supervision have a better grasp of the detrimental consequences which accompany Parkinson's Law. The introduction of perpro on affluence displays an un- usual effect in that a high perpro score increases the prob- ability that high priority should not be given for personal values and ethical standards, while the same combination of affluence and perpro results in increasing the probability that popularity with students should be used as a salary priority. Both priorities, personal values and popularity with students, are not representative of elements included in ideal types of either the bureaucratic or professional model. A high perpro score in an affluent college/school is also more likely to result in lower priority for service in the community. 109 The effect ofia high perbur score is to cause a less positive direct relation between affluence and most of the salary priorities, the lone exception being the priority popularity with students. One of the most interesting effects is noted when perbur and legbur are compared. Perbur and legbur, whuue commonality is based on type of activity supervised, have the same negative effect on academic advising and job coun- seling (most likely undergraduate student Oriented) while they have the opposite effects on research and publication (most likely discipline oriented). When a perceived need for closer supervision from either a hierarchical or a col- legial source is introduced, the effect is to reduce the positive correlations for the local oriented activities, but the effect is selective for the discipline oriented activi- ties. A low perceived need for closer hierarchical super- vision of research and publication activities has the same effect on the relation between affluence and discipline oriented activities as a high perceived need for collegial control has for these activities. In more affluent colleges/ schools a perceived need for supervision is more selective regarding source of supervision for professional oriented activities than for bureaucratic oriented activities. Attitudes toward source of supervision and type of activity supervised do constitute discriminating conditions which influence the nature of the relationship between the D organizational variableaffluence and individual beliefs 110 regarding the salary priorities which selected activities should receive. Affluent units are most likely to resist closer super- vision from any quarter than are less affluent units, and affluent units are more likely to promote both local and professional activities. A paradox exists here for both the advocate of strict [5‘ organizational supervision and the individual who believes a zero sum effort exists which must be divided between teach- ing (local) or research/publication (professional). In- creased affluence appears to lead to increaSed emphasis i given to both local and professional activities. Although it may be disquieting to the bureaucratic oriented adminis- trator and more expensive for the parent organization, it is greater affluence for the subunit which leads to an ef- fort to meet all three charges given to higher education-- instruction, research and service. Blau's conclusion re- garding the importance of affluence as a major factor is supported. Affluence is the only structural variable in this re- search which has a direct relationship with both local and professional priorities. The most obvious conclusion to be drawn from this section is that the development of a "uni- versity" must be attended to $2 2222 rather than in a patch work fashion, and, as in most instances, one dimensional solutions, e.g., paying higher salaries or hiring more highly qualified staff, do not lead to a balanced university. 111 Size Blau writes, "The size of an academic institution un— questionably has a predominant effect on its character. . . ."4 Size is operationalized for this study by the num- ber of full-time equivalent faculty members in each college/ school. Although the present researcher still intuitively strongly believes in the importance of the effect of size on an organization, this study does not indicate empirical sup- port for this position. Utilizing multiple regression tech- niques, size did not have a significant direct effect on any of the supervision indices. The only salary priority activity, popularity with stu- dents, on which size has a significant direct effect is the priority most susceptible to influence, being significantly affected by four out of five of the structural variables. The placement of size in the causal sequence was not detri- mental to its probability of being identified as a signifi- cant factor since more instances of significant direct ef- fects occurred after the position occupied by size than occurred before the position occupied by size. Even in the Yule's Q analysis, size has a significant effect on only two activities, service in the community and service in the uni- versity. The introduction of the supervision indices as control variables on size did produce the greatest number of significant differences. A number of reasons may be advanced as to why size does not appear to be the important variable it was originally 112 hypothesized to be. The most acceptable reason still ap- pears to be the confounding effect of the other structural variables. The units of analysis are not identical but rather distinctively different. One of the difficulties of . this situation is that traditionally ”productive" variables, i.e., producers cf significant results, may not appear as singularly important as they appear in studies investigating similar organizations. This "difficulty" also represents an advance since, as Perrow suggests, an approach which recog- nizes that