F‘; i' 4 .‘ ' I.»- esra'n “6.92 Al ' ~Y .- é“ . . ' . .- I' E: .l ' 'tv I .l. .‘S l Jaw: AW“ 21:... ~. LU o 1. wk (0 I" v. ‘0" ' . l tyn'k‘. . " “2.: tr E35?" F. . 4 1.! .. , 1 * ('38 't ' «g. - “$3,. :5“ ‘1. 1 . . . r ‘R ‘ v ,.. .. | J .-,-,-~\4..v.. r 1...”: 'I-vu "Vf‘A'vJ\‘”,',DV v ' xv‘ .. ‘.'.‘ . w . 7.4.3.44; "v 4! my I ,. - ff‘u'h' . .H ‘. .—:. .., "v H <~ ’ ..- “.3 I. ’ '-"~I -.4 .',',. 1» .q - _ J- . Ink-"W'Hd‘, ns't ‘ m ’IgI:--rn"_' \ ‘3" ' -I¢ a | a 75:. ,3. .4. ’. I 55:1 ‘ 3.; @5514. g ‘ ‘ 'u‘ H. ' h" J‘ l‘K‘? 3" a :‘b "35%.? 4 5?:‘Q‘1‘oh '.' Sikl .‘J I. . '4‘. ifififigm «'lxr-fu ’3 ' ' .' "V q‘fi‘i‘v" '>.r'.‘)::€,‘.'.3fl ‘ “3&9. "s ‘9'v‘v ”‘Qfi‘a “if I wm’axc ~' 4. 453‘.” , .4. u'.‘ ‘ ”V 'o \ w it. v '- ~1 ¢~ . 1""lv'uv‘?!v v~~‘ ‘0 u - 1 Hi.” 4 ' I ‘g '..‘ .o..- - vuy. ’ 'vul P. ‘L I IV .v'5{r . . fr ' o ‘I I .. {- 01 b\ I I ' is ' I hung" I 7 ll:"\ . ' ‘3» ,' , u .- .r A a V I. ‘ 9"".fil“. 1w . "H~c.~ O‘v H. w c ~‘4::.;.\ L: :1. I‘ll fig I o {h‘ n ." “5" 7443' 1 . 2%?“ ' ' ~ ‘ “.3 -‘ J; '1' V'. ym’C'R' ' . :. .55 ‘-. fitwb $3 5.2.: - V.“ '3, ' I UT“? . ,A '- .r..."'~‘ o... . ‘v . r v \ 1 u v | ‘ . avail: 45:1 3,... J :' ' :3. 4 | v . [L‘Hu-f‘ ‘ |~<1 iii-{1:815 LIBRARY will llll/lI/l/II/l/l/ W m MW University This is to certify that the thesis entitled ORGANIZATION AND SERVICE: A CDRRELATIDNAL STUDY OF SOCIAL SERVICES FOR THE ELDERLY presented by John C. Jeppesen has been aceepted towards fulfillment of the requirements for M” - degree in _ {3 fizgéy Major professor Date ”An/73 I / 0-7 639 ORGANIZATION AND SERVICE: A CORRELATIONAL STUDY OF SOCIAL SERVICES FOR THE ELDERLY By John C. Jeppesen A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1978 ABSTRACT ORGANIZATION AND SERVICE: A CORRELATIONAL STUDY OF SOCIAL SERVICES FOR THE ELDERLY By John C. Jeppesen The purpose of this research was to investigate, through correlational methods, the variables attributed to characteristics of (l) a service coordination organization, (2) the interaction between the organization and key service providers, and (3) the elderly clients served by these organizations. More specifically, the research examined the characteristics of the staff at the Information and Referral centers (components of the service coordination organization), the interaction of these staff with Department of Social Services and Social Security Administration (key service providers) agencies, and how these combined character- istics, along with those of the elderly clients of this study, were related to the service received by, and the attitudes of these elderly clients. A final sample of 15 staff members of the service coor- dination agency and 74 elderly clients (which had been contacted by the service coordination agency either by phone, or in person) were included in the data collection. These staff provided John C. Jeppesen information on job satisfaction, negotiation latitude, job design, exchange with supervisors, and interaction with key service pro- viders. Information about the elderly clients included data on receipt of service, response level, elderly client demographics and elderly client attitudes. While the relationships between variables of each data set were examined separately, the basic research question was “what relationship do these organizational/ interorganizational and elderly client variables have to the fundamental outcomes, receipt of service, response level, and elderly client attitudes?" Pearson correlations and discriminant function analyses were used in this research. The results indicated that clients who tend to receive services appear to have fewer self-help capabilities, have less contact with their own children, know more about social services, and tend to be more satisfied with these services. 0f the service coordination staff, those who tended to be more satisfied with work, have more frequent contacts with coworkers, and have fewer optional learning experiences had clients with better general response to agency contact. From study of client attitudes it was found that clients seem to have better attitudes about key ser- vice providers when the service coordination staff member who contacted them had more learning about these agencies. Finally, greater intraoffice communication (feedback, participation, and information) was positively related to staff having greater interorganizational orientation. This research suggests that often-cited elements of the work environments in non-service John C. Jeppesen organizations are also quite relevant specifically to social service organizations, and that, along with key client character- istics, they are also important for service outcomes for clients. ACKNOWLEDGMENTS I wish to thank the members of my committee, Dr. William S. Davidson 11, Dr. Esther Fergus, and Dr. Eugene Jacobson, for their guidance throughout this research. Their comments and support were invaluable. Also I would like to thank Dr. Robert Calsyn, who was very helpful in the planning phase of this study. My thanks also go to Bob Dolsen of the Region IV Area Agency on Aging, for his cooperation and appreciation for applied community research. Further, I extend my thanks to all the members of the Region IV information and referral centers. Their hard work, patience and good humor were an inspiration. Special thanks must go to Martin Kushler, who was also pursuing research in the Region IV area, for his continued companion- ship and his appreciation for teamwork. Finally, thanks to my dear friends, for knowing the differ- ence between support and advice. The encouragement they provided will long be remembered. ii TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES . Chapter I INTRODUCTION AND AN OVERVIEW OF THE CHARACTERISTICS AND NEEDS OF THE ELDERLY AND THE ORGANIZATIONS WHICH SERVE THEM . . . . . . . . Characteristics and Needs of the Elderly State Level Findings . . . . Substate Findings . . Organizational