ANALYTICAL SKILLS AND THE UNIVERSITY TRAINING OF EDUCATIONAL ADMINISTRATORS (1)

1 JOURNAL OF EDUCATIONAL ADMINISTRATION – May 1974 ANALYTICAL SKILLS AND THE UNIVERSITY TRAINING OF EDUCATIONAL ADMINISTRATORS (1) JAMES G MARCH To use universities effectively in the development of critical administrative skills, we need to attend to the problems posed by the context of decline and by the nature of educational organizations, managerial work, and university comparative advantage Five analytical skills are identified as satisfying such criteria: The skills of analyzing expertise, coalitions, ambiguity, time, and information If we can improve the capabilities of educational administration to deal with experts, to solve problems in the absence of goals, to treat data from a decision perspective, to manage conflict and coalitions, and to allocate time, it will be an impressive set of contributions appropriate to the traditions of academe INTRODUCTION During the next ten years many university faculties in educational administration will revise their curricula The process is ordinary and continuous; the changes are incremental The natural drift of institutional change assures a sequence of enthusiasms It does not, however, assure attention either to the nature of leadership roles in contemporary educational organizations or to the distinctive competences of universities and their limitations I believe we can better The contemporary problems of education are familiar They inhabit the professional journals, the daily press, and the bookstands at airports Our schools are too rigid, too casual, not innovative enough, too inclined to adopt every new panacea, not sufficiently concerned with the problems of poor people, too likely to substitute social standards for intellectual standards, not adequately attending to the problems of multiculturalism, not integrating diverse cultures into a common social and ethical system They are failing to educate many children; they are failing to prepare students for jobs; they are failing to keep drugs, crime, and venereal disease out of the school The fist is important; it is long; and it is inconsistent The problems in education are the business of educational administration Insofar as education is failing, the educational administrator is subject to indictment; and schools of education are proper codefendants As a result, educational administration has become a focus for complaints identifying general problems of educational administration, however, is a dubious undertaking The management of learning involves widely varying activities by many different individuals and groups in numerous social institutions The activities range from defending budgets before legislatures to scheduling buses; from negotiating union contracts to administer-indiscipline for adolescents; from psychotherapy to plumbing The positions vary from assistant principal to superintendent; from budget analyst to project officer; from county school officer to Minister of Education; from assistant to the dean to college president The institutions vary in size, in resources, and in complexity Ordinary professional conservatism suggests caution in identifying generic needs for educational administration It is not a single field; different jobs require different skills Although I think we can say a few things that apply to many relatively high level positions in relatively complex educational organizations in the United States, I would be cautious about extending them to other positions, or to less complex institutions, or to other institutional or cultural contexts We need also to be wary of the ideology of administration It is an ideology that appears to be widely shared among modem administrators in business, public affairs, and education It is linked to a rich collection of cultural and philosophical traditions that are important to us It is seductive and often misleading We can characterize that ideology by the following set of beliefs: If there is a problem there is a solution If there is a solution it can be discovered by analysis, and implemented by skill in interpersonal relations or organizational design The solution to a problem requires the identification of underlying causes and the discovery and implementation of solutions are duties of the administrator If a problem persists, it is due to inadequacy in an administrator's will, perception of problems, analysis, skill with people, or knowledge of organizations Inadequacies in an administrator can be corrected through proper administrative training Such beliefs form a familiar basis for administration, administrative theory, and administrative training The beliefs are attractive They comprise a faith of hope They encourage persistence in the face of adversity; they encourage commitment to continuing administrative development Nevertheless, there are conspicuous problems with parts of the faith The existence of a problem does not necessarily imply the existence of a solution Analysis is only one of several techniques for the discovery of solutions There are many solutions that cannot be implemented with the best of human relations skills or organizational design Conflict is sometimes not susceptible to 11 resolution" The complexity of implementation may considerably outrun either the complexity of the solution or the capabilities of participants The removal of root causes may be one of the least likely ways to solve a problem Problems persist for many reasons that have nothing to with the administrator Many administrative inadequacies are quite immune to training interventions The implications are conservative We should have only modest hope that a change in administrator training or behaviour will ameliorate the problems of modern education Those problems respond to many factors; only a small proportion of them are amenable to administrative control The argument suggests a style that is pessimistic about great drama, but ambitious about making marginal improvements that are perceptible It also suggests that we should not be over-seduced by a recitation of problems or needs The improvement of training for education administration must build on our knowledge of the kinds of organizations we have, the kinds of work that managers do, and the kinds of capabilities that universities have Our analysis includes constraints and resources, as well as needs THE CONTEXT OF DECLINE The context of educational administration constrains the prospects for dramatic administrative success That context has too many dimensions for me to enumerate here It includes aspects of the political, technological, social, economic, and international worlds I will limit mention to three elements of the context that seem to me particularly important to educational administration in the United States over the next decade: (1) Education is a declining industry (2) Social expectations with respect to education have changed (3) Administrative careers and professional esteem are subject to doubt Each of these aspects of the educational context has consequences for administrative roles Education as a declining industry Education is a declining industry The rate of growth in enrollments has been reduced markedly; in many settings we see an absolute decline in the number of students Although schools are still being built to accommodate shifts in the population or to replace existing facilities, the number of classrooms being abandoned is increasing at a faster rate Budgets continue to rise, but without increase in productivity The fundamental reasons are well-known American education has a clientele of the young It has grown rapidly in the past by a simultaneous increase in the proportion of the young in education and in the number of young A declining birth rate, in conjunction with a near-saturation of the education rate, assures the industry of a decline in patronage within standard educational programs A declining industry has certain regular administrative characteristics Management tends to age This is particularly true when decline follows a period of rapid growth Management is relatively young because of the pattern of promotion and mobility during the growth period It is locked into place by a lack of opportunities A high rate of management turnover can be sustained only by an increase in involuntary exits As a result, we can plausibly expect a gradual aging of administrators (at all levels) over the next decade At the same time, decline produces a loss of managerial vitality There are fewer chances for advancement There are fewer resources There are fewer occasions of success The aging of employees (not just management) tips the cynicism scale further in the direction of doubt about the utility of commitment Organizational goals and personal self-interest diverge Finally, decline results in an oversupply of qualified administrators The process by which administrators are developed is not controlled soon enough, or fast enough, or hard enough, to avoid a large output of possible administrators at a time of lessened demand for their services The result is a backlog of would-be administrators doing other things Such a backlog simultaneously reduces enthusiasm and increases pressure for more administrative superstructure, in order to provide career opportunities Although the major shock of discovery should pass shortly, the reality of decline is likely to persist for some time Declining industries need administrators too; but they provide a setting somewhat different from the one that is implicit in some of our fantasies Shifts in social expectations Recent decades appear to have seen a significant shift and expansion of social expectations with respect to education This includes a shift in the kinds of people to be educated, a shift in the kinds of educational outputs to be produced; and a shift in the range of consequences attributed to education Concern with the educational role in racial and sexual inequities, in the problems of the cities, and in the uses of technical knowledge are the latest manifestations of a long-run increase in the social saliency of educational institutions The process has been steady over many years It appears more dramatic in the last decade or two partly because we are living now and partly because recent extensions of social claims coincided with a decline in the industry Educational systems have not been successful in meeting these expectations For example, American education appears to have had little impact on racial and sexual inequalities; it has not changed trends in the quality of life in cities or in the character of social mores; it has not made politics conspicuously more responsive nor citizens more active; it has not assured that graduates are employable; it has not even made many recent inroads in eliminating ordinary illiteracy in reading, writing, and arithmetic Numerous educators have observed that education cannot solve the complex of problems confronting society In particular, there has been an attempt to impress on community leaders the limits of education As John Goodlad observed in 1970: If you want to really eliminate unemployment, you create jobs If you want to really eliminate slums, you clear up slums, but you don't hold education responsible for getting it done.