-u-u‘ " -1n-n - lJi'J.'_mt . .r -‘ 'I I ' a . U .45..»- .“UII 4 _ ... , .‘ ‘2' I I..‘. . .. . . I II' 1"” I"? ‘J 'I". . .‘ .\ ‘ . ‘..‘ '- . H‘ I" :gV ‘,‘ 4 III"I I"? -, a . '.'I\I"I IV .‘ . ' ."‘ I “1‘ '.. . 3.01.1} “I". ‘ “ I.I..‘.."v" l I III I . ‘.I ' ‘ ii.“ .'\"' I 'I' In V, l‘ -. l I W231. m“:IVI:‘\ . 'v ,x “1‘ I I' '1‘ A i r“ J» I ”€38: "IKS' ’5‘ II'I I" ‘ . ‘ J‘Wt I ‘ ' M. '.. .I . . I‘ II:I‘:I IS."- a I I II'{I I Dr I n ."'."‘ I IH'W | HMg It ILI'IWIH' II . '1 -.u n -I' "I IIIIIII I. .I. I 4II all \ “Ski: ;>II.I£§ III.“ ""‘ ”4'5“: .4.1~,.;\.‘ _ I Iipxg- - . 35$. :- “:13. "‘5 ‘HI'J' (:v: \T ‘U'nnlz‘zh I' V__1I(\_ I y.?' 3.3%” U'fi'hw ”"5 w 9.55: 3M”. {SKI MIMI, “ ’4 ”.3? i" fl.y{' {:vkv 2'76 “\z‘nf.‘ “‘1“ { III." 5" 2:.“ 3'" on. II ::I$l;'a "J. '15:“ ':/‘.:I IX!" I I .774; if» 2“ . . J‘v.’¢._A lul‘III' In” I I..|‘..|:O ." .-> "J' .III‘JII. .‘ “ . lllH'II'VI . 6' ... I . I \‘II Q“ . 4% I 'I\"I.l'l‘.l {In \ I I| “5*?" _ "' ‘U‘q ' I . ,.‘ L' . .“1 ll::. f. I .I . 25".}: ”I.“ 4'“... If: I I ”I ’. III" .I’rl' VII .4 I431.” I 0W? '5‘ I'M“ \ I'I'u . lfllllllllliilfllaflllillllfllllflflllsflllljlljlljfll LIBRARY Michigan State University This is to certify that the thesis entitled Communication and Organizational Change: A Case Study and Empirical Analysis presented by James Allen Taylor has been accepted towards fulfillment of the requirements for Ph. D. degree in Communi cation 1/ n /‘ MagiZwLafi.L/f1i¢¢c""" Major professor Date July 11, 1977 0-7639 COMMUNICATION AND ORGANIZATIONAL CHANGE: A CASE STUDY AND EMPIRICAL ANALYSIS BY JAMES ALLEN TAYLOR A DISSERTATION SUBMITTED TO MICHIGAN STATE UNIVERSITY IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF COMMUNICATION 1977 ABSTRACT COMMUNICATION AND ORGANIZATIONAL CHANGE: A CASE STUDY AND EMPIRICAL ANALYSIS BY James Allen Taylor This dissertation reviews the limitations of organizational theory in tenms of communication and change behaviors prescribed by the bureaucratic school, the human relations school, rationality theorists, and organizational development theorists. A reconceptu- alization of the change process is proposed. Essentially the author argues that change is a perceptual process which is least disruptive when perceived deviations from prior organizational norms and beliefs are minimized. Five major hypotheses are tested, along with numerous cor- relaries in a time-series field test of the model. Data are gathered from a state-wide system of administrators in Special Education. Research used network analysis to separate population into communication roles. Metric-multidimensional scaling was used to study attitude change over time. Some support for the model is found, however, implementation difficulties prevented precise observations. Accepted by the faculty of the Department of Communication, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Doctor of Philosophy degree. WVW' Director of Dissertation Guidance Committee: ‘ ODLLAJL’/" , Chairman I a, / a '/ éffiér) tweet/9 I ,‘Lf’LJ IL 3 ~ 7 / fl ’ ' 71' (6”.«(1143/ V v (75M \ v. I ‘1 ‘ i. ..._. — (ILC[I‘;/,CC./ “"C.( I‘) w-~a . /. ii This dissertation is dedicated to my wife Judy, and my daughter Katherine, since without them it would not have been written. iii ACKNOWLEDGMENTS I would like to thank Dr. Nancy Kaye, Dr. Ronald Nutter, and Dr. Leonard Burrello for their assistance and the support they provided through STANSE. I would like to thank Dr. Richard V. Farace for his remarkable patience and skills at motivating behavior. I know this has not been an easy one for him. Thanks also to Dr. Michael Moore for support, Joyce Bauchner, for inspiration, and Dr. George Barnett for a number of things. Finally, I owe a debt to my wife which is literally unmeasurable. How much she disliked much of the graduate student life, I will never really know. Still, she pushed and provided and refused to let me quit. Thank you all. iv TABLE OF CONTENTS LIST OF Tums . O O O O O O O O O O O O O O O O O 0 I O O 0 LIST OF FIGURES O O 0 O O O O O O O O O O O O O O O O O O O I INTRODUCT ION . 0 C O C C O O O O O I O O O O O O O O O O O 0 CHAPTER 1. CHAPTER 2. CHAPTER 3. CHAPTER 4. CONCEPTUAL APPROACHES TO ORGANIZATIONAL CHANGE AND COMMUNICATION . . . . . . . . . . . . O The Context of Change . . . . . . . . . . . .1 Bureaucratic Organizations. . . . . . . . . 2 The Rational Organization . . . . . . . . . 3 Communication and Change: The Organizational Development Perspective. 1.4 Organizational Change: A Communication Perspective . . . . . . . . 1.5 A Cybernetic Approach to Change . . . . . . 1.6 Statement of the Problem. . . . . . . . . . 1.7 Research Setting. . . . . . . . . . . . . . l 8 Summary . . . . . . . . . . . . . . . . . . RESEARCH “PMSES 0 O O O O O O O O O O O O O O ETHODS O O O O O O O O O O O O O O O I O O O O O 1 Introduction. . . . . . . . . . . . . . . . 2 Network Analysis. . . . . . . . . . . . 3 Metric Multidimensional Scaling (MMDS). . . 4 Message Strategy. . . . . . . . . . . . . . 5 Message Selection . . .'. . . . . . . . . . 6 Research Design . . . . . . . . . . . . . . 7 Summary . . . . . . . . . . . . . . . . . . uuwuwwu RESULTS AND INTERPRETATION . . . . . . . . . . . 4 1 Response Rate . . . . . . . . . . . . . . . 4 2 Network Analysis Results. . . . . . . . . . 4.3 Message Dissemination Activities. . . . . . 4 4 Hypotheses. . . . . . . . . . . . . . . . . 4 5 Summary . . . . . . . . . . . . . . . . . . vii 17 23 30 39 42 53 57 67 67 67 73 77 79 86 87 89 89 103 116 119 155 CHAPTER 5. 5.1 Conclusions- 5.2 Future Research- 503 sumary- o e o 0 APPENDIX A: APPENDIX B: APPENDIX C: APPENDIX D: APPENDIX E: APPENDIX F: BIBLIOGRAPHY. CONCLUSIONS AND DISCUSSION. INSTRUCTIONS AND MMDS QUESTIONNAIRES- INSTRUCTIONS AND NETWORK.ANALYSIS QUESTIONNAIRE SAMPLE MESSAGE. WHOLE SAMPLE PAIR-WISE MMDS SAMPLE SIZES- KEY COMMUNICATORS PAIR-WISE MMDS SAMPLE SIZES NON-KEY COMMUNICATOR PAIR-WISE MMDS SAMPLE SIZES- vi 156 156 166 170 171 178 196 199 202 205 208 LIST OF TABLES Rationale for Concept Selection . . . . . . . . . . . .. Percent of Returns by Roles . . . . . . . . . . . . . . Percent of Returns by Intermediate School District . . Sample Response Rate for Three Waves of Metric Multidimensional Scaling Data . . . . . . . . . . . . Random Split-Half Reliability Coefficients for Three Waves of MMDS Data . . . . . . . . . . . . . . Sample Sizes for Key Communicator and Non-Key Communicator Groups for Three Waves of Pair Wise Data Split-Half Reliability Coefficients . . . . . . . . . . Comparison of Sample Nonmanipulated Dissimilarity Estimates for Key Communicators and Non-Key Communicators in Wave Two . . . . . . . . . . . . . . Percentage of Respondents in Role by Network . . . . . Entire Network . . . . . . . . . . . . . . . . . . . . Bridges in Network and STANSE . . . . . . . . . . . . . Time 1 Dissimilarity Matrix: Complete Sample . . . . . Time 1 Factor Coordinate Matrix: Complete Sample . . . Time 2 Factor Matrix: Complete Sample . . . . . . . . . Time 3 Factor Matrix: Complete Sample . . . . . . . . . Time 2 Dissimilarity Matrix: Complete Sample. . . . . . Time 3 Dissimilarity Matrix: Complete Sample. . . . . . Distances from Concept STANSE for Target Concepts and Messages Concepts . . . . . . . . . . . . . . . . Time 1 Dissimilarity Matrix: Key Communicators. . . . . Time 2 Dissimilarity Matrix: Key Communicators. . . . . vii 81 92 93 95 96 101 102 103 105 106 107 109 110 117 118 121 123 124 125 126 Time Time Time Time 3 Dissimilarity Matrix: Key Communicators. 1 Dissimilarity Matrix: Non-Key Communicators. 2 Dissimilarity Matrix: Non-Key Communicators. 3 Dissimilarity Matrix: Non-Key Communicators. Changes in Message Pairs for Key Communicators and Non-Key Communicators at Time 2. . . . . . . . . . Dissimilarities for Message Concepts Between Key Communicators and Non-Key Communicators at Time 3. Change in Message Concept Pairs Among Key Communicators Time 1 to Time 3 . . . . . . . Change in Message Concept Pairs Among Non-Key Communicators Time 1 to Time 3 . . . . . . . . . Rounded Discrepancies for Key Communicators and Non-Key Communicators at Time 1. . . . . . . . . . Comparison of Message Concepts for Key Communicators and Non-Key Communicators at Time 1. . . . . . . . . . Rounded Discrepancies Between Time 1 Key Communicators and Time 2 Non-Key Communicators . . . . . . . . Comparison of Dissimilarities Matrix for T2 Key Communicators and T3 Non-Key Communicators . . . . Obtained Correlations Time 1 to Time 3 . . . . . Calculations Between Key Communicators and Non-Key Communicators Factor Structures: Change Score Difference for Key Communicators and Non-Key Communicators for Message Concept Pairs: Time 1 Factor Coordinate Matrix: Key Communicators. Time Time Time Time Time 2 Factor Coordinate 3 Factor Coordinate 1 Factor Coordinate 2 Factor Coordinate 3 Factor Coordinate Matrix: Key Communicators. Matrix: Key Communicators. Matrix: Non-Key Communicators. Matrix: Non-Key Communicators. Matrix: Non-Key Communicators. viii Three Points in Time . Time 1 to Time 3 . 127 128 129 130 132 133 134 135 137 137 138 139 140 141 144 145 146 147 148 149 150 LIST OF FIGURES Differing Effects of Change. . . . . . . . . . . . . . . . . . . . 21 Model of Control Process for Cybernetic Model ofOrganizationalChangeooo-oo-o-oo..-coo... 35 Proposed Model of the Direction of Change. . . . . . . . . . . . . 43 Organizational Chart of the State Technical Assistance Network in Special Education- . . . . . . . . . . . . 50 Distance Between Concepts: Optimal Relations . . . . . . . . . . . 59 Distance Between Concepts: Non-Optimal Solutions . . . . . . . . . 59 Motion Vectors as Lines of Force . . . . . . . . . . . . . . . . . 83 Optimizing Criteria for Message Selection. . . . . . . . . . . . . 35 Random and Ideal Scree Lines . . . . . . . . . . . . . . . . . . . 97 Scree Lines for Whole Sample Data. . . . . . . . . . . . . . . . . 93 Scree Lines: Key Communicators Only. . . . . . . . . . . . . . . . 99 Scree Lines: Non-Key Communicators Only. . . . . . . . . . . . . . 100 Message Strategy for STANSE: First Two Positive Dimensions. . . . . . . . . . . . . . . . . . . . . . . 112 Representation of Increasing Variance-Explained for 3 Points in Time . . . . . . . . . . . . . . . . . . . . . . 120 Changes in the Distance Between STANSE and My Job: Three Points in Time . . . . . . . . . . . . . . . . . . . . . . 122 ix INTRODUCTION This doctoral dissertation is concerned with the nature of organizational change, the nature of organizations, models which have been proposed as either adaptive or predictive mechanisms for organizations experiencing change, and the role of communication in.management approaches to change. Communication and change cannot stand as constructs independent of production processes. Organizational change affects directly the kinds of inputs which are selected and processed, and the nature of outputs. Secondly, change, particularly planned change, takes place not only within the organization as a series of adoptive and coping mechanisms, but is also interaction with the environment, the recognition of environmental cues, and internalizing those inputs into the organizational planning process. Third, it is held that change takes place at both the individual and the system's level within organizations. This dissertation is not concerned with changes which are direct impositions of the environment -- e.g., changes which result from economic recession, acts of God, war -- and hence are not subject to the control of organizational managers. Rather we are interested in purposive change -- the planned introduction of innovation into an organization. Therefore, this thesis focuses on traditional models of organiza- tional structure, the relationship posited by these models between structure and change, and theories on the diffusion of change within the organizational environment. Each of these theoretical elements will be examined for the implications it holds for organizational communication scholars. It is worthwhile to note that while management theory has emerged as a social science of its own, change and change behavior v are fundamentally communication problems. At the most simplistic level change cannot be perceived if it is not communicated. Since the value of theory ultimately rests on its predictive utility, and predictions necessarily imply periodic change, there can be no theory of organization which does not deal with communication phenomena. As Barnard has stated (1938, p. 91): "in an exhaustive theory of organization, communication would occupy a central place, because the structure, extensiveness and scope of the organization are almost entirely determined by communication techniques." This thesis does not introduce a new theory of communication in organizations. It seeks to identify some of the functions of communication under conventional organizational models. From these models, principles will be abstracted which will be utilized in the formulation of an intervention strategy, to provide managers with the tools to control and coordinate communication aspects of the change process. This intervention strategy offers a distinctly novel approach to the implementation of change in complex organizations. The dissertation includes a discussion of the nature of change, an introduction to a communication perspective on bureaucracies, an examination of rational organizations, communication perspectives on the diffusion of innovations, and a cybernetic model of organizational change. Five major hypotheses, derived from the model, and numerous corollaries are evaluated in a time-series field test conducted on a large organization. CHAPTER 1 CONCEPTUAL APPROACHES TO ORGANIZATIONAL CHANGE AND COMMUNICATION 1.0 Egg CONTEXT QF‘QQANQE It is such a truism of modern society, that change is inevi- table, continuous, and omnipresent, that it almost seems trivial to discuss the gross features of change taking place in the world today. For purposes of this dissertation, however, social change, and partic- ularly organizational change, must be placed into context. Toffler (1970) argues persuasively that the next 20 years will bring an explosion in the rate of change taking place in the institutions which govern society. Drucker (1975) argues that we have entered a period of "post- industrial" expansion for which the only appropriate historical analogue is the late 19th century. Drucker indicates that the period between 1870 and 1914 saw the growth of a major new industry every decade. The metals, oil, chemicals, electric power, automobiles, fibers, telephone and aircraft industries emerged during that period as dominant economic institutions, and as principal consumers of raw~material. Similarly, this period ushered in the expansion of information and communication industries--which has continued unabated to this day. Similarly, the 19508 ushered in a new era, not only of economic expansion, but expansion in the diversity of alternative industries. For example, in the 19503 we saw the emergence of the computer industries, in the 19603 the advanced space industries; both decades witnessed the development of the nuclear industry. These decades have also seen rapid growth in information technology. Information technology has grown during both these historical epochs concomitant with low institutional stability. As organizations developed and expanded, they became subject to new institutional uncertainties. These uncertainties, arising from unstable sources of supply and instabilities in demand, aroused the need for greater and greater amounts of information exchange, both between the organization and its environment, and among organizational members. As the un- certainties associated with rapid growth and expansion increased, tradi- tional approaches to organizing human behavior--assumptions derived from laissez-faire economics--became increasingly incapable of meeting the challenge of organizational change. Drucker notes that some organizations have grown proportionately with these changes in the basic structure of societal and global socio- political relationships. He characterizes these organizations as inno- vative companies, and argues that while such organizations are very different from one another structurally and functionally, they possess common characteristics in their managing philosophies which distinguish them from "managerial companies." By managerial organization, Drucker indicates organizations which are directed through a traditional bureaucratic structure. By innovation, Drucker means a strategic alteration in either organizational form or function characterized by planning. The principle characteristics distinguishing the traditional "managerial organization" from the inno- vative organization is the way intra-organizational communication is managed. Traditional managerial organizations require that members minimize their communication contacts, conform communicatively to a rigid hierarchical structure which specifies the individuals with whom persons may communicate and the kinds of information they may receive, and which minimizes the variability of informational inputs from the environment. Huse (1976) indicates that the organization which deals with change most successfully is the organization which structures itself in such a way that changes in society and unanticipated changes in the organiza- tion may be observed and responded to. In this sense, the innovative organization restructures its internal communication network to maximize the number of alternative inputs any single employee may receive, con- sonant with productivity values. Metaphorically, this means that one is put in a position of the tree climber who can see the trees and the forest. Brewer (1971, p. 479) describes the position of Blau: "communication flow in organization hierarchies is the combined result (1) of the structurally induced communication needs of managers and operating personnel; and (2) of the opportunities that the organizational structure provides for communication between them." The differences between the traditional managerial organization and the "innovative company" lie principally in the flexible communication alternatives provided by the organizational structure. It is increasingly clear that all organizations are subject to what Ruse and Bowditch (1973, pp. 379-389) call ”the accelerating pace of change." Huse and Bowditch argue that change, as an observable phenomena, occurs most rapidly and with the greatest consequence for organizations in five principal macro-level areas: 1. 2. Knowledge: 90 percent of the scientists who ever lived are living today. The rate of both knowledge generation, and the utility of information is increasing. Hence, $25 value g§_information declines rapidly 52g ghglneed £23.22mmunicative efficiency i§_ increased. Rapid Product Obsolescence: As new knowledge is acquired old products and processes are rapidly eliminated or are rendered obsolete. This imposes a tremendous demand on workers whose skills rapidly become obsolete, and on organizations for increased flexibility and communication efficiency. The Changing Character 9; the. Labor Force: The 0.8., and the rest of the world continue to become more organized, better educated, and more dependent on service workers. This has two coma munication implications. First, the sophisti- cation with which the average worker can process informational inputs is rapidly increasing. Workers demand more and more information both about their jobs, and about the behavior of the firm (Bureau of National Affairs, 1975). Second, service institutions are essentially information processing organizations. As the number, size, and proliferation of service organizations in- creases, the gross quantity of information which is circulating within the society, and particularly information which is circulating between organiza- tions, is growing. This means that the efficiency of communication technology is being continuously upgraded, and the amount of redundancy that organi- zations can tolerate for informational inputs is rapidly decreasing. Huse and Bowditch note that, ”younger, more mo- bile, more highly educated workers show an increas- ing desire to 'do their own thing.'" (p.211) Organi- zations are more sensitive to the political conse- quences of their behavior. Government has shown a greater willingness to impose policy--environmental, racial, and ethical--on free enterprise organizations than in the past. This means that the communication environment which the organization must cope with is itself growing. It is not only that they must cope with this environment, but the organization must participate interactively in the deliberations over policy at local, state, and national levels. For those managers educated in the 19503 and before, the fact of the changes, never mind the question of implementation, has had tremendous implications for their job perfor- mance, their perception of the world, and their perception of the nature of organizations. Boulding (1973), in fact, cautions that in measuring organizational attitudes, one must be careful of deviations, since it is always possible that norms themselves are shifting. 5. Increasing Internationalization 9;,Business: Huse and Bowditch argue that the multi-national charac- teristic of international business imposes alien norms, expectations, legal systems, and political reference systems which affect the behavior, struc- ture, concerns and information-seeking of the modern firm. In addition, multi-national enter- prises impose a new set of constraints upon the ethnocentric character of communication between members of the firm. There are two ways change may be viewed: from the decision-makers to the environment, or from the environment to the decision-makers. Implicit in this distinction is an assumption made by two great schools of organizational theory about the nature of uncertainty. The tradi- tional, bureaucratic model of organizational behavior assumes that the data upon which organizations make decisions, and the consequences of organizational decisions, are known, understood, and predictable. This is, by definition, behavior under certainty. The alternative is to assume that the organization does not have full access to all data which affect decisions, cannot control the environment, and that organizational decisions have unforeseen consequences. This is 22f havior Eggs; uncertaint , and is the characteristic assumption of the rationalist school of organizational theory. 1.1 BUREAUCRATIC ORGANIZATIONS In the classical, structural-functional view of Weber (1947), the carganization is a hierarchy of supervisor-subordinate relations in which all activity derives from the decisions made at the top. It is a centralized model of organizational behavior which operates from a closed-system perspective OMiller, 1965). By "closed-system” it is meant that the organization secures from the environment a set of stable, known resources, processes these resources through an internal system in which all relationships are well-defined, and produces an output which is highly redundant, and relatively inflexible. Having defined the system of inputs, throughputs, and outputs, the bureaucracy attempts to minimize all other environmental input. Downs (1967) defines a bureau (bureaucracy) as a collection of related large organizations characterized by full-time workers who depend upon the organization for their income; promotion and hiring based exclusively upon merit; and the organization is not evaluated by outside markets. In the traditional view, government, heavy industry, suppliers of raw material, and the suppliers Of capital are not subject to market constraints. This view is supported by empirical research. Woodward (1965) looked at over 100 organizations, and found that highly bureaucratized firms tend to cluster around traditional enterprises such as heavy industry, mining, and capital suppliers, which are not subject to wide variations in demand. Bureaucracy is more than just a system of classification. Downs treats bureaucracy as a continuous variable depending upon the degree to which an organization adopts a formal bureaucratic system. In the .Aston studies, Hickson, Pugh and Phesey (1969) found that the degree of structure, or the degree of standardization and formalization of 10 of the rules and procedure3--i.e., bureaucracy--correlated highly with organizational size and found that differences in bureaucratic structure, especially differences in structure across nation states, are a function of type of product manufactured, technology, and size (Hickson, Pugh and Phesey, 1969; Inkson, Pugh and Hickson, 1970). All organizations are somewhat bureaucratic. The bureaucratic model implies that communication necessarily flows upward through an organization, and is screened by a series of gate-keepers at each successive level within the hierarchy. Without this screening process, all information collected from all points in the organization would flow to the decision makers, and the organization would bog down in a ‘morass of information overload (Farace, Monge and Russell, 1977). The nature of that gate-keeper relationship thus becomes crucial to the development of an efficient bureau. If a bureau remains relatively small, and the organization is fractioned into a minimum of levels, the amount of redundancy allowed to flow up through the hierarchy protects decision-makers from uncertainty. As the number of sub-units increases, however, the amount of information screened by each gate-keeper also increases, and hence, information which flows up contains increasing amounts of equivocality. As the equivocality of input increases, as the technology of the firm increases, and as the stability of markets decreases, the amount of communication taking place within a bureaucracy will increase, and the equivocality of those communication events will correspondingly increase. Thus it is not surprising that as organizations have become dominated ‘by’advanced technology, the bureaucratic model, and its implicit come Inunication network array, has been largely abandoned. 11 The bureaucratic model of management is insufficient by itself because it cannot meet the demands of rapid institutional change. Thompson (1967) argues that organizational structure, particularly the impediments to the free flow of ideas within bureaus, inhibits innovation. Downs (1967) points out that if a bureaucracy is large, is dominated by conservative management, has a stable budgetary basis, and exists in a politically stable environment, it will be highly resistant to change. Change, Downs (1967) notes, is facilitated by rapid personnel turnover, tolerance of a diversity of opinions, and the organizational tendency of bureaucracies to aggrandize, or to engage in organizational imperialism. Such characteristics are anathema to the bureaucrat. The bureaucrat emphasizes tenure, merit, authority, and responsibility to a well-defined organizational mission. Emery and Trist (1955) argue that organizations respond to the demand for change in the environment. They claim that as the environ- ment changes, so must the organization. If the management model held by the organization fails to provide the flexibility required for the organization to respond to external demands, the organization will either change its managerial structure, or cease to exist. I have noted that we are living in a time of rapid change and high environmental uncertainty. Environmental uncertainty interfers with an organization's formulation of its rules and procedures, since the environment in many instances changes at a greater rate than any organization can make policy adjustments. Lawrence and Lorsch (1969) point out that when the environment is changing rapidly, one of the consequences is differentiation in management styles as a cognitive 12 and emotional variable; i.e., as the environment changes, the organiza- tion tolerates a greater degree of variability among its personnel, and hence the bureaucratic model becomes increasingly irrelevant. Thompson (1969) points out that creativity (or operationally, the degree of innovativeness in employees) is a function of five conditions which are antithetical to the bureaucratic model: (1) psychological security and freedom; (2) a greater diversity of inputs available to any single individual; (3) internal commitments to the search for solu- tions; (4) a certain amount of structure or limits to the information search situation; and (5) a moderate amount of benign competition. To the bureaucratic manager personal freedom is irrelevant, and, given a reliance upon rules, undesirable. Communication inputs are minimized in the pursuit of routine work processes; decisions are made at the top, and hence, the search for solutions is constrained, and competi- tion within the organization is minimized and discouraged. The initial theoretical response to the weaknesses of bureaucratic management systems was first articulated by Chester Barnard (1938), and has been subsequently elaborated by Herbert Simon (1958), and James March (1965). They developed the principle of organizational "rationality" based upon the need for flexible decision-making and communicative interdependence among organization members. 1.2 Egg RATIONAL ORGANIZATION The rational organization conceives of itself as a system of relations, both productive and communicative, which are derived from system goals. A goal is a statement of organizational purpose to which behavior and policies of the organization refer. It is a fundamental indication of an organization's expectation of a future state. 