Structures, Objectives, and Programs for the Elderly . Organizational Structures . Organizational Objectives . . Key Programs for the Elderly. The Role of the Area Agency on Aging II AN HISTORICAL EXAMINATION OF THE LITERATURE ON ORGANIZATIONAL AND INTERORGANIZATIONAL THEORY AND RESEARCH, RESEARCH RATIONALE. . . Organization of Background Literature Organizational Theory and Research. Leadership and participation . Exchange between leader and subordinates Job design and job satisfaction Need for broadened scope in organizational research . . . . Summary and research direction Interorganizational Theory and Research Environmental characteristics . Exchange between organizations Summary and research direction Research Rationale . . . iii Page vi . viii —l OKOGJCDOS b-bN d Chapter III METHODS . . Setting . . Logistics . Participants . . Elderly Clients Outreach Workers . I and R Directors Design . Procedures . Overview of Measures and Their Administration Administration of organizational and inter- organization measures . . Administration of elderly client measures The Measures and Their Development . Organizational and interorganizational data Elderly client data . . . . . IV RESULTS Framework for the Exploration of Relationships Within and Between Data Sets . . . . Relationships Among Organizational/ Interorganizational Variables . General correlational findings . Relationships Among the Elderly Client Variables Relationships Between Organizational/ Interorganizational and Elderly Client Variables Prediction of outreach response Prediction of response level Prediction of senior citizen' 5 Opinions about the AAA. Prediction of senior citizen's opinions about DSS and SSA . . . V DISCUSSION . . Organization of the Discussion . Correlational Findings . . General organizational/interorganizationalo correlations . Specific organizational/interorganizational correlations . . . . . . Elderly client correlations Discriminant Function Analyses on "Behavioral Outcome” Measures . Outreach response . iv 78 80 9O 93 97 99 104 107 107 107 107 109 111 112 113 Chapter Discriminant Function Analyses on "Client Attitude" Measure Conclusion . REFERENCES . APPENDICES . A. Memorandum of Agreement . B. Tab1e Bl . Page _..|_a ._a._a (Tl-b 121 130 132 134 LIST OF TABLES Table l Measures 2 Final Data Reduction Solutions for Each Measure of the Organizational/Interorganizational Data 3 Final Data Reduction Solutions for Each Measure of the Elderly Client Data . . . . . . . . . 4 Significant Pearson Correlations for Three Key Organizational/Interorganizational Variables 5 Significant Pearson Correlations Between Director's Interorganizational and Job Design Variables 6 Significant Pearson Correlations Between Outreach Worker's Interorganizational and Job Design Variables . . 7 Significant Pearson Correlations Between Director's Interorganizational Variables and Their Outreach Worker's Job Design Variables . 8 Significant Pearson Correlations for Three Key Elderly Client Variables 9 Summary of the Wilks Lambda and Roas V Scores for the Variables in the Discriminant Function for Outreach Response l0 Capability of the Discriminant Function to Predict Outreach Response . ll Summary of the Wilks Lambda and Roas V Scores for the Variables in the Discriminant Function for Response Level . . . . l2 Capability of the Discriminant Function to Predict Client Response Level vi Page 56 73 77 82 85 87 88 92 96 97 98 . 100 Table Page l3 Summary of the Wilks Lambda and Roas V Scores for the Variables in the Discriminant Function for Senior Citizen's Opinions about the AAA . . . . . . lOl l4 Capability of the Discriminant Function to Predict Client Opinion of the AAA . . . . . . . . . . . lO3 l5 Summary of the Wilks Lambda and Roas V Scores for the Variables in the Discriminant Function for Senior Citizen's Opinions about DSS and SSA . . . . . . . l05 l6 Capability fo the Discriminant Function to Predict Client Opinion of DSS and SSA Agencies . . . . . . l06 vii LIST OF FIGURES Figure Page 1 Relative Importance of Specific Problems for the Elderly . . . . . . . . . . . . . . . . 5 2 Information and Referral System . . . . . . . . 48 viii CHAPTER I INTRODUCTION AND AN OVERVIEW OF THE CHARACTERISTICS AND NEEDS OF THE ELDERLY AND THE ORGANIZATIONS WHICH SERVE THEM In the past ten years public attention has increasingly turned toward the growing proportion of the elderly in our country and the social service organizations required by their particular needs. While some investigators have detailed the needs of the elderly and others have done research which begins to relate to the organizational settings in which service delivered, very little research has coordinated the study of both elderly service recip- ients agg_the organizations providing the service. This paper provides a study of'11eedy elderly in a three county area of Michigan and the community-based organization which served to coordinate social services appropriate to their needs. Chapter One of this paper gives background information on (a) characteristics and needs of elderly clients and (b) the actual organizational structures, objectives, and programs which addressed key needs of the elderly. Chapter Two continues with a summary of literature on organizational systems: This includes an examination of the research on (a) organizations and (b) interaction between organizations. This research suggests characteristics which appear to be related to performance of people working in the organizations. Performance of staff members in the community-based service 1 organization was important since it seemed to have direct effects on the service received by elderly clients. In Chapter Three of this paper the research (a) setting, (b) logistics, (c) partici- pants, (d) design, and (3) procedures are discussed. Chapter Four contains the results of the study and Chapter Five follows with a discussion of the results. Characteristics and