(3) The sentiments are amplified by standard interpretations of such things as the Coleman Report and the Jencks assessment of education as an instrument of redistributions The self-denial, while mostly correct, comes suspiciously late As long as our social problems appeared to be solvable and being solved, educational leaders did not resist responsibility for the success Much of the growth of our educational establishment (and particularly our system of higher education) has been rationalized in terms of the educational contribution to everything from prosperity to social progress to moral reform Now we are not so sure Failure has lead to a recognition of the limits of responsibility It is a classic story The contrast between school administrators in 1970 and school administrators in 1950 is the contrast between the response to foreign competition by American steel industry administrators in 1960 and 1920 Although we may believe, as I do, that the present posture is closer to the truth, the earlier conceits have not made the job of an educational administrator any easier The failures, in combination with a general retreat from claims of responsibility, will almost certainly reduce demands on schools over the coming few years Much of the widely-noted social pressure on schools is likely to decline Social expectations adapt to social beliefs about the capabilities of institutions That does not make the context much better Indifference is different from hostility, but it is not a less constraining administrative position The educational establishment will have to pay the price of having persuaded society that education is not a general solution to social ailments Managers who have become adept at dealing with confrontation will have to adapt to a world in which most of the audience is elsewhere Educational administration programs that have only recently embraced conflict management as the fundamental fact of administration will discover the conflict muted by ennui and institutionalized into bureaucratic procedures Administrative careers and professional esteem Our present career system for pre-collegiate administration has been developed during the 20th century It is built on a coalition of local schools of education with local schools and local administrators Certification through formal education, the strong tendency for advanced training in educational administration to be part-time, and the development of regional baronies of career management based in schools of education have produced a system that relatively few observers are willing to defend; but which no one knows how to change decisively A career in educational administration is a highly local career for most people Promotional opportunities depend on personal reputation within the district and with local barons; but there are no clear-cut standards for establishing that reputation, or claiming it It is a "buddy" system, complemented by a social validation of judgments and formal educational entry criteria The career system is complicated by the problem of esteem Educational administration is a profession It has a specialized literature, an apparatus of associations, and a system of entry through formal certification Relative to many occupations, educational administration enjoys considerable social prestige and substantial income expectations; but it consistently suffers from some natural comparisons: -Business administration is apparently viewed as having more competent individuals and better comprehension of relevant techniques It is often suggested that we introduce more of the techniques of business administration into education; almost never is the converse proposed -Doctors (M.D.'s), lawyers, architects, and engineers are accorded more systematic acknowledgement of professional standing They are more likely to be assumed to know something about their subject matter They are granted the right to speak a technical language rather than a jargon -In the informal academic status hierarchy the Ed.D is a lower ranking degree than the Ph.D and the Ph.D in Education is a lower ranking degree than the Ph.D in most other fields Doctoral students in education are (by test scores and academic performance) often systematically less qualified academically than doctoral students in other fields; faculty members in education are (by the criteria of the academy) often systematically less distinguished than faculty members in other fields The legitimacy, or relevance, of these comparisons may be questioned Most of them are based on arbitrary distinctions of no great inherent persuasiveness They have, however, social significance They are part of the social structure of beliefs within which educational administration operates The educational administrator has standing within the educational establishment that depends on his bureaucratic position and his informal career network His position outside of the educational establishment is rarely as high as that of individuals with whom he naturally compares himself So long as education was a growth industry and most of the social expectations about education could be fulfilled, the career system and the problem of esteem were masked by successful growth The mask has been removed by decline and the social certification of failure Moreover, localism, limited esteem, and a baronial system of career management are not conducive to the innovative leadership that we are regularly advised is required in education Quite the contrary, they seem likely to encourage the recruitment of individuals who are relatively uncreative and to extinguish administrative creativity if it should arise The third stage of development The complications produced by these contextual elements are profound We are apparently entering the third of a series of natural phases in the history of a social institution The first stage is a period of dynamic growth Social expectations rise; the institution is able to meet those expectations; there is excitement, expansion, and self-confidence The second stage is a period of conflict Social expectations outrun the capabilities of the institution; there is frustration, anger, and recrimination The third stage is a period of neglect Social expectations decline; the institution is able to meet many of the reduced expectations; there is indifference, passivity, and stagnation During the first two phases of the institutional life-cycle, we generate significant elements of superstition in our beliefs about administration and the ingredients of administrative success During the period of growth, most problems have many good solutions Administrators (and their teachers) come to believe that they understand what it takes to make a successful administrator Since many different things work about equally well, different schools of experts arise Each is subjectively confident; each is able to cite experience to support their confidence During the period of conflict, the half-life of a problem is forever No matter what you do, the problem remains Administrators (and their teachers) keep looking for the secret without finding it Schools of administration, foundations, and governmental agencies keep supporting solutions, quickly abandoning one for another without changing their luck until finally as the institution enters the period of neglect, they decide to give up entirely The move from one period of conflict to a period of neglect is not a happy one Decline, social expectations, and professional esteem feed upon each other A declining institution encourages an administrator to accept a limited social role and reduced esteem Clients who are persuaded of the institution's disabilities reinforce parochialism and decline The movement is neither grand nor dramatic; but it is a movement that is unlikely to produce the exciting administrative-situation-of-theyear Although I believe the process is in many ways inexorable, I not, propose we abandon the field in self-pity Administrative glory is largely determined by broad exogenous factors; heroes and villains are made by events; but administration is the profession of leadership, the art of intelligent coping with an arbitrary fate It is a minor matter; but our prospects for human control over events are built on collections of minor matters Nor should universities abandon tile effort to make administrative training more appropriate for modern educational administration The system we have evolved for certifying educational administrators through the granting of a doctorate is, in my opinion, a curious system The doctorate is a silly prerequisite for an administrative position I would not argue for perpetuating that system as it now exists I believe, however, that there is a place for university training in administration The university is not a good training ground for all administrators or all administrative skills; but it does have important special functions An active and imaginative university effort will become particularly critical as government agencies and foundations become disillusioned by unspectacular results and reluctant to maintain a long-term investment in the marginal improvements that might be feasible THE CONTEXT OF TRAINING I take a training perspective What are the kinds of things that a university program in educational administration should attempt to teach? The viewpoint is not heroic I assume the educational apparatus will change