13 Goals provide both guidelines to organizational action, and the test by which the satisfactoriness of any behavioral proposition may be judged. From goal statements, management derives policy. Rationality, then, is operationally defined as the degree to which any policy, attitude or behavior is instrumental to the achievement of a goal. The organizationis management seeks to minimize the uncertainties surrounding the achievement of a given goal state. Rather than assuming that all inputs relevant to the achievement of a goal are known, the managers assume that such inputs are, to some degree, unknown. The organization is organized in such a way that uncertainties associated with the goal state are minimized over time. Rationality and rational behavior refer to the means by which goals are realized (Simon, 1958, p. 40). The individual employee acts rationally when, in the judgment of others, his behavior increases the probability that a goal will be realized. Thus the rules for acting within the organization are not fully defined by pre-existing organizational communication structure, but are derived more from the demands imposed upon the individual by the goal itself. In this sense, organizations which are "rational" unfreeze their communication structure. As Simon (1958) notes, a theory of rational, goal-pursuing organi- zational behavior must deal with the differences between individuals. Implicity a goal demands that certain objectives and actions be under- taken in order to realize a goal; these objectives and actions may not be the same as those an individual holds for him/herself. Thus entering into the process of goal selection are such variables as internal value systems of the individuals who make up the organization, the concept of organizational norms, ethical considerations, and variations in indivi- duals' personal conlnunication effectiveness. 14 Barnard (1938) and Simon (1959) shift the focus of organizational theory from formalized mechanisms of control to deliver known and well- understood services, to maximizing directed behavior at the individual level in pursuit of goals which are themselves subject to change over time. The environment is an interactive external element which is not assumed to be stable and which is not assumed to be under the control of the organization. The emphasis is on the interplay between techno- logical, sociological, cognitive and psychological factors of behavior within the firm, and on the nature of extent of equivocality external to the firm.(March and Simon, 1958). March and Simon (1958) point out that the fundamental difference between bureaucratic models and rational models is that the bureaucratic 'model assumes that the organizational members are passive instruments, whereas the rational approach assumes that the individual brings to the organization attitudes, values, and goals which are themselves part of the resource pool upon which the organization draws in developing solutions (Krupp, 1961). Within the rational approach, the organization also brings to the individual a set of values and norms which Taylor (1975) has shown have a strong effect on an individual's perception of the work environment and the nature of appropriate behavior within that environment. The rational approach may be characterized as an "open systems" model of organizational behavior (Miller, 1972). "Open systems" implies that the organization attempts to maximize the absolute number of total inputs available to decision-makers. After having received this input, decision-makers then apply patterns, usually casual, to the analysis of the data, and from these patterns abstract strategies which are applied as productive processes. 15 The open systems concept has been explored empirically by Burns and Stalker (1961), Chandler (1962), Emery and Trist (1965), Woodward (1965), Thorelli (1967), and Lawrence and Lorsch (1969). In brief, these researchers have argued that a comprehensive understanding of organization functioning and behavior requires an examination of the variety of energy transfers (inputs), both within the organization and between the organization and its environment. They have found that organizations and their management styles vary along a continuum from closed and mechanistic, to open, organic-organizational structural designs. As we would expect, from the comments of Drucker and others, this continuum tends to reflect the degree of innovativeness of organi- zations (Lawrence and Lorsch, 1969). Simon notes that as organizations become "more rational," in rough correspondence to the continuum noted above, the organization adopts a model which is essentially divided into two sub-parts: (1) a theory of motivation which explains the decisions of people who participate in and remain in organizations; and (2) a theory of decision- making within organizations comprised of such people. In the motivational theory formulated by Simon, it is postulated that the motives of individuals can be divided into inducements and contributions. Inducements are positive rewards desired by the members of an organization; contributions are participant inputs to the organi- zation's productivity, but generally have negative utility to partici- pants. In other words, inducements are the benefits one receives from 'work in an organization, and contributions are the things one gives up as part of the exchange. To the extent that this equation is 16 maximized, each individual maintains his position and productivity within an organization. Inducements refer to more than monetary benefits. There are also benefits such as vacations, health plans, etc., and perquisites such as the size of the office, the extent and range of control, responsibi- lity and social experiences. In order to maintain the balance between the individual's inducements and contributions such that motivation to produce is maximized, the organization is compelled to allow employees to "negotiate" the discrepancy between their internal value system and the value system implicit in the choice of organizational goals. Thus, communication between subordinate and supervisor becomes less a process of simple directives from the supervisor to the subordinate than a negotiating process on the most mutually beneficial supervisor-subordi- nate relationship. From this perspective, organizational change takes on a new meaning. A change in predominant organizational goals means a change in the pre- dominant assmmptions an individual makes about his/her work. The organi- zation management's response must be an increased willingness to provide information which relates to the goal-state of the individual. To the extent that organization managers can communicate the relevance of a change in goals to the instrumentalities (inducements) of the individual, the disruption caused by change will be minimized. Secondly, to the extent that an organization experiences a great deal of change in a short period of time, the amount of energy or communication necessarily expended will be proportionately greater. 17 1.3 COMMUNICATION QNQ‘QQANQE: TEE ORGANIZATIONAL DEVELOPMENT PERSPECTIVE Organizations traditionally have not looked upon change as a con- tinuous process. Despite the best efforts of organizational change theorists, change continues to inhibit system performance. Organiza- tional researchers, concentrating largely on group process models of change behavior (Bennis, 1966, Dickson, 1966, Marrow, 1967), have developed techniques of consultant intervention--organizational develop- pgppr-which are employed by organizations undergoing either shifts in behavior or shifts in policy. These intervention strategies emphasize the need for interpersonal communication among managers, both laterally and horizontally, to maximize congruity between organizational goals and the personal goals of individuals acting to achieve the organization's goals. Organizational development theorists emphasize the socio-personal processes associated with management. They look to the importance of "good supervision and leadership," defined as a supervisor's ability to obtain willing cooperation from a subordinate. From Simon, this would be the ability to utilize maximally an employee's inducements. Organizational development theorists (acting as change agents) attempt to train managers in the principles of effective communication: the importance of recognizing group norms and behaviors, the importance of inter-group relations, and the importance of what has been called the informal communication network (Bennis, 1966). Typically, the organizational development consultant encourages knowledge and understanding of these process variables by the use of some form of group confrontation (Huse, 1976). One of the features 18 of the organizational development approach is its systematic approach. It may not incorporate many members of the system into the intervention activities, but the consulting change agent is interested in estimating the effects of intervention at one point on the overall system. Lewin (1948) developed a three-step systems model in which the organization experiences a period of "un-freezing, changing, and re- freezing." In the un-freezing stage the organization's members come to recognize the need for change, or the fact that it is changing indepen- dent of its own inclinations. When an organization is "un-freezing," managers are coming to understand that change i; taking place, or the need to intend to change. This period is usually characterized by an increase in the frequency of informal communication within the system. .As system members try to reduce their uncertainty, they seek out organi- zational others, most likely opinion leaders, in order to identify the range of alternative solutions, and the appropriate expectations which are associated with the change. During the "changing" phase, the organization experiences any adjustments in structural re-alignment required to deal with the change, or which are themselves the manifestation of change. After the organi- zation changes it freezes into a new form for some unspecified period of time. Lewin looked upon change as not necessarily continuous, but cyclic and sinusoidal. Other organizational system intervention specialists, particularly Lippitt (1975), drew upon Lewin to develop a parallel model of consultant intervention for "planned change"; the resultant model is outlined below: 19 Phase 1: Scouting. In this stage the change agent and the client develop a relationship but remain uncommitted. Phase 2: Entry. The client and the consultant develop a contract stipulating their expectations, mutual goals, the role of the change in the organi- zation, the methods to be employed. Phase 3: Diagposis. The change agent identifies (1) the problems of the client; (2) the goals of the client as a response to the problem; (3) the resources the client can draw upon in solving the problems; and (4) the resources the change agent can apply to the change situation. Phase 4: Planning. Having identified the problem and having identified the resources, the change agent develops the steps to be taken in instituting a planned change. At this point the change agent attempts to identify the problems the change will entail, and techniques for coping with these problems, particularly human problems. ‘gpase 5: Action. In the action phase the planned change is implemented, and intervention strategies, particularly communication strategies, are set forth. Phase 6: Stabilization and Evaluation. Following the implementation of a change the change agent and the organization attempt to stabilize the situation and evaluate the effect of the change. Phase 7: Termination. The change agent leaves the system. Implementation of this model of system-wide organizational develop- ment has created perhaps as many problems as it has solved. First, organizational development is not based upon a theory of organi- zational change, but rather is designed to respond to acute situations. A change agent is rarely called prior to an organization's recognition, at least a tacit recognition, that dysfunctional change has already taken place. Such recognition is usually based upon an increase in the level of conflict, a drop in employee morale, and/or increasing instability in the labor force (Schein, 1967). 20 Secondly, the choice of which individuals are to be represented in the change effort is made largely by the client, and not by the organizational specialist. Hence, key individuals, who themselves are resistant to change in pursuit of their own goals, are left out of the change negotiation process. Third, and most significantly, these models do not treat change as a continuous process. These models treat a specific change as terminal, both in its development and in its conse- quences. Finally, evaluation efforts tend to be one-shot quasi-experi- mental or field-study designs, and hence, little experimental valida- tion exists. Since the models do not take a process view of change, organiza- tional implementation efforts have often failed (cf., Litwin and Stringer, 1964; Lawrence and Lorsch, 1965). I would argue that this is largely because in the organizational development model, most manage- ment innovation specialists do not see organizational change as an adaptive process, involving the negotiation of intrapersonal value states, organizational value states, and environmental value states, all within the context of some hierarchical structure. Another perspective is provided by Lawrence and Lorsch (1969), who argue that change is a continuum which varies from changes in interaction patterns (least severe) to changes in key personnel (most severe) (Figure l-- Lawrence and Lorsch, 1969, p.187). Differences in interaction patterns, or restructuring, result in only modest behavior change; i.e., the obser- ved changes here will be in terms of the formal communication network-- the formal hierarchy--and are not significant for the organization. However, Taylor (1975) and Danowski and Farace (1974) have shown that an individual's formal communication patterns have long term consequences 21 HGCOdUSM o>auqnmoo Ilwwsooo hwoumuuu scams u mausoABSocw mo unoaoomano mmuuouauo souuooaou 3oz madman nanosecond usonomuao mandamus ouosou 332 1i. ouauosuuu huwuonusm one wooed mo unemoa>ao 3o muaouwoun auscuuuoooo o>aosousn scaumounsaaoo mo ufiosmmno Amaoumuo 30c n.0uo undamaged .muowooo .noosuoa sequenuouooo women soz .mma .n .aoma .zoauoa one canvass; scum A.ouo.so«usu~«mme.uosoa.uomao>oanomv ownmno uoa>msoo Huumoamoosm A, Amll 3 anon—ammun— aus~e> o u souumuooquo unencummm acouuouoo no oaou usououuan 4‘ masons u0fi>msoo umoooz uuoo um.~oco«uoao one o>uuwswou we was ooouoa museum swamoo mo muommmm wseuommfin H muowfim ouomumu mucosa 22 for attitude development and the individual's perceptual set. While this kind of change may represent the least amount of overt behavior change, from the network perspective it is likely to have the greatest long term effects. At the second level of their change hierarchy, Lawrence and Lorsch place changes in role expectations. Operationally, this is a change in the activities performed on the job. While it is clearly a step up in difficulty from network restructuring, role redefinition is not indepen- dent of network structure. If one changes the job, hence changing the uncertainties associated with the job, one changes the information search behaviors, and the network. At the third level of behavioral consequence, Lawrence and Lorsch argue for the effect of different orientations and attitudes. Never ‘mind the weak relationship between behaviors and attitudes (Siebold, 1976), it is difficult to operationalize the distinction between chang- ing values and changing role states. At the fourth level, Lawrence and Lorsch look to variations in selection criteria and replacement of incumbents (particularly powerful incumbents) as resulting in the most fundamental behavior change. First, in any organization, particularly a large organization, replacement of powerful individuals does not often result in a significant redefinition of'a firm's mission, and hence, in the short run it does not change nmach behavior. In the rational organization individuals are instruments .for'the achievement of goals, which are themselves instruments for the achievement of individual's goals. It is a cyclic process, and to the extent that norms about the organization are well institutionalized-- 23 through agreement among large numbers of individua13--the system will be only marginally affected by personnel changes. More fundamentally, motives are operationalized, as attitudes toward inducements. That is, a motive to perform an act emerges when an inducement approaches some threshold as a limit. The key point here is that Lawrence and Lorsch's list is not a list of different changes calling for different strategies, but a list of factors or variables which may all be present to a greater or lesser degree, in any change. They assert that behavior change is associated with a continuum ranging from cognitive (objective) changes to emotional (subjective) changes; we can hypothesize that to the degree that a change involves the replacement of personalities-~as opposed to the replacement of task--the emotional consequences of change will be greater. This means that the change agent would have to be more concerned with communication about intrapersonal instabilities than about interpersonal instabilities within the overall responding work force. The two concerns are not easily distinguished. Changes in a worker's intrapersonal set (motives, attitudes, inducements), if negative, will carry over into the interpersonal communication behaviors. Their mes- sages will change, their frequency of contact will change, and the peeple with whom they interact will change. This means that the posi- tion of Lawrence and Lorsch is not predictive of outcomes. It does not specify the manner in which changes diffuse, and the structures which might be designed to minimize disruption. 1.4 ORGANIZATIONAL CHANCE: A CQIMUNICATION PERSPECTIVE Communication research scholars, in the field of the diffusion of innovations, provide a second model which can be applied to the organi- 24 zation. Katz, Leavin and Hamilton (1972. p- 69). defined the diffusion of innovations as a process which may be characterized as the: 1. acceptance, 2. over time, 3. of some specific item--an idea or practice, 4. by individuals, groups or other adopting units, linked, 5. to some specific channels of communication, 6. to a social structure, and 7. to a given system of values or culture. I take the perspective that the diffusion of innovations should be one of the principal activities of a manager's efforts. The organiza- tional context, of course, differs somewhat from the societal context to which Katz, et al. (1972) referred. First, the manager may be required to make an independent decision to adopt. The decision to choose an innovation, then, is made by only one individual. Thus, the issue of acceptance, or instilling changes of policy, procedures, and product innovations into the employees' perceived inducements and con- tributions equation, becomes the focus of a manager's diffusion efforts. The earliest articulation of an innovation model of change in organi- zations seems to be Graham.Wa1as' (1914) formulation of the construct of creativity. Walas identified four phases in the development and imple- mentation of a creative or innovative idea: (1) preparation (assembling) the inputs, identifying the problem; (2) incubation (the unconscious or pre-conscious combining and re-combining of internalized components); (3) illumination (sudden insight into the solution); and (4) verifica- tion (testing through communication). This is rather similar to the Rogers and Shoemaker (1972) model of the diffusion of innovations. Taylor, Farace and Monge (1976) summarized Rogers and Shoemaker's process as follows (p.12): 25 1. Initially a subset of individuals within a culture must become aware of the existence of the new social object; 2. The culture must exhibit interest in the innovation; 3. There must be an opportunity to evaluate the charac- teristics of the innovation; 4. There must be a trial of the innovation within the social system; and 5. The innovation is formally adopted. Adoption, as Taylor, et al., have pointed out, is "the point at which the innovation may be said to have been integrated into the body of belief which defines appropriate behavior." For the manager, this means determining the a priori inducements which are, at the time of adoption, balancing an employee's contributions. Any change which is not in some way related or evaluated in the light of pre-existing expec- tations about the nature of the job, or the work an employee performs, will very likely change that balance in favor of contributions. As we have noted, when contributions grow greater than inducements, the employ- ee's motivation drops. Consequently, either his/her productivity drops such that contributions equal inducements, or the employee leaves the firm. Crucially, then, the manager must communicate the relationship between (1) past inducements, both psychological and financial, (2) new inducements or rewards associated with adopting change, and (3) the con- tributions an employee has already been sustaining. The amount and kind of information needed to clarify the relation- ship between inducements, contributions and a new innovation is depen- dent upon the degree to which the recipients of the innovation perceive the innovation to differ from their existing belief and/or behavior system. The manager analytically determines the a priori relationships between 26 an innovation, past inducements, and past contributions. Once that relationship is established, pre-existing communication channels within an organization can be utilized to maximize the probability that change will be adopted with a minimum of social disruption. This argument can be restated as a series of propositions: l. A manager may "adopt" an innovation unilaterally; 2. to the extent that an innovation is integrated into the body of belief (within an organization) about the appropriate balance between inducements and contributions, social disruption will be minimized; 3. to the extent that a manager communicates that relationship, adoption is rendered more probable; and 4. the greater the discrepancy between prior behaviors and new behaviors, assuming no change in perceived inducements, the greater the resistance to change. As Taylor, et a1. (1976) have pointed out, these propositions have several implications. First, the organization's management may choose to adopt an innovation without notifying the rest of the hierarchy and install it, in effect, by fiat. However, the installation of a change, per se, does not guarantee its "adoption." Second, only when an innova- tion is related to the corpus of belief about appropriate behaviors, and the distinctions minimized, will organizational members change their be- havior in some permanent manner. Third, as Newman (1965) pointed out, the uncertainties associated with organizational change are such that the consequences of altering beliefs and exchange equations within organi- zations are not always predictable. This means that the subordinate not only needs to be sensitized to the nature of the change, but also to the need for information search activities, and reporting on those activities in order that the consequences of change may be fully identified. This 27 is rational behavior. The manager must be capable of reconciling an innovation both to the context in which it is to be employed and to the social system.which will be artifactually altered. In order to meet the demands imposed by the adoption of innovations, the manager must monitor the process of change while simultaneously faci- litating the introduction of the change. We have noted that conventional organizational development theorists do not emphasize the continuous measurement of change. They make the assumption that all the variables, and all the elements associated with change, are either unknown initially, or never will be known. It is the case that the manager can a priori define some critical elements associated with the goal-state, which will be changed by the adoption of an innovation. The implication here is that the manager takes a "systems" view of the organization, and of the organi- zation's expected behavior changes consonant with the change effort. The manager attempts to monitor and incorporate many factors imping- ing upon an adopting individual, factors which arise from the constraints imposed by the organizational environment, external and internal. In this way the balance between inducements and contributions at the individual level, and the balance between programmatic pursuit of a goal state and institutional uncertainties at the organizational level, can at least be approximated, if not maintained. This conceptual discussion of organizational change stresses that any change in organizational practices and behaviors occurs over time. Hage (1974) argues that it is possible to view changing organizations as systems which can be modified by cybernetic controls. According to Hage (cf., Monge, 1974), a cybernetic system is characterized by: 28 1. a measurable goal state 2. realizable parameters around that goal state 3. a system of control for maintaining the system within those parameters 4. feedback to verify that the system.is within those parameters 5. a regulating mechanism which keeps the system within those parameters Conventionally, cybernetics have been employed in the development of self-regulating production processes, but Hage (1974, p. 27) notes that the cybernetic theorist "start(s) with the simple assertion that the system of variables is a production process with inputs, throughputs, and outputs," and that there is no reason to assume that the concept of production needs to be limited to the creation of a well-defined product. In other words, in the management of individuals, it should be possible to maintain, through the adjustment of the inducements-contributions equation, a set of behaviors which maintain a steady state performance which over time collectively constitute the realization of a goal state. Organizational development models, since they do not provide for continuous feedback, do not enable us to develop cybernetic models of organization. In order to monitor cybernetic systems (or any systems model) data must be gathered at multiple points of time, and the data must be regularly and repetitively fed back to the manager 30 that ad- justments in the inducements-contributions equation can take place. we do not suppose that managers can automatically make changes in finan- cial inducements; what instead is proposed is that the manager can adjust, through the input of information, the intrapersonal satisfaction elements which Simon (1958) indicates are part of the set of inducement factors. 29 By gathering, from employees, attitude data concerning both the organization and an individual's job, a manager can input information, through opinion leaders, which relates activities performed on the job to organizational goal-states and to the intrapersonal goal-states of individuals. Once a range of satisfaction has been established (or a satisfaction parameter has been fixed) managers can determine when to take corrective action to restore inducement-contribution balance. Communication is the means by which the manager can make regular adjustments in the instrumentality-contribution equation, or the balance between rewards and performance. Since the value of inducements is, in effect, culturally defined, the manager can take advantage of the organi- zational culture to redefine the value of various kinds of inducements. The re-definition process essentially involves relating objects which are known to be valued to objects which are not known to be valued, and to stress that relationship through communication provided over a con- siderable period of time. The key to this process, as Taylor, et al. (1976) note, is that change, both in terms of organizational behaviors and attitudes of organizational members, occurs through identifiable system structures and communication acts. Huse (1976) points out that organizational change takes place both in formal and informal relationships within a larger social system. That is to say, there is a complex of social structures which is related to the change process. This complex enables employees to judge the critical factors associated with their work which become either inducements or contributions. In the next section, I will describe a conceptual and methodological approach to the change process which has its roots in the organizational development models we have discussed, but which 30 provides for the manager of change a communication approach to balancing interpersonal inducements and contributions among the labor force. 