Needs of the Elderly The subpopulation of the elderly "consists of a hetero- genous composite of individuals with differing physical character- istics, economic and social situations, and needs or desires (Office on Services to the Aging, 1974, p. xxii)." Often, however, the diversity of the elderly is forgotten and these persons are simply identified as those 60 years of age or older. Further, places of employment often perpetuate arbitrary definition of the elderly through the custom of requiring mandatory retirement at the age of 65. While these conventions of definition help identify the elderly, the social policies which use these definitions have been cited as needlessly arbitrary (Schultz (citing Slavick), 1966; Withers, 1974). Because of an increasing life expectancy of our general population it has become ironic that these fixed age-determined definitions have identified a rapidly changing subpopulation. In 1900 life expectancy at birth was 45.6 years for men, 48.4 for women; in 1920 it was 54 for men 55.9 for women; and in 1970 it was 67.1 for men and 74.8 for women (OSA, 1974). Thus, as life expectancy has increased the elderly have come to represent a growing pr0portion of the total population. For Michigan the increase in and projections for the percentage of the total population the elderly represent are as follows: 8% in 1910, 8% in 1930, 11% in 1950, 12% in 1970, and 15% in 1990 (OSA, 1974, p. xvii). Therefore, a growing interest in the needs of the elderly seems warranted--if only due to the growth of this constituency. As is true of other subpopulations within society, elderly persons also have social service needs. When the elderly person approaches sources of the service they need, however, they often are hindered by several problems with the service delivery systems. Traditionally, the systems of service organizations for the elderly have been comprised of individual organizations, each specializing in a limited range of services. For convenience in administration of service programs the criteria by which a person qualifies for a service has been standardized. While these standard criteria assist the service agencies, they pose a problem for the elderly person who has several interrelated needs: each need, though related to other needs, must often be approached through separate service sources. In each of these contacts the person may be associated with the problems with which that agency can deal while the person as a whole is largely ignored. This selectivity of problem identifica- tion makes it easy to ignore related or coincident problems, and most agencies are not equipped to deal with them. Thus, whatever the particular need of the elderly person there seems to be a general need for service coordination. The service coordination organization studied in this paper was selected as the focus for research since it was designed to meet this general problem. Findings on the specific needs of the elderly have been compiled for the State of Michigan and for the substate regions selected for the study presented in this paper. These findings allowed the ultimate selection of elderly persons who participated in the study. State Level Findings In 1974 the State of Michigan Office on Services to the Aging (OSA) developed a Comprehensive Plan on Aging_based on a systematic survey of 3,000 noninstitutionalized elderly. Figure 1 below shows the responses of the peOple when asked to identify one or more of eleven particular problems as those which were most salient to older Americans. These findings suggested that greatest attention be given to problems most frequently selected as most important. As is clear in Figure 1, Income and Health were among the most frequently mentioned problem areas. Therefore, the re- search described in this paper was directed to include elderly per- sons enrolled in specific income and health assistance programs. Sgbstate Findings Basically, research on the needs of the elderly at the local level had reiterated the findings of the state survey above. Studies of the needs of the elderly had been done by a variety of groups for the three counties (Berrien, Cass and Van Buren) which were selected for research discussed in this paper. Financial FIGURE 1 RELATIVE IMPORTANCE OF SPECIFIC PROBLEMS FOR THE ELDERLY Income 1 138% Crime 31% Health Transportation :43 35 Nutrition/Food 15% «35? 5% Consumer Protection 11-12% S< E395 Spare Time Activities I 11-122 .J ‘3 Housing ‘_]11-12% Employment Opportunities ‘ )8% Age Discrimination : 18% Getting More Education 1 [5% 4 l I J 0 10 20 30 40 PERCENTAGE OF ELDERLY RESPONDENTS IDENTIFYING PARTICULAR PROBLEMS (Adapted from OSA, 1974, xix) assistance to elderly citizens, whether needed for basic living expenses or for medical expenses, was consistently reported as an essential need in each of these counties. Berrien county reported only 23 percent of the polled elderly said their resources for the future were adequate. Only 22 percent of the elderly Cass county residents had monthly income over $300. And in Van Buren county, not only did the elderly often need basic income or health assis- tance,but they frequently could not afford transportation to agencies where this assistance could be found. All findings were reported in the Annual Plan of the Region Four Area Agency on Aging (1974). Assessment of the needs of the elderly has therefore re- vealed that income and health assistance are among the most fre- quently cited problems. The following section discusses the organizational systems which exist to address these needs. Organizational Structures, Objective, and Programs for the Elderly Increasingly, National and State Government have recognized that a basic problem in services to the elderly has been fragmented, specialized services provided by a variety of agencies. This tends to discourage the elderly from'seeking all the services they need. This problem is