over time, but that educational institutions approximately as we know them will persist There will be schools, school districts, and colleges I assume that the activities of administrators will be modified in the future, but that recognizably administrative roles will continue to exist in recognizably bureaucratic organizations I assume that university schools of education will continue to develop and change, but that the fundamental role of the university in the training of administrators will not disappears A professional training program should provide opportunities for intellectual growth along several dimensions that are only loosely tied to immediately usable administrative skills A university-trained educational administrator should be an educated person, capable of interpreting and elaborating ideas from the culture of intelligence He should have an elegant appreciation of society, its problems, its heterogeneity, and its heritage The graceful wisdom of an educated person, however, is not enough An administrator needs to be competent both in the fundamental skills of administration and in the basic technology of the institution he administers; and universities need to provide education in those competencies if they are able to so efficiently Administration is a job, and it calls for talents Some of those talents are learnable The kinds of basic skills that a university program in educational administration should attempt to teach are the skills that are useful, will be used, and can be taught The trilogy is not redundant We need to examine the problems of educational administration, and to be attentive to those problems But we also need to be concerned with what educational organizations are, what educational administrators do, and what universities are good at Schools as organized anarchies Educational institutions are organized anarchies By this I mean that they frequently exhibit three important, and troubling, features: First, their goals are problematic It is difficult to specify a consistent set of goals Instead, goals seem to shift over time; they seem to vary from one part of the organization to another; they seem to be stated in terms that are hard to translate into action There is conflict over goals, and the conflict is not resolved easily Although it is sometimes possible to impute goals to the organization by observing behavior, such amputations appear often to be unstable or to define goals that are not acceptable to all participants in the organization The decision process seems to reflect more a series of actions by which goals are discovered than a process by which they are acted upon Speeches on goals express platitudes that are not useful administratively Second, their technologies are unclear Although we know how to create an educational institution, to staff it, and to specify an educational program for it, we not know much about the process by which it works It does work, at least in some senses Students seem to change Moreover, we can duplicate our results If we recreate the procedures in a new school, they will often have approximately the same outcomes But we have remarkably little capability for designed change in the system We not, in general, know what will happen if we make changes; we not, in general, know how to adapt the standard system to non-standard students or situations New occasions require a new set of trial-and-error procedures, either in the school or in an experimental laboratory Third, participation in the organization is fluid Participants come and go Students, teachers, and administrators move in and out There is even more turnover in other participants or potential participants Parents, individually and collectively, are erratic in their involvement; community leaders sometimes ignore the schools, sometimes devote considerable time to them; governmental agencies are active, then passive All of the potential actors in the organization have other concerns that compete with the school for their attention Thus, whether they participate in the school depends as much on the changing characteristics of their alternatives as it does on the characteristics of the educational organization involved Organized anarchies are not bad organizations They are not unusual Indeed, they are quite common Decision situations involving problematic goals, unclear technology, and fluid participation are familiar to all types of organizations They do, however, pose some problems for administrators, and particularly for our standard theories of administrative action The manager of an organized anarchy requires a modified theory of decision Almost all of our current ideas about intelligent decision making relate actions to preexistent goals that are clear and widely shared A manager is directed first to determine his goals and then to act Such behaviour is often possible and useful, but it is not always so When he wanders into the domain of coalition decision making or goal-free problem-solving, the administrator is little aided by our current ideas about problem-solving.7 The manager of an organized anarchy requires a normative theory for the allocation of attention In a system in which participants wander in and out, outcomes frequently depend on the timing of involvement Time and attention are key resources that need to be managed.(8) The manager of an organized anarchy requires a new set of ideas about management The fundamental postulates of management emphasize control and coordination of the organization with respect to a set of well-specified objectives Where goals and technology are unclear and participation is fluid, it is not clear what to make of the usual managerial precepts What managers do? One of the persistent difficulties with programs for reform in the training of administrators is the tendency to try to improve managerial behaviour in ways that are far removed from the ordinary organization of managerial life Unless we start from an awareness of what administrators and some idea of why they organize their lives in the way that they do, we are likely to generate recommendations that are naive Henry Mintzberg has recently taken a careful look at the literature on managerial time allocation and attempted to identify how managerial time is organized and what roles a manager plays within an organizations The studies he reviews not significantly involve educational administrators, but attempts to make similar studies of executives in educational institutions show similar patterns (10) With respect to the organization of time, these studies indicate that executives work long hours The exact number of hours per week varies, but it is substantially greater than the number of hours worked by most non-executives in most organizations Much executive work is brief, disconnected, incomplete, and initiated by others Most time is spent in verbal interaction with others Executives tend to move to problems and information that are specific rather than general, concrete rather than vague, solvable rather than impossible, and currently pressing and local rather than distant in time and place Relations with subordinates and with individuals who are not in a direct line relationship within the organization occupy substantially more time than relations with superiors A considerable amount of time is spent in scheduled meetings Mintzberg also tries to identify the roles that a manager performs His breakdown is based on data that are more difficult to interpret and, therefore, subject to greater doubt than the data on the allocation of time; but his categories are more carefully based on empirical observations than others in the literature and have some substantial implications for management training From Ws analysis, Mintzberg deduces eight basic sets of managerial skills that would improve managerial performance in doing better what they now (1) Peer skills: The ability to establish and maintain a network of contacts with equals (2) Leadership skills: The ability to deal with subordinates and the complications of authority, power, and dependence (3) Conflict-resolution skills: The ability to mediate conflict, handle disturbances, and work under psychological stress (4) Information-processing skills The ability to build networks, extract and validate information, and disseminate information effectively (5) Skills in unstructured decision-making The ability to find problems and solutions when alternatives, information, and objectives are ambiguous (6) Resource-allocation skills The ability to decide among alternative uses of time and other organizational resources (7) Entrepreneurial skills The ability to take sensible risks and implement innovations (8) Skills of introspection The ability to understand the position of manager and its impact on the organization Although the process of deduction is not always clear, Mintzberg's judgments are informed by as good an understanding of what executives actually as we are likely to obtain If we are going to consider ways of improving the effectiveness of managers in their behavior, as opposed to their "function" or "purpose" deduced from "needs" or "problems", we might intelligently start from here In particular, if we wish to identify tools of analysis that will be used, they will need to be tools that match the pattern of managerial activities The distinctive competences of universities The university does some things badly Such things it clearly should not It does other things well, but not well enough It cannot all of the things that it does relatively well Indeed, it probably ought not to some things that it does better than any other social institution The university should attend primarily to activities where it has a distinctive competence For example, recent commentaries on administration, not only in education but also in business, the military, and public bureaucracies, have emphasized the importance of three talents: (1) the talent to deal effectively with people (2) The talent to manage conflict (3) The talent to mediate between the organization and the broader society Because of the importance of such talents, there have been suggestions that training programs for administration should include an amount of training time proportional to the importance of each talent The suggestions depend on an implicit assumption that the talent-return