1.5 A CYBERNETIC APPROACH IQ CHANGE A system member's response to an innovation depends upon the degree to which that innovation is congruent with existing inducements. There- fore, we will say: pp organizational innovation gill pg_adopted £p_phg extent ngp i; g perceived £2 possess attributes congruent 3111:}; £135 conceptions p§_ppggp_pppk, For the purposes of managing the dissemina- tion of an innovation within an organization, I will dismiss monetary inducements and status perquisites from the list of inducements directly controllable by the manager. It is true, as many unions have found in their negotiations (Chamberlain and Kuhn, 1974), that technological change within organizations is often accompanied by either changes in the salary structure for specific job categories, and/or reassignment of employees to positions of higher status as a result of the technologi- cal change. These variables are subject to company-wide policy which may or may not be directly under the control of any given manager, but are assumed to be adjustable as a part of the change process. An object becomes an inducement when the object is a pre-condition for performance on the job. The degree to which an object is a pre-con- dition to performance varies with attitudes about the object. It is the attitudes which can be changed through message input. In this sense an inducement can be an attitude about an attribute of the job from which the individual derives satisfaction. Studies of informal communication networks (Proctor and Loomis, 1951; Danowski, 1974; Taylor, 1976) have 31 shown that these attributes are defined through interpersonal processes which take place informally within the organization. That is not to say that systems managers cannot define some key attributes and key sources of satisfaction as part of an individual's job description, but as we have noted, the value of a job's attributes is culturally determined. On the surface, the assumption that an innovation will be adopted, to the extent that it's related to those aspects which an individual defines as cricial to his/her motivation to perform, seems relatively straightforward. It does not seem to reflect any radical departure from the commonsense notion that persons tend to adopt those things that are compatible with their own existing role view, and to shy away from those things that seem foreign or alien. What makes this assumption interesting is the logic that underlies an individual's concept of the work--his/her job-~and the process by which an innovation comes to be seen as compatible with the job concept. The concept of the job as a psychological and cognitive phenomenon has been studied by Taylor (1915), Hare (1967), Thompson (1969), and Pekar and Barrack (1976). The results of these largely base-line re- search efforts show that a person's job, his/her performance, the job's status and its characteristics play an important role in determining an individual's overall life satisfaction. In this view, the job becomes .a central referent by which an individual defines his/her position vis- iawvis other actors in the social system. The individual observes other individuals responding to him/her and from that extrapolates his/her position in the world (Hare, 1967). 32 Like anything else, the concept of the job is open to change. Adopting a new machine, new technology, or new accounting practices fundamentally changes the way in which people behave on the job, and hence, their perception of their own position in the world. For people in the world of work, the concept "my job" typically is built up from a large repertoire of prior message experiences. Some of these occurred in early childhood socialization (through the mass media, and interper- sonal sources). Some are culturally (or sub-culturally) delimited, and others are specific to the work history of the individual. The aggregate of all such message events yields the reinforcement schedule which lends power (instrumentality) to certain attributes. The power of the attribute lies in its degree of relatedness to the individual's overall psychologi- cal conception of self. Barnett, Serota and Taylor (1974) indicate that for certain atti- tudes, the message history is so large that the attitude is "massive." This means that the number of messages which collectively have defined that attitude, its magnitude, and the relationship between the attitude, other attitudes, and the attitude's contribution to the definition of self is very large. I would argue that a person's perception of the job, and the attributes of the job which represent inducements for the individual, are very massive. They result from a great history of infor- mational stimuli. This suggests that the concept "my job," as a psychological phenomr enon, and the instrumental attributes which are related to it, are not likely to be altered substantially by the volume and kinds of messages which a manager can produce in a change project. Given that one's con- 33 cept of "my job" is not readily changeable (or movable), how then can the manager take advantage of this cognitive phenomenon in a change effort? I would suggest that the manager begin by determining those attri- butes of the job which are instrumental to an individual's performance, and selecting from that set the subset which on the average is "most instrumental." Then the job of the change manager becomes one of representing, through a series of messages, the degree to which an innovation possesses those attributes which are most instrumental in a person's job performance, and in their definition of self. We can restate this as two propositions: 1. To the extent that an attribute is instrumental to performance on the job, it will be closer to a person's definition of self. 2. To the extent that an innovation possesses attributes which are themselves close to individual's definition of self, the proba- bility of adoption is increased. These propositions provide the key to constructing a cybernetic model of organizational change. Monge (1974) argues that a cybernetic system essentially consists of (l) a phenomenon to be controlled--in this case an innovation; (2) the variables that affect the phenomenon-- in this case a set of individuals representing alternative transmission units, each of whom is possessed of a set of instrumentalities or induce- ments and contributions; (3) a set of information about the system; (4) a goal-state which defines appropriate system level behavior; and (5) a control mechanism which monotors the degree to which this system is approaching a goal. 34 Cybernetic approach requires a system of gathering and monitoring data about the system and feeding it back. It means that one knows the parameters which are indicative of acceptable performance. Thus, the control center has the ability to respond and make changes in the system as a result of feedback originating from the system's members. The model as outlined below (Figure 2) shows the manager in the center point of the system. At the initiating point the manager has information about attitudes which are instrumental to employee perfor- mance and motivation. Secondly, the manager has information about the nature of the change, and some prediction about the behavioral and attitudinal consequences of the innovation to be adopted. Finally, the manager has a goal-state, hypothetically a non-disruptive adoption, but clearly the goal is related to some overall expectation of system perfor- mance. The initial responsibility of the manager is to provide informa- tion which defines the innovation in terms of those attributes which are instrumental for the employee. W3 note that the manager should ideally identify opinion leaders, or key communicators (operationally, indivi- duals who are highly linked within the communication network), and allow them to pass information through the system interpersonally. Bennis (1966) notes that information about the organization is much more likely to register on individuals' perceptions if it is transmitted interperson- ally. The model implies that opinion leaders, in some sense, define the values for the rest of the system. Taylor and Bauchner (1977) have pro- posed a test of the theoretical assumption that opinion leaders set the agenda on values for the rest of the organization. It is also possible 35 On— uu com 3 Ill-l. osmuugojs — scuuonacumuo — — ouooooa sausage — oo=~s> nouonsoo l 33>: 1: commune: nuancn 1:306:2— amouuuum smash» .u .omcuaoz nooam> .o nouooooun .u F 035165. ououusuum ~o3p0uou ouauoouum fiqauom soda-wuoomwo coquoroccH ouuouom A sou coeds ii‘ owsuso —uco«uonucsmuo mo demo: ouuuouoozo new nooooum gowucoo mo demo: N m¢=UHm couscou«>:m— a mafia 36 that opinion leaders become so because they have values which themselves are‘p‘ppgppi valued by the rest of the system.members. By inputing this information, the relationship between an innovation and attitudes of system members (which are instruments of performance) is clearly defined. This has two effects: first, in a co-orientation sense, the manager and the group of subordinates experience perceived agreement on the nature of those system values, experience understanding about those system values, and become more accurate in their prediction of the relative magnitude of each other's position vis-a-vis the objects related to the attribute. Second, as Farace, Monge and Russell (1977) note, accuracy between supervisors and subordinates enables communication to take place at a high level of efficiency. This is because two indi- viduals in a supervisor-subordinate relationship can avoid problems of communication which stem from a lack of understanding of the meaning of important objects which are the subject of the communication act. Farace, Monge, and Russell (1977) caution against the existence of "monolithic consensus." By monolithic consensus they mean that there is agreement about the meaning of all objects, or agreement about the ‘magnitudes of attitudes toward those objects among all members of the system. Certainly I am not advocating such a position here. What I am advocating is that for objects that are instrumental to a worker's per- ception of his job, the manager must recognize the importance of that inducement, and propose the innovation in the context of that inducement. By continuously measuring inducements or attributes, and by applying appropriate time-series analytical procedures, the manager can determine the degree to which change in the perception of the innovation, and 37 change in the behavior of the system.vis-a-vis the goal-state, is appro- priate both in its magnitude and its direction. That is to say, the manager can ask the question, "Is this goal being met?" or alternately, "Are we tracking on our goal-state?" If the results of data show that the answer to that question is pp, then the manager adjusts both the information levels and perhaps the kind and quantity of transmission channels the manager alternately employs. If the answer to the question is es, the manager continues current activity levels and imposes no change. Thus, in essence, this model proposes four key factors which will enable the manager to overcome many of the present deficiencies of current organizational change models.' (1) Time series analytical methodologies that provide for the con- tinuous monitoring of critical inducements throughout the change process. Thus, changes in trends can be quickly spotted, and program efforts can be redirected accordingly. (2) The change process should be conceived of as a complex phenom- enon best represented in a multi-variate configuration, so that precise changes in key variables and their interdependent consequences can be tracked over time. This allows the manager to both make changes in the volume and quantity of information being input into the system, and assess- ‘ments of the unanticipated (uncertain) consequences of an organizational innovation adoption. (3) The objective of an adoption is to bring into congruence the instrumentalities by which people obtain satisfaction from their work and the innovation itself. To the extent that the innovation is perceived 38 to possess attributes which are highly instrumental in an individual's definition of self, adoption will occur more rapidly. To the extent that information specifying the relationship between the innovation and the attributes is diffused interpersonally, the information will be more efficient. If the discrepancy is large, significantly more communication resources will have to be expended. (4) The focus of communication campaigns will be to link key instrumental attributes of persons' perceptions of their jobs, and the innovation; this can best be done by disseminating linking messages and separately gathering independent validation that the innovation in fact reflects the attributes the message purports it to possess. The innovation must be related to a person's perception of self, be- cause a person's perception of self cannot be rapidly changed to reflect the attributes of an innovation. The model does not cause change. The model is efficacious for the installation of planned change. By "efficacious" I mean that the model is one of a number of tools at the disposal of a manager planning the implementation of change. This model is, however, particularly valuable. It provides a basis for linking specific changes to goals of the organization and goals of the individual. It provides criteria for specific messages «of great power and utility for sponsoring perceptual changes. It calls for multiple measurements of change to allow for strategic changes in the message strategy in response to structured feedback. It allows for the observation of organizational structure,and sensitivity to organi- zational structural context in the initiation of change efforts. 39 Finally, and most crucially, it requires that managers be sensitive to the needs and expectations of system members. Causing changes between attributes without providing evidence of the behavioral validity of the claim defeats the implementation effort. Similarly lack of funding, failure to sustain message flow, and policy changes can all unhinge change effort prior to adoption. 1.6 STATEMENT 93 Ill; PROBLEM Organizational change, to be rendered consistent, requires the continuous transmission of messages to guide the behavior and perceptions of system members. The proposed model stipulates that an observant manager, operating a "transforming" system, must dedicate some propor- tion of the communication message load toward the reduction of stress responses to perceived variation. By reducing stress, the manager re- duces disruption and increases the probability of adoption. If the manager has some fixed notion of the tolerable rate of change for the system under consideration, the manager can fix tolerable upper limits--i.e., the rate at which change can be imposed without significant dislocation of production processes. The goal is to push the system to the upper limits of tolerable deviation without substantially reducing production norms. This addresses the way an innovation fits into a continuing cyber- netic monitoring system. Generally, cybernetic controls are imposed upon stable systems. In this case, we are considering the imposition of change upon a theoretically controlled system, and pushing the system to its theoretical limits. 40 Having placed demands to change upon the system, one of three options seems most likely. Either (l) the system will reject the inno- vation and return to prior normative behavior; or (2) the system will adOpt the innovation and return to prior values; or (3) the equilibrium ranges will be altered. Of the three options, the third seems most likely, particularly when the innovation represents a major shift in organiza- tional processes. Imposed change from either external or internal sources can redefine the acceptable tolerances within which the systemlmay operate. That is, significant shifts in the demands of the environment for altered output imposes a new set of constraints upon system behaviors. The manager must "re-tune" the system while maintaining satisfactory performance levels. The program or process which represents the internal response to the demand for change is the innovation. Inducing adoption requires an increase in the amount of communication resources which are directed at new processes, and a consequent reduction in the communication re- sources directed toward conventional behavior. The research question related to the argument above, which this dissertation addresses, is: To what extent can the systematic input of messages re-direct the perception of organizational members such that an innovation comes to be perceived as normative? The research reflects an attempt to systematically organize and implement a communication strategy which optimizes the directed effect of each message. It is proposed that organizations consist of bureaucratic hierarch- ies, rational goal pursuing activities, and informal communication net- works. By identifying the informal leaders through a reliable method, by providing opportunities for managers to participate in implementation 41 processes, and by creating a uniform goal-directed communication stra- tegy, change can be effected. First, the manager of the dissemination effort selects out those in- formal communication others who control the flow of information. Re- search by Jacobson and Seashore (1951), Likert (1961), Guetzgow (1965), and Farace, Monge and Russell (1977) has indicated that such individuals powerfully affect the perception of the system regarding new processes. Information directed toward these "key communicators" can induce some change at a minimum of resource cost. If the messages are tied to the prevailing instrumentalities of the individuals, there is a high probability that change in the key communicators will be reflected in their ongoing interactions with network others. As shown in Figure 3, changes in key communicators at one point in time are likely to produce changes in non-key communicators at subsequent points in time. The model shows that between T1 and T2 non-key communicators should minimally change (0), inasmuch as information does not filter down. Informational stimuli are provided to key communicators which cause change (A) between T1 and T2. Between T2 and T3, non-key communica- tors change to the position of key communicators at T2, hence the zero sign on the diagonal between X2 and Y3, and the delta sign between Y2 and Y3. A zero for the line between X1 and Y1 indicates an assumption of initial equilibrium. Thus, by redefining the perceptions of key communicators, it is possible to strongly affect entire organization. The problem is ini- tially isolating key communicators; i.e., separating them from normal relational processes. Since this cannot easily be accomplished, it is 42 expected that some information will continuously filter down. There- fore, the manager chooses points in time to monitor the system which are wide enough to allow change, yet small enough to detect differences between the two groups. The precise length of time is dependent upon the extensions, on potential for impact of the innovation, and the manager's ability to produce and introduce innovations, the criticality of rapid change, and the availability of funds and personnel for research purposes. Further, the manager recognizes the need to incorporate information about the innovation in memoranda slated for general distribution. This reduces the potential that non-key communicators will perceive obsequiousness vis-a-vis the innovation. The purpose of this research is not to test the cybernetic nature of the system, since that would require involvement lasting over a per- iod of years, but to test the efficacy of the innovation implementation strategy. This strategy was tested in an organization at three points in time, over a 19-month period. 1.7 RESEARCH SETTING The proposed model of organizational change was tested within the state of Michigan, Department of Educatidn, Special Education Services Unit, Mr. Murray Batten, Director. Included in the State-wide services organization are administrators in intermediate districts, and administra- tors at the local district level. All members of the system.are tied to a common funding base, common legislative codes, and overlapping admin- istrative jurisdictions. The State Department and its corollary units at the district level provide administrative support for Special Education students. No teachers 43 FIGURE 3 PROPOSED MODEL OF THE DIRECTION OF CHANGE T1 T2 T3 stimulil stimuliz l key communicators x1 % ax, -——x 9x3 . ’l o o W W Y1 0 Y2 A #3 non-key communicators 44 were directly employed in the research effort, although administrators in rural districts might well spend a percentage of their time as teach- ers. The focus of the research effort is on managers -- progrmm managers or system.managers -- and their collective responses to an innovation. Considerable effort has been expended on explicating differences and similarities between bureaus, rational organizations, and change- oriented organizations. This effort was undertaken to provide the reader with a sense of the magnitude of effort involved in transform- ing bureaus into modern systems-oriented orgnnizations. The present effort is directed at introducing a mechanism for participatory decision- making into the Special Education administrative framework. The inno- vation we will consider is a system for gathering the perceptions of administrators remote from Lansing, and systematically imputing that data into State Department decision-making processes. Change has been a profound and continuing problem for administra- tors in special education since 1968 (Kay 1976, p. 1) notes: These forces for change have affected the individual child level in the form of alterations in the procedures for determining eligibility and placement. Changes in the scopes and type of program for the handicapped have also occurred at the individual child level. At the system level these changes have affected the nature of the relationship between general and special education. At the state level, both intermediate and school system relationships have been altered from a service relation- ship to a monitoring and compliance one. These changes have affected the way in which local and intermediate units relate to each other and to the State Department of Education. Other areas that have been affected include the philosophy of special education, resource allocation, staffing patterns, and training. In 1972, in response to increasing parental litigation and in 45 compliance with new legislation (P.A.l98), the State Department of Education created the Special Education Simulation Consultation Project. This group developed goals for special education through 1980. Among the recommendations was one which indicated the need for an organization which could systematically address administrative needs in special educa- tion, respond to those needs, and provide input into state decision- making processes. As Burillo (1975, p. 9) noted, the fundamental pur- pose of this project was to determine the steps necessary to "stimulate the development of more integrated service delivery models for handi- capped children within their own school buildings." My personal observations are that special educational administra- tors traditionally apply local standards to special education. In more sophisticated and wealthy districts, this means richer and more variega- ted special education. In rural districts particularly, and poor dis- tricts generally, the quality of special education is substantially lower. Parents, through the courts and through political pressure, have compelled the State Department of Education to require more uniform standards of education and a more equitable distribution of services. Districts have been required to provide individual instructional plans for each special education student; a program to reintroduce the student to normal academic careers (mainstreaming); a precise method for identifying non-organic learning disabilities; and a method of accounting for teacher producti- vity. Finally each program was subordinate to statewide standards and the review and approval of the State Director of the Special Education Services Agency (SESA). 46 Fundamentally, this policy shift affected the administrators of special education who are responsible for funding and program direction. First, they had to conceive of themselves operating in concert with a wider political and cultural reference base. Specific local inclina- tions were made subservient to state priorities. The legislature passed laws which set minimal standards and required the application of regional planning models. Each child within the special education milieu was to have an individualized program which, to the extent possible, allowed the child to enter into the regular curriculum. This meant a reduction in traditional separate classrooms and a higher em- phasis on counseling and therapy. Secondly, the special education administrator was compelled to be less a special educator and more of an administrator. Interviews with special education administrators which I conducted indicated that special educators derive most of their job satisfaction from working directly with the children. Planning, needs assessment, evaluation, and respond- ing to the requirements of the State left little time for active class- room involvement. Special education administrators perceived the changes in their own role performance negatively--or, imposed. In the logic of the model explicated above, planning and administrative functions were highly discrepant from those attributes of the job closest to the core psychological definition of self among these administrators. As confusion and consternation mounted, the state adopted the recome mendations of the Special Education Simulation and Consultation Project, and adapted a specific strategy for the encouraging reforms in special education administrative practices (Kay, 1976). The department, 47 responding to a proposal from faculty of the University of Michigan, allocated funds for the creation of Project STANSE -- the State Technical Assistance Network in Special Education. STANSE is an administrative innovation which involves the provision of organizational training at the district level and the creation of a specialized management task-force which is dedicated to identifying and solving administrative problems in Special Education. The management task force was composed of personnel from the local, intermediate and state levels. Individuals were selected after peer nomination and review by a "blue-ribbon" panel which included a repre- sentative of the State Director, permanent STANSE staff, and prominent persons in the field. The management task force -- State-wide Manage- ment Task Force (SMTF) -- was the key element in the STANSE concept. Prior to STANSE, there existed no formal mechanism for moving percep- tions, information, and contributions from the bottom.of the system to the top without reference to complicated procedural protocols. New State legislation (P.A. 198) and Federal legislation (P.L. 94-142) required that the State provide a mechanism for planning which utilizes input from all administrative levels and which is independent of the resource base of the district or level. This means that local rural districts must be included in state planning processes on an equal basis with large districts that command disproportionally larger amounts of funds. STANSE included local district administrators, intermediate dis- trict supervisors, and State Department personnel on the SMTF. Monthly meetings, regional meetings, and meetings at the State Department emphasized communication across hierarchial levels aimed at state-wide 48 goal formulation, accountability and technical assistance. In short, STANSE was the vehicle for promoting rationality in Special Education administration. Thus, change/diffusion efforts shifted from what Havlock (1973) called "natural diffusion" to what he refers to as "natural communication network utilization." The notion of utilizing a communication network as a method of encouraging area planning -- or coordinated behavior -- requires some degree of formal organization. The State, therefore, funded the creation of the state-wide technical assistance network in special education, project STANSE. STANSE was conceptualized as both an inno- vation in itself -- i.e., a new organization which would link all special education administrators regardless of system level -- and as an arena within which innovative planning activities could take place. STANSE organized itself as a participatory decision making organi- zation. Figure 4 reflects the overall pattern of institutional and organizational relationships. A project director fulfilled project management responsibilities. He was assisted by a project coordinator who directed the research staff, and coordinated the day-to-day planning activities of the wider special educational administrative group. Addi- tional project staff consisting largely of three organizational develop- ment trainers worked with the SMTF. No position could be taken, and no decision could be made which did not reflect the consensus of the SHTF staff, and the State Department. STANSE focused on analyzing three administrative phenomena (Kay, 1976, p. 9): 49 l. The interaction processes within and between the three organizational units, with particular emphasis on management functions at the three levels; 2. Communication within and between levels of the system; and 3. Identification of specific problems and recommendations for solutions. The idea was to use the SMTF as linking pins in the sense of Likert (1961). The SMTF, through their contacts in the field, would specialize in the development of procedures for identifying the problems, and methods for attacking those problems. Each month, STANSE would hold an SMTF meeting at a different location throughout the state. SMTF members, usually between 25-30, would attend the meetings, and discuss problems of mutual interest. Over time, the SMTF split into sub-committees, each of which was dedicated to a specific concern of special education administrators; e.g., State reorganization of special education administrative units, implementation of new laws, and identifying new curriculum programs. The results of STANSE meetings was to be "products" and positions. A product is some item, such as a comparison of all compliance procedures in administrative law affecting special education administrators, which could be disseminated to the field. A position is an advocacy state- ment regarding some issue in the field of special education administra- tion. First, the results would be communicated directly to the state director for consideration and state planning, state decision-making, and the selection of state fiscal priorities. Secondly, the results would be communicated back down to the field for "coordinated implementation." 