obviously aggravated for the elderly since "advancing age is frequently associated with decreased financial status, increased isolation, poor health and other problems (OSA, 1974, p. 1)." Accordingly, governmental planning perspective has changed from an emphasis on independent services specialization to more comprehensive service system planning. This perspective has recognized the contributions the elderly can make to their communities and it has encouraged more sensitive views of their practical problems. For example, if the problem was inade- quate nutrition, related needs for programs in income assistance, education, and transportation became important. The organizational structures which have provided a framework for development of services to the elderly is discussed below. Organizational Structures In 1965 the Older Americans Act was passed by Congress (PS 89-73). It established the Administration on Aging (AoA) within the United States Department of Health, Education,and Welfare. In the 1973 Amendments to this Act the AoA was given independent status as a Federal agency and has since developed service programming under Titles 111 (community service) and IV (nutrition). Title III affected Michigan by allowing designation of a State agency on aging. While Michigan had established statutory agencies and coun- cils for several years they mainly operated as intergovernmental advisory groups to State government. It was not until Public Act 106 of 1973 that the present Office on Services to the Aging (OSA) was created. The OSA developed its role, according to the Amend- ments of the Older Americans Act of 1969, as the designated agency mandated to do comprehensive statewide "planning, coordination, and evaluation ofservices to older citizens (OSA, 1974, p. 2)." In 1973 further amendments required development of sub- state agencies to actuate the planning, coordination and evaluation functions at the local level. These substate agencies, called Area Agencies on Aging (AAAs) were developed in response to Title III community services requirements. In Michigan the OSA established thirteen AAAs which had responsibility for as many State Compre- hensive Planning Regions. (There was one exception to regional boundaries in the Upper Penninsula.) This AAA network had been in effect a little over a year and had not yet reached full potential when the current research took place. In addition to the administrative similarity of the AAAs to the OSA one AAA (in Area Four) also began the complementary function of actually contacting the elderly in "outreach" efforts. As suggested previously, this substate agency was selected for the investigation detailed in this paper. Structural characteristics of the Area Four AAA are detailed in the methods section. Organizational Objectives All AAAs in Michigan were mandated to do comprehensive planning, service coordination,and provide advocacy for the elderly in their respective regions. While they were directed to coordinate existing resources and insure their access to the elderly,these functions were generally limited to " . . . providing access to information and referral services . . . (OSA, 1974, p. 10, italics added)." In the Area Four AAA the basic goals were consistent with the above with the exception that their goal was "to provide infor- mation and referral/outreach to senior citizens and to disseminate information on services available and to ensure that all elderly in the area received as much financial assistance as required and/or available (Region Four Area Agency on Aging Annual Plan, 1974, italics added." Other goals were to establish emotional security in the elderly, assure access to transportation, augment health ser- vices, repair and maintain homes, provide meals, procure legal services, develop employment opportunities, and establish means for education and participation. While these objectives suggested several direct services roles, the only direct service mandated was the information and referral/outreach aspect based on a special contract with the Area Four office of the Michigan Department of Social Services (DSS). Local DSS offices were eligible for Title XX money for this function and had, in this one Region,elected to contract it out to the AAA. Other AAA objectives were fulfilled through coordination and planning with local services agencies. Key Programs for the Elderly, The organizational structures and objectives of the Office on Services to the Aging and its Area Agencies on Aging were designated to accommodate and initiate action for the betterment of the senior citizen. Among the most important programs directed at needy senior citizens were the Income and Health related pro- grams provided by agencies other than local AAAs. In helping quali- fied elderly persons enrolled in these programs,the area Four AAA provided general information and referral. Since the people selected for the current research were those enrolled in income and health related programs, the characteristics of these programs have importance for this research. Both key programs important to this research came from programs resulting from the 1965 Social Security Amendments of Congress: Old Age Assistance (0AA), Aid to the Blind (AB), Aid to the Disabled (AD), and Medical Assistance (MA). In Michigan the Department of Social Services (DSS) provided these programs locally and contracted the supporting services not provided by their own case workers. It was generally found that although people could 10 get these services by applying through DSS, many did not seem to take advantage of this assistance due to the "welfare stigma" and the requirements to sign over limited possessions in order to receive public assistance. Others pointed to the lack of work incentive in these arrangements. These shortcomings brought about current programmatic changes. The 1973 Social Security Amendments brought about the