on training investment is constant across domains of training The assumption seems implausible The amount of time to be spent in a training program on different forms of training depends not only on the importance of the talent involved but also on the return (in terms of the value of talent improvement) per unit of time spent in training We have, I believe, some ideas about how to improve the capabilities of administrators with respect to the three talents cited above However, the technologies are not well-developed; and some of them involve exposure to experience more than the specification of formal knowledge or formal procedures As a result, there is likely to be a rapidly decreasing marginal return to the investment of time in training in some of these areas, and a comparative disadvantage for educational institutions There are numerous areas of administrative training in which the university has an absolute advantage As a partisan, I may exaggerate; but I think the list of domains in which the university has a simple advantage is long: Universities as good a job as anyone at most aspects of management training They better at providing the basic knowledge, at identifying general problems, at isolating and providing broad experience in the necessary interpersonal and intellectual skills, at discussing value issues, at encouraging risk-taking and innovation, at building social and personal sensitivity, at exposure to conflicting ideas and sentiments, and at building a sense of self-esteem I cannot demonstrate the absolute superiority of the university in these domains, but I believe the argument is plausible I believe, in fact, the list could be made considerably longer Despite this, the university does not have a comparative advantage in all of these domains It ought not to be a general purpose social institution, or even a general purpose participant in administrative training It ought to pursue its special role The university has a distinctive area of competence It is the domain of the intellect What the university does best, relative to other institutions, is to develop new knowledge and its implications It is an intellectual institution Except in a few areas of scholarly activity, universities are the site for the vast majority of the basic scholarship that is done in the United States This is true throughout the biological, social, and physical sciences; in the humanities; and increasingly in the arts It is true in almost every area of applied knowledge: law, medicine, engineering, education, business administration, criminology, social welfare, politics In their history, universities have done other things They have provided entry to the religious or commercial establishment; they have managed major sportsamusement systems; they have been centers for social experimentation in individual mores and social philosophy and for socialization into proper behavior But their primary claim is an intellectual one Their faculties and students are smarter in intellectual terms and know more of the things that are learned in scholarly ways than most other people They read, write, and think more and better than most other people They have the time and the organization to good research and good thinking Since the problems of the world are not necessarily amenable to intellectual solutions, the social value of a university can shift over time With such a shift, we would expect some shift in the comparative advantage calculations also Some nonintellectual activities would become more appropriate for the university However, shifts in the problems of the world are almost certainly much greater over time than shifts in the comparative advantage of the university Most of the time, regardless of shifts in social needs, a university should retain approximately the same basic specialization In a similar way, the advantage for the university in the training of admin- 10 istrators is primarily in the intellective domain It is in providing the research basis for intelligence and in teaching the intellective skills of management Other institutions are likely to have a comparative advantage in other training areas As the mix of administrative needs shifts, the primary role of the university is to provide the intellectual base of new skills; it is not to attempt to provide all of the training For some administrative jobs and some administrators, the university is a relatively inefficient training center Some important administrative tasks have relatively small intellectual components It may be considerably more vital that the administrator be strong, or loving, or energetic, or sensitive, or charismatic, or a member of a particular social, ethnic, or sexual group We can recognize the importance of such tasks and the legitimacy and value of such attributes without accepting the proposition that the university should provide either the training or the certification for them FIVE CRITICAL SKILLS OF ANALYSIS If we are going to use universities effectively in the development and dissemination of critical administrative skills, we need to attend to the problems posed by the context of decline and by the nature of educational organizations, managerial work, and university comparative advantage Not all problems can be solved, and certainly not by university programs in administration But there are some things that are relevant, appropriate to the things managers do, consistent with the kinds of organizations in which they them, and intellectually demanding I will mention five different analytical skills that seem to me to satisfy the criteria They include: (1) The analysis of expertise The management of knowledge (2) The analysis of coalitions The management of conflict (3) The analysis of ambiguity The management of goals (4) The analysis of time The management of attention (5) The analysis of information The management of inference Each of these skills is important to the problems facing the administrator; each is consistent with organized anarchies; each can be adapted to the things that managers do; each is appropriate to academe Others might identify a different list Mine has no necessary uniqueness It is one person's view of some important problems and skills in administration and administrative theory Each of the skills is linked closely to the everyday requirements of managerial life, but each is clearly an intellective skill Each involves a technology of analysis and thought Each is based on the intellectual development of a scheme of analysis Each is a tough problem, incompletely understood Each demands a program of research as well as a program of training and testing For each we have a start The analysis of expertise Much administration involves managing the taking and giving of advice Administrators deal with experts, and are experts They give advice, and they take it Both aspects are important Indeed, both are important in a single relationship Consider the problem of dealing with individuals who have greater knowledge in an area than the administrator does but to whom he is not willing (or permitted) to delegate full responsibility for action The situation is a daily one Educational administrators confront experts on teaching, curriculum, finance, management science, research, construction, bus schedules, and the rest of the technical knowledge necessary to schooling The level of expert sophistication often considerably exceeds that of the administrator Our problem is to specify a set of procedures for coping with differentials in technical knowledge Imagine the following experiment: An administrator is presented with two individuals, each claiming to be 12 expertise by comparing his observations with the observations of other experts In effect, we assume that the errors made by different experts are independent Thus, we infer that if several experts say the same thing, it is likely to be true An expert who says many things that other experts not say is less likely to be reliable than is an expert who usually agrees with other experts (4)The distribution of confidence assumption In any body of knowledge, there is variation in the degree of confidence in which different beliefs are held Individual experts who not exhibit such variation are likely to be less well-informed than those who do; the relative confidence an expert has in different statements is likely to be more reliable across experts than any absolute level of confidence he may express (5) The interconnectedness assumption Expert knowledge tends to be interconnected rather than discrete There are theoretical interconnections and contextual interconnections We infer that experts who report knowledge in terms of a series of independent bits of knowledge are either experts in a domain in which the knowledge is less reliable than others, or are themselves less reliable experts Clarification and elaboration of such assumptions would be of substantial significance to improving the way in which managers receive advice from experts They would also help in the giving of advice I have suggested that the relationship requires an assessment of what the expert can say that is relevant to the decision problem and an assessment of the degree of confidence that can be placed in what he says Although there are frequently some elements of conflict between expert and non-expert, the relation is nevertheless an interactive one The expert needs to have the right questions posed in order to provide the right answers; he needs to solicit help in understanding his own confidence in his beliefs, particularly since the terms of confidence relevant to the manager may be quite different from the terms of confidence familiar to the expert in his other activities The expert and the manager also need to be wary of situations in which the assumptions by which the manager forms inferences may be quite incorrect Knowledge is not necessarily homogeneous; density of knowledge is not necessarily correlated with precision; the errors of experts may not be independent at all; the variations in confidence across experts may be quite meaningful; knowledge may be quite unconnected Finally, the analysis of expert testimony (giving or taking) requires an awareness of the expert and the manager as actors The taking and giving of advice are performances with certain rules that simplify the process at