50 FIGURE 4 Organizational Chart of the State Technical Assistance Network in Special Education STANSE Project Director State Department Special Project A Education Coordinator I Services Director - Research Staff State Management Task Force Intermediate Local School School District District Administrators Administrators Arrows indicate assigned direction of authority 51 "Coordinated implementation" meant that a large number of districts heretofore operating independently would adopt uniform management and operating protocols, and hence develop interdependency. The SMTF could "cause" such adoption by virtue of the prestige of its membership and the rationality of its strategies. STANSE began its activities in late 1973. All administrators in the state were invited to join the SMTF. Thirty-five of five hundred and sixty-four administrators agreed to join. The STANSE attempted,using the Delphi technique and various other T-group (Bennis, 1969) strategies,to effect a high degree of cohesion and establish working goals (STANSE Goals, 1975). Kay (1976) indicates that initial diffusion efforts were based upon an explicit application of the work of Havlock (1973) and Rogers (1971). They argued that once problem solvers (administrators) became acquainted with "need" to change, change would occur systematically and logically. The problem was, first of all, that "need" to change was a function of the sophistication of the district and its geographic isolation. Since early efforts were sponsored by the larger districts, smaller rural districts and their own specialized concerns tended to be under- represented in decision-making. In pursuing the research effort, STANSE staff and the SMTF were to be kept explicitly informed of research results, and the interpretation applied to such results. Staff members accepted responsibility for planning and disseminating messages, with editorial assistance from Dr. Farace and myself. Plans called for extensive dissemination of information throughout the course of the 52 project. As I have noted, the model requires that energy be invested in information and message dissemination continuously if an innovation is to become adopted. The research design called for infrequent formal position papers, bulletins, and explanations to be disseminated to the network, and many specific messages to be disseminated through "key communicators" iden- tified within the organization. Staff were to use a combination of mass channels and interpersonal channels to maximize the overall visibility of their information, and hence, improve the probability of attention and subsequent response. The model calls for repetitive measurement events within the organization, and information decisions based upon observations from the measurement event. Therefore, three points in time were studied during the nineteen months of the research reported here--a measure- ment approximately every six months. After measurement and analysis, the researchers discussed results with staff and SMTF, and decided upon a message strategy and appropriate implementation strategy. STANSE staff, then prepared messages, such as in Appendix A, referred them for approval to the SMTF, and disseminated the messages. At crucial junc- tures (usually concerning a crisis of some sort), SMTF members would be directed to telephone or contact personally three "key communicators" and either deliver a specific piece of information or ask for a response to an SMTF proposal. Between the first point in time and the second point in time, STANSE staff and SMTF members cooperated fully with the procedure. No precise record was kept, but at least seven formal position statements were distributed to all members of the organization, three telephone 53 messages were disseminated, and numerous bulletins and newsletters were released. Results will show that considerable positive change was effected during this period. Between Time 2 and Time 3, however, STANSE lost the staff member who served as the liaison between Dr. Farace and myself, and as a result, the message campaign ceased abruptly. The researchers were told that efforts were proceeding normally, but we found out that between September of 1976 and April of 1977, not a single formal message was authorized or disseminated. As will be shown, the change in level of effort had a profound effect on the outcomes of the STANSE project. Dr. Farace and I were asked to participate in STANSE change efforts in January of 1976. We were engaged as consultants, and charged with providing assistance in the development of a state-wide communication strategy. The model discussed here was developed early in the project. My participation in the STANSE program included measurement, ‘message design, group counseling, and organizational development. These responsibilities allowed considerable editorial control over the context, timing, and distribution of promotional developmental messages. 1.8 SUMMARY Change and innovation processes in organizations are important social and managerial phenomena. It is argued that communication re- search methods and paradigms offer insight into the mechanics of the ‘process of organizational change. Bureaucratic and rational schools «of organizational theory are contrasted in terms of their treatment 54 of change, and the role of communication in management under the alter- native models. In bureaucracies change is antithetical to the assumptions of stability and control which are fundamental. Communication is assumed to be upward, structured, and characterized by relatively inflexible relationships and communication rules. The rational school admits to flexibility in communication relation- ships and rules, as a function of the rational goals of the firm. Change is treated as a constant to be dealt with as a part of normal managing behavior. Rational management models, it is argued, led to the emergence of organizational development, and its humanistic assumptions about the behavior of people in the world of work. Organizational development theorists made explicit the importance of attitudinal attributes in human performance, while taking a simplis- tic view of the change process. Drawing from communication attitude research it is argued that attitudes toward the job, and related pheno- mena are the elements in an inducement-contribution balance which can be modified through communication strategies. A model is proposed which holds that the key to organizational change lies in the recognition of the discrepancy between prior behavior, and new behavior imposed by a change. It is argued that this discrepancy is a perceptual variable which depends upon the attributes associated with the definition of self in the job, and the degree to which those attributes may be assigned to innovative objects within the environment. By identifying the attributes associated with work in a specific environ- ment, which are also close to the definition of self, it becomes possible 55 to design a communication strategy which, over time, will reduce the perceived discrepancy between the objects of change, and prior behavior. This is accomplished by assigning attributes which are close to self, to change object in an over-time communication campaign. It is proposed that this constitutes a cybernetic model, because the relationships between change objects, self, and attributes of the job can be measured and adjustments in the communication strategy can be introduced at various points in the change effort. By so doing, the manager controls the perceptions of employees about change objects, the organization, and messages which positively and adversely affect those balances -- without having to alter monetary inducements. The model is held to be efficacious because it takes into account the needs and expectations of message audiences in the planning and for- mulating of communication strategies. Communication is held to be the mechanism by which concepts become "meaningful" in the organizational environment, and programmatic approaches to the development of specific attitudes have been successful in the past. Secondly, the manager controls the process in terms of key organizational goals, hence, change efforts have long-term beneficial consequences. A research design is proposed in which hypotheses derived from the model can be tested. Network analysis is identified as a method for identifying key actors in the communication system, and time-series metric-multidimensional scaling is proposed as an acceptable method for observing change and developing message strategies. A research setting is identified and described, and the consulting role of the research staff, including the author, is elaborated upon. It is noted that the 56 author had responsibility for the design and dissemination of messages directed at the achievement of non-disruptive change in the proposed organization. CHAPTER 2 RESEARCH HYPOTHESES The discussion in Chapter 1 suggests that change occurs most reliably and predictably when: l. the discrepancy between an innovation and the prevailing beliefs, attitudes and values of organizational members is minimized; 2. the attitudes, beliefs, and values which define a member's perception of the job are linked to the innovation; 3. the attitudes, beliefs, and values linked to an innovation are instrumental to the perfor- mance of the job; 4. when communication is specific in establishing linkages between attitudes and innovations; and, 5. when the organization during the change effort is treated as a cybernetic system. Thus, extraneous information and noise is minimized, consensus is encouraged, and a programmatic-~or, goal-defined--change is rendered more probable. "More probable" is stated so that the concept of resistance to change is explicit. It is possible that the concepts or social objects which are targets for redefinition are so "massive" that no amount of information can induce change. A caveat to this research which must be borne in mind is that in changing the relationship between object and self, we are attempting to redefine the "meaning" of the object. The redefinition of meaning occurs in a generalized cultural setting. Culture may be taken to the common, shared, ritualized experiences of .an.aggregate of people (Gillham, 1972). This definition can be extended to normal information experiences. Through interactions among a fixed 57 58 group of people or common information experiences, people come to observe which objects are important to the largest proportion of the population and the attributes by which those objects are known, defined, and hence, made meaningful. As Woelfel (1972, p. 10) notes, ...if the concept of culture is to have any meaning, then there must be some central tendency of opinion around which individual beliefs may be seen to cluster themselves more or less cohesively. Culture is thus the tendency of individuals engaged in common ritualistic-~or repetitive, symbolic--behavior to define phenomena in similar terms, and to understand the meaning of terms in a common way. Without such common systems of meaning, communication would be problematic, or at least extremely tedious, and cohesive interdepen- dent behavior would not be possible. In short, without consistent shared definitions, organizations, like cultures, would not exist. Thus, the meanings of attributes in terms of relation to self are idiosyncratic, i.e., dependent upon individual experiences, but the commonality of individual experiences across a culture provides that some meanings are largely culturally determined. Organizations, by isolating individuals into structured communication roles and relationships, develop unique systems of meaning which can be called "culture." So also do professions--e.g., law, medicine, teaching. In the present case, we are considering an organization which has been imposed upon a professional network. From the point of view of the system members, both STANSE and its primary goals may be taken to be the innovations. Innovation is operationally defined as concepts and objects for which meaning is largely undefined or inappropriately 59 defined within a given system of relations -- in this case the concept "STANSE." The job of the change process is to affect a definition of the objects such that the dissimilarity between object and self is minimized. In my view, this is best accomplished by minimizing the dissimilarity between innovation, objects and those attributes of social objects which are also close to self. That is, to define an innovative object, or to express the relationship between an innovative object and objects which are already important to people as very similar (or very close). Operationally this implies a triangulation of relations (see Figure 5). To the extent that the innovation is close to an attribute of self, it is ideally close to self. Figure 5, however, describes an optimal relationship. It is also possible that this distance between innovation and attribute could be equal to the distance between attribute and self (as in Figure 6) while self and innovation can remain far apart. Figure 5 Figure 6 Distance between Distance Between Concepts: Optimal Concepts: Non-Optimal Relations Self Solutions Innovation .Attribute Self Attribut nnovation :1 __. ___ In the case of Figure 5, the dimensionality, and hence the comp plexity of the defined relationship, is reduced with respect to that of Figure 6, but the innovation remains relatively unimportant in the definition of self. Hence, the function of information reducing the 60 distinction between self and innovation is to provide linkages (defini- tions) which unambiguously establish similar meaning among concepts for self, attribute, and innovation. By establishing such a relationship, the number of dimensions of meaning for the innovation is increased (in this case from one dimension to two dimensions), and hence relations are more complex while dissimilarity is reduced. Assuming the distribution of information which indicates specific relation between an innovation, the attributes of self, and self, we should find: H1: Over time, the number of positive dimensions within which the relationships among key concepts can be represented will increase. A positive dimension is a factor which contributes variance to the solution. We should test the hypothesis that decreasing the dissimilarity between self and innovation should increase productivity vis-a-vis the innovation. This is, however, impossible in the present case. No prior data exists inasmuch as the organization is a new one, and hence any activity would reflect an increase. However, I will document some of the products which STANSE developed, in my concluding remarks. The conceptual discussion above also suggests that individuals differ in their receptiveness to communicate as they differ in the degree to which attributes are present and salient. Rather than moving information through an organization via mass channels such as "memos" and training aids, it is preferable to utilize the existing informal information structure, and allow messages to "disseminate." 61 The function of the linking pin (Likert, 1967) in an organization is to link large numbers of individuals whose relationship (or communi- cation) pattern form dense clusters (or groups). Farace, Monge and Russell (1977), in their work with networks, suggest that these link- ing pins may exist in two communication network roles: Liaisons: Individuals who are not themselves members of groups, but who link groups or cliques. Bridges: Individuals who are members of groups or cliques, who are connected to members of one or more other groups. After isolating these individuals, information can be dissemina- ted through interpersonal channels which will augment information dis- tributed via other media. Even without specifically addressed informa- tion, the individuals tend to access information earlier and use information most efficiently (Allen and Cohen, 1969). In effect, it is hypothesized that "key linkers," or "key communicators," by virtue of their unique communication network roles of positions, are organiza- tional "opinion leaders." The dissertation presents an examination of the question: To what extent can the impact of specific message stimuli cause an innovation to become integrated into the job perceptions of the members of an organ- ization? This question can be translated into the following working hypotheses: H2: The magnitude of the attitude A toward innovation I will be significantly reduced between T1 and T3 for the whole population. 1) The second hypothesis indicates that implementation of a message strategy will eventually produce key changes in the perceptions of the whole system 62 during the three points in time. This is crucial to the demonstration of the effectiveness of the proposed change strategy. Successfully changing the aggregated perceptions of the organi- zation is, however, contingent upon producing and disseminating an appro- priate flow of targeted information. H3: To the extent that information is directed toward key communicators, (K), the magnitude of their attitudes with respect to advocated positions will be less than that of non-key communicators (N) at T2 and T3, or: Ilsa: AT K < AT N and H31: ‘ AT3K < AT3N and it follows H3c= ATIK > ATZK > AT3K and H3d: Ar111 2 ATZN > AT3N 2) Hypothesis 3, and its corollaries indicate that the magnitude of change on indi- vidual experiences at each point in time is, in part, contingent upon their communication role. The model of imple- mentation is time-phased to suggest that key communicators change their perceptions, alter the content of subsequent interactions, and then change occurs throughout the re- mainder of the organization. Since we are proposing a two-step flow of information, i.e., to the general group through key communicators, change in non-key communi- cators should lag behind changes in key communicators across time inter- vals, or: 63 H4: The magnitude of attitude (A) toward the innovation held by non-key communicators (N) at successive time intervals (t+l, t+2, t+3...t+4) will approach the magnitude of attitude (A) held by key communi- cators at previous time intervals (t, t+l, t+2,... t+n-l), or: “45‘ ATlK = ATlN H413: ATZK < ATZN H : A A 4c T K = 'r N 2 3 This is to say, that assuming that the system is nearly at equilibrium initially, changes in key communicators will be followed by changes of equal magnitude for non-key communicators. Should these hypotheses be upheld, there remains the question of casuality. Do people enter roles as linking pins or key communicators because they are similar to the population they represent, or are changes in their attitudes predictive of subsequent change in the larger organization? A theory developed by Lewin (1961) illustrates the latter point. Organizations go through periods of change which begin, diffuse, and end. Change is associated with an initial flurry of activity, a period of substantial change, and a period during which the change becomes fixed. In the Lewin perspective, and also in the Rogers and Shoemaker (1971) perspective, change is brought about by a cycle involving early adoption, testing, and installation. It can be argued that central to the process is the opinion of key communication linkers. If they support a change effort, they are more likely to pass along information about the change. 64 By examining specific patterns of obtained correlations, it will be possible to review the efficacy of key communicators to facilitate change. At T1, we expect to observe only small differences between key communicators and non-key communicators. Key communicators and non-key communicators, through normal attitude formation processes should be relatively homogeneous. By inputting information which takes advantage of key communicators' positions in the network, key communicators should change, and those changes should be observed in changes in the non-key communicator group over Time 2 and Time 3. By continuing to direct messages through key communicators, it should always be the case that key communicators change more than non-key communicators or: HSa‘ XTK " xTK> xIN " XTN 3 l 3 l and H : r r r 5b T EN :> T'KN 2* T KN l 2 3 Given a time lag between key communicator change and the rest of the network, we should observe. and H : r = r ‘_. 5c TIKN TIKIZN :> T KI N 65 This argument means that key communicator's attitudes from a previous point in time predict the attitudes of the rest of the network at a subsequent point in time better than non-key communicators' atti- tudes can predict key communicators' attitudes. Finally, the attitudes of key communicators and the rest of the network will always be more similar under conditions of relative informational (change) stability (T1). An opposing hypotheses for each of these conjectures would consti- tute a casual argument for the validity of the argument that key communi- cators occupy their positions because they are representative of the general position of the organization. If values, beliefs and attitudes of the larger information network cause key communicators to adopt minimally discrepant positions, we should find: : r r and r r 6 TINTZK > TlKTZN TZNTBK > T2163N and H6a‘ XTN " xTN > xTK ' xTK 3 l 3 1 Thus, the attitudes of network members are better predictors of the attitudes of key communicators than the attitudes of key communicators are of the rest of the network. In neither case should we find: XT3K ' leK = XT3N ' leN or the complete absence of differential rates of change. 66 The hypotheses test the assumption that change can be systematically introduced through message developed on the basis of measurements taken at an earlier point in time. We test the assumption that the system is initially at an equilibrium, and information is constructed and input into the system. The researcher controls the perception of systems members with respect to the innovation, by directing the largest pro- portion of messages at the key communicator directly. Ideally, subse- quent measures would cause changes in the message program which would cause the organization to continue changing in an optimal direction. CHAPTER 3 METHODS 3.1 INTRODUCTION The analysis of the hypotheses requires: a) a measurement of communicator influence--or a measure which enables key communicators to be unambiguously discriminated from the balance of the population; b) a measurement of the dissimilarity between an innovation and perceptions of self by members of the system; c) a measurement of the dissimilarity between key job attributes and members of the system; d) a technique for determining a reliable set of key attributes of the job; e) a method of examining changes in relationships between the attributes over time; and f) a methodology for identifying an effective message; i.e., a message which is likely to optimally reduce perceived dissim- ilarity over time. It is key to this effort that measurements maximize the total amount of information which can be derived from a single measurement event; and it is important that measurements be independent, particular- ly measurements of cognitive structure and network roles. The technique used to isolate key communicators cannot rely upon methods used to iso- late dissimilarities. If that were the case, the research efforts would be confounded by tautology. Therefore, two separate methods were chosen: network analysis and metric multidimensional scaling. 3.2 NETWORK.ANALYSIS The concept of communication systems as being comprised of networks, or patterns of interpersonal interactions, was originally noted in the 67 68 work of Weber (1947). Weber noted the necessity of prescribed communi- cation roles as a method of achieving control over production processes. By specifying the hierarchy of relations--or the chain of command through which information should flow--Weber argued that extraneous data would be eliminated and production efficiency would be maximized. Downs (1969), in studying the actual functioning of bureaucracies, noted a strong tendency for communication within bureaucracies to follow subformal or non-prescribed information pathways. A network is, therefore, the system of overlapping dyadic relation- ships, both formal and informal, which collectively constitute the sum of all possible communication pathways within a bounded social system (c.f., Richards, 1947a). Organizations differ from non-formal social systems in the extent and reliability of assigned or prescribed infor- mation pathways. In an organization, the assignment of authority and responsibility tends to govern, to a greater or lesser extent, the nature of the network. In this discussion, a "pathway" will be called a link. Two indi- viduals are linked if: a) they communicate directly, or b) there exists a set of communicators between person A and person B such that information can flow between A and B. Thus, individuals are highly linked if they are: a) linked directly to many other individuals; or b) connected to a few individuals who are highly connected to many other individuals. (Monge and Lindsey, 1974) Network techniques have been described in research literature for more than two decades (Schwartz and Jacobson, 1977). The particular 69 technique employed here was first described by Jacobson and Seashore (1951), and subsequently elaborated by Weiss and Jacobson (1955). Richards (1974a, 1974b, 1974c) wrote a computer program (1976) which is predicated in part on the Jacobson and Seashore formulation. Richards' analytical methodology uses dyadic interaction frequencies in order to detect underlying organizational structure. Individuals are asked to indicate how frequently they interact with other individuals within a given organization or bounded social system. This data are arrayed into an N x R data set where N is the number of nodes or persons in the organization, and R is the number of reported contacts. Cluster analysis techniques enable the data set to be decomposed, and based upon one's position in the re-ordered data structure, and the pattern of one's links, an individual is assigned to a communication role. Thus, the model allows for the classification of individuals based upon the extent, frequency, and duration of their communication behavior (Farace and MacDonald, 1970; MacDonald, 1971; Monge and Lindsey, 1974). Nodes, or individual members of the organization, are assigned one of five possible communication role states (Richards, 1975)- Inasmuch as STANSE is a very "flat" organization, or it lacks many hier- archical levels, net link strength is probably a better indicator of overall importance than formal occupational position. The roles are: 1) Group member: a node with more than some minimal percentage of interaction within the group (here, equal to 50.01%); and constructed such that no single node or link can be removed from the group and cause collapse of group structure; and the group is linked such that a path exists which connects all group members by some minimal number of steps. 70 2) Bridges: nodes which are members of groups, but which are also connected to another group, and hence, link two or more groups. 3) Liaison: nodes which link two or more groups but are not themselves members of any group. 4) Isolates: nodes which have no links (Type 1 isolate) or nodes which are connected to only one other node (Type 2 isolate). 