combining of 0AA, AB, and AD into a single pro- gram, Supplemental Security Income (SSI). Instead of DSS, the Social Security Office became the administering agency. The remaining Medical Assistance (MA) program continued to be adminis- tered through the Michigan Department of Social Services. The Role of the Area Agency on Agigg, The transfer of three basic assistance programs from DSS to SSA, though theoretically beneficial, introduced immediate difficulties for potential recipients. Since many of those affected were those currently the concern of the OSA and the AAAs, these organizations were the logical facilitators in this transi- tion. Continued coordination activities were maintained for the MA recipients. During this period changes in service procedure meant an additional effort by both DSS and SSA employees. Almost by default, it became the AAA's job to assist the qualifying elderly. In doing this, AAAs negotiated agreements to share data and techni- cal information with county DSS offices and local SSA units. 11 The importance of the coordinating function of the AAAs was reemphasized when reviewing local needs assessments. The reader will recall that, in Michigan, the greatest proportion of the elderly (38%) said income was a problem for aging Americans. Also, a large proportion (24%) cited the importance of health problems. Further, the counties targeted for this research (Berrien, Cass and Van Buren) reported that, together, 36,647 residents were 60 years old or more and, of those, 31% were below the poverty level as established by CEO guidelines (Region Four Area Agency on Aging, 1975b, p. 3). Both $51 and MA programs provided needed financial assistance. In addition to the regular AAA functions, the Area Four AAA was responsible for Information and Referral to the elderly in its region. In order that it carry out the general goal of aiding the elderly the AAA had to both attend to the management of its personnel, and also interact effectively with the other service agencies in the community. These two issues were the focus of this research. The author believed that the way internal and external operations were performed might have influence on service received by the elderly. For this reason the following section presents a review of the organization literature which provides the foundation for the research of the service-coordination agency, the AAA. CHAPTER II AN HISTORICAL EXAMINATION OF THE LITERATURE ON ORGANIZATIONAL AND INTERORGANIZATIONAL THEORY AND RESEARCH, RESEARCH RATIONALE Organization of Background Literature The theories and research on organizations and their inter- organizational environments discussed in this part of the paper provide the empirical basis for researching the Area Four AAA as a service-providing organization. From this base the correlational research of the relationships within and between characteristics of the elderly citizens and the AAA service providers has been developed, implemented, and analyzed. This review of literature is organized in two sections. First, for the purpose of looking at the AAA (as a focal agency where the elderly are concerned), impor- tant conceptions in organizational theory are reviewed. Second, since the AAA is also very much concerned with its involvement with major agencies which serve the elderly, (i.e. the Social Security Administration and the Michigan Department of Social Services), that which will be termed interorganizational theory is also dis- cussed. For each of these areas brief discussion of the theoretical foundations of current research is followed by a treatment of con- cepts important to this research. 12 13 Organizational Theory_and Research The basis for the actual research presented in this paper comes fromeiwell established literature in organizational theory and research. These writings have helped to identify the social pheno- menon of work organizations and have related study of both the organization as a whole, and as "a body of persons organized for some specific purposes (Webster, 1968, p. 1033)." Early writings in organizational literature observed organ- izations as established management systems. In the early l900s Max Weber (1958, translation) emphasized that the ideal organization, a bureaucracy, was the best means to ensure efficient pursuit of organizational goals. Notable dimensions of such an organization included division of labor, detailed hierarchy of authority, a sys- tem of personnel rules and rights, impersonal interpersonal rela- tions, a system of work procedures, and promotion based on merit (Hall, 1963, p. 33). Thus, early conceptions or organizations included definition of an idealized, rule defined,structure designed to be impartial, fair, and efficient. Clarification of the limitations of the organizational ideal has served to suggest other conceptions and studies of organizations. In practice, the ideal organization seemed to be suitable, but only under ideal conditions of a money economy, capitalistic system, Protestant ethic, and large size (Blau, 1956, p. 34). Specifically, when a worker could be expected to do the required work, by the rules, because he was paid, then the system seemed to work. However, as modern technologies changed at ever 14 increasing rates it became clear that the fixed ideal of a bureau- cratic organization fell short of the changing demands on organiza- tions. Among these, Bennis (1969) recognized inadequate attention to the personal growth and development of organizational members, the "informal" side of the organization, and emergent or unantici- pated problems. Also, there seemed to be need for new systems of control and authority, more adequate means of resolving intraorgan- izational conflict, and fewer constraints on communication patterns and the exchange of innovative ideas (pp. 436-437). In recognitioncfl'the diveristy of organizations and the people working in them, organizational theory and research has like- wise become highly varied. Even so, many of these researches have in common a central concern of how the organization should relate to the worker in order to direct work toward organizational goals (Bennis, 1969). In developing research of the AAA, organization literature specifically dealing with the relationship between organization and worker was ofinterest both in its own right, and also because these relationships might effect the way workers work toward the organization's goal, helping the elderly get needed services. Thus, the discussion below reviews the organization/ worker issues of leadership, participation, exchange, job design, and job satisfaction. Research demonstrating the need for broadened scope in community research follows discussion of each of these issues. 