the cost of sometimes confounding it So long as the rules of the performances are satisfied, we are inclined to certify the "expert" as expert and the “manager" as manager Since socialization into the performances is a part of the training to become one or the other (or both), most experts know how to play expert and choose to so; most managers know how to play manager and choose to so An expert or manager who does not know how to play the role, or chooses not to, will sometimes confuse the process This is particularly a problem with an expert who, though entirely competent as an authority in his field, is inexperienced in the role of giving advice; or with a manager who, similarly, is inexperienced in taking advice Expertise in the relation is an important aspect of success The analysis of coalitions An educational manager is a builder of political coalitions The process is not identical to the process in governmental affairs, but key elements are the same Different individuals and groups bring different interests and objectives to the organization The interests vary in their mutual compatibility partly as a 13 consequence of the form they take, but also partly as a consequence of the alternative combinations of policies and options that are provided Administrative skills at coalition building and assessment are the skills of developing and exploiting non-uniqueness in coalition solutions The technology of coalition-building can be moved from a relatively pure art form practiced with various levels of success by different managers to a learnable skill The technology requires more development, but the rudiments are there The rudimentary base for the techniques lies in a simple perspective on decision making: We assume that any decision can be made if there is enough support for it within the constituencies of the organization In order to secure support, various trades are undertaken A policy proposal is modified to include (or exclude) items of concern to key potential supporters; agreements are made to trade support on future issues for trade on current issues; arrangements are made to encourage third party intervention A market in political horse-trading is established; an omnibus policy is developed This perspective on decision making has been developed over many years of research into political negotiation Although most of the ideas were' originally spawned within political science for the analysis of governmental politics, they extend easily to other organizations What makes them interesting from the point of view of educational management is not only that they appear to capture some significant truth about organizations, but also that they have been specified with enough precision to allow some efforts at simple skill development We can describe any situation (according to this paradigm) in terms of a set of actors, their capabilities for controlling the final outcome (their power), their agendas (what they want), and their resources (what they have) We can prepare a table that lists actors, agendas and resources, For example, we can imagine a city council of five members deciding on a new police chief and a city manager considering his options Suppose the manager assesses the situation in the following way: The political manager becomes an organizer of mutually attractive exchanges In order to serve as such a broker, he asks four simple questions: (1) Is there at least one winning coalition and associated program? That is, can we find a program that will secure enough support to assure adoption within the system under consideration? If there is not, the best manager may be better advised to spend time elsewhere In the city council example, the question is whether there is at least one possible decision that would secure three votes Council Member (Power) Votes Agendas Resources Adams I Chief must be from outside A promise from Dawson to vote with Wm provided it does not violate D's agenda Brown I Chief must be a woman Willing to support anyone else on anything else in exchange for support on this None Will support anyone in exchange Carlisle 14 for future vote Dawson Edison I Chief must be an experienced man None Chief must be a member of an ethnic minority or Tom Smith, the present assistant chief A commitment from Adams to support E's candidate if it is not an insider An inspection of the table indicates that such a possibility exists An experienced male outsider from an ethnic minority will secure the votes of Adams, Dawson, and Edison (assuming no other more preferable alternative exists) (2) Is the solution unique? Is there only one possible outcome given the agendas and resources? Since most of the possibilities for management stem from non-uniqueness, the political manager generally looks for situations in which the outcome is not completely determined by the political constraints The answer in the city council example is "no" There are several possible outcomes Indeed, the outcome that will occur depends considerably on what alternatives are presented Any of the following alternative police chiefs could be chosen under some circumstances: An outside, white woman (Adams, Brown, Carlisle) An experienced outside minority man (Adams, Dawson, Edison) An outside minority woman (Adams, Brown, Edison) An inside minority woman (Brown, Carlisle, Edison) In two cases the coalition is created by Brown entering into a trade with Carlisle, promising to support some future vote (3) Is any solution excluded? Are there any possible (and possibly attractive) solutions that cannot be achieved? Policies that cannot secure necessary support, or combinations of supporters that cannot be made? In the city council example, we ask whether there is any kind of chief who could not be elected There is An outside white man cannot be chosen Neither can Tom Smith (nor any other inside white man), an inexperienced outside minority man, nor an inside white woman (4) Can the options be changed? Can we modify the possibilities if we consider various ways in which the situation can be altered? In particular, are there some combinations of policy issues that make new winning coalitions possible? The winning coalition may be made problematic if the definition of alternatives is subject to administrative or political control For example, we might ask what would happen if Edison were willing to trade a future vote to Carlisle Then, Tom Smith and an experienced outside minority man become possibilities More generally, we would want to consider what would happen if we combined the choice of police chief with some other issue, for example the choice of city attorney Coalition management involves asking these questions in a real world situation and exploiting the indeterminacy of the solutions At least in many situations, there are several potential winning coalitions Thus, there is a role for leadership The situation in educational institutions is analogous to the city council situation There are, however, some differences Decisions are not always - or even normally - by formal vote It may be more important to have some supporters than others The exchanges may be somewhat more subtle, or across a different range of behavior than simple policy support For example, trades often occur across organizational boundaries The general theory of trades (or bribery) is not well understood, but we know some important things about it The technocratic, relatively mechanistic, approach to coalition building 15 appears to define an administrative skill of significance It is clear that the technique is used effectively in large elements of organized life in our society It needs to be regularized, codified, and improved It is a skill that can be learned and used by an educational administrator It requires a consciousness about political negotiation that can be developed The analysis of ambiguity Educational administrators are involved in decision making under ambiguity That is, not only are they uncertain about the consequences of alternatives They are also unclear about their goals, uncertain of their technology, and unaware of their alternatives They need help in defining how decisions are made under such situations Our usual management recommendations are classic: If you not know what your objectives are, take the time to identify them If you not know what your alternatives are, search until you find them If you not understand your technology, undertake research (or hire consultants) to establish the cause and effect connections in your activities Such recommendations are not wrong; but they are frequently inadequate Experience with efforts to establish goals is often positive It generates attention and interaction As a side effect, people worry about the organization and each other; they talk to one another But the process is slow and the output rarely precise enough or stable enough to be of much decision help The idea of establishing goals first and then acting has some limitations as a model of behavior The search for alternatives is often important, but it is rarely improved by the simple injunction to search What are the techniques of search? Where alternatives lurk? How you invent new alternatives? How you find alternatives that have been invented? How you avoid the creation of pseudo-alternatives as ways of making a preferred alternative look good? Recently, there have been some efforts to invent some ideas for goal-free problem-solving They occupy some of the current literature in psychotherapy, problem-solving, and leadership The techniques are not well established, but they are close enough to imagine that something is already known and further analytical tools are likely to be possible Probably the most difficult domain of analysis is developing some ideas of how we can act without knowing the objectives of our action In terms of our standard theories of intelligence, such a conception is nonsensical We are deeply committed to the idea that goals come first and actions should be related to those preexistent objectives Elsewhere (12) I have tried to suggest some reasons why such a conception is overly limiting and some ideas about the possibility of acting without a good reason and the possibility of discovering goals through behavior A major concern in any such effort is how managers might plausibly relax the requirements of consistency