5) Other: nodes which fail to meet the criteria for role assignment. Of particular interest here are the classifications of bridges and liaisons. Bridges and liaisons allow information to flow between large groups of people. A bridge could serve within a group as the source and receiver of information from the rest of the network. By exercising discretion, the individual who is a bridge can allow or block information flow, and hence influence the course of events. Similarly, liaisons, by virtue of their position between two or more groups, significantly influence the flow of information. Likert (1961) noted the position of the individual who links large chains or groups of other links. He described such an indivi- dual as a "linking pin" and attributed to that individual authority independent of hierarchial position. Since such an individual could exercise choice in allowing infor- mation to flow, and since information is the mechanism of control and coordination within an organization, a linking-pin could effectively manipulate information flow to personal ends (Likert, 1961). Of course, this capacity varies from organization to organization, and depends upon the degree of autonomy, flexibility, and coercion within a system (Whittmore and Yovits, 1973). In the case of STANSE, with its reliance 71 on participatory decision making and its basically ad hoc nature, we can assume that coercive control is minimal and flexibility and autonomy are maximal. Therefore, in STANSE, linking pins, or operationally, liaisons and bridges, should exercise a great deal of influence on atti- tudes and behavior within the organization. Guetzkow (1965) argues that liaisons are important to the function- ing of a communication network, and further that they are aware of their prominent role. Weiss (1956) supports the argument that liaisons have higher relational reciprocity, and hence, are acknowledged by others as important. Wagner (1972) indicated that liaisons function as sources of organizational news. Schwartz and Jacobson (1977) found that network liaisons (as opposed to hierarchically determined liaisons) are aware of their influence, have highly discrepant information acquisition behaviors, know critical information earlier, and have higher influence than other members of the organization. Essentially, Schwartz and Jacobson (1977) demonstrate that liaisons embrace both dimensions of informational leadership described by Katz and Lazarsfeld (1955). Katz and Lazarsfeld argued that individuals could be gatekeepers, or in positions to influence the flow of information to others, in two senses (pp. 113, 119): 1) as information transmitters to influence the distribution of information; and 2) as opinion leaders, who influence the interpretations applied to information and its subsequent use by a group. Allen (1969) provides support for the findings of Schwartz and .Jacobson (1977). He found that some individuals have highly discrepant environmental information searching behaviors. These nodes maintain 72 diverse contacts with individuals outside the organizationr-personal contacts, journal subscriptions, and participation in membership groups--and are accorded high status as a result. In addition, in the scientific laboratory, Allen found that the "technology gate- keepers" were more likely to publish, held significantly more patents, and.in general, they tended to occupy higher supervisory positions than peers of the same age and with the same educational background. While the research literature on the role of liaison is relatively sparse, there is a consensus on the importance of the role. Here, that important function is also accorded to bridges. As we noted, bridges link two or more groups while being members of communication groups or cliques themselves. Studies (Festinger, Schacter and Back, 1951; Danowski, 1974; and Taylor, 1975) indicated that cliques have strong influences on the attitudes and beliefs of members about the function of the social system of which they are a part. Findings of Allen (1969) indicate that the gatekeeping function maintained by bridges within their clique give them many of the opportunities and functions of liaisons. Communication networks can be constructed according to simple frequency of interaction, or they can be made specific to content (Farace, Monge and Russell, 1977). The present research is concerned with the effect of information on changes in the perception of a node's job, and changes in the perception of innovations. Specifically, STANSE adopted the dissemination of information about change or innovation as a goal; information about planning as a goal; and information about the nature 73 and function of the role of special education administrators as a goal. The SMTF, the management task force group, felt that the rate and magni- tude of change in the job required special educators to redefine their definition of role performance. Therefore, three networks were studied (Appendix B): 1) A network based upon interactions about the concepts of change and innovation. 2) A network based on discussions of planning. 3) A network based on discussions of the day-to- day responsibilities of special education administrators. These data were utilized to identify and assign communication roles. For purposed of this research, key communicator is operationalized as any person who functions as a liaison or bridge in one or more of these overlapping communication networks. 3.3 METRIC MULTIQIMENSIQNAL SCALING (MMDS) Metric multidimensional scaling (Woelfel, 1972: Serota, 1974; Barnett, Serota and Taylor, 1974, 1976) provides a methodological frame- work for the evaluation of change in cognitions over time (Woelfel, 1976). Based on the psychological work of Gullickson (1946) and Torgerson (1951, 1958), multidimensional scaling uses judgments of distance or dissimi- larity between concepts (or stimuli) to place concepts in a spatial re- presentation. Multidimensional scaling generates a picture or map which represents the relationships among a set of objects 01, 02, . . . , On. The method utilizes a symmetrical data matrix whose rows and columns correspond to objects 01, 02, . . . , On' The ijth cell contains the dissimilarity, 74 or the observed differences between object 0i and object 0 In general, j' the smaller the distance or dissimilarity between objects Oi and Oj’ the greater the perceived relationship between the two concepts. The method of data collection employed here utilized measures of the perceived dissimilarity for (N) x (N-l) / 2 object pairs, where N is the number of concepts. The metric method differs from the nonmetric method (Shepard, 1962; Kruskal, 1964). Nonmetric MDS employs proximity judgments based on an ordinal scale; metric MDS utilizes proximities measured on at least an interval scale (Torgerson, 1958; Tucker and Messick, 1963). Thus, the metric method allows one to apply subtraction and multiplication rules to data. The significance of the MDS technique lies in its power to represent various influences in the projection of psychological structures simul- taneously (Taylor, Barnett and Serota, 1974). According to Torgerson (1958, p. 248): . . . the notion of a single underlying continuum is replaced by a notion of an underlying multi- dimensional space. Instead of considering the stimuli to be represented by points along a single dimension, the stimuli are represented by points in a space of several dimensions. In- stead of assigning a single member (scale value) to represent the position of the point along the dimension, as many members are assigned to each stimulus as there are independent dimensions in the relevant multidimensional space. Each numr ber corresponds to the projections (scale value) of the points on one of the axes (dimensions) of the space. The number of dimensions is equal to the number of stimuli or concepts. Data are aggregated and averaged into a distance matrix which is trans- 75 formed into a scalar products matrix. This matrix is factored using a direct iterative unstandardized procedure. Factoring yields a coor- dinate matrix consisting of orthogonal axes with rows which are the projections of concept locations on the axis or dimension. The procedure allows for the rotation of multiple data sets. At each point in time, spaces are rotated about the centroid of the pre- vious point in time to a least-squares best fit criterion to provide precise approximations of concept motion over time. From this time series approximation, it is possible to apply curve fitting and other methods to describe relational changes in the set of concepts. The procedures for generative metric MDS analysis are described in complete detail in Barnett, Serota and Taylor (1974) and Woelfel and Barnett (1974). However, a brief description is provided here. Sub- jects are given a complete (N) x (N-l) / 2 set of paired comparisons. They are asked to make a judgment of the form: If x,and y,are u units apart, how far apart are concepts a and b? The respondent is, therefore, asked to provide a distance estimate with reference to a standard referent. Although the precise perceived size of that standard may vary, the Law of Large Numbers and the Central Limits Theorem allow confidence in the representational value of the aggregate mean. Barnett (1974) indicates that pair-wise judgments of this form become reliable with samples as small as 35 respondents. Metric MDS has been used to study differences in cultural response to television characters (Wigand and Barnett, 1975); political attitude formation (Barnett, Serota and Taylor, 1976); the development of belief 76 systems (Danes, 1976); occupational choice (Gordon, 1977); mass com- munication processes (Woelfel and Barnett, 1974); and cognitive com- plexity in language development (Barnett, 1975). Using another computer algorithm.but the Torgerson solution, Stager, Schultz and Klein (1966) were able to account for cognitive complexity, and gradepoint averages in student judgments of art. Goldstein, Blackman and Collins (1966) studied army leadership characteristics while Jones and Yancy (1972) investigated student-fac- ulty relationships and found them to be a function of status, political ideology, and methodological interests. The interpretation of MMDS spaces rests not so much on the distri- bution of projection on the axes, as in conventional principal-components factor analysis, but on the relationships between points. Barnett, Serota and Taylor (1976) show that the distances between concepts in the resulting factor space can be utilized to predict voting behavior. By summing the distances for a concept representing the average self- position ("me") and candidate concepts, they were able to predict the results of an election within one to two percent of accuracy. Woelfel and Taylor (1976), in a study of educational administrators, were able to develop strategies which greatly decreased perceptions of the appro- priateness of change among educational administrators. This resulted in dramatic increases in interest in products and hence, sales for the or- ganization which commissioned the study. By examining carefully the relationship between points in space, it is possible to detect which concepts or stimuli are significant for the individual and which are not. In addition, by examining the same 77 relationships it is possible to abstract a message strategy which will maximize the probability of change. 3 .4 MESSAGE STRATEGY This effort is directed at the determination in measurement of the efficacy of a message which will induce a positive acceptance of the concepts of "innovation," "planning," and "STANSE" among special educa- tion administrators. By conducting a network analysis, key communicators were identified. Messages could then be transmitted through the SMTF interpersonally to the key communicators, and then to their field. In addition, memoranda and "white papers" advocating positions on the inno- vation were distributed to the entire network. The success of a message strategy is dependent upon the method of selecting a concept, and the method of selecting a message. If a concept set is chosen which does not reflect important or instrumental attributes and job-related phenom- ena, measurements will not provide acceptable representations of judg- mental criteria. Hence, derived messages will have minimal chance of accomplishing change in people's perceptual sets. Concepts were identified and selected in an analysis using members of the SMTF under the direction of Dr. Farace and myself. Thirty-two SMTF members attended a two-day workshop conducted in November of 1975. They were arranged into five-person groups. Each group was asked to submit a list of concepts related to ggganizational aspects of the job of special educatiggfadministgators, sources of satisfaction, difficul- tigs on the job, factors affectingfthe job, and changes they had recently observed. 78 The lists were then gathered and collapsed into a single list of 40 concepts. The concepts or items were accepted only if they could be reduced to a single word or a short phrase. The list of 40 con- cepts was grouped into a single questionnaire. The questionnaire required SMTF members to judge the importance of each item for his/her job on a seven-point Likert-type scale. Those items which were rated as most important, and which had the largest standard deviation, were retained for further consideration. Items were selected for importance, because we were interested in at least having a number of items which we knew were "close to" peOple's perceptions of their jobs. Items with large standard deviations were selected because those concepts which exhibit low variability are unlikely to be changed or subject to much subsequent change by the kind of message campaign we were able to Final concept selection was made in a series of meetings with the STANSE personnel. Several concepts were dictated by the research effort. These were: 1) My job 2) STANSE 3) Planning 4) Change Concepts selected from data provided by SMTF members were: 5) Management systems 6) Efficient 7) SESA (State Department of Education, Special Education Services Area) 8) Planning 9) Frustrating 10) Mainstreaming 11) Influence 12) PA 198 (Public Law Affecting Special Education) 13) Helpful 79 14) PAC (Parents' Advisory Council) 15) Efficient 16) Collective bargaining Explanations of individual concept choice criteria appear in Table 1. It should be noted that in repeating the measurement three times, some concepts were eliminated or replaced. In the administration for Wave 2, the concept "collective bargaining" was eliminated. Prior to Wave 3, STANSE staff and SMTF members required that some unused con- cepts be eliminated and replaced by some new issues which had developed during the intervening time (1 year). Thus, data analyzed in this research consists of a subset of 12 of the total number of concepts, and these 12 appeared in all instruments (Appendix A). The final concept list used for this report is: 1) My job 2) Child centered 3) STANSE 4) Management systems 5) Efficient 6) SESA 7) Planning 8) Change 9) Mainstreaming 10) Influence 11) PA 198 12) Helpful All strategic messages employed only concepts selected from this list. The remaining concepts represented the specific interests of either SMTF's subcommittee, or STANSE's staff, and were purely infor- mational. Copies of the three instruments can be found in Appendix A. 3.5 MESSAGE SELECTION Sherif and Sherif (1967) have shown that a receiver's response to a message is a function of social judgment. In general, responsiveness 80 to a message is a function of the dissimilarity between a receiver's attitude and that attitude articulated in the message, the receiver's involvement with the topic, and the degree to which the message pos- sesses structure in its organization (c.f., Bettinghaus, 1969). It has been argued here that interest in a message, particularly a message about an innovation, is aroused by relating the attributes of the object of the message (innovation) to attributes of the receiver. That is to say, messages will be maximally efficient when highly dis- crepant perceptions of social objects are related to common, comfor- table judgments about the self. MMDS provides a method of determining which attributes are close to the receiver and, hence, their approximate relationships. Woelfel, Cody, Holmes, Fink and Taylor (1975) have developed a procedure first suggested by Taylor, Barnett and Serota (1976), which provides criteria for the optimization of a concept's motion through an MMDS space over time. The object is to cause a concept to converge with the average} self-position of the sample, or the "my job" concept in the present case. In general, the lower the perceived similarity between self and an object, the greater the importance of that concept as a determinant of behavior and attitudes (Marlier, 1976; Woelfel, 1976; and Danes, 1977). The procedure suggested by Woelfel, et al., 1975, uses vector addition to determine which set of attributes will, when linked to an object, cause a predicted line of motion through the space, and at subsequent points in time minimize the discrepancy between the object and the perceived self. Figure 8 provides a hypothetical case. 81 TABLE 1 RATIONALE FOR CONCEPT SELECTION* RATIONALE CONCEPT 1. My Job 2. STANSE 3. Management Systems Child-Centered Efficient SESA Planning Frustrating Change The concept is needed to identify the respondents' self perception. This self perception was used to compare the interrelationships of all other concepts. To gather data on variability of the organizational innovation. To pro- vide base line data for further research as dissemination plans were developed. As managers, how did the respondents view a management system in relation- ship to their job. Represented the organizational aspect of the delivery system. Concept also represented the concept of the providers of the ser- vice. Wide variability was also con- sidered. ‘ The concept represents the receiver of the services of Special Education. Had questions about the relationship be- tween the providers and the consumers. A positive attribute (descriptor) for other concepts. Part of the delivery system which pro- vides leadership, sets policy, and ‘monitors the system. Question of per- ceptions about the effectiveness and relationship with other concepts was considered. Would provide data as to how related planning is to a manager's job. A negative attribute (descriptor) for other concepts. Had wide variability among SMTF. Would provide data about the relationship between change and the manager's job. CONCEPT 82 RATIONALE 10. ll. 12. l3. 14. 15. 16. Mainstreaming Influence P.A. 198 PAC Helpful Labeling Collective Bargaining High variance and much emotional discussion. Represents a program alternative and potential for Special Education/general educa- tion interface. A positive attribute (descriptor) for other concepts. Low variability. Provides basis for all that Special Education is required to do. Parent Advisory Committees represented the parent involvement that is critical to Special Education planning and imple- mentation. A positive attribute (descriptor) for other concepts. High variability and high emotional responses. Represents the philosophical considerations within Special Education. Very high variance. Represents poten- tial concerns related to programming, training, philosophy, and funding. *Also appears in Kay, 1976, p. 219 83 The attempt is to move object A through the space, roughly along the dotted line towards "me." By indicating that A_is B, and A.is 9, "force" is exerted which causes simultaneous motion in both directions (Figure 8). FIGURE 8 MOTION VECTORS AS LINES OF FORCE The actual vector may deviate slightly; however, in general, when the number of concepts is large, a message can be identified for which the cosine of the angle between the desired vector and the actual vector will approach 1.0. These criteria were employed as a method of con- veniently identifying a message strategy. Therefore, we should observe the greatest changes in both the dissimilarity matrix and the factor matrix in terms of the concepts employed in the message. In evaluating the hypotheses, data from concept pairs will be used indi- vidually, as well as the entire dissimilarity and factor matrices collectively. 84 Figure 9 provides a more precise depiction of the message choice process. For any given vector combination, five different pieces of information are available as criteria for selecting among the many combinations which emerge from the analysis. In the figure, Line A is the maximum length of a motion vector. If a message strategy works as predicted, motion continues until such time as vector forces are equalized and further motion is problematic. Line B is the actual distance between the concept one wishes to move (START) and the con- cept one wishes to move it toward (TARGET). Interrelatedness of concepts is directly related to closeness of concepts to one another. If two concepts lie in the same position on all dimensions, the two concepts are, by definition, identities. Line C is the point at which the distance between the TARGET and the START concepts are mini- mized for a particular strategy. This line is large or small depending on-G, the cosine of the angle between the predicted actual motion vector (Line A) and the desired optimal vector (Line B). Line D is the distance by which it is theoretically possible to exceed the target for a particular strategy. For purposes of this research, the innovation STANSE is the start concept, and the average self perception of the job, MY JOB, is the target. The factor and the distance matrices have never been used in pre- cisely this manner. Since we are anticipating fundamental changes in the way factors associated with the job of special educators are per— ceived and evaluated, it is reasonable and logical to look at the collective impact of change on all cognitive spacial relationships present in the data set. Thus, correlational analysis of key communi- 85 FIGURE 9 Optimizing Criteria for Message Selection TARGET A - Resultant length B - Length of target concept vector C - Target to right angle point D - Distance the concept can exceed the target -0-- Cosine of the angle 86 cator and non-key communicator judgments will be conducted on factor spaces across time. From data set to data set, correlations of the separate orthogonal factors will be obtained and average across the n-dimensions. Both the average correlation will be calculated, and a second correlation coefficient weighted by the mean explained variance of each constituent factor will be generated. Since the size of the factor loadings for. each variable and the sum of the squares for each factor are covariants, sumply comparing entire matrices would bias the coefficients upward. This correlational analysis is appropriate in this case since interest is directed toward the gross amount of change occurring from one point in time to the next, and between key communicators and non-key communi- cators. A continuing problem with MMDS as well as other large sample re- search methods is statistical significance. No acceptable method exists for evaluating the significance of difference between two (or more) coordinate matrices rotated to least squares minimizing criteria. Given the theoretically potential range of an MMDS distribution (00 to t-test differences of means tests bias the interpretation of discre- pancies in favor of finding "significant differences." In this re- search significance tests are not reported. 3.6 RESEARCH DESIGN The design used in this research is a three-stage quasi-random panel design. Three points in time are measured and no control groups are used. Initially, a network analysis was conducted and an MMDS questionnaire was distributed. A second point in time was conducted 87 using the MMDS measurement roughly six months after the initial measure- ment, and third point in time was conducted six months later. Following each measurement, the researchers would report back to the SMTF and the STANSE management message strategies and techniques for disseminating 'messages. It was left to the staff to conduct the actual dissemination exercises, although the researchers participated in criticizing the actual messages released. The first point in time involved a census of the entire organiza- tion. Since network analysis requires that all members respond, the entire organization was used for the first wave. The second wave was a 20% random sample of the entire organization. Random assignment was used to fill out key communicator and communicator cells. This means that the sampling procedure pooled all respondents, and was based upon an "equal opportunity" without replacement criterion. The third wave was again a 20% random sample of the entire population without replacing second wave respondents. Thus, no respondent was measured more than twice, and many respondents were only measured once. Questionnaires were administered by mail with telephone and letter follow-ups. Administration periods involved between one and two months. 3.7 SUMMARY Research reported here involved the use of network analysis and metric multidimensional scaling. Network analysis was employed as a method of operationally detecting key communicators and non-key communi- cators. Concepts relevant to the job of special educational adminis- trator were identified and incorporated’into a 12-concept MMDS instrument. These instruments were administered at three points in time along with 88 a single administration of the network analysis instrument. A Spatial configuration derived from the first point in time was employed to develop a message strategy, and this message strategy was operationa- lized by the STANSE management staff. The results of subsequent measurements were fed back to the STANSE staff and utilized in mon- itoring the continuing change as suggested in the proposed model. CHAPTER 4 RESULTS AND INTERPRETATION This chapter is divided into three principal sections. The first section will be an analysis of response rates, dissimilarity matrix and factor structure reliability for three waves of complete pairs data. The second section is a discussion of the results of the network analysis, the message strategy and dissemination procedures. The third section is an evaluation and an analysis of the hypotheses offered above. 4.1 RESPONSE RATE In deve10ping its organizational posture, STANSE managers, or the SMTF, conceptualized their task as one of creating a conducive environ- ment within which change could take place. In their view, the goal of the SMTF would be accomplished if and only if STANSE became positively associated with change and became crucially instrumental for the per- formance of administrative functions in special education. They conceived of their task not only as one of facilitating the implemen- tation of these STANSE goals, but one of introducing the notion of change as a permanent attribute of the job of special educators. Secondly, recognizing that recent thrusts in federal law have placed greater and greater burdens upon special educators for planning, SMTF personnel indicated a need to emphasize the increased role of planning in special educational administration activities. Therefore, 89 90 the analysis of communication networks focused on three principle issues "(ééé Appendix B). lRespondents were asked to indicate how frequently they interacted with all possible other special education administrators in the following areas: 1. Discuss Change: New Ideas, Programs, Procedures; 2. Discuss Planning: Programs, Services, Implementation; 3. Discuss My Job: Day-to-day Responsibilities. Respondents were asked to estimate how frequently they talked with any of five hundred and twenty-nine individuals identified as being members of the organizational communication network. The principle discriminating criteria used to separate organizational members from non-organizational members was formal identification of a professional administrative role within the State of Michigan in the area of special education, or special education services. It should be once again emphasized that STANSE was an innovation designed to represent the separate interests of all special educational administrators, and to provide input into State-wide planning processes for these individuals through the action of SMTF. Over time, administrators would come to understand STANSE itself as being a key component of the job of special educational admin- istration and, hence, adopt this innovation. Five hundred and twenty-nine questionnaires were distributed in the first wave of data collection. Each respondent received a cover letter, a network analysis instrument with a set of instructions, and a fifteen concept MMDS questionnaire (105 pairs). Respondents were included in the network if they were: 1. ISD Directors: Directors Intermediate School District Programs; 91 2. ISD Supervisors: Senior administrators of Intermediate District Special Education Program; 3. LEA Directors: Directors of Special Education Programs at the local district level; 4. LEA Supervisors: Supervisors of local district programs within specialty areas (e.g. programs for the blind, programs for the multiply handicapped, etc.); 5. SESA Director: Director of the Special Education Service Area for the State of Michigan; 6. SESA Staff: Michigan Department of Education personnel associated with special educational planning; 7. STANSE Staff: Full- and part-time professionals employed by the STANSE Project. Table 2 provides rate of return figures for each of the groups named above by occupational category for Wave One. Three hundred and eighty- six individuals returned complete network analysis instruments out of a possible five hundred and twenty-nine, or a return rate of 73 per cent. In general, larger school districts (Macomb, Washtenaw, and Muskegon) showed a somewhat lower rate of return than did smaller districts (Table 3). Due to requests for increased confidentiality of the data, identifying respondent codes were eliminated for subsequent waves after the identification of whether respondent was a key communicator or a non- key communicator and, thus, are not reported. Network analysis results are based on a less than one hundred per- cent rate of return. Therefore, the network analysis was performed using unreciprocated links in order to maximize the total number of usable links within the communication network. This means that non-respondents were admitted to the network if they were identified by any respondent as a contactee. Return rates for the metric‘multidimensional scaling measure ments for each of three waves is presented in Table 4. 