15 Leadership and participation. While the research on lead- ership and participation could easily fill several volumes, dis- cussion of a few well known studies demonstrates an historical basis for continued research in this area. Whilemost studies have been done in actual work organizations, one of the first investigations employed a nonorganizational setting using the laboratory paradigm. This study by Lewin, Lippet, and White (1939) helped to emphasize leadership and participation issues. In contrived laboratory situ- ations various types of adult leaders were assigned to groups of 10 year old children involved in hobby clubs. While the results were mixed, they did find that children who experienced each leadership situation, had pronounced preference (19 out of 20 responses) for the "democratic" (guided group participation deci- sions) leadership over the "authoritarian" (leader decisions) leadership. This study was a landmark in studying individual behavior in a group situation. Generalization to real-life, formal work organizations, however, would definitely be inappropriate. Never- theless, the findings showed that different styles of leadership and the different amounts of participation required of the group member in each leadership situation could possibly effect the out- comes for group members. Certainly, the value of doing research with formal organi- zations should be recognized. How workers deal with procedural changes is one important question which was studied in early studies. Through many visits to the Area Four AAA, it was apparent 16 to the author that staff members frequently had to adjust to many changes in procedure. Since 75% of the Operating budget came from the local government, fluctuations in fundings required changes in operating procedures. Also, because the AAA's only direct service was Information and Referral to other service organizations, any changes affecting the other service agencies affected the AAA. Coch and French (1948) studied overcoming resistance to change in a pajama making factory. As with much of American indus- try, methods of doing jobs often changed with the requirements for new products. The researchers observed that when a change in procedure was introduced, production, as measured by units per hour, reduced drastically, and it took several weeks for the production rate to return to the pre-change level. The researchers saw this as a motivational problem and prOposed an experimental design involving variations of democratic procedures in handling groups to be introduced to a change in procedure and then checking for differences in subsequent production rates. Over a 40 day period, reactions to the change showed that "rate of recovery is directly proportional to the amount of participation, and that turnover and aggression are inversely proportional to the amount of participation (p. 524)." It was generally concluded that the more workers were allowed to participate in negotiating the changes. the better was their recovery in production rates. This study had several methodological weaknesses. Basically, comparison group sizes, task assignment and raw material availa- bility were not held constant across groups. Thus, serious doubt 17 was cast on the reported conclusion. In an attempt to replicate results, French, Israel, and As (1960) did not find significant production differences when worker participation was varied in a Norwegian factory. Even so, these studies prompted more research on worker participation. Morse and Reimer (1956) examined the effects of the type of decision making in an organization on the satisfactions and productivity of workers. Subjects were rank-and-file women clerks in a non-union organization having four divisions with comparable functions. In a year-long experimental period two experimental program changes were introduced. For two divisions the role of upper level management in decision making was increased and main- tained while in the other two divisions the role of the rank-and- file clerical worker in decision making was increased and maintained. The authors found that in comparison of pre- and post-satisfaction questionnaires the rank-and-file group increased significantly, while the upper-level decision making group decreased significantly. Productivity, as indexed by clerical costs, was shown to have increased for both groups,but the upper-level decision making group did significantly better than the rank-and-file group. While it would appear that greater rank-and-file participation yields higher satisfaction and comparatively less production gain, reported staff reductions during the study tainted the results of productivity. The authors suggested a larger theoretical framework in future research on changes in production procedures. 