in order to experiment with doing things for which they have no good reason and with alternative conceptions of themselves I think an analysis of the problems of decision making under ambiguity suggests some things that a manager can reasonably do: (1) He can treat goals as hypotheses He can experiment with imagining that he wants something other than what he now believes he wants in order to test his beliefs (2) He can treat intuition as real It is not obvious that intuition is any one thing, or anything It is, however, an excuse for acting differently and an alternative way of consulting our intelligence (3) He can treat tradition as meaningful In our efforts to glorify the ability of an individual to comprehend his world, we sometimes overlook the attractive coercion of traditional rules as a device for discovering goals (4) He can treat memory as an enemy Rationality and consistency require a 16 memory, and for most purposes good memories make good decisions But the ability to overlook, or forget, is also useful (5) He can treat experience as a theory He needs to revise what he has learned yesterday on the basis of his interpretations of today, and to try alternative rewritings of history All of these are adaptations of devices more familiar to discussions of behavioral pathology than they are to our usual understanding of managerial behavior, but I believe they provide some clues on how to introduce a more inductive version of managerial goals The general ideas extend easily to the problems of educational organizations: We need to reexamine the functions of management decision One of the primary ways in which the goals of an organization are developed is by interpreting the decisions it makes, and one feature of good managerial decisions is that they lead to the development of more interesting value premises for the organization As a result, decisions should not be seen as flowing directly or strictly from a pre-existent set of objectives Managers who make decisions might well view that function somewhat less as a process of deduction or a process of political negotiation, and somewhat more as an opportunity for upsetting preconceptions of what the organization is doing 1) We need a modified view of planning Planning in organizations has many virtues, but a plan can often be more effective as an interpretation of past decisions than as a program for future ones It can be used as a part of the efforts of the organization to develop a new identity that incorporates the mix of recent actions into a moderately comprehensive structure of goals Procedures for interpreting the meaning of most past events are familiar to the memoirs of retired generals, prime ministers, business leaders, and movie stars They suffer from the company they keep In an organization that wants to continue to develop new objectives, a manager needs to be relatively tolerant of the idea that he will discover the meaning of yesterday's action in the experiences and interpretations of today 2) We need to reconsider evaluation As nearly as I can determine, there is nothing in a formal theory of evaluation that requires that the criteria for evaluation be specified in advance In particular, the evaluation of social experiments need not be in terms of the degree to which they have fulfilled our a priori expectations Rather we can examine what they did in terms of what we now believe to be important The prior specification of criteria and the prior specification of evaluation procedures that depend on such criteria are common presumptions that restrict the serendipitous discovery of new criteria Experience should be used explicitly as an occasion for evaluating our values as well as our actions 3) We need a reconsideration of social accountability Individual preferences and social action need to be consistent in some way But the process of pursuing consistency is one in which both the preferences and the actions change over time Imagination in social policy formation involves systematically adapting to and influencing preferences It would be unfortunate if our theories of social action encouraged leaders to ignore their responsibilities for anticipating public preferences through action and for providing social experiences that modify individual expectations 4) We need to recognize the role of duty and obligation Recent work on organizations has emphasized theories in which individual action is intentional and self-interested Such theories account for important aspects of behaviour in organizations; but they ignore the significant ways in which action follows a logic of obligation rather than a logic of choice By developing systematic systems of duty we introduce a strain between current conceptions of self-interest and the claims of duty As a result, we can encourage the reconciliation of the two 17 through creative elaboration of goals 5) We need to accept playfulness in social organizations The design of organizations should attend to the problems of maintaining both playfulness and reason as aspects of intelligent choice Since much of the literature on social design is concerned with strengthening the rationality of decision, managers are likely to overlook the importance of play This is partly a matter of making the individuals within an organization more playful by encouraging the attitudes and skills of inconsistency It is also a matter of making organizational structure and organizational procedures more playful The managerial devices for maintaining consistency can be varied We encourage organizational play by permitting (and insisting on) some temporary relief from control, coordination, and communication I am not sure where we go from here Clearly, we need a better idea of how we should decide which among all of the possible foolish things we might is most likely to lead to attractive new conceptions of ourselves and our organizations We need better notions of the costs and benefits of playfulness The problems of decision making under ambiguity, however, are conspicuous problems for educational administrators They are not a residual category; they are a major category They require the development and use of new kinds of techniques The analysis of time Educational administrators, like other managers, operate under time constraints Time is a scarce good As a result, the procedures by which time is allocated are of some significance to the management of an enterprise Recent research has indicated that many administrators (including those in education) have the subjective sense that their time is badly allocated As I noted earlier, Mintzberg has described management time allocation as it appears in a series of recent studies: The manager feels compelled to perform a great quantity of work in an unrelenting pace [His activities] are characterized by brevity, variety, and fragmentation Interruptions are commonplace [He I gravitates to the more active elements of his work the current, the specific, the well-defined, the nonroutine activities [He spends] most of his time in verbal contact [Nonline] contacts generally consume one-third to one-half of the manager's contact time Subordinates generally consume one-third to one-half of the manager's contact time On the order of 10 percent of his time [is spent] with his superior (13) Students of executive behavior have noted the extent to which the allocation of time is dictated, out of the control of the executive A President's priorities are set not by the relative importance of a task, but by the relative necessity for him to it He deals first with the things that are required of him next Deadlines rule his personal agenda.14 The latent absurdity of being the executive leader of an organization that does not know what it is doing haunts the presidential role That sense of absurdity is somewhat ameliorated by a pattern of attention that reinforces a feeling of reality for the role This means a busy schedule, the press of work, the frequent reminders of the fact that one is the president, the attention to minor things one can (14) At the same time, there have been efforts to develop techniques for the management of time In effect, these efforts start with an economizing view of time allocation The manager considers alternative uses of his time in terms of the return each provides The cost of attending to one problem is not attending to another The 18 broad implications of such a view of a "rational" allocation of time are contrasted with a broad view of the allocation of time generated by the behavioral process within an organization On the basis of that comparison, some areas of over- and underallocation of time are identified; and some techniques for introducing shifts in the allocation are proposed The analysis is close enough to being useful to warrant more development even though we may have some doubts about the kinds of recommendations that are usually generated now If we take a conventional view of a manager's role, we would expect his high leverage activities (i.e., those with a relatively large return per unit of time spent) would include long run planning, nonprogrammed tasks, general consideration of overall objectives, systematic surveys of the organization for potential problem areas, and the establishment of major strategies Despite the widespread acceptance of a view of executive leadership that lists such matters as being of first concern, most studies of actual management indicate that such activities represent relatively minor claims on the total time in a manager's day The usual analysis then proceeds to suggest some ways in which the activities that are apparently over-allocated (e.g., ceremonial activities, short run decisions, idle chatter, interpersonal interaction, deadlined commitments, minor actions, etc.) can be reduced in favour of the under-allocated activities The methods are ordinarily some mixture of conscious rationing by the manager to replace the reactive pattern of much of his life, and the development of staff to perform some of the functions to which the manager currently attends The problem is probably more subtle than current solutions suggest The solutions seem significantly to misunderstand what a manager does with his time and why he does it For example, conventional enthusiasm for time spent on longrun planning and overall objective-setting seems to