92 TABLE 2* PERCENT OF RETURNS BY ROLES ISD Directors 53 out of 58 = 91% ISD Supervisors 98 out of 148 = 66% LEA Directors 102 out of 140 = 73% LEA Supervisors 104 out of 152 = 68% SESA Director 1 out of l = 100% SESA Staff 19 out of 21 = 90% STANSE Staff 9 out of 9 = 100% TOTAL 6 out of 529 = 12;21% *Also appears in Kay (1976) p. 222 93 TABLE 3* PERCENT OF RETURNS BY INTERMEDIATE SCHOOL DISTRICT Intermediate School District Percent returned SESA, STANSE Staff 87% Ingham ISD 80% Kent ISD 80% Macomb ISD 77% Muskegon ISD 67% Oakland ISD 60% Washtenaw ISD 43% Wayne ISD 75% Allegan ISD 100% Alpena-Montmorency-Alcona ISD 100% Barry ISD 100% Bay-Arenac ISD 100% Berrien ISD 73% Branch ISD 64% Calhoun ISD 87% Cass ISD 67% Char levo ix- Emit ISD 100% Cheboygan-Otsego-Presque Isle ISD 100% Eastner Upper Peninsula ISD 100% Clare-Gladwin ISD 100% Clinton ISD 72% Delta Schoolcraft ISD 100% Dickinson Iron ISD 100% Eaton ISD 87% Genessee ISD 73% Gogebic-Ontonagon ISD 100% Traverse Bay Area ISD 100% Gratiot-Isabella ISD 100% Hillsdale ISD 100% Copper Country ISD 100% Huron ISD 100% Ionia ISD 100% Iosco ISD 100% Jackson ISD 100% Kalamazoo Valley ISD 50% Lake ISD 100% Lapeer ISD 100% Lenawee ISD 100% Livingston ISD 34% Manistee ISD 50% Marquette-Alger ISD 82% Mason ISD 100% Mecosta-Osceola ISD 100% Menominee ISD 100% Intermediate Schogeristrict Midland ISD Monroe ISD Montcalm ISD Newaygo ISD Oceana ISD Ottawa ISD Coor ISD *Also appears in Kay (1976) p.222 94 Percent returned 100% 67% 50% 50% 100% 80% 100% 95 TABLE 4 Sample Response Rates for Three Waves of Metric Multidimensional Scaling Data gamma; Instruments distributed 529 150 161 Instruments returned 326 97 120 Percentage 73 65 75 For Wave One, an average of 293 observations were obtained per cell; for Wave Two an average of 91 observations were obtained per cell; and for Wave Three an average of 115 observations were obtained. Complete samples for each data wave are found in Appendix D. Note that for Waves Two and Three, respondents were randomly sampled from the total pool of respondents to Wave One, without re- placement. This means that all Wave One respondents were returned to a sample pool, from which two simultaneous random samples were pulled. Individuals were classified on the basis of network analysis data into either non-key communicator or key communicator roles, and were assigned either to the Wave Two sample group or the Wave Three sample group. The population from which the samples were drawn consisted of 529 individuals. This means that no single individual could have been drawn more than twice. Additionally, personnel changes were included in Waves Two and Three if a personnel change caused the replacement of an individual on a titled basis. Reliability of Dissimilarity Matrices Random split-half reliability tests were conducted on data from each point in time (Table 5). The "split-half" procedure required that a given sample be randomly divided into two sub-samples of equal size. 96 Mean dissimilarities were then computed and a correlation coefficient of means calculated. Wave One data obtained a split-half correlation co- efficient of .96. Wave Two, with 97 respondents, was the least reliable data set (R = .81), but was within acceptable tolerances. TABLE 5 Random Split-Half Reliability Coefficients For Three Waves of MMDS Data Wave 1 Wave 2 Wave 3 Sample Size 389 97 120 Split-Half Correlation Coefficient* 96 .81 .86 *Random split-half correlation of mean dissimilarities. Barnett (1974) indicates that scree tests are an apprOpriate reliability indicator of MMDS factor spaces. Studies by Klahr (1969), Spencer and Ogleby (1973) and Stenson and Grow (1969) computed spatial representations of random data input proximities (dissimilarity estimates). The results of these three studies indicated that to the extent that the shape of a stress curve deviates markedly from that pictured in Figure 10, the representation of dissimilarities will be increasingly random. Thus if a stressed distribution shows a sharp elbow in the curve, that distribution is likely to represent reliable and accurate proximity estimations. Barnett (1974) showed that for the MMDS algorithm employed here, input of random data produces an increasingly flat stress distribution. The analysis in Figure 10 was performed by Barnett (1974) on random data. The curve labeled "Ideal" is suggestive of a data set which reflects underlying structure. (Note the axis for Figure 10 and subsequent 97 FIGURE 10 RANDOM AND IDEAL SCREE LINES DIMENSIONS SSHHIS 40 35 30 25 20 15 10' TIME 2 98 FIGURE 11 SCREE LINES FOR WHOLE SAMPLE DATA TIME 3 TIME 1 DIMENSIONS SSHHIS 40' 35' 30' 25' 20' 15’ 10' TIME2 SCREE LINES: 99 FIGURE 12 KEY COMMUNICATORS ONLY DIMENSIONS 40 35 30 25 SSERICLS N O 15 10 TIMEZ SCREE LINES : TIME 1 TIME3 100 FIGURE 13 NON-KEY COMMUNICATORS ONLY DIMENS IONS 101 stress tests indicates variance explained by the specific dimension on the axis labeled "Stress"). The curves represented in Figures 11, 12, and 13, should be compared to the curve labeled Ideal in Figure 10. For each constituent data set a stress curve is provided. Figure 11 gives stress curves for total sample data; Figure 12 gives stress curves for key communicator groups only; Figure 13 provides stress curves for non-key communicator groups. Clearly the figures in all cases support a high degree of apparent under- lying proximate reliability. In fact, despite relatively small samples of key communicators in Wave Two and Wave Three, the shape of the stress curves consistently conform to the "Ideal" pattern indicated by Spencer and Ogleby (1973). Table 6 provides sample sizes for key communicator groups and non-key communicator groups for Waves One through Three. In general, non-key communicator groups provided high random split-half reliability coefficients on the dissimilarity matrices. TABLE 6 Sample Sizes for Key Communicator and Non-Key Communicator Groups for Three Waves of Fair Wise Data 14229.; Real Lara}. Non-Key Communicators 252 87 87 Key Communicators 80 10 33 Population 529 518 401 The sample figures indicate that the Wave I sample was drawn from an organizational population of 529 individuals. Between Wave I sample draws and the sample draw for Waves II and III, eleven individuals were replaced or left their positions. These new individuals were Egg added to the population, hence a population of 518. The Wave II and Wave III 102 samples were drawn randomly from this group, hence, the population available for the Wave 111 sample was 401 persons. Note that ten staff members were measured at each point in time, hence, the dis- crepancy in the figures when they are compared with those in Table 4. Complete sample sizes for key communicator data is provided in Appendix E. Complete sample sizes for all possible pairs for non-key communicators are provided in Appendix F. In general, non-key communicator groups are acceptably reliable (Table 7). Split half reliability coefficients for each response group are given below. TABLE 7 Split Half Reliability Coefficients* Time 1 Time 2 Time 3 Key Communicators .87 .20 .76 Non-Key Communicators .96 .83 .81 *Random split-half correlations of mean dissimilarities. Key communicator responses, due to small sample sizes, are consider- ably less reliable. The only group which is unacceptably weak in terms of reliability coefficient is Wave Two key communicators. The small sample here is due to a combination of chance factors, a small draw due to the random sampling procedures, and a lower response rate (See Table 4). Unfortunately, there was no statistically acceptable way of boosting Wave Two key communicator response rate. The researcher was, at the time, principally concerned with the acceptability of the sample as a random draw, and did not discover the shortage of key communicators until several months following Wave Two questionnaire administration. 103 With the possible exception of Wave Two key communicators, it is believed that samples provide sufficient latitude of variance to permit a high degree of reliability in the estimate of concept similarities. Wave Two key communicators' responses will be treated as reliable for two reasons: scree tests for Wave Two key communicator point to an acceptable underlying factor structure; and an examination of unmanipu- lated dissimilarity judgments (Table 8) between my job and unmanipulated concepts indicates that the Wave Two responses for key communicators are correctly patterned. TABLE 8 Comparison of Sample Nonmanipulated Dissimilarity Estimates for Key Communicators and Non-Key Communicators in Wave Two Non-Key Communicators Key Communicators My Job and Management System 29.29 36.40 My Job and Efficient 27.18 31.00 My Job and Change 27.30 22.40 My Job and Influence 38.77 42.00 My Job and P.A.l98 15.24 18.40 The pairs My Job and SESA, and My Job and Mainstreaming were consi- dered manipulated since they were utilized in several important messages. t-tests revealed all results to be significantly different (p mo unwouom mo:0m>=o00m Hauafiu: 000 <0 oocoo0wo0 0:0Ewouumc0mz owcwsu 0:0ccm0m meow uno0o0wwm .m0m .uwz oncoum oououuoo 000:0 sou. 0: 00 00 00 0 0 umooaoo 111 innovation, STANSE, is 66.4 units. A "unit" is the measure of dissim- ilarity. It is established by the respondent as a function of the statement: If a history class and a math class are 100 units apart, how far apart are: (concept A and concept B). Respondents are instructed that the differences between a history class and a math class taken together add up to 100 units. They are also told that while there are differences between a history and a math class, there are also similarities (e.g. setting, material, context, etc.). Hence, without saying so directly, the 100 unit criterion is supposed to be a "middle range" value. The task of identifying a message is based on the assumption that concepts converge with other concepts with which they are publicly associated. By identifying a small number of concepts which are dis- tributed about the concept "My Job" it should be possible to create a message system which would increase the perceived similarity between the job of special educators and STANSE. Messages, as mentioned, are selected on the basis of the degree to which a projected line through multidimensional space correlates with the optimal "perfect" vector. Secondary criteria are distance traveled to optimizing point, and the estimated distance by which concept motion exceeds optimality. Analysis was performed on a complete data set from Wave One; 790 possible messages were generated. Of these, 19 messages met the three criteria referred to above. These are presented in Table 13. The final selection of a message was based upon reasonableness vis-a-vis the organization's mission. STANSE management were given considerable 112 mu0c= 00.00 a monwum00 0.0 n :0: : 1' II vmuoucoo 000:0 ll \ v 0 :\ \ \ \ 000. 0: n a n b - . :\\r\\4 . . . _ 0 083% {$055300 1 0200020209 m>HHHOom 039 H0000 "mwz<80 mom Vumh HO uGOUHOh mo=0m>cow00 HawaHo: 000 <0 uoaoa0wc0 0:0Eouuums0oz owcoso 060::o00 mmom auaafiuauum .u0m .umz omcnum vapoucoo 000:0 non 0: 00 00 00 0 0 umoocou 118 00.0: 00.000: 00.0: 00.00: 00.0 00. 00.0: 00.00: 00.0 00.0: 00.00 00.00 00.00 00.0: 00.: 00.00: 00.000: 00.0: 00. 00.0: 00.0: 00.0: 00. 00.: 00.: 00.0: 00. 00.0 00.0 00.0 00. 00.0: 00.0: 00.00 00.0: 00.0: 00. 00.0 00.0: 00.0: 00.0 00. 00. 00. 00. 00.: 00.: 00. 00.000 00.0: 00.0 00.0: 00.0 00.0: 00.0 00.0 00.0 00.0: 00.: 00.: 00.: 00.0 00.000 00.0: 00.0 00.0 00.: 00.0: 00.0 00.0: 00.0: 00. 00.0 00.0 00.0: 00.0 00.000 00.: 00.0 00.0: 00.0 00.0: 00.0: 00.0: 00.0: 00.0 00.0: 00.0: 00.0 00.0 00.000 00.00 00.00: 00.0 00.00 00.00: 00.: 00.0 00.0 00.0: 00.0 00.0: 00.00: 0002<0 08000200 00 00.0 00.00 00.0000 00.0000 00.00 00.0 00.0 00.0: 00.00: 00.00 00.00: 00.00: 00.00: 00.00: 00.0 00.0: 00.0 00.0 00.0: 00.00 00.0: 00.00 00.0 00.00: 00.00 00.00: 00. 00.0 umeH<2 0080<0 mam<8 00.00 00.00 00.0000 00.0000 00.0: 00.00: 00.00 00.0 00.00 00. 00.0 00.00: 00.0: 00.: 00.00: 00.0 00.00 00.00 00.00: 00.00: 00.00: 00.00 00.00: 00.00 00.0 00.00: 00. 00.00: 0 0208 oucm0um> mo unmouo0 mo=0m>com0m anuauue 000 <0 oo=o=0us0 0:0amuuunu0o2 owumso 0:0cco00 «mom uuo0o0uwm . 93 . um: omaoum oououcoo 000:0 non 0: 00 00 00 0 0 UNOUGOU 119 4.4 HYPOTHESES Hypothesis 1: Over time, the number of positive dimensions within which the relationship among key concepts can be represented will increase. Tables l3, l4, and 15 provide factor coordinate matrices for each MMDS measurement. At time 1 (Table 14) seven positive factors are generated and four negative factors. A negative factor is a factor with an eigen root which results from the square root of a negative number. The dimensions are not "real" in the mathematical sense, but they are useful in the interpretive sense. A negative factor results when there is inconsistency in the data. For example, consider the estimates below: 1. A and C are 25 units apart 2. A and B are 19 units apart. 3. C and B are 229 units apart. There is no straight line which can link B and C; hence, a curve is imposed, a negative eigenroot emerges, and the dimensionality expands - from two planes to three. Negative factors indicate inconsistency of judgment - or, the absence of information establishing meaningful inter- relationships. At Time 1, negative variance accounted for 19.98% of the variance, reducing the total variance explained by the factor matrix to 80.02%. At Time 2 (Table 14), eight positive dimensions emerge from the factor analysis. Four negative factors reduce total variance explained by 13.87% or 86.13% of the variance in the sample at Time 2 is accounted for by the factor structure. At Time 3 Table 15) 10.9% of the variance is accounted for by the four negative factors, or Time 3 data account for 89.1% of the variance in the sample. Figure 14 shows the trajectory of the increasing positive explained variance. 120 Figure 14 Representation of Increasing Variance-Explained for 3 Points in Time 90 85 80 75 t; t; t: I The trajectory demonstrates the increasing positive variance established across points in time. As noted, increased positive variance is an indication of underlying consistency in respondent dissimilarity estimates. One suspects that the trajectory has reached an apogee at Time 3; however, the factor structure has undergone significant change during the course of the research effort. Because the judgment situation is complex, MMDS pair-wise instruments are relatively immune to test-retest sensitivity (Barnett, 1976), and since no respondent was required to provide more than two sets of data, changes in variance explained are regarded as an indication of an increase in underlying relational structure among the concepts. This means that the concepts have become more meaningful, and relationships are better understood (Barnett, Serota and Taylor, 1976) 121 00.0 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 0500000 00 00.0 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 000 <0 00 00.0 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 o0.550000 00 00.0 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 0:0Emouumc0m2 0 00.0 00.00 00.00 00.00 00.00 00.00 00.00 00.00 owcoso 0 00.0 00.00 00.00 00.00 00.00 00.00 00.00 0:0acm00 0 00.0 00.00 00.00 00.00 00.00 00.00 omom 0 00.0 00.00 00.00 00.00 00.00 uco0o0000 0 00.0 00.00 00.00 00.00 .m00 .u02 0 00.0 00.00 00.00 umcmu0 0 00.0 00.00 wououcuo 00050 0 00.0 now 02 0 mm. 0:0. 0:0. .0. .0: M .0. m 0. m. M 0. 0.00% 0002<0 08000200 "0008<2 0800400200000 0 0208 00 000<8 122 Hypothesis 1 is supported by the data. Hypothesis 2: The magnitude of the attitude A toward Innovation will be significantly reduced between Time 1 and Time 3 for the whole population, and the Innovation I will converge with those concepts with which it is associated in messages. Tables 12, 16, and 17 provided dissimilarity matrices for the whole population (Time 1) and sample measures (Time 2, Time 3). Since the matrix is symmetrical, with D 1, only the lower half of the 11"”: total dissimilarity matrix is provided. Distances along the diagonal are always equal to zero since, by definition, the dissimilarity between a concept and itself is an identity. The matrix is square symmetrical with column numbers (1-12) indicating the same concepts as listed along the rows. At Time 1, STANSE was 66.04 units from."My Job". At Time 2, the distance between STANSE and "My Job" was 54.22 units, a change of 11.82 units in the predicted direction. At Time 3, STANSE shifted in the opposite direction, to 69.74 units from the concept "My Job", or a change of 15.52 units away from the definition of self. Figure 15 gives a graphic representation of the observed changes. Figure 15 Changes in the Distance Between STANSE and My Job: Three Points in Time 70 65 6O 55 50 123 oo.o wm.mm do.m¢ oo.o mn.om oo.o m~.¢¢ nn.m¢ oe.mm oo.o O‘I mo.md N~.mm mo.m¢ -.mm oo.o w. mm.c~ en.mm 00.nc o~.~¢ ¢~.n~ oo.o "I mum2HHM2 NH HH oH N H umoocoo 129 oo.o Ho.ms ~m.a~ oo.o mm.nm oo.o mm. mm. md.mm mc.¢¢ co.mn oo.o O‘I mmo9 Am > Ar31< H3C proposes that concepts associated in a continuous flow of mes- sages should minimize the magnitude of attitudes across successive time intervals. Obviously, this means that a message input effort can sustain the directionality and the degree of change over time. In the present case, this simply did not occur. Table 27 provides comparisons of key communicators at each point in time for the message pairs and the pair STANSE and "My Job". 134 TABLE 27 Change in Message Concept Pairs Among Key Communicators Time 1 to Time 3 WWW STANSE & Child-centered 61.34 60.10 89.72 STANSE & Planning 25.75 23.70 36.10 STANSE & Helpful 43.32 30.30 66.63 STANSE &:My Job 61.68 60.30 85.00 The data show that some change occurred between Time 1 and Time 2, although the changes that did occur were in the predicted direction. Between Time 2 and Time 3, however, large changes did take place, all of which oppose prediction. The key communicators changed negatively overall with respect to STANSE, and with respect to the positions advocated in the messages. H3c is not supported by the data. (H313) ATIN " ATZN > AT3N The proposed model predicts that change for secondary receivers (non-key communicators) will occur in the same general direction as the change for key communicators, but less rapidly. Since information is filtering down through the key communicators, it is likely that little change will occur between Time 1 and Time 2, and substantial impact should be observable by the third point in time. Table 28 provides the results for non-key communicator group for manipulated message pairs from Time 1 through Time 3. 135 TABLE 28 Change in Message Concept Pairs Among Non-Key Communicators Time 1 to Time 3 2221.; 2322.2 use; STANSE & Child-centered 74.54 61.81 75.69 STANSE & Planning 28.40 28.08 34.87 STANSE & Helpful 44.92 48.64 60.56 STANSE & My Job 67.45 53.51 63.68 Again, the inequalities across the three time periods are in the opposite direction of prediction. As with key communicators, some pre- dicted changes between Time 1 and Time 2 occurred; however, between Time 2 and Time 3 the direction of change is strongly reversed. The amount of change between Time 1 and Time 2 for the pair "My Job" and STANSE is somewhat surprising (13.94 units). It is possible the secondary effect of filtered information is greater due to the credibi- lity of the key communicator sources. More remarkable, however, is the size of the shifts between Time 2 and Time 3 for all message concept pairs. The changes average approximately 10 units, but more importantly, the shifts represent changes of twenty percent or more in every case. Hypothesis 3D is not supported by the data. Hypothesis 4: The magnitude of attitude A toward the Innovation held by non-key communicators (u) at successive time intervals, T+1, T12, T+3... T+n will approach the magnitude of attitude A held by key communi- cators at previous time intervals (T, T+l, T+2... +n-1). 136 This hypothesis and its corrolaries stipulate that attitudes of key communicators will change, and through normal attitude development or socialization processes, they will cause the attitudes of the balance of the system to change. It is, in effect, a two-step flow hypothesis (Lazersfeld and Barrelson, 1951). One difference between this model and the classic two-step flow model, is that directed information which flows through key communicators changes both the attitudes of the key communicators themselves, as well as the probability that informa- tion will be disseminated to secondary receivers. (BAA) ATlK " ATlN This hypothesis tests the assumption that key communicators and non-key communicators initially hold identical perceptions of the dissimilarity of the concepts. As proposed, the dissimilarity is an indication of the degree to which concepts are interrelated into a system of meaning; the observation of identity will allow us to assert an equilibrium state at the initiation point in the change effort. Table 29 provides rounded differences for all possible pairs of concepts between the matrix for key communicators at Time 1 and the dissimilarity matrix for non-key communicators at Time 1. 137 TABLE 29 Rounded Discrepancies for Key Communicators and Non-Key Communicators at Time 1* 1 2-7 3-6-13 4-7-5-2 5-2-4 3-13 6-6-8-3-6-11 7-2-3-3-12-7-16 8-6-5-3-14-2-5-4 9-5-6-3-6-9-1-4-4 1-10-3-3-4-6-4-3-4-4 11-4-9 4-9-7-3-4-3-12-4 12—3-1-1-9-3-6-8-8-8-5-11 1 2 3 4 5 6 7 8 9 1o 11 12 Diffij ' K1]- Nij The results clearly indicate that key communicators perceive the concept set to be somewhat more interrelated than do the non-key com- municators. The correlation between the two dissimilarity matrices (.93) indicates that, overall, there is a high degree of association. It is possible that the differences indicate different perceptions of the meaning of the criterion pair; however, it is more likely that key com- municators report smaller values, because they have a greater understanding of the relationships among the concept sets. Table 30 provides a compar- ison of perception of message concepts between the two groups. TABLE 30 Comparison of Message Concepts for Key Communicators and Non-Key Communicators at Time 1 Key Non-Key Communicators Communicators Difference STANSE & Child centered 61.34 74.54 -13.20 STANSE & Planning 25.75 28.40 - 2.65 STANSE & Helpful 43.32 44.92 - 1.60 STANSE & Me 61.68 67.45 - 5.77 138 These findings support the contention that, while the groups are close in their perception of the concepts, they are not identical. Therefore, H4A is not supported (H413) A'rm -“-'- ATZN This hypothesis holds that the attitudes and perceptions of non-key communicators at Time 2 should be very similar to the perceptions of key communicators at Time 1. This is because the information should not have had sufficient time to filter down to impact the entire net- work and, hence, if key communicators are related in the way prOposed by the model, to attitude formation processes, we should find that there should not be much change between Time 1 and Time 2 for key communicators and non-key communicators. TABLE 31 Rounded Discrepancies Between Time 1 Key Communicators and Time 2 Non-Key Communicators l 2 - 5 3 - 8 0 4 - 8 - 2 0 5 - l - 5 2 - l4 6 - 4 - 7 - 2 - 7 - 7 7 - 7 - 5 - 3 - 13 - 12 - 9 8 - 5 - 13 - 5 - 10 - 14 + 3 - 6 9 - 12 - 4 - 7 - l3 - 13 - 7 - 8 - 9 10 - 8 - 6 +11 - 7 - ll - l - 3 + 5 - 12 ll - 8 - 12 + 5 - 12 - 15 - 5 - 7 - ll - 11 - 9 12 - 4 4 5 15 7 10 8 2 4 - 3 - 12 1 2 3 4 5 6 7 8 9 10 11 12 Table 31 provides a matrix of the rounded discrepancies between the dissimilarity matrix for key communicators at Time 1 and the dis- similarity matrix for key communicators at Time 2. A negative number 139 indicates that the value for key communicators was smaller than the value for non-key communicators. The results do not support the hypothesis. Again, non-key communicators systematically report larger distances than key communicators. In addition, both groups have changed considerably their evaluation of many of the objects within the space. Hac‘ ATZK ' AT3N Table 32 gives differences of mean scores for dissimilarity matrices between Time 2 key communicators and Time 3 non-key communicators. Again, we are proposing serial change, in which non-key communicators change subsequent to key communicators. TABLE 32 .Comparison of Dissimilarities Matrix for T2 Key Communicators and T3 Non-Key Communicators l 2 ~11 3 ~ 3 ~15 4 ~ 8 ~ 5 ~ 2 5 ~ 2 ~ 1 ~17 ~ 1 6 ~19 ~17 ~ 9 ~14 ~24 7 ~ 6 ~ 6 ~11 - 8 ~ 8 ~23 8 ~ 8 ~ 7 ~10 ~16 ~ 8 ~41 0 9 ~14 ~15 ~ 9 ~20 ~ 9 ~35 ~13 ~10 10 ~ 3 ~20 ~25 ~12 ~17 ~23 ~16 ~30 ~30 11 ~ 3 ~16 ~11 ~ 0 ~13 ~21 ~ 8 ~22 ~11 ~18 12 ~ 3 ~ 5 ~30 ~ 4 ~12 ~37 ~ 5 ~15 ~13 ~17 ~11 l 2 3 4 5 6 7 8 9 10 ll 12 In this comparison we can see that the discrepancies are aggravated. The average discrepancy has grown from 7.6 to 14.3, and hence, the attitudes of key communicators are even less of a predictor than pre- viously. Part of the differences are attributable to the differences 140 TABLE 16 OBTAINED CORRELATIONS TIME 1 TO TIME 3 T1 T2 T3 \, l\ .88 .65 ' q :5 .59 / / mm. me. oh. mm. a owouo>< voustoz mm. 54. Ne. HN. u owoho>< me.a A-V «a. H-V mm.w A-V we. am.33 A-V an. mo.~3 H-V as. «3 on.m H-V so. ow.4 A-v 34. 44.n 3-3 44. A-V 3m.4 H-V 44. 33 54.3 H-V No. «4.3 A-v mm. 44.4 A-V 4N. A-v 43.~ H-V 4a. 03 «N. A-v AN. an. A-v o4. A-V N3. a-v 44. NN. A-V an. 4 4o. 4N. A-v no. 43. A-V no. 43. A-v 43.o co. m 43.~ 44. H-v 44.3 34. ~4.~ «N. wo.~ no. a 1. .m~.4 34. 44.4 ma. 44.4 43. mm.4 3n. 4 “m 34.4 4a. 43.4 44. A-V oh.4 4o. A-V 33.~ 34. m mm.a 34. 44.~ so. A-V mm.o3 4m. A-v ch.» 53. 4 mm.33 Na. m~.m3 m3. n4.M3 n3. A-V 44.43 44. n mn.m~ as. o~.- ma. 3-3 a~.n~ Na. 4~.n~ mm. N 44.4m no. ma.~4 Na. am.on as. 4n.am ma. 3 soa3mHmmm u u u oocoHuo> a Na comuoom Na coauoom Nx acmuoom uo N GOoumom mucuoom ~z3x mzmx ~z~x 3z3V3 oEHH cH oucHom oousa "nouauoauum uOuoom mu0uooHcsesoo hoxacoz can unauooHcsaeoo hog coozuom ocoHuoHaoHou «m mHm<fi 142 44. 44. 44. mm. H owwuo>< 3433343443 mm. Ho. 44. on. u owouo>< 43.433-4 44. 44.333-4 44.3-4 44.333-4 34. 44.433-4 44. 3-4 43 44.4 3-4 44. 44.4 3-4 44. 43.4 3-4 44. 3-4 44.4 3-4 44. 3-4 33 44.4 3-4 44. 3-4 44. 3-4 44. 44.4 3-4 44. 3-4 43.4 3-4 44. 3-4 43 43. 3-4 44. 44.4 44.3-4 44. 3-4 44. 3-4 43. 3-4 44. 3-4 4 44. 44. 3-4 44. 44.3-4 44.4 . 43. 43. 3-4 4 44.4 43. 3-4 44.4 44.3-4 44.4 44. 3-4 44.4 44. 3-4 4 43.4 44. 43.4 44.3-4 44.4 44. 3-4 44.4 43. 4 44.4 44. 44.4 44. 43.4 44. 44.4 43. 4 44.4 44. 3-4 34.43 44. 44.4 44. 3-4 44.4 44. 3-4 4 44.43 44. 44.43 44.3-4 44.43 44. 3-4 44.43 44. 4 44.44 44. 44.44 44. 44.44 44. 44.44 44. 4 43.44 44. 34.44 44. 34.44 34. 44.44 43. 3-4 3 H H H R New GOmHmwm Nun flOm-Hmmm Nun fiOm-ummm Nun flOmumwm mucuumm 42443 4232 443 343 33,3443 143 wn.o No. mm. H owouo>¢ 443:43o3 Hm. 04. an. n owmuo>¢ 43.43-4 34.3-4 44.333-4 44. 44.4 3-4 44.3-4 43 43.43.4 44. 44.4 3-4 43.3-4 34.4 3-4 44.3-4 33 44.3-44 44.3-4 44.4 3-4 34.3-4 44.3 3-4 44. 43 44. 34. 34. 3-4 44.3-4 44. 44.3-4 4 44.3-4 43. 44. 44.3-4 44. 43. 4 44.4 34. 44.3 44. 44.4 33.3.4 4 43.4 43. 44.4 44.3-4 44.4 44.3-4 4 44.4 43.3-4 44.4 44.3-4 44.4 44. 4 34.4 44. 44.4 43. 44.4 44. 4 44.43 44. 44.43 44.3-4 44.43 34. 4 44.44 44. 44.34 44.3-4 44.44 44.3-4 4 44.44 44. 44.44 44. 44.34 44. 3 .H u H 1.! xx :Omuoom Nx coauoom Nx comuoom ououoom 4242 4444 4244 144 in reliability between the two measures, however, the patterns of gross negative attitudes shift established earlier seems to have asserted itself once again. H4cis not supported. Hypothesis 5: (A) n3K ' lek > 212311 ' XTIN This hypothesis indicates that mean changes on a manipulated attitude should be greater between Time 1 and Time 3 for key communicators than changes for non-key communicators. Ironically, this hypothesis is supported. Table 33 provides a comparison of the change score for each of the message concept pairs and an average change score for key com- municators and non-key communicators. TABLE 33 Change Score Difference for Key Communicators and Non-Key Communicators for Message Concept Pairs: Time 1 to Time 3 Key Non-Key Communicators Communicators Difference STANSE & Child-centered 28.94 1.05 26.99 STANSE & Planning 10.35 6.47 3.88 STANSE & Helpful 23.31 15.64 7.67 STANSE & My Job 23.32 3.77 19.55 Mean Changes 21.26 6.73 15.52 Considerably greater amounts of change were experienced by key communicators than were experienced by non-key communicators. It is noted that obtained changes were not in the desired direction; however, the hypothesis is predicated on the unique information resourses of key 145 mo.oHn NN.4- mm.Na 00.: 00.: mm.H NH.0 00.0 0N.w wn.0H H0.0N HH.wm oocmHum> no ucoouom 4m.¢4wu HH.00N- Ho.- 00.000 m¢.wH4 wo.m¢HN mo.mN0- 0m.wm- Nm.4HH H0.mO0 oN.m¢¢H 00.NNmm mosHm>come 0m.n- m4.0n 04.0 04.N Ho.- 00.0 00.0: 0H.0n m4.m- nN.m 0o.m 00.0- HnmaHom NH 0a.au 00.HH- N¢.Nn m0.H No.1 00.0: Ho. 0N.Nn $0.0 0N.N ¢¢.ON- 0N.n 00H <4 HH 0m.N Nw.H 44.0 4H.- Hc. mm.- 4H.NH 4<.N 0N.¢u 00.NN 4m.N- NH. oocousmH oH No.0 mw.mu Hm. oH.- Ho. 04.: 00.H- MH.0H mN.wu 0m.HH- N0.0u 40.0Hu meEmouumchZ m Ne.mu 00.H Hw.n NN.H- No.1 0N.m 04.0 H¢.H HH.oH om.0H- No.4 H0.- owamco w Nm.4u oH.N 00.0H 0m.Nn No.1 40.0: no.0: Nw.mn no.1 Nw.0u HN.0 m¢.N memcmHm 4 m0.mn 0H.w mN.0- no.N- Ho.- No.N 00.01 00. 0m.0u 00.4 0H.MN- 00.HN moom 0 40.4- 4m.mu M4.0- 0m.Hu No.1 04.Hn OH.1 40.: Hn.n- 4m.0 00.0N 00.0: acoHonmm n om.n Om.HH 0N. ao.m oo.- N0.u 0m.¢n Ho.w m0.oH aH.n 0m.NH om.oH .oam .uwz 0 00.0H 0N.0- 40.: H0. 00. ¢O.H- 00.N 0N.m- mm.Nu mm.NHu Nm.0 H0.4m oocmum m 40.0 o4.HH Q4.Nn 0H.H No. 0m.Nu Nm.N mm.HH- 4m.¢- m0.oHu 04.01 H4.