particular configurations of variables and per- spectives (not theories) have valid and nonvalid applica- tions dependent of the type of organization being investi- gated is superior to an approach which, if by no other means than Eiéir forces results into "the" universal model or "the" universal theory.5 The effect of the faculty of larger schools/colleges believing there should be closer supervision of bureaucratic activities by the legislature is to significantly increase the probability that the activities, academic advising, job counseling, personal values and service in the community, will be given higher salary priority while the effect of the faculty in larger schools/colleges not believing there should be closer supervision of bureaucratic activities by the legis- lature is to significantly increase the probability that the activities popularity with students, publications, research and service in the university will be viewed as higher pri- orities for salary considerations. 113 The effect of the introduction of legbur as a control variable on size is similar to the effect legbur has as a control on affluence. In the larger schools/colleges an attitude which fosters rejection of hierarchical control over bureaucratic activities results in a greater likelihood that discipline-related activities will receive higher pri- orities for salary than will local activities. The probable interpretations presented for affluence are also applicable here. Faculty of the larger colleges/schools who believe in closer supervision of professional activities by faculty peers caused a positive increase in the probability that personal values be considered a higher faculty salary pri- ority. Belief in closer supervision of professional activi- ties by faculty peers reduced the relation between size and the priorities academic advising, job counseling, popularity with students and publications. The above findings are surprising in that peer super- vision is expected to be the preferred method of supervision in a professional organization such as a university, but the Vfindings indicate that perception of the need for closer supervision from this source causes a reduced priority for both local and discipline oriented activities. One assumption implicit in the data collection instru- ment, which should be made explicit, is that the question does ask "should the faculty . . . be under closer super- vision," thereby placing both those who feel the present 114 actual amount of supervision is satisfactory and those who feel it excessive into the same category--plus we have no measure of the actual level of supervision. Given this situation, it is possible that the individuals scoring high on the supervision indices are non-discriminatory seekers of supervision, thereby not predictable in terms of a logical selection of priorities. This possibility appears small if F‘ _ -O the patterns of responses for the other structural variables and supervision indices are explored. What appears to remain as a vocation linked priority for salary considerations among the larger college/school ; faculty advocating closer collegial control is some as yet undefined intangible. A not new criticism of the collegial system. . The effect of the faculty of larger colleges/schools believing there should be closer faculty peer supervision over bureaucratic activities is to increase the probability that every activity but popularity with students be given consideration in salary determination. This finding, in conjunction with those where legbur and perpro were used as control variables, is a surprise. The type of activity, bureaucratic, appears to exert a greater influence than the source of supervision on outcomes. The results of control- ling for perbur approximates, with exceptions to be noted, more nearly those where legbur is controlled than where perpro is controlled. 115 The opposite effects of controlling for legbur and perbur occur on the activities of publications, research and service in the university, where the effect of legbur is negative and the effect of perbur is positive. The reverse effect on publication and research is in the direc— tion expected since collegial supervision is more likely identified with discipline oriented activities. The combination of the faculty in larger colleges/ schools perceiving a need for closer collegial supervision of bureaucratic activities is the only combination which improves the probability that both local oriented and disci- ' pline oriented activities be given higher priorities when salaries are considered. This study indicates that size, pas ss, is neither the beneficial nor deleterious factor proposed in many other studies. Although the particular correlations were not significant in this study, the finding comes closest to Blau's conclusions: the direction indicates size increases the priority given to research while decreasing the pri— orities for other educational activities. The introduction of the attitudinal variables as test variables on the relation between the structural variables and the activities considered for faculty salaries indicate the importance of considering both structural and attitudinal dimensions when investigating professional organizations. The size related association most affected by the introduc- tion of structural variables is the relation between size 116 and research controlled by perbur, faculty peer supervision of bureaucratic activities. The zero order correlation be- tween size and research is a non-significant .21, but the condition of a high score on perbur gives a size related significant correlation of .36, while the condition of a low perbur results in a significant-339 correlation. Any class of