18 Early research on the relations between management and the worker addressed questions about interactions in a somewhat piecemeal fashion. These studies addressed problems in deciding on preferred methods of leadershjp_and approaches to pgrticipation of workers in decision-making. Democratic or worker/participant arrangements seemed to increase worker satisfaction but effects on productivity were less clear. Theories discussed below expand on these issues with examination of leader, member, and group inter- actions or exchanges as key determinants in how well an organization functions. To recapitulate, examination of such determinants be- comes important to research of the AAA organization since they may effect clients. The concept of "exchange" between leader and sub- ordinates in an organization may be used to identify several focal research efforts. Exchapge between leader and subordinates. George Homans (1958) postulated the fundamental concept, that social behavior may be regarded as exchange between persons. Generally this view esta- blished that there is an "economy" of relationships, based on economic and behavioristic notions, that persons put forward effort (cost), in interaction, in order to receive (benefit) some- thing of value to them. If this interaction behavior is useful to the individual, it should persist. Alternately, a change in an, interaction behavior is likely to occur when the profit in current behavior is minimal. These principles give broad definition to the research reviewed below. 19 Fiedler (1967) developed a theory of leadership effective- ness offering a "contingency model" which attempted to define the relationship between leadership styles and the favorableness of the situation for the leader. These extensive studies were based on a measure of the leader's perception of his least preferred coworker (LPC). Leaders who had described least preferred coworkers in generally complimentary terms were said to be people-centered (a high LPC score) while those using generally negative terms in describing least preferred coworkers were said to be task-centered (a low LPC score). Favorability of situations for leaders involved leader/ member relations, amount of power in the leader's role, and amount of task structure. Fiedler's book, A Theopy of Leadership Effective- pg§§_(l967), summarized leadership effectivenss research in a multi- tude of contexts. The general ‘finding was that low LPC leaders tended to have groups which worked well in situations either highly favorable or highly unfavorable to the leader and high LPC leaders seemed to have groups which worked well in situations moderately favorable to the leader (Fiedler, 1967, p. 169). These findings were based on median rank-order correlations between leadership style and group performance using data from development. In a major test of these findings Graen, Alvares, and Orris (1970), compared a sample of these leadership effectiveness studies (antecedent) with several other such studies (evidential). From two statistical procedures assessing the comparability of results in these two sets of studies it was found that the evidential studies (Fiedler, 1965, 1966, 1967; Hunt, 1967; Graen, Orris & Alvares, 20 then in press; Mitchell, 1969; and Hill, 1969) cast doubt on the antecedent findings. Using a procedure offered by Winer (1962), the authors compared correlations between leader's LPC and group performance for the two sets of studies (antecedent and evidential). The analysis of variance summarizing the findings indicated that main effects (group atmosphere, task structure, and position power) were not significant, but some interactions were. Graen and his coresearchers concluded "something systematic is operating (p. 292)." Also, when only the evidential correlations were examined none of the effects were significant. Fiedler's conclusions were not supported. The second statistical procedure, a comparison of ante- cedent versus evidential correlations within cells using a t-test between means of each, revealed five significant differences out of seven tests. The authors applauded these studies of the contingency model of leadership effectiveness, but concluded the model lacked sufficient empirical support. The Ohio State Leadership Studies were initiated by Shartle (1950). Interest in various aspects of leadership (Hemphill, 1949) yielded several factor analytic studies (Halpin and Winer, 1957) which resulted in the identification of two key factors: "Consider- ation" and "Initiation of Structure." Consideration included leader behavior involving the demonstration of concern for workers in allowing them to take part in decision making and encouraging interpersonal communication. Initiation of Structure included behavior in which the leader defined the tasks at hand, who should do them and how they were to be executed. The Leadership Behavior 21 Description Questionnaire (Hemphill, 1950; Hemphill & Coons, 1957) and the Leadership Opinion Questionnaire (Hemphill, Seigel, & Westie, 1951; and Fleishman, 1957) were developed to measure these dimen- sions. A review by Abraham Korman (1966) summarized the findings of studies using the Leadership_Opinion_Questionnaire and the Leadershiprehavior Description Questionnaire. The former was a Likert-type attitude scale designed to measure how the supervisor thinks he should behave in his role as a leader; the latter measured the perceptions of the leader by those he supervised. Korman's review included the reported correlations between "Consideration" and "Initiating Structure" and the organizational criterion variables of interest in each study for each of the instruments above. First, for the Leadership Opinion Qpestionnaire studies by Bass (1956, 1958); Flieschman and Peters (1962); Oaklander and Flieshmann (1964); Parker (1963); and Spitzer and McNamara (1964) reported correla- tions for both variables which predominantly did not reach statis- tical significance (the .05 and .01 levels were used). Second, the Leader Behavior Description Questionnaire studies seemed to have a more consistent relationship between performance and a) "Consideration" (positive) and b) "Initiating Structure" (negative) but as Korman