¢N- monouaoo mHHsu N 40.0H 4m.m- 00.: Nm.Hu «o. 0m.N ON.01 4H.m- NH.NH 0w.m 0m. ow.0Hu non 42 H H.3- -3-3- .413- N m H. m .4. m .m M H mum-ma H oEHH wow oooam HocoHocoEHmHuHsz onuoz m :H moHnoHum> NH 00 mommaHmuooo ooHHHou mxoa mo acoouom Nm.NOHH: m0.0mc: No.: om.H¢m 00.000 04.4oHN 4m.0N0: oH.HN: N¢.4¢N 0m.Nn0 04.00NH Hm.0Hmm mosHm>come 40.0H: Hw.m: O0.: 00.H: #0. 04.0: 00.0: mm.NH m4.H: wo.NH 0m.m 40.: HawaHom NH Hm.o: oH.: 4H.HH 4H. Ho. 00.0 mm.¢- 40.0: 00.0: mm.m HN.¢H: 00.0: on NH 00 mouochuooo ooHHHou mMOH NH 00 nouochuooo 04.: 00. H0. NH.00H: 00.0: 40.: 00. H0. 40.0 N0.: 00.: 00.: m4.H: 40. 00.0 H0. 00.H: 00. 00.N 00.: n0.m: H0. mm. H0.: 0H.N H0.: Nm.N: 00. .4- m mMOH NO uCOOhwm mosHo>come HomaHom 00H <4 oocooneH onanouuocHoz owsoso wancon omom ucoHonum . 444 . 4443 oosoum wououcou vHHso 004. .4: NH HH 0H N H HNOUGOU ooHHHm0 148 04.0H: 00.0: N0.H: 00. N0. 00.N 04.0 NN.4 0H.0 00.NH 40.0N 00.00 oucoHuo> mo ucoouom 00.NO0H: 00.00H: 0N.40 N4.N00 00.000H HN.000N 40.0N0: H0. 40.0Nm 00.000 00.0NOH 00.NNNO moaHo>come 0H.0: 00.0H: 00.0 N0.: N0.H 00.0 00.0H H0.0- N4.0: 0H.N: H0.0 00.0: NH 00.0H: 00.HH: 00.N: 00.: 00.: H0.0: 00.N: H0.: 0N.0H 00.0: 0N.0N: N0.4 HH H0.0 0N.H: 00.0 N0. 0N. 00.0: 00.NH: 00.0H: 40.0H: 00.0H: 0H.0: 0N.N oocosHmcH 0H N0.N Hm.0: 00.: H0. N0.H: 00.N: 40.H: 00.0H 00.0H: 0H.0H 0N.NH: H4.0H: wcHBmouuocHoz 0 HN.0: 0N.H 04.: 00.: nn.H N0.0H N0.NH: 00.0: H0.0 00.0H 4N. 00.N: 0 40.0: H0.4 04.0 00.: 40.H: 4N.0: 0H.: 00.H 00.0 0N.4 0H.NH 00.0: 4 0N.0: 04.NH H0.: N0.: 0N.H: N0.m 00.4 00.: 00.0: 00.N: N0.4N: 00.4N 0 0H.HH: 00.N 00.0: 00.: 00.N: 04.: H0. N0.N: 00.0: H0.0: 00.Hm 04.0: usoHoHuuw 0 00.0: 00.0 00.H: H0.: 00.0 NH.H: 0N.H: 00.0H N0.: 00.0: 00.NH 00.HH .440 .uwz 0 0H.HN 04.0: 00.H: 00. 00.: 00.H: NN.N 0H.0: 0H.0 00.HH 00.0H 40.00 m 00.0H 00.0 04.0: 00. 04.H 0H.0: 00.0 4H.0: H0.0 00.0 40.0: N0.Hm: wououcou 0HH50 N 00.0H 04.H: 00.H 00. 40.H: 00.0 H0. 00.0 00.4 0H.0H: 0H.N: 0H.0H: H MH- Hn3. 0.3. N. m M m m 44. m .4. ..3. www.80- H oEHH mom oooam HmconaosHvHuHsz oHuuoz o :H ooHnoHuo> NH mo oouochuooo ooHHHoo mmOH0¥ozoz 44.4 44.43 44.43 44.44 44.44 ooca3u4> 44 4444444 44.4433 44.3444 44.444 34.3443 34.4444 44:34>:4434 44.4- 44.4- 44.4 33.4 43.43- 3444344 43 44. 44.43 44.4- 44.44- 44.43 443 <4 33 44.4 43.4- 44.44 44.3- 44.4 4444:3443 43 44.4 44.43- 44.43- 44.43- 44.44- 443544444434: 4 44.4 44.4- 44.- 44.- 44. 444444 4 44.4- 44.3 44.43- 44.4 34.4- 44344434 4 44.4- 44.4- 43.4 44.44- 44.44 4444 4 44.4- 44.4- 44.4 43.44 44.33- 344343444 4 33.43 43.4 43.4- 44.43 44.33 .444 .44: 4 44.4- 44.4- 44.43- 44.43 44.34 444444 4 44.4- 44.4 43.4- 44.4- 34.44- 44444444 43344 4 44.4 44.43 43.4 43.4 44.4- 444 42 3 .m .4 .m _M .H 4444440 N oEHH you oooam HoconcoEHvHanz oHuuo: o :H moHnoHuo> NH mo mouochuooo ooHHHou "xHMH mo ucuouom 00.400: 00.00: 00.00 00.000 40.000 00.0000 00.000: #0.: 00.000 00.004 00.0000 00.0000 mmaaw>cowwm 00.0: 00.0: 0H.H: 00.: 00.H: 00.0: 00.0: 00.0 40.00 00.0 00.4 00.00: #50040: 0H 00.00: 00.0: 04. H0. #0. 00.0 00.0 00.0H: 00.0 H0.: 00.00: 00.0 000 <0 as 00. 00.H: 00.: #0. 00.H: 00.0 00.0: 00.0 00.0H: 04.00 00.00: 00.0 mucosgmau 04 00.H: 00.4: 04.H 00.: 04.: 40. 00.0 00.04 00.0: 00.00: 04.00: 00.00: wadauouumcuwz 0 00.0: 00.0 40.4: 00.: 00. 00.4: 00.0: 00.00: 00.00: 40.00: 00.0 00.: owcuso 0 40.0: 00.0 00.: 00.: 00.: 00.0 00.0: 00.: 00.0 04.0: 04.00 04.0: wcwcamam 4 00.0 00.4 40.: 00. 40.H 00.0: 0H. 04.4 00.0 HH.0 00.00: 00.00 «cum 0 00.0: 40.0: #0.: 00.: 00.0 00.~ 0H.: 00.0 00.0: 00.00 00.00 00.00: 4:0000000 0 00. 04.0 00.0 #0. 00.: 40.0: 00.4 00.0: 00.0 00.NH 40.00 00.0 .440 .402 0 00.H~ 40.4: 00.: 00.: 00.: 40.H 00.0: 04.0 00.0 00.00: 00.00 00.00 «4:040 0 40.04 04.H 00.0 00.: 00. 04.H 00.00: 00.0: 00.0 00.4: 00.Ha: 00.00: wououcoo vauso 0 00.00 00.0 00.0: 00.: H0.: 00.: 00.00 00.0: 00. 00.4 00.4: 00.0H: non 0: a .44 NM. .44 0 .0 M 34. m .0 m .4. .3. 4.4.4900 0 4309 you oumnm Hwao«4:oE«0au~=: 04444: a :4 40404444> 04 mo muuucuvuooo 0944040 mmofi 1'1“sz > ”T3101 Figure 16 depicts the obtained correlation pattern for various com- binations of matrices described in the hypotheses. The correlations used in the figure are weighted average Pearson correlations between individual orthogonal factors. Table 34 provides the individual cor- relation coefficient for each of the orthogonal factors (or the cosine of the angle between factors). The average Pearson correlation co- efficient is provided and an average coefficient weighted for the average percent of the variance explained by each factor is provided. Simple correlations, rather than path coefficients are provided because we are comparing factor matrices, rather than scores on individual variables. The average weighted correlation coefficient is used in the figure. Complete factor matrices are found in Tables 35 and 40. The Hypothesis SB indicates that the correlation of factor structures for key communicators at Time 1 with non-key communicators at Time 1 will exceed (i.e., be more predictive than) the correlation between factor structures at Time 2, which will, in turn, exceed the correla- tion between key communicators and non-key communicators at Time 3. This hypothesis is based upon two considerations. First we are imposing unique data and experiences on key communicators which should 152 cause them to change at different rates than non-key communicators and, hence, reduce the homogeneity between themselves and the rest of the network. Secondly, as they change, key communicators will be attending to novel information stimuli which would cause them to have less "message" or communication time to fulfill their normal communication behaviors. This implies that individuals have finite amounts of time to devote to communication behavior. In this view, a new communication demand is more likely to involve replacement of a prior activity than the addition of more functional time for relevant communication. At time 1, the weighted average correlation for key communicators' and non-key communi- cators' data matrices is .88; at Time 2 the weighted correlation coeffi- cent is .70; and, at Time 3, the weighted correlation coefficient is .65. The unweighted correlation coefficient provides somewhat equivocal dis- conformation. Time 1 average unweighted correlation is equal to .71; at Time 2, average unweighted correlation is equal to .42; and at Time 3, the average unweighted correlation coefficient is equal to .47. It is held that the weighted average correlation coefficient is the better indicator, since it takes into account the fact that certain fac- tors account for most of the structural variation of concepts within the overall factor space, while other dimensions, particularly the zero dimen- sions, account for no variance, or small fractional amounts of variance. Thus, hypothesis 5B is regarded as confirmed. (H50) rTKNngKTN?rTKTN 1 1 2 2 3 153 This hypothesis indicates that the pattern of change in the factor structure should induce changes across points in time for key communi- cators and non-key communicators to nearly equal one another. That is, changes in key communicator factor structures from Time 1 to Time 2 should be reflected in changes in non-key communicator factor structures from Time 2 to Time 3. The average weighted correlation coefficient for rT KN is equal to .88; the average weighted correlation coefficient for rT KT N is .85, an approximate identity. The weighted average cor- relation goefficient for rTZKT3N is .66, a considerable discrepancy from the other two correlations. Hypothesis 5C is, therefore, not supported by the data. (HS ) r r D TIICI‘ZN > TINTZK and r r T2KT3N > TZNT3K This hypothesis indicates that in every case, the correlation across points in time between key communicators and non-key communicators should be greater than the correlation across points in time between non- key communicators and key communicators. That is, key communicators should be better predictors of the future position of non-key communicators than non-key communicators should be predictive of key communicators. Dis- confirmation of these hypotheses was argued to be support for the position that the larger bulk of the network sets the agenda for key communicators. Hence, disconfirmation of these hypotheses constitutes some confirmation for Hypotheses 6 and Hypothesis 6A' 154 (H6) and r r r r TINT2K> TlKTZN T2m3x > TZKT3N and X -X 31 (H6 ) - A T3N TIN > XT3K TlK First of all, we have already determined that the changes for key communicators across the three points in time were greater than the changes for non-key communicators across the three points in time; or Hypotheses 6, 6A are dismissed. The correlation pattern observed in- dicates that rTIKIZ’ rTlKTzN had a weighted correlation coefficient of .81, while rTlNZK had a weighted correlation coefficient of only .39, indicating strong support for Hypothesis SD. The correlation coefficient for rTzKT3N was found to be .66, while the correlation coefficient for rTzNT3K.was found to be .59 or, again, a difference in the predicted direction. Therefore, H5O and HSC1 are regarded as confirmed, and H6A is regarded as disconfirmed. Note that we did not find that changes across the two groups were equivalent, or that there were differential rates of change. Thus, the data do support the notion that key communicators are good predictors of the changes to be observed across subsequent points of time for the network. The fact that the raw pattern of change deviated from the prediction about the direction of change and the effect of the messages does not reduce the value of the finding that key communicators do seem to be influential, or highly related, or indicators of the direction of change within an organization. 155 4.5 SUMMARY Overall, the hypotheses provides equivocal support for the model. It is found that substantial and predicted changes do occur between Time I and Time II, however, between Time II and Time III changes oppose prediction. In general, it is found that key communicators did experience more changes than non-key communicators, giving support to the concept of two-step leadership. CHAPTER 5 CONCLUSIONS AND DISCUSSION Organizational change may be the consequence of external or internal pressures, however, it's felt impact is at the personal level. Organi- zational members are responsible for altering their job-related behaviors in conformance with an alteration in production strategies or institu- tional policies. It has been argued that change will occur most successfully when managers consider the consequences for subordinates and provide messages which clearly relates the objects changed to the self-perception of the subordinates. Implicitly this means that one's self-perception is tied to one's occupational status and one's productive motivation -- or, one's balance of inducements and contributions. This argument was expressed as three central propositions: 1. An organizational innovation will be adopted to the extent that it is perceived to possess attributes congruent with the salient major inducements and contributions involved in individual's conceptions of their work. 2. To the extent that an attribute is instrumental to performance on the job, it will be closer to individual's definition of self. 3. To the extent that an innovation possesses attri- butes which are themselves close to individual's definition of self, the probability of adoption is increased. 156 157 These three statements provide the basis for building a model of change management. The function of the manager is to develop strategies which increase the probability that salient attributes of novel processes become generally known; to identify those salient attributes; to monitor the progression of the system in terms of an adoption goal; and to shift strategies as conditions emerge and change. This process requires, at a minimum, a method of ascertaining critical values in the work force, a method of monitoring employee perception over time, and a method of deducing message strategies from the analysis of employee perceptions. In short, it requires a method for identifying and operationalizing a communication program for change. Once the manager institutes the information or persuasive message effort in conjunction with training and structural changes appropriate to a particular innovation, the manager's role is to continue to gather data -- feedback -- until such time as the change is either institu- tionalized or it is replaced by something else. The program is predicated on the assumption that when two social objects are related to one another in a message, they come to be seen as more similar. This is a rather basic premise for language develop- ment and understanding. However, the validity of applying the premise to particular situations is constrained by such factors as credibility of the source, direct salience of the change, and distortion of the channel. This research attempts to exploit these potential constraining var- iables. Key communication linkers were identified as message bearers to control source credibility and significance (although no attempt 158 was made to differentially establish the source credibility of particular actors.) One innovation was selected which was arguably very significant-- the restructuring of administrative relations within a state system of special education administration, although the degree to which the change was perceived as significant was not examined for fear of re- ducing the effectiveness of subsequent messages. Finally, through the use of key communicators I attempted to exploit interpersonal channels-~the most effective communication channel although the channel most subject to distortion. To control distortion, interpersonal messages were supported by memoranda signed by the State Director of Special Education Services. Hypotheses Table 41 provides a summary assessment of the hypothesis tests. It was found that key communicators did change to a substantially greater degree than did non-key communicators. Between Time 1 and Time 2, the results are much as predicted, however, between Time 2 and Time 3 the direction of change has shifted, and the results oppose prediction. Overall, STANSE did not converge either with "My Job" or with the concepts with which it was publically associated in the message strategy. The results did, however, show that in all cases, key communicators changed significantly more than did non-key communi- cators, and the pattern of hypothesized correlations was generally obtained. 159 Summary of Hypotheses Hypotheses Hypothesis 1: Over time the number of positive dimensions within which the relationship among key concepts can be represented will increase. Hypothesis 2: The magnitude of attitude A toward the innovation will be significantly reduced between Time 1 and Time 3 for the whole population, and the innovation I will converge with those concepts with which it is associated in messages. Hypothesis 3: To the extent that information is directed toward key communicators (K) the magnitude of their attitude will be less than that of non-key communicators (N) at T2 and T3. ATZK < ATZN 113,, AT3K < AT3N ATlK > ATZK > AT3K H3d ATIN 2 ATzN > AT3N Hypothesis 4: The magnitude of attitude A toward the innovation held by non-key communicators (N at successive time intervals, t+l, t+2, t+3... t+N) will approach the magnitude of attitude A held by key communicators at previous time intervals (t, t+1’ t+2000t+N-1)o H48 ATlK = ATIN H41, AT K < ATZN ATZK = AT3N Result Supported Not Supported Partially Supported Supported Not Supported Not Supported Not Supported Not Supported Not Supported Not Supported Not Supported 160 Mass Hypothesis 5: Key communicators will, in general experience greater amounts of change, and changes in key communicators will predict subsequent changes in non-key communicators. H53 iT3K — f'rlx > 3&3N — 'iTlN H51) rTlKN > rTZKN > rT3KN H5 c H5d rTlKTzN > rTlNTZK and rT2KT3N => rTzNT3K Hypothesis 6: Non-key communicators will experience more change than key communicators, and changes in non-key communicators will predict changes in key communicators. H63 3T3N - 2T1»: > §T3K - RTIK and Result Supported Supported Partially Supported Supported Not Supported Not Supported 161 Two factors together may account for the predicted changes between Time 1 and Time 2, and the absence of predicted changes between Time 2 and Time 3. First, rather than insisting upon a precise information dissemination effort directed at key communicators, SMTF members changed the research plan and emphasized memoranda to the entire network. Three attempts were made to individually contact key communicators by the SMTF, however, the bulk of dissemination effort shifted from the inter- personal channel to mass channels. Key communicators were in effect bypassed. While key communicators did not respond negatively at Time 2, it is also clear that they did not provide whole-hearted enthusiasm. In fact, non-key communicators were more positive about STANSE at Time 2 than were key communicators. It is possible that key communicators, operating in the position of gate keepers, withheld judgment on the innovation until such time as they could observe its manifest effects. Since their interests had been aroused deliberately by STANSE, and since they had expected an informal role in STANSE activities, the sudden end of the message campaign was associated with a negative shift between Time 2 and Time 3. This shift in attitudes was also re- flected in changes in non-key communicators between Time 2 and Time 3. We know from the network analysis that key communicators were highly linked to the STANSE group. Since STANSE was releasing no information, except that which filtered from the group to the network on an inter- personal basis, it is possible to argue that the shift in attitudes between Time 2 and Time 3 was initiated and sustained by key communi- cators. This is supported by the pattern of correlations. Across the 162 three points in time, key communicators were better predictors of non- key communicators at subsequent points in time, than were non-key communicators. Secondly, the model is predicated on the continuous expenditure of communication resources. Had a cut-off in message flow been anti- cipated as a part of the research design, it would have been hypothe- sized that ending a message campaign before the point of adoption would have caused the system to reject the innovation and return to normative values. Clearly, insofar as key communicators were concerned, the innovation was rejected. An interesting observation is that the concepts associated with the innovation did not change substantially in relation to "My Job" and, hence, this lends support for the pro- position that rejection results in stabilization at pre-existing levels. In fact, the hypothesis (H1) that increased information about the concepts would increase the positive dimensionality of the factor space was not rejected. This seems to indicate that following rejection of the innovation, the system returns to an even more cohesive equili- brium. The meaningfulness of all concepts, including the innovation, is more clearly established. This may provide evidence that successive failures to diffuse innovations will continuously increase a system's resistance to change. This possibility should be borne in mind by others using this method. We observe that the predicted correlational pattern was upheld (HSa’ H5b). Associations between key and non-key communicators did diminish, particularly between Time 2 and Time 3. While the correlation 163 is strong, (.65) at T3, it is interesting to note that key communi- cators at Time 1 were better predictors of non-key communicators at Time 2 than were non-key communicators at Time 1. It seems that between Time 1 and Time 2 both groups were responding to the stimuli and, perhaps, to one another. Between Time 2 and Time 3, key communicators changed in their perception of STANSE, while non-key communicators remained relatively stable (.78). I would argue that this is because key communicator expectations had been aroused, which were not subsequently satisfied. If we accept the argument that between Time 1 and Time 2 key communicators withheld judgment, and then between Time 2 and Time 3, they observed a lack of action and a lack of input for their own part, it becomes possible to understand the reason for the radical shift in their attitudes. Were there to be a fourth point in time, the model would predict that non-key communicators would approach the attitude of key communi- cators at Time 3. Assuming that no new information was provided to the system, key communicators would stabilize their attitude set. Non-key communicators would then approach the attitude of key communicators. This would be an indication of a "re-freezing" of the organization. This expression, re-freezing, may be taken in its most literal sense, since the increase in variance explained by the factor structure points to subsequent difficulties in inducing change. While the remaining hypotheses were essentially unsupported, it is contended that the results provide some support for the model. Between Time 1 and Time 2, changes were observed which corresponded to prediction. This was the only period during which messages were actively being dis- seminated. Following Time 2 measures the researchers indicated to STANSE 164 and SMTF personnel that, while they had begun to obtain change in the desired direction, the rate at which change was occurring needed to be accelerated. This conclusion was based on the fact that the dis- tance between STANSE and "My Job" was still one of the largest dis- similarities in the data matrix, and the distance was associated with a large standard deviation (40). The response of the SMTF was complacency and a shift in emphasis from communication in the three levels of the organization to working on internal "product" development and internal relationships. In effect, after proposing to the network that they were representative of the field in State planning, they ceased to perform that role. The key communicators were in the best position to detect this change. They had all been individually contacted, and were led to expect that they were to play an important role in the determination of SMTF activities and priorities. Having aroused the expectations of key communicators, subsequent judgments about the innovation were dependent upon the observation of subsequent action. When these ex- pectations were not satisfied, a boomerang effect took place, and rapidly STANSE diminished in organizational acceptance. Another factor which undoubtedly affected the outcomes of the change effort was re- lated to the inability of SMTF to determine its institutional relation- ships with a given concept. It was emphasized in the consultation- implementation process that if one is to say one is associated with a given concept, one must engage in publicly observable activities, as well as messages,which support the tendered association. The researchers indicated to STANSE that they needed to engage in specific activities 165 (called functions) which were demonstrative of the concepts Child- centered, Planning, and Helpful. We suggested they develop a set of routine activities such as reviewing the Federal rules and regulations on programs for special children, synthesize these and report them to the field. we suggested they present "friend of the court" briefing to SESA's special education policy committees and planning agencies. We suggested they conduct, or authorize, an analysis of administrative Options for reducing overhead, and hence provide more direct support for the children who were the ultimate beneficiaries of their activi- ties. While SMTF sub-committees agreed to work on these or related problems, they did not provide the organization with continuing infor- mation about their efforts in these directions. STANSE staff concentrated on the development of "products" associated with the problems listed above. These efforts resulted in the creation of a comparative evaluation of legal and administrative requirements of State Law P.A. 198 and Federal Law P.L. 94-142. Secondly, they developed a state-wide human resources inventory -- a listing of specialties among special education teachers in the state, and persons with specialties. The administrator of a particular student with an unusual disability could identify a teacher with exper- ience in that disability. Third, they produced a position statement on the State Department reorganization. Using planning sessions con- ducted with SMTF members, STANSE produced a critique of the state reorganization plan, which precisely paralleled subsequent criticism's from a public blue-ribbon commission. Had SESA managers been aware of the implications of STANSE recommendations, management would have gone 166 into the committee with a much more realistic program. In addition, if the state-wide network of special education administrators had been aware of the STANSE findings on a continuing basis, pressure would have been applied to force management at the state level to respond to SMTF criticisms. This list of "products" is provided to show that while STANSE was not communicating, they were engaged in some valuable activities. As the model indicates however, acceptance of an innova- tion depends as much on the perception of the innovation among organi- zational members as it depends upon the objectively determined contribution of the innovation to the organization's success. Perception of organi- zational success develops through communication. And the absence of communication between Time 2 and Time 3 caused STANSE to be rejected as an innovation despite a rather impressive list of accomplishments. This rejection is ultimately reflected in the final distance between STANSE and organizational members. The future of STANSE has not been firmly decided. While SMTF members have indicated continued support, management at the State Department has apparently decided that STANSE will be dissolved. The dissolution of STANSE, given the requirements of P.L. 94-142 for a state-wide planning system, indicates that another organization, of a similar nature will arise to replace STANSE. The sad fact is, that this organization will face greater resistance in attempting to develop state planning protocols. 5.2 FUTURE RESEARCH It is held that the model has been given preliminary support by this research effort. Several obvious research problems were overlooked 167 in an attempt to test the entire system. First, this research design should be replicated in a more controlled situation. The organization used for this research equivocated on message strategies, and did not fully commit to the research effort. An experimental organization would be inadequate. The model assumes that an organization has existed for sufficient time for normative practices to have developed, and for an innovation to represent true options which counter historical practice. The replication should be conducted with additional research efforts directed at the organizations historical response to change, its message costs, and the perceived turbulence of the environment. Such research ought also to carefully document message dissemination efforts, and direct the bulk of communication resources at key communicators. Second an experiment needs to be conducted on the method of message selection. Cody (1976) found some support for the method, as did Barnett, Serota and Taylor (1976). A precise experiment should be conducted. Ideally, such an experiment would compare the efficacy of a "perfect" message strategy selected using the vector summation tech- nique, with a randomly composed message (the potential of which could be established using the vector summation technique). In addition, control group receiving no messages should be designated. While this would not solve all the problems, it would get at the issue of the differential value of using an optimizing routine over normal compo- sitional methods. Third, a method for weighting the differential force of different concepts needs to be developed. Assuming that the vector summation 168 technique can be experimentally validated, there is no reliable method for establishing the differential effect of different individual concepts. This is a two-edge problem. First, if a concept is highly effacacious of change (or more instrumental), the predicted vector will vary from the obtained vector as a ratio of the difference in strength between the two concepts. If the difference is large then the message could produce change which is considerably less than optimal, and possibly dysfunctional. In addition, if the concepts are differentially subject to change because one is more "massive" than another, it could be the case that associating a non-massive concept with an innovation will cause the concept to move toward the innovation rather than the other way around. It is suggested that examination of variances for pair dissimilarity estimates could provide a method of determining concept mass a priori. Essentially a variance is an estimate of the degree to which a sample disagrees about the true score of an indicator. If the ratio of a known variance to the average variance for a data set is determined, the com- parative size of that number would be an indicator of pair instability. Larger variances would indicate disagreement within the population about the meaning and the relationship between the concept pairs, and small variances would support generalized population agreement. In general, it would be hypothesized, that the larger the variance, the greater the potential for change in the concept relationship. Information, is then defined as the medium which induces change in both the absolute dissimi- larity, and in the variance around a given dissimilarity estimate. Consistently, it was observed that key communicators reported lower values for pair distances than non-key communicators. We know from the 169 research of Allen and Cohen (1969) that network liaisons seek and receive more information than non-liaisons. Here we have extended the functional definition of a liaison to bridges. The question is, do key communicators report smaller values because their information behaviors cause them to have a greater understanding of the inner-relationship among measured objects? Finally, considerable research needs to be directed to the estab- lishment of the parametric properties of the "control system". It has been stated that a cybernetic system is governed by a control mechanism which uses upper and lower variable range limits to determine the amount and kinds of input required to maintain the system within tolerable deviations. It has also been argued that an innovation effort should push the system's performance to the upper ranges of established toler- ances. It remains unknown how much information produces how much change. Secondly, we do not know what variables impact upon that functional relationship. It is suggested that an experiment be conducted which utilizes concept variances, information input (as a function of cost), and different combinations of media to establish some expectation probabilities of how much communication energy it will take to cause a movement of g_units of concept‘g with a variance 82. A corollary to the problem of fixing the amount of information required to effect a predictable rate of change, is the question of how much change can a system absorb. As suggested in the literature review such an answer would be dependent upon system structure, perceived envi- ronment turbulence, success in past adoptions, generalized organizational inclination to change, resistance of key communicators to change, and the nature of the dissemination effort. This problem could be addressed 170 through a stochastic process model which posits the functional relation- ships among such factors as organization attributes and a dependent variable consisting of an index of profit, growth and stability of productive processes. 