variables which can result in conditional relations of the above magnitude cannot validly be ignored if the in- tention of the researcher is to understand the organization and its operation. The findings related to size indicate the validity of Hall's approach of using both structural and attitudinal dimensions. Faculty Salaries Blau writes, "Money buys a lot of things, and it helps in getting others that cannot be bought for money."6 Faculty salaries are a major area where money helps. Using multiple regression, faculty salaries are found to have significant direct effects on legpro (negative) and legbur (positive). The evidence tends to indicate that the higher salaried colleges/schools are more likely to favor closer hierarchical supervision of bureaucratic activities while being more likely to oppose hierarchical supervision of professional activities. Faculty salaries is one of the most productive struc- tural variables if measured in terms of the number of times it directly affects salary priorities. Faculty salaries 117 has a direct effect on the priorities given to student advising (negative), job counseling (negative), popularity with students (positive), publications (positive), research (positive), service in the community (positive) and service in the university (negative). The data support Blau's find- ing that higher salaries are positively associated to disci- pline oriented activities. With the exception of the priority, popularity with students, the direct effect of high faculty salaries is to increase the probability that colleges/schools will promote discipline oriented versus local oriented activities for higher salary priority. The effect of introducing attitudes toward supervision as a control variable on the structural variables and be- liefs about salary priority was surprisingly uniform. Per- ception of the need for closer hierarchical supervision of bureaucratic activities reduced the probability that higher salaried celleges/schools would place priority on academic advising, job counseling, personal values, service in the community or service in the university. Perceived need for closer collegial supervision of pro- fessional activities increased the probability that higher salary colleges/schools would put more priority on academic advising, job counseling, and popularity with students. Perceived need for closer collegial supervision of bureau- cratic activities lessens the probability that either per- sonal values or services to the university would receive 118 more priority in salary considerations. The imposition of supervision indices as controls appears to affect the local oriented priorities to a greater extent than the discipline activities. Again the evidence indicates that type of activ- ity versus source of supervision exerts a greater influence under these conditions. Complexity The results of complexity in the multiple regression analysis were not very productive in terms of illustrating significant relations. Complexity did not have a signifi- cant effect on any of the supervision indices, and its only direct effects (negative) are on the priorities personal values and popularity with students--the two activities least likely associated with either the bureaucratic or professional model. Complexity, similar to size, does not play as large a part as had been predicted. Looking at the effect of com- plexity on salary priorities, the zero order correlations indicate more complex colleges/schools are less likely to put higher priorities on personal values, popularity with students, service in the community and service in the uni- versity while they are more likely to put higher priority on research. The more complex schools/colleges appear less likely to give priority to local activities while they are more likely to promote discipline-oriented activities. The introduction of legbur as a control variable has a positive effect on the relation between complexity and the 119 priorities job counseling, personal values, service in the community and service in the university while it has a nega- tive effect on complexity's relation with publications. In more complex colleges/schools perceived need for closer hierarchical supervision of bureaucratic activities indi- cates a higher priority for local-oriented versus discipline— oriented activities. The effect of the faculty of more complex colleges/ ‘schools believing there should be closer peer supervision of professional activities is to lessen the probability that popularity with students, publications, research, service in the community and service in the university will receive higher priority, while increasing the probability that per- sonal values will be given higher priority. The effect of the perceived need for closer collegial supervision does not appear to affect the original complexity-priority relation in any patterned manner. Perceived need for closer collegial supervision of bureaucratic activities does not appear an influential con- trol for the relation between complexity and salary priori- ties. The presence of perbur in more complex colleges/ schools decreases the probability that popularity with stu- dents will receive support as a salary priority While in- creasing the probability that service in the community will receive support as a salary priority. Size and complexity are often dealt with together. In this study complexity, per ss, appears to be a stronger and 120 more consistent influence than size. Complexity is signifi- cantly correlated with five of the eight activities while size is significantly correlated with only two of the eight activities. In terms of being affected by the supervision indices as test variables significant conditional