me excessive I think a case can be made for the proposition that activities normally used to accomplish such goals are rarely worth significant amounts of administrative time Similarly, I suspect that many current advocates of time management overestimate the importance of thinking time as something distinct from sequential attention to crises The model of thinking as an activity that is inconsistent with acting seems unlikely to capture very much of the real process of management These potential quarrels with the details of some current recommendations, however, should not discredit the effort to develop a good analysis of how time is used and how it might be used better We need to develop routine procedures for reporting time allocation We need to understand the rough character of the return to time investments Most managers, for example, seem to operate on the implicit assumption that managerial time on a problem or in a single contact has a strongly negative first derivative with respect to time Such a conception seems to be quite possibly correct and to have considerable implication for time management It suggests that managers should try to encourage many brief contacts rather than a few long ones At the same time, we should be able to move from an improved understanding of the decision process to a clearer idea of how an educational administrator might spend his time if he wants to accomplish a program For example, one analysis of college administration lists the following key attributes of decision making in organized anarchies: (1) Most issues most of the time have low salience for most people A major share of the importance an issue gains is related to its symbolic significance (2) The total system has high inertia It is hard to start and hard to stop (3) Any decision can become a garbage can for almost any problem What problems are attached to which choice is heavily dependent on timing (4) The decision process is easily subject to overload (5) The system has a weak information base Histories and current statuses are obscure 19 Time analysis points to the two-sided nature of time as a scarce resource On the one hand, since time is scarce, it needs to be rationed On the other hand, since time is scarce, the willingness to provide time places one in a strong position Thus, when a set of simple tactical rules is derived, many of them become rules for managing one's own time, exchanging time and energy for acquiescence, or managing the time of others Specifically, the analysis leads to eight tactical rules, of which four are connected directly with time management: (1) Spend time Since time is scarce, the willingness to provide time considerably increases the likelihood of being present when things happen and provides a resource that permits exchanges (2) Persist Do not assume that the mix of individuals and interests that decided things yesterday will return today If defeated today, return tomorrow If victorious today, beware of a new combination of people tomorrow (3) Overload the system When time is scarce, an overload on the system produces decisions that are more in the control of relatively fulltime participants (4) Provide garbage cans Manage the time of others Provide opportunities for them to exercise problems somewhere away from your concerns Although such recommendations are clearly subject to qualifications, they indicate some of the possibilities for the systematic management of time based on an analysis of the role of time in the organization Finally, it should be possible to develop some clearer ideas about the impact o f deadlines on time utilization and techniques for using deadlines Weiner has shown how deadlines affect the organization of time in a public school decision Particularly where there is a conflict of interest He has suggested some strategies for the use of deadlines that operate to aid allies and confound opponents Managers not only impose deadlines; they also react to them The description of managerial time allocation makes it clear that deadlines are an important factor in the time-pacing of managerial life This opens up the possibility that a manager might be provided a strategy for creating and responding to deadlines on his own behavior Such a strategy would build on some notion of the time distribution of response to deadlines of various sorts, an area of research that is quite feasible but barely begun The analysis of information Decision making in education involves data There are data on student performance, exposure, background, and capability There are data on teacher training, assignment, reputation, and cost There are data on facility utilization, cost, expected life, and design There are data on administrator experience, training, responsibilities, and performance There are data on birth and mobility rates; on experience with alternative programs; on aspects of other school districts; on the relation between educational output and the weather, social structure, cultural tradition, physical surroundings, diet, and investment There are data; the data are often relevant to some decision; they are usually incomplete or subject to some error or indecisive; as a result, they require some kind of inferential judgment The obvious answer is an understanding of statistics It is a correct answer Statistics, and particularly management statistics, are well-developed domains of analysis They not qualify as "new developing" techniques in the same sense as the other cases However, educational administrators are notoriously under-trained in statistics; insofar as they are trained in statistics, they tend to be trained in the wrong kind of statistics It is difficult to defend the present ideology or practice of quantitative training 20 for educational administrators The dominant ideology is that administrators should (as a part of their doctoral training) learn the quantitative techniques of the research scientist - most commonly those of the educational psychologist The secondary ideology is that, as future chief executives, they not need any serious exposure to formal skills in quantitative analysis They can always hire a statistician Neither ideology seems defendable to me To suggest that senior administrators should not have the professional skills to deal with data is to handicap them in any serious effort they might want to make to manage or lead the organization I have argued elsewhere some of the reasons why the technology of decision analysis is not, in itself, adequate for a manager I will repeat that assertion here But an ability to make inferences from data is certainly necessary At the same time, however, it seems to me inefficient, and often quite misleading, to base the statistical training of managers on the statistics of behavioral science If the student goes far enough and achieves a high enough degree of sophistication, he can see that the logic of decision and the logic of research are similar; but the recognition of that fact is not trivial It is almost never a part of the elementary training in statistics provided the administrator and reinforced by reading he may in the professional educational literature Insofar as statistics is taught to educational administrators, the teaching tends to emphasize description, experimental design, and hypothesis testing at an elementary level Despite the fact that such a focus has been built into the sacred beliefs of the profession, various Departments of Education, and innumerable schools of education, they are usually the wrong kind of statistics for the problem faced by an administrator Consider the following simple problem in administration: A consultant has suggested to the school district that it should adopt a team-teaching open classroom concept in an elementary school The consultant reports that recent research has indicated such a program produces significantly better student performance than does the standard program The administrator wants to decide whether to implement such a recommendation in an elementary school the district is about to build If the administrator has some awareness of the statistics he might have learned, he will raise some relevant questions about the studies on which the recommendation was based He will ask how well controlled they were; how large the sample was; what kind of schools and students were in the sample; whether the study has been replicated For the most part, however, he is likely to miss the fact that the idea of “statistical significance” - and the standard operating procedures for using it - is not oriented to the problem he faces The standard procedures for testing for a significant difference between alternative treatments are familiar They involve asking how likely it is that the observed difference in outcomes could have occurred if, in fact, there were no differences That is, we assume two universes with equal means and (usually) equal variances and ask what proportion of the time random samples of a certain size drawn from these two identical universes would exhibit differences in means as great as our observations The significance level for a test as it is normally learned, is an essentially arbitrary convention It can be any value between 0.1 and 0.5 without major cries of pain from the audience Noted in any good elementary statistics book, but usually forgotten in further exposition and by the student are three conspicuous reasons why the standard conception of hypothesis testing is limited in application to administrative decision situations: (1) The procedure effectively ignores Type-II errors The statement is not true in a formal sense The theory and some presentations of it note features of Type-II errors and the power of a given test Even more, there is an attempt to press toward tests that are relatively robust But the bias is clear In the procedures 21 that are used in ordinary statistics in ordinary educational research, the chance of rejecting a favored hypothesis that is, in fact, true are usually substantially higher than the probability of accepting a favoured hypothesis that is false The justification for this bias in research studies is, in fact, fairly complex; although it is usually stated rather briefly and as though it were self-evident But there is no reason to believe that administrative decisions require the same attitude toward the