5.3 SUMMARY This dissertation examined a model of organizational change as controlled by the systematic introduction of information designed to adjust attitudinal instrumentalities. A model was proposed which linked acceptance of an innovation to the degree to which the innova- tion is integrated into the job perceptions of organizational members. The model was tested on an organization consisting of 539 administra- tors of special educators working at three levels. Network analysis was used to identify key communicators and non-key communicators. A key communicator is a person who links large groups of other people. The results showed that when information is used in the way dictated by the model, predicted changes did occur. When information was absent, a boomerang effect was observed, and the innovation was not adopted. APPENDIX A INSTRUCTIONS AND MMDS OUESTIONNAIRES 171 SIAIE CW AUCHKEAN DEPARTMENT OF EDUCATION WWW... lensing, Michigan 48902 warm: 1'st my tuuau JUN“? 7. 1975 “scans am Visa have” Inuuwbrzvxxorrna d 8mm I '"‘“*"” AHNZTIAIHLLEI Nit m [was IAIIARAILINflHOUCHELLE or. not. a. nun? Dear PartiCiPantr aaaaaxx:.aonsnrs W 0110 sroczwsrza. sx There are many problems facing us in our attempt to deliver oomsmmvbc. W .1.0 no meaningful Special Education programs and services to handicapped ‘children and youth. Perhaps no one is as familiar with these concerns as you are. One serious problem that we all share is in the area of communication. We simply do not have a complete, effective and efficient system which allows everyone to be aware of the state of affairs in Special Education. My staff and I feel this deficiency seriously undermines our ability and yours to act effectively in many situations. The many written and verbal messages concerning the lack of communication testifies to the seriouSness of this issue. In order to reach a solution to this problem I have asked Project STANSE to develop procedures which would accomplish two tasks: 1. Provide an up~to-date description of the communication network we presently have. This information would include who is communicating with whom and a description of the topics discussed and, 2. Develop procedures to "test” the network so that by August. 1976, we would have a better organized system. This task will entail contacting you more than once this school year. In order to achieve these goals we have prepared the attached questionnaire. I realize that the questions are somewhat more difficult and time-consuming than conventional questionnaires you are probably more familiar with. However. the tvpe of questions asked here allow us to achieve a great deal of accuracy and reliability. In addition, these techniques allow us to measure changes which take place during the course of the STANSE Project. measurements of change will provide educational planners with precise estimates of future needs. This task is a complex one. but with your help it can be successful. I urge you to complete the enclosed questionnaires and return them immediately. If you do not send in this information. your vital information will. of necessity, be omitted. In the past we have shared the successes and the problems associated with Special Education in Michigan: I feel confident that the future will have us working together even more effectively. Your assistance is absolutely necessary. All individual responses will be kept in the strictest confidence. However. it is necessary that you sign the questionnaire for our record keeping purposes. All names will be converted to code numbers and no one except the staff directly concerned with processing the data will ever see the original questionnaires. All participants will be provided with a summary report explaining the results of the study and outlining the implications of this research for the future of special education in Michigan. If you have any questions about these questionnaires please feel free to call Pat DuPort. Nancy Kaye or Evan Peelle collect at 313- 763- 3411. As staff of Project STANSE they can be of assistance to you: 1 . M'C‘H'Gptfl Sinc‘re yflW:—— :3" \ .77“ 3"sz Murray . Batten 1f72 SURVEY OF NEEDS IN SPECIAL EDUCATION: THE PLANNER'S PERSEPCTIVE Instructions to Respondents The following questionnaire asks you to give us your opinions on a set of ideas or concepts that are related tospecial education. We would like you to give your opinions by telling us how different pairs of concepts are. The way you will do this is to estimate how far apart two concepts are. Distance between concepts is measured in units, so that the more different two concepts are, the more units apart they are from each other. To give you a "yardstick" to enable you to express how far apart two concepts are, we will say that a Special Education Classroom is 100 units different from a General Education Classroom, or a Special Education Classroom and a General Education Class- room are lOO units apart. In other words. all the differences between a Special Ed- ucation Classroom and a General Education CTEEsroom together account for lDO units of difference. The idea is for you to tell us your opinion of how many units apart the concepts which follow are from each other. Remember, the more different two concepts are from each other, the larger the number of units apart tfiEy are. If you think any pair of concepts are more different than a Special Educa- tion Classroom and a General Education Classroom. you would write a number larger than 199. If you think two concepts are less different than a Special Education Class- room and a General Education Classroom you would write a number smaller than 100. If you think two concepts are identical. that is, they are the same thing, you would write a “0." FOR EXAMPLE, when completing a similar questionnaire, an educator was instructed that "one room schools and large high schools" were 100 units apart. He was then asked to estimate the distance between: MY SCHOOL AND RELIGIOUS EDUCATION He determined that since his public primary school did not offer any courses or programs in the area of religious education. but since ”moral education" was a small part of their curriculum, the two concepts were 90 units apart. After completion, the sample line looked like this: MY SCHOOL AND RELIGIOUS EDUCATION 90 Three key definitions need to be kept in mind as you begin to make your distance estimates: l. STANSE means Statewide Technical Assistance Network in Special Education. a state and federal funded project. 2. SESA means the State Department of Education (Special Education Services Area. 3. P.L. 94-l42 means the new Federal Law for the handicapped. We realize that you might feel that your estimates are not perfectly accurate for every pair of concepts. Remember, there is no one right answer. Providing your own best estimate of the distances betweer each pair will be sufficient for our purposes. If you do not recognize or cannot give an estimate for one pair, leave the space blank. Please ignore the numbers adjacent to each pair in the boxes. They are used for coding your responses for the computer. If you have any questions, or you need any help in responding to this questionnaire, you may call collect. Project STANSE. and ask forlwaroaret, Thom..n (313-763-34ll). ]fi73 Page 1 IF A SPECIAL EDUCATION CLASSROOM AND A GENERAL EDUCATION CLASSROOM ARE 100 UNITS APART, HOH FAR APART ARE: Child-Centered and PAC STANSE STANSE STANSE STANSE STNASE STANSE STANSE STANSE STANSE STANSE STANSE and and and and and and and and and and and Manag-ment System Efficient SESA Planning Frustrating Change Mainstreaming Influence P.A. 198 Helpful PAC Management Management Management Management Management Management System System System System System System Mainstreaming Management Management Management System System System and and and and and and and and and Management System and Efficient and SESA Efficient and Planning Efficient SESA Planning Frustrating Change Influence P.A. 198 Helpful ,PAC Do Not Hrite in These Spaces 01-08) 09-17) 0214 (18-26) 0304 (27-35) 0305 (36-44) 0306 (45-53) 0307 (54-62) 0308 (63-71) 0309 (72-80) 0310 (01-08) (09-17) 0311 (18-26) 0312 (27-35) 0313 (35-44) 0314 (45-53) 0405 (54-52) 0405 (53-71) 0407 (72-80) 0408 01-08) 09-17) 0409 (18-26) 0410 (27-35) 0411 (as-44) 0412 (45-53) 0413 (54-62) 0414 (63-71) 0506 (72-80) 0507 1fl74 Page 2 IF A SPECIAL EDUCATION CLASSROOM AND A GENERAL EDUCATION CLASSROOM ARE 100 UNITS APART, HOH FAR APART ARE: Efficient and Frustrating Efficient and Change Efficient and Mainstreaming Efficient and Influence Efficient and P.A. 198 Efficient and Helpful Efficient and PAC SESA and Planning SESA and Frustrating SESA and Change SESA and Mainstreaming SESA and Influence SESA and P.A. 198 SESA and Helpful SESA and PAC Planning and Frustrating Planning and Change Planning and Mainstreaming Planning and Influence Planning and P.A. 198 Planning and He.pful Planning and PAC Frustrating and Change Frustrating and Mainstreaming Do Not Hrite in These Spaces (01-08) (09-17) (18-26) (27-35) (36-44) (45-53) (54-62) (63-71) (72-80) (01-08) (09-17) (18-26) (27-35) (36-44) (45-53) (54-62) (63-71) (72-80) 0608 0609 0610 0611 0612 0613 0614 0708 (01-08) (09-17) (18-26) (27-35) (36-44) (45-53) (54-62) (63-71) (72-80) 175 Page 3 IF A SPECIAL EDUCATION CLASSROOM AND A GENERAL EDUCATION CLASSROOM ARE IOOUNITS APART, HON FAR APART ARE: My Job and ChiId-Centered My Job and STANSE My Job and Management System My Job and Efficient My Job and SESA My Job and Planning My Job and Frustrating My Job and Change My Job and Mainstreaming My Job and Influence My Job and P.A. 198 My Job and Helpful My Job and PAC Child-Centered Child-Centered System Child-Centered Child-Centered Child-Centered Child-Centered Child-Centered Child-Centered Child-Centered Ch i l d-Cente red Child-Centered and STANSE and Management and Efficient and SESA and Planning and Frustrating and Change and Mainstreaming and Influence and P.A. 198 and Helpful Do Not Write in These Spaces (01-08) (09-17) 0102 (18-26) 0103 (27-35) 0104 (36-44) 0105 (45-53) 0106 (54-62) 0107 (63-71) 0108 (72-80) 0109 (01-08) (09-17) 0110 (18-26) 0111 (27-35) 0112 (36-44) 0113 (45-53) 0114 (54-62) 0203 (63-71) 0204 (72-80) 0205 (01-08) (09-17) 0206 (18-26) 0207 (27-35) 0208 (36-44) 0209 (45-53) 0210 (54-62) 0211 (63-71) 0212 (72-80) 0213 176 Page 4 IF A SPECIAL EDUCATION CLASSROOM AND A GENERAL EDUCATION CLASSROOM ARE 100 UNITS APART, HON FAR APART ARE: Frustrating and Influence Frustrating and P.A. 198 Frustrating and Helpful Frustrating and PAC Change and Mainstreannng Change and Influence Change and P.A. 198 Change and Helpful Change and PAC Mainstreaming and Influence Mainstreaming and P.A. 198 Mainstreaming and Helpful Mainstreaming and PAC Influence and P.A. 198 Influence and Helpful Influence and PAC P.A. 198 and Helpful P.A. 198 and PAC Helpful and PAC Do Not Write in These Spaces (01-08) (09-17) 0811 : : : : ‘_’ I (18-26) 0812 (27-35) 0813 (36-44) 0814 (45-53) 0910 (54-62) 0911 (63-71) 0912 (72-80) 0913 (DI-08) (18-26) (3644) (45-53) (54-62) (63-71) (72-80) (09-17) 0914 (27-35) 1012 1011 1013 1014 1112 1113 1114 (01-08) (09-17) (18-26) (27-35) JI77 Page 5 IF A SPECIAL EDUCATION CLASSROOM AND A GENERAL EDUCATION CLASSROOM ARE 100 UNITS APART, HON FAR ARE: Labeling and my Job Labeling and Child-Centered Labeling and STANSE Labeling and Management System Labeling and Efficient Labeling and SESA Labeling and Planning Labeling and Frustrating Labeling and Change Labeling and Mainstreaming Labeling and Influence Labeling and P.A. 198 Labeling and Helpful Labeling and PAC Special Education Classroom and General Education Classroom Thank you. Do Not Write in These Spaces (01-08) (09-17) (18-26) (27-35) (36-44) (45-53) (54-62) (63-71) (72-80) (01-08) (09-17) (18-26) (27-35) (36-44) (45-53) (54-62) (63-71) APPENDIX B INSTRUCTIONS AND NETWORK ANALYSIS QUESTIONNAIRE 178 STANSE Project February 9, 1976 figs Two If you have any questions, either about the instrument or STANSE Project, please feel free to call Evan Peelle, Patricia DuFort or Nancy Kaye collect at (313) 763-3411. Sincerely, Leonard C. Burrello Project Director, STANSE Enclosure 179 E-----—--1 Common cat n your Name 1 o To is Discuss Change: Discuss Planning: Discuss My Job: New Ideas. Programs. Day-to-day Programs. Services. responsibilities Procedures Implementation PAS. FOUR NON HS 84 STATE DEPARTI-ENT ”814d" Batten, Hurra Barter Jan Back, Theodore ”Ted" seltron, 1 dia Blair, Har‘ Braccio, John Bryant, l Chapocl, Frederic "Fred" Collins, Michael ”Mike" Barbara Djorling, Donaldson, Bert Ensign, Aroelia Sates, Robert Hamlin, Leonora "Lee" Eodson, Diane Howard, Thomas ”Tam" Mikrut, Marsha Honk, George fit ‘ R' , James on, 22 Barrel DuFort Feldis t Do ”an in lin Lanson J Nutsnr Ronald "Ron" Paella, Evan 1180 pms________] a _—i Communication ic Discuss Planning: Discuss My Job: Programs, Day-to-day Services, responsibilities Implementation FOUR NON HS Discuss C hange: New Ideas, Programs. Procedures P A.S. Harru Richard Larru Butler Carnanw Cook Barn Lelbridve Fuller David Davis Halco Gear Robert Haarer Dave Herbert Benson Hobbs, Dorochu Jones, ann Konrad, Doris Lilly, Albert UoJlone, Fred Michaela, Gary Mikel, Richard [9‘4 Z len, Leo Price, Virginia Ouitiquit, Gary Raynior, Sherry Romsek, Helen Scandary, Emma Jane Shiffardecker, Duer Spicknall, Harrold Stevens, David , Charles Pat Vorce, Bruce via, , Jean I 7: , Kenneth , George 181. FITOB ______ _ _ I 3. Communication ic Discuss Planning: Discuss My Job: Programs. Day-to-day Services. responsibilities Implementation FOUR NON HS Discuss Change: New Ideas, Programs. Procedures PAS. 41 KENT Allen, Velma Bantam, Barbara Birch Ed Boulter . Collins, Norman "Norm" 115 Cramp, Robert "Bob" Dckuyaen, Case Ekster, Barbara Elders, Darrel For, Patsy Gogoleski, Shirley Jacob, George Ke y, Steve Kt r, Kelly t1 ‘Lm Noortnoek Joe Gre Oudman Mar orie Pattison Osmun Saur Barbara S: VanderVecn Jo Anna J. Richard Veenendall Marie Waltz Jack warren, Fount t John 182 Communication. ic Discuss C hange: NEH I 6285 0 Programs. Procedures PAS Discuss Planning: Discuss My Job: Programs, Day-to-day Services, responsibilities Implementation . FOUR NON HS Contest. he Hechlii Hessler Foul Dale Johnson Kitchen LaPlante Francis "Fran" Mason Paul Richard I Elizabeth Susan MC t te .‘lohan Ma Moore, Morreale, Paul Mosher, Carolyn Nouland, Fred Pellegrino, Vito "Bill" Powers, Thomas Rittgers, Philip Sauer, Thomas "Tom" Sheehy, Joyce Siebert, Harold "Hal" Smith, James "Jim" Smith, Shirley Seemik, Frances Troff, Fredric wheeler, Jim white, Albert "Al" Hilaon, Robert "Bob” ' 183 EEiPB .._._..._._....-j 6. Communication is Discuss Change: Discuss Planning: Discuss My Job: New Ideas, Programs. Day-to-day Programs, Services, responsibilities Procedures Implementation PAS FOUR MON HS Freeman, Gerald "Jerry " Garveliuk, r Haddad, Joseph Hogarty, Robert "Bob" Hallock, George Harvey, Robert Heard, Dorothy Henderson, Paul Henzie, Xenwood Jahnke, Warren Jenkins, Hobart Kassner, Fred Kirchhoff, Lucille Rotting, Charles "Chuck" McGhe J Miller hfilliam Mol John Hon Ne Kin Jos Parker Fred Jack Pillar Be Donald "Don" Rouell, James Shrusbree Al Smith S Sundbe lor Janice Pearson Place las Hart Geo Hilliam I" 184 E§e________] . g g a tion ic Discus s C hange: New Ideas. Programs. Procedures PAS Discuss Planning: Discuss My Job: Programs. Day-to-day Services. responsibilities Implementation FOURHON HS Thams g L Paul I t Villiam 81 LE h Fri WASHTEHAW Ba Anderson Ansted Al Robert "Bob" Hank Erdlitc, Kathleen "Xa Fitch, Gae Barnes DeYounc French, Sophie Grant, Dorothy Gregerson, Harvey Helber, Paul Keene, Jane Lauhon, Carol Morey, Richard "Dick" Melikan, Christopher Parkis, Michael Rezmierski, Virginia Schroeder, Anna Shea, Iona Stephens, Eleanor Sturm, Barbara Turner, Hazel Va len, Jamie Vt , Susan An Barnard l85 Froe________] a. Communication ic Discuss Chat c: Discuss Planning: Discuss My Job: New Ideas. Programs. Day-to-day Programs. Services. _ responsibilities Implementation Procedures PAS' FOUR NON HS Barton, Helen Beall, Charles Billin s, Doroth Brown, Andrea Bryden, William Carnahan, Robert Carr, Irene Christinidis, Fred Grain, Bob D'Alessandro, Gena ' Darnell, Gwendolyn "Gwen" Decker, Donald "Don" Ldetiker, Robert "Bob" Draper, Ingrid Durbin, Mary Lou G Gusto son Guano Al Shir H dusiewicz Ruth Marsha Haves Hilton Holt Ja e Robert Johnson Kenneth "Ken" 186 1-05 -—--———-1 Communication ic Discuss Change: Discuss Planning: Discuss My Job: New Ideas. Programs. Day-to-day Programs. Services. responsibilities Procedures Implementation PAS FOUR MON HS Kaselemis, Gus Hell Clark Kokovich Anthon Lockwood, Elizabeth "L' Loudenslaoer, William Luedtke, Leonard Luke, Walter MacG re gcr, Donald Mandell, 7%eodore "Ted" Martin, Lee fihthey, John McCarthy, Ann McGuire, Donald Miller, Laurence Montambeau, Roy Heumes, Margaret Page , EdJin "Ed" Rcfford, Ray Euehle, George , Russell Sevrey, Denna I , Nancy S iewicz, He ivan, Tarbut Tnabman Vedder Barbara Haters Vatson Matters Heiderhorn Hubert 187 FF”..-___--.J _ g 10. Commun cation ic Discuss Planning: Discuss My Job: Programs. Day-to-day Services. responsibilities Implementation FOUR NON HS Discuss C hange: New Ideas. Programs. Procedures PASI Weiner, Milton Uri t, Charles Zimmer, Yvonne 03 MEGAN Fi trick, Hike Gutshall, Robert ”Bob" wildfong’ Lisa ALPENA- PDNTMO RENCY -ALCONA Baker, Herbert 05 BA RRY Blackmore, william Lowe, Fred 09 BAY ARENAC Claes, Ruth DuBois, Mary Luce, Robert "Bob" Vilhe Vil He 12 BRANCH Block Rel Brnan Eloise Carlton 188 l-OB _ _ ______ J 12. Communication ic Discuss Change: Discuss Planning: Discuss My Job: New Ideas, Programs. Dey-to-day Programs, Services. responsibilities Procedures Implementation PAS FOUR MON HS LePres, Tom Mont , Robert Hobinson, William Silver, Robert Barbara Sloan, Smathers, Philip Tower, Paul 1 3 CALHOUN Agne, Ma: Anderson, Elizabeth cempbell, Jerry Clark, Keith Lattmer, Dale Duff, Phillip Eilers, Henry Garrett, Hilliam 14 C Homer Carl Linda 15 CHARLEVOIX-EMNET Caldoell Thomas "Torn" T lor Patricia ”Pat" Jean Shores Trafe let, Arnold 189 E?s_____-__] u. ication ic Discuss C hange: New Ideas. Programs. Procedures PAS Discuss Planning: Discuss My Job: Programs. Dey-to-day Services. responsibilities Implementation FOURUON HS 476 Gerald Waite 17 EASTERV U?PER PENINSULA (curepswa, nncxxnaw) LUCE. Christensen, Dorthea Je I Gallo er Goldtho. t "Tim M J Greene Time Leach, Jay Fe , Garry 18 CLARE-GLADWIN Raymond, Jack 19 CLINTON Belzer, Eonald "Ron" Ellsworth, Roy Malitz, Howard Gerald Powell, Don Ives tar, Schwartzhopf, Larry All Dabs Hosie Judu Kowalk Duane 19C) Fi—oe __ __ ______ l a la. Communication ic Discuss Change: Discuss Planning: Discuss fly Job: ‘ New Ideas, Programs. Day-to-day Programs. Services. responsibilities Procedures Implementation PASf FOUR MON HS Peabodu Hildred C Don S Niebe Phi U Beauvoir Cartwriaht Mariorie t Uiertz Hill Hil Prins Raske ’n. 0“ James Jan David Ri Donald Russell Donald Sabourin, Robert Smith, Vi Sventko, Joe Tow, Vernon white, Phil Williams, David 27 GOGEBIC-ON‘IONAGON Horpela, Heine "Bill" 28 TRAVERSE BAY AREA (GRAND TRAVERSE. BENZIE. WSKA, LEELANAU. ANTRIM) 015 Asiala, Richard "Dick" l2l Draper, Fran 202 264 428 Hansen, Rose' Laird, Ralph Shikoski, Tom 191. FW9_-_----_ . .: 2L ic Communication Discuss Planning: Discuss Hy Job: Programs. Day-to-day Services. responsibilities Implementation FOUR HON HS Discuss Change: New Ideas.. Programs. Procedures PAS. 30 H 042 Blair J (HOUGHTON, BARAGA. KEWEENAW) X lo Yennant Albert Gene 34 IONIA Haverhate Jean 35 IOSCO Rod Eid Foster Charon Gieser Dennis Gillette, Gene Rollin, Herman 38 JACKSON Barre tt, Dave Beech, Rey Cbntat, Michael Kekke, Robert Neiswander, Lucylee Rutz, Shirley Schesky, Casimir "cash" Stewart, Dorothy Van Schoick, Betty Jo Young, Christopher ”Kit" 39 KALAMAZOO VRLLEY Ball, Sandra . Galdwell, Elliott 192 Pl-OB .. ------- ] 15. Communication Tb ic Discuss Change: Discuss Planning: Discuss My Job: New Ideas, Programs. Day-to-day Programs, Services. responsibilities Procedures Implementation PAS FOUR HON HS Coin, Larry innlee Decent, Engle, John Guarino, Robert "Bob" Henderson, Donna McCann, Richard Miller, C. Villiam Roellchew, Betsy 392 Ross , Marian Stevens, Grace 4 6 LENAWEE ids Geo 4 7 LIVINGSTON Farabee David Fisher Dave 7‘ J ov ce 5 l MAN ISTEE Dittnon Thomas Fitch Robert 52 MA Ahola Allan Bond Bill Hammer Art "Tom" -ALGER Jones, William Bob Lindberg, Kenneth I Horlen Bet 193 PTOB___--__-] 27. Communicat ic Discuss Change: New Ideas. Programs. . Procedures PAS. Discuss Planning: Discuss My Job: Programs. Day-to-day Services. responsibilities Implementation FOUR NON HS 329 Mulder Harru 72 C002 (Crawford. Oscoda, Hthanc Richard 73 SAGINAH Acker Jane Vi Anderson Bell, Bert Bleseh, Ge Bush, Loretta Dundas, Chris Pearn, Xayte Fitzpatrick,. rite Harshman, Robert "Bob" Hoffman, Gerald Hueffner, Margaret Mulka, Stanley Faulgher, Lowell Pe ters , Larry Pistons, Kathleen 194 P1~oe________] ' za. Communication ic Di scus s C hangs: Discuss Planning Discuss My Job: New Ideas. Programs. Day-to-day Programs. Services. responsibilities Procedures Implementation PAS FOUR HON HS Tomlinson Joann 7s T Barnard Ruth Camcbell Larrv' Ri James hovda fire 0 76 SANILRC Hollis Joe McNeilZ Titus, Harold Ken Vilson, Larry 76 SHIAWASSEE Paul 1, Gary Bin En Green, James ‘I Jeffries, r Richard, Chester 79 TUSCOLA Benscoter, carolyn 195 P 1'08 -------- 1 l6. Communication ic Discuss Plannin Programs. Services. Implementation FOUR HON Discus s C hange: New Ideas. Programs. Procedures PAS. Discuss Hy Job: Day-to-day responsibilities HS Richer Thomas Schaefer June John Rose Ha Sormunen Zenti S3 MASON Barton Peter 54 HECOSTA-OSCEOLA Kitchell Elizabeth 55 MENOMINEE X the Jon Lun rhausen Ann Peltier, Ronald "Ron" Seidl, Louis 56 MIDLAND Frazee, Michael Marvin, Lynn Mayer, Charles 58 IDNROE Gampbell, Mary Donahue, James Heath, Bob Jansen, Cornelius "Corey" 64 .Robinson, Ivan 7O OTTAWA APPENDIX C SAMPLE MESSAGE 196 ‘4; 1’ 4 g, -_-. STANSE PROJECT “£900. «h‘ 130 5 th 1‘“ t St t 2:. }Pnr“f,~‘b\ Ann A2301“. Hizhigahe648109 id \ l/l \\‘ \ May 10, 1976 Dear Colleague: Enclosed is a two page report which gives a description of the work of STANSE, and reports some key points from the questionnaire send to you recently. We are also enclosing a list of members so you can identify persons you can contact if you wish to have input. Sincerely. STANSE Project 197 STATE AND FEDERAL PLANS MEAN DOLLARS The State and Federal Plans have become even more important since the passage of the Education for All Handicapped Children's Act (P. L. Sh-IQZ). The development of a comprehensive plan is being tackled by another STANSE committee. This plan will be useful to you because it will supply guidance about planning requirements, field input, responsibilities, programs and service delivery. A procedure to obtain field input into the develOpment of this plan is being completed. The design of a systematic way to get your input into the plan will help assure that your issues and needs are represented in the plans. PRACTICAL PLANNING SCHEDULES The fifth committee is developing a practical planning schedule indicating yearly planning activities as they relate to state and federal requirements. This report will include federal, state, intermediate and local schedules of tasks, timetables, data needs, etc. This will help you to have the necessary information on hand so that you can fulfill your planning responsibilities. REPORT ON QUESTIONNAIRES In January, you received a package of three instruments (now known as the "green monsters "). Over 70% of you completed them. An initial review of the data shows that both the reliability levels and percentage return rate mean that the information provides a useful guide to effective planning. The data show that the most important concern of Special Education administrators is to serve the needs of handicapped children. He also found that members of Project STANSE reflect the range and diversity of views that are found around the state. STANSE members represent the field with a high degree of reliability. We were also able to substantiate that there are distinct groups and key people who communicate with each other about the topics of planning, change and our job. Information does flow to and from STANSE members as the project proceeds. Our task now is to use our network more efficiently and to improve our communication with each other. A more.complete report will be,forthcoming as soon as the analysis is completed. STANSE will continue to contact you in various ways to get your input. 198 Project STANSE is an organization which helps Special Education Directors and Supervisors. STANSE is made up of 36 Directors and Supervisors, drawn from intermediate and local levels around the state, from the State Department, with a support staff of nine from the University of Michigan. One goal of STANSE is to become more child-centered in its planning activities. Another goal is to develop recommendations which help children and thereby help you. Using your input, we have identified five target areas of concern to you. Each target area is being dealt with by a STANSE committee. These target areas reflect problems that are giving direction to STANSE's planning. Decisions will be made which reflect your views concerning services offered to handicapped children. . WHAT'S SPECIAL EDUCATION TO DO? One committee is defining the children to be served by Special Education, and the programs and services the children need. The State Board of Education a few months ago urged that responsibilities for delivering services be identi- fied more clearly. So, recommendations will be made to increase understanding of the responsibilities of Special Education and of other education departments and agencies. By providing this input, STANSE will help administrators be more effective in their planning for children. STATE SPECIAL EDUCATION SERVICE AREA PLANS TO REORGANIZE Hurray Batten and his staff are seeking ideas for reorganizing so that they can be more effective. A second STANSE committee is giving reactions and suggestions to the State Department's (Special Education) reorganization, based upon what members of the field report they need from the state. in addition, this committee is looking at the implications that the state changes have for intermediate and local levels. We are advocating that more child- centered responses be supplied by each level. HEARING EACH OTHER AND RESPONDING The need to make meaningful input into the State Department's planning and decision making is being addressed by another committee. This group is developing a way for the field to let the State Department know what‘s going on in their districts, as well as to get accurate information in return. The procedures developed for input will affect your interactions with the state peeple and the quality and quantity of information exchanged. STANSE is helping you by suggesting ways the state can more effectively respond. APPENDIX D WHOLE SAMPLE PAIR-WISE MMDS SAMPLE SIZES 199 men 0 mmw mwm Now mam NNN mi mam com omw com col mom com mom can opm 54 _ meme HmwNwm mmcocmma was: mmwzucwma Rum omm mum NAN Nwm owm to I mom now mom haw com mam mum EH mam mmm mmm wwm mom Non cum mom mx-eoz F meFF ”mmNFm mFQEmm mszicFma Faca_m: OmF