relations appeared twenty times in the size relations while only thir- teen times in the complexity relations. These results support Hall's schema of both attitudinal and structural dimensions as legitimate areas of investiga- tion while also supporting Anderson and Waskov's conclusion that complexity rather than sheer size is a more valid ex- planatory variable in organizational analysis. Faculty Qualifications The measured impact of faculty qualifications utilizing multiple regression techniques is slight. There is no sig- nificant effect on the supervision indices while student ad- vising (positive) and job counseling (positive) are the only salary priorities which are significantly affected. The evidence indicates that higher faculty qualifications tend to promote activities more likely directed at undergraduate students as salary priority. The Yule's Q zero order correlations indicate that more highly qualified colleges/schools tend to promote aca- demic advising, job counseling, service in the community and service in the university as activities which should be given priority when considering faculty salaries. The evidence indicates that higher faculty qualifications increase.the 121 probability that local oriented activities will receive pri- ority over discipline oriented activities. The use of the supervision indices as control variables on the faculty qualification and activity relationship re- sulted in only two categories, hierarchical supervision of bureaucratic activities and collegial supervision of profes- sional activities, showing significant changes. The effect of both supervision indices on the zero relationships is the same on the significant activities in common, i.e., there were positive changes in popularity with students and negative changes in personal values, service in the community, and service in the university for legbur and perpro. The effect of perceiving a need for closer hierarchical supervision of bureaucratic activities appears to cause a 2 reduction in priority for local oriented activities while increasing the probability that higher priority will be given to research oriented activities. Dichotomization of the colleges/schools according to faculty qualifications appears to be the structural.cate- gorization schema least likely to be influenced by atti- tudinal dimensions. The faculty qualification to activity zero order cor- relations were unexpected since they indicate that the higher qualified faculty are more likely to favor student oriented over research/publication activities as priorities for salary consideration. The majority of doctorates in 122 the university are, at least nominally, research oriented doctorates, but this does not appear to be the activity most emphasized in the more highly qualified colleges. The find— ing is unexpected because of the above reason of the type of faculty training and also because of the popular belief that the more qualified and prestigious units are professionally oriented and consequently less dependent and less responsive to local activities. It is interesting that the combination of the struc- tural variable high qualified faculty and the attitudinal variable high legbur is most likely to result in lower pri- ority for local activities and higher priority for profes- sional activities. A paradox appears since it seems that the perceived need for closer supervision of bureaucratic activities from an external source is associated with a greater likelihood that professional activities be given greater priority than local activities. The posSible reason for this unexpected finding may be‘ that the university is primarily a teaching institution with less emphasis given to research and service. If this find- ing is not due to some uniqueness in the data collection setting, some major revisions will have to be made in terms of the relations presently given in the literature. The findings certainly do not indicate any inherent conflict between professional and bureaucratic dimensions. 123 Salagnyriority Activities Conclusions The data thus support Blau's findings in some areas while not supporting them in others. The importance of af- fluence is supported while the importance of size is not supported. Blau states "The size of an academic institution unquestionably has a predominant effect on its character,'7 but the only significant effect found in these data was for the variable popularity with students, and even the non- significant effects did not appear to point in any specific direction of influence. The effects among and between the undergraduate activities, discipline activities and service activities could be no more random than if designed to ap- pear random. Blau states that "Investing financial assets in high faculty salaries bears interest, as it were."8 The data support Blau in terms of discipline oriented activities such as research and publication: however, for undergraduate ori- ented activities such as student advising and job counseling the interest Blau refers to appears paid rather than received. Higher salaries are negatively related to these latter cate- gories. This finding is interpreted as supporting Blau's con- clusions since he finds high salaries lead to an impersonal atmosphere which has detrimental effects for undergraduates. The data does not give significant direct effects to support or reject Blau's conclusions on complexity but the pattern of non-significant effects supports Blau's conclusion that a high degree of complexity allows greater specialization 124 which is conducive to the promotion of research and detri- mental to undergraduate teaching. The data do not support Blau's finding that higher qualified faculty's "prevailing orientations . . . tend to be toward scholarly research . . . which curtails their com— mitment to the local institution."9 The significant direct effects are positive for student advising and job counseling, while the non-significant effects are negative for research and publication while positive for all other local activi- ties. As with the size related finding it should be noted that the data collection site is a teaching oriented insti- tution which may affect the finding due to selective recruit- ment of faculty oriented to teaching rather than research. General Conclusions The conclusions of this study have been stated in study specific terms up to this point of the study. Since one of the main functions of research is to produce general state- ments whose applicability exceeds the specific parameters of the phenomenon investigated, the following will consist of statements on higher levels of abstraction. It must be cautioned that while support for these statements is indi- . cated from the findings of this study, they are considered more as hypotheses to be further tested rather than defini- tive statements of sociological "law," if such indeed exists. 1. When considering perceived need for supervision, supervision from a source internal to the organi- zation is preferred to supervision from a source external to the organization. 125 The relationship above is obviously not a new addition to organizational theory, being a specific restatement of the more established "out-group” and "in-group" preferences. 2. When considering perceived need for supervision, supervision of activities which can be unambigu- ously operationalized is preferred to supervision of activities which can not be unambiguously oper- ationalized. Unambiguously defined obligations allow the setting of standards of performance, albeit minimum standards, by which personnel can evaluate their own performance. The individual thus is not as liable to the possible capricious, personal whims of a supervisor, since performance is organizationally defined, as is the case in most bureaucracies. Ambiguously operationalized activities, however, definitionally involve phenomena which are liable to multiple interpretation, and- even if an interpretation is agreed on, which of multiple treatments is to be applied varies, particularly as viewed from a later time perspective. A job whose performance con- tains uncertainties which must be handled by personal judg- ment decisions is not the type of job which the job holder wants closer supervised. It should be noted that closer 'supervision does not necessarily equate with more consulta- tion or information exchanged. 3. When considering perceived needfor supervision, source of supervision is more influential than activity supervised. It is logical to assume that expert and knowledgeable supervision of any activity would not be opposed to the extent that non-expert and non-knowledgeable supervision of 126 any activity would be opposed. The Board of Regents is not considered expert by many in either the disciplines taught at the university or the machinations of the university bureaucracy. 4. The effect of organizational resources has more influence on perceived need for supervision than does the effect of organizational structure. The literature has stressed size and/or complexity as major, if not E22 major, variables which determine organiza- tion perception. While the results of this study indicate the importance of size and complexity, the results also indicate that the effect of resources is more influential. It is logical to expect this result since effective versus efficiency criteria are more likely applied in a resource wealthy versus a resource poor organization. Personnel working by efficiency versus effective criteria in a common purpose organization are more likely concerned with maxi- mizing performance on highly visible tasks important to the organization, in anticipation of more abundant reward. In an efficiency oriented organization pressure is generated for all to perform at visible tasks, while in effective oriented organizations less pressure is generated as long as the "work" is accomplished, hence less supervision is deemed necessary. Involvement in long range, low visibility, or ”nonsense" activity does not threaten the main task of the wealthy organization since it is using its "surplus” to fund the activities. Neither size nor complexity guarantees 127 surplus, although the presence of resources is usually a prior reciprocating condition to both size and complexity. Bennis writes that one of the changes which will occur in the future is accentuating the conflict resulting from different goals in the same organization, the reason being "professionals . . . who tend to identify as much with the supra-goals of their profession as with those of their im- mediate employer."10 Blau states, "The coexistence of some conditions that have opposite and some that have parallel effects on the two functions is the source of the teaching- research dilemma because it shows that the conflict between the two cannot be resolved without depriving either of im- portant benefits by simply separating them into different institutions . . . ."11 Two categories earlier identified as (1) most likely undergraduate oriented and (2) most likely discipline ori- ented allow conclusions to be advanced along this line. 5. General organizational factors promote different goals by promoting differential reward priorities. The multiple regression sections show that increases in complexity or faculty salaries tend to depreciate undergrad- uate oriented priorities and service activities while appre- ciating discipline oriented priorities. Increased faculty qualifications