relative importance of Type I and Type II errors that research decisions do, or that different decisions call for the same posture (2)The procedure will identify some statistically significant differences that are not administratively important The likelihood that differences of a particular magnitude will be identified as statistically significant increases with sample size Suppose there is a real difference of about one point on enough samples, such a difference will be identified as statistically significant achievement tests due to team-teaching and open-classrooms With a large enough sample, such a difference will be identified as statistically significant With a small sample it will not Since the administrative significance of a difference is likely to turn more on its estimated magnitude than on its significance, the administrator is often welladvised to be more conservative about acting on "significant" results from large samples than on "significant" results from small samples This rule is distinctly counter to his usual teaching (3) The procedure will treat some administratively significant differences as statistically insignificant Quite frequently, the administrator is asking a different question He does not want to know whether one treatment is significantly better than the other in the special way in which those terms are used in statistical hypothesis testing He wants to know what is the best estimate of the probability distribution of outcomes stemming from the alternative programs He wants to combine these estimates of benefits with estimates of costs, and the usual question of statistical significance simply does not enter These points are statistically elementary, but they are not learned (or usually taught) They are well-publicized in the literature but not built into the institutional structure What underlies this failure in analysis for administrators is the curious relation between administration and research in the social system of education The educational administrator is supposed to be certified as an educational researcher This makes him vulnerable to the methodology of the research establishment; he accepts it On the other hand, he doesn't ordinarily use it The research is often not relevant to his problems and the results are often stated in a form he cannot use For example, relatively few administrators (or students of administration) appear to be aware that the statement that there is "no significant difference" between two treatments (or groups) is notably different from the statement that the two treatments are probably equal The research/administration relation needs to be strengthened by the development of genuine decision analysis skills among administrators; we need firstclass analysts of decision data The technology of decision statistics exists It needs to become a part of the basic administrative repertoire and of the research concerns of programs in educational administration CONCLUSION We are in a period of doubt about education We are doubtful about educational institutions; we are doubtful about administration; we are doubtful about formal training for administration; we are doubtful about applied research in administration The doubts are well-founded There are limits to what educational institutions can There are limits to the impact of leaders on events There are limits to the effectiveness of formal, training in developing leaders There are limits to our capacity for developing useful applied research Our pretensions are absurd in many 22 important ways Doubt is different, however, from despair The impact of knowledge and intelligence on the course of events is perceptible We need not retreat from a belief that there are things to be discovered and learned that are important to the management of education As university programs change, they should ask to be judged by their ability to contribute to the uses of intelligence in administration, The role is a limited one It emphasizes a special set of talents and the development of a special set of skills University programs are preeminently programs for developing the applications of analysis to administrative problems They are designed for the smart and the reflective Not all leadership requires analysis, smartness, or reflection; all leadership requires many other' things But university programs probably best contribute to society when they what they can without claim to universality, and without apology I have tried to identify five critical analytical skills They are, I believe, central to the job of most high-level educational administrators The management of expert knowledge, the management of conflict and coalitions, the management of ambiguity, the management of time, and the management of inference from incomplete information are all consistent with the criteria I suggested earlier They are relevant to the problems an administrator faces; they are appropriate to educational organizations; they are attentive to the way an administrator defines Ws own job through his behavior; and they are consistent with the comparative advantage of the university Each of the skills involves a high level of sophistication at three vital points First, each requires considerable managerial sophistication Fitting the ideas to a particular administrative situation is a subtle task It requires an elegant appreciation of both the formal analysis and the organizational setting Second, each requires a high order of sophistication in the development of the tools They are not simple modifications of instruments already easily available Third, each requires an ability to use analysis playfully To generate creative solutions to administrative problems Analysis is not a game for the serious, although it is a serious game Educational administration has not been noted for its creative contributions to management analysis It has been pictured, with some justice, as managerially parasitic If the profession is going to lead in the refinement of instruments for modern administration, it will require a cadre of superbly trained analysts The cadre can be developed We require a faculty that is willing and able to show significant intellectual capabilities, and students who are willing and able to work on the boundaries of applied technical knowledge We require a reform in our ambitions and our expectations As professors of educational administration, we might start by taking ourselves and our students seriously The research and development of administrative technology is not a trivial job University programs in educational administration have a responsibility to develop those skills and to train administrators who have the ability to use them Major problems in analysis remain to be solved for administrators; major techniques remain to be developed; major limitations remain to be discovered The task is large enough to occupy our talents for a long time The intellectual challenge is significant; the technical demands are impressive; success is not certain The job needs a high order of imagination, technical knowledge, and self-confidence If we develop techniques and training that improve the capabilities of educational administration to deal with experts, to solve problems in the absence of goals, to treat data from a decision perspective, to manage conflict and coalitions, and to allocate time, it will be an impressive set of contributions appropriate to the traditions of academia We will not change the fundamental course of major events; but we will make the marginal improvements in technique that distinguish a profession 23 REFERENCES This Paper was read as the Seventh Annual Walter D Cocking Memorial Lecture at the August 16, 1973, meetings of the National Conference of Professors in Educational Administration, held at Bellingham Washington It has profited particularly from comments by James R Glenn, Jr Creamer, D and Feld, B Some Innovations in the Training of Educational Leaders New York Conference Board 1972 p.2; Mitchell, D P and Hawley, A Leadership in Public Education Study: A Look at the Overlooked New York Academy for Educational Development 1972 Ch.l Cited by Mitchell, D P and Hawley, A op cit p Coleman, J S Equality of Educational Opportunity Washington Government Publishing Office 1966; Jencks, C Inequality New York Basic Books 1972 Mayhew, L B Changing Practices in Education for the Professions Atlanta Southern Regional Education Board 1971 Cohen, M D., March, J G and Olsen, J P "A Garbage Can Model of Organizational Choice" Administrative Science Quarterly XVII, March, 1972 pp 1-25 March, J G "Model of Bias in Social Action" Review of Educational Research XLII, March, 1973 pp 413-429 Cohen, M D and March, J G Leadership and Ambiguity: The American College President New York McGraw-Hill 1973 10 Mintzberg, H The Nature of Management Work New York Harper and Row 1973 11 Hemphill, J K and Walberg, H J An Empirical Study of the College and University Presidents in the State of New York Albany Regents Advisory Committee on Educational Leadership 1966 Cohen, M D and March, J G op cit 12 Baldridge, J V Power and Conflict in the University New York John Wiley 1971; March, J G "The Business Firm as a Political Coalition" Journal of Politics IOUV, October, 1962 pp 662-678; Wirt, F M and Kirst, M W The Political Web of American Schools Boston Little, Brown and Company 1972 13 March, J G "Model Bias in Social Action" op cit 14 Mintzberg, H op cit p 17 15 Neustadt, R E Presidential Power The Politics of Leadership New York John Wiley 1960 p 155 16 Cohen, M D and March, J G op cit p 150 17 Cohen, M D and March, J G op cit pp 206-216 18 Weiner, S S Educational Decisions in an Organized Anarchy Stanford Stanford University Doctoral Dissertation 1972 ... analysis of expertise The management of knowledge (2) The analysis of coalitions The management of conflict (3) The analysis of ambiguity The management of goals (4) The analysis of time The management... and the legitimacy and value of such attributes without accepting the proposition that the university should provide either the training or the certification for them FIVE CRITICAL SKILLS OF ANALYSIS... that university schools of education will continue to develop and change, but that the fundamental role of the university in the training of administrators will not disappears A professional training