appreciate undergraduate oriented activities and service activities while depreciating discipline orié ented priorities. Increased affluence tends to appreciate all three undergraduate, discipline oriented and service priorities, while size appears to have little effect on any 128 of the priorities. The organizational variable affluence appears to be the only variable which does not present a dilemma, but rather a problem, a problem which can be solved without having to choose between opposite purpose results. Affluence tends to be positively related to higher priorities for undergraduate activities, discipline oriented activities and service activ- ities. The creation of more affluent units is the only orga- nizational variable in this study which has a direct positive effect on all the activities investigated. Future Research In this section suggestions are advanced for further research in this area. The findings in this study did not support a number of findings, particularly Blau's findings in his study of higher education.12 Contrary to Blau, more affluence, and more highly qualified colleges/schools placed higher priority on undergraduate activities and lower priorities on research and publication than did their less prosperous and less qualified counterparts. Size and complexity have been used--validly and successfully--in organization study, but basic parameters have usually not been explored beyond the dichotomization larger/smaller and more/less. The effects of both size and complexity may vary within absolute ranges (plateaus) in accord with some mathematical function more complicated than a simple linear or parabolic function. 129 Some variable, or set of variables, has caused organization X to be larger and/or more complex than organization Y. Perhaps the effect later attributed to size or complexity is merely an accelerated result of the original causal vari- ables being impacted through the modified organization. The tracing Of the development of the original variables, while difficult, appears possible using multiple regression and/or path analysis techniques. The tracing of the original vari- ables over a sample which varies through a wide range of Isize and complexity would allow detection of either plateaus of effect, if they exist, or the continuing causal variables, if they exist. Attitudes concerning supervision relative to both source of supervision and type of activity supervised mediate in a significant fashion the effect of structural variables on faculty salary priorities. The present study does not inform us as to whether or not the presence of a specific attitude toward supervision is the result of selective recruitment into the organization or the result of experience in the Organization. More simply, is the effect because of social psychological or sociological variables? Explanation and/or modification of an organiza- tion certainly takes a different tack dependent on the origination of the mediating variable. . The findings do provide support for the position that the consideration of both the factors does provide for more accurate determination of which salary priorities will be 130 advanced. If the conjunction of the two sets of variables affects an area as important as reward priorities, there is a strong probability that other organization processes are also affected. The composition of the supervision indices is explora- tory and at the time the questionnaire was developed, this particular use was not anticipated. Original intentions were directed at source of supervision only, and type of activity was developed later. The combinations originated here appear capable of being developed into useful measures. Development would be assisted by measures of actual styles of supervision and preferential choices relative to the actual situation. Attention to structural variables and criteria of evaluation, e.g., efficiency or effective, relative to the development of supervision also appear to be fruitful areas for further investigation. 10. ll. 12. 131 Footnotes Peter M. Blau. The Organization of Academic Work. New York: John Wiley and Sons. 1973. p. 279. William H. Starbuck. "Organizational Growth and Devel- Opment” in James G. March (ed.), Handbook of Organiza- tions. Chicago: Rand McNally. 1965. pp. 451-533. Blau, 9p. cit., p. 251. Ibid., p. 252. Charles Perrow. Organizational Analysis: 5 Sociologi- cal View. Belmont, Calif.: Wadsworth. 1970. p. ix. Blau, as. BEE-r p. 237. £2iQ°I p. 252. £2i§-v p. 251. 121a,, p. 255. Ibid. Ibid. Ibid. APPENDIX A Data obtained from documents can present a number of difficulties for the researcher. This study attempted to minimize the biases introduced by utilizing this data source. The major sources of potential bias explicitly dealt with were: (a) (b) (c) (d) authenticity--the documents are official reports of the organization being studied; credibility--the documents are public information and as such are examined from above by the Board of Regents and the legislature in the capacity of funding agencies, and from below by the colleges/ schools and the departments as funding recipients; this potential scrutiny makes deliberate misrepre- sentation less likely; statistical interpretation--the information taken from the report was raw data with all statistical techniques applied by the present researcher; verbal interpretation—~although the danger exists that the present researcher may have misinterpreted the data categories, this danger appears slight since the information gleaned was all quantitative, consequently lowering the likelihood that misinter- pretation occurred through connotative confusion over terms. 132 MICHIGAN STATE UNIV. LIBRARIES ll!WI“HIM""WWHllllmllfl"WWI 31293104632264