pointed out "there is a great degree of inconsistency (C. F. Bass, 1957; Hemphill, 1955), even among studies using some— what similar p0pulations (Halpin and Winer, 1957; Halpin, 1957)." Of these two sets of studies, only two (Bass, 1956, 1958) were 22 predictive field studies, the remainder had concurrent measurements. In general, findings showed mostly low to moderate correlations. While these studies indicated the deficiencies in the significance of "Consideration/Initiation of Structure" findings some important exceptions should be noted. In educational studies, which seem more similar to the human service context of the AAA, some interesting findings were reported. Halpin (1956) studied school superintendent leadership as described by staff, board mem- bers, and the superintendent himself. Staff members and board members seemed to agree within their respective groups but it is interesting that the two groups perceptions were significantly different. For "Consideration," staff saw the superintendent as having less than either the board or the superintndent and for "Initiation of Structure," board members perceived him as having more than either staff or the superintendent. Hills (1963) reported 872 teachers' perceptions of their principals. "Consid- eration" and "Initiation of Structure" were found to be highly correlated to both the principals' representation of teachers to his superiors and to the schools' public. While these two studies were not predictive field studies, their findings suggest that in a human service setting, where people from different sectors have influence, perceptions of those in management may be different between groups and they may be related to how leadership is per- ceived regarding interactions with others in matters important to the worker. 23 A recent series of papers by George Graen (1972, 1973) and collaborators (Cashman, J., Dansereau, F., Graen, G., & Haga, W. J., (no date); Dansereau, F., Graen, G., & Haga, W. J., 1975; Haga, W. J., Graen, G., & Dansereau, F., 1974) contributed considerably to researching the exchange between organizational members. In review- ing Max Weber's (1947) theory about bureaucracy Graen (1973) and Dansereau, Graen, and Haga (1975) noted that this theory suggested 1) that established roles of interaction between superivsor and subordinate are an outcome of fitting the right people at the right levels, 2) that jobs and people's abilities are stable over time and 3) that superiors behave the same to each subordinate they supervise. They pointed out that this "fixed model" is definitely not supported by research demonstrating that when employee and supervisors share tasks and feedback ("enriched jobs") employee turnover was reduced (Ford, 1969) and that most employees (38%) express role uncertainty. Replacing this "fixed model,‘I Graen suggested that when members enter an organization a "role-making model" would best describe initial confrontations, working-through, and eventual integration with the working group. Thus the "role- making model" acknowledges the visable placement of personnel but it also provides for investigation of the information establishment of working roles between those in an organization. Dansereau, Graen, and Haga (1975) tested this "role-making model" in a study of the role-making consequences for managers when a university housing division reorganized. Researchers mea- sured "the extent to which a superior is willing to consider 24 requests from a member concerning role development" (e.g. "Negotia- tion Latitude,“ as reported by managers, p. 10) at the first, third, and eighth month of this process. In the first month the distribu— tion of responses was split in the middle, creating heavily equal sized groups into those IN (high negotiating latitude) and OUT (low negotiating latitude) with their supervisors. Reports from mana- gers, their superiors, and observers were analyzed by a One-way ANOVA for repeated measures. INs reported more attention (p = .08) and support (p = .060). OUTs had significantly more (p < .01) communicating and administration with him. However, the group by time interaction showed diminishing significance in difference (p < .02) between INs and OUTs over time. Final analysis of the manager's reaction to their job situation (role outcomes) demon- strated that INs had significantly (p < .03) more job satisfaction and, more specifically, they were more positive about a) intrinsic outcomes of their work (p < .01), b) personal interactions with their supervisor (p < .001), c) their supervisor's technical compe- tence (p < .01), and d) the value of rewards for their performance. Further, a significant (p < .03) group by time interaction for the value of performance rewards suggested that INs were more positive about this value over time, and that this difference increased over time. These results support the concept that when an employee enters a new position in such an organization, something more com- plex and informal than mere "plugging into a position" may be observed. Effectiveness in negotiating a role involving exchanges 25 with a supervisor seems to predict the nature and character of manager-supervisor relationship behavior and also job attitudes. With specific reference to the continuous internal adjustments made necessary in the AAA by the requirements of multi-service coordina- tion, consideration of Negotiation Latitude and related variables would seem warranted. Job design and job satisfaction. In researching the idea that roles in working are established, change, and are reestablished, several investigators have underscored the concept of the need to create and maintain work settings which integrate staff needs (Goldenberg, 1971; Sarason, Zitnay, & Grossmann, 1972; Sarata & Reppucci, 1975; Reppucci, 1973). Staff needs may be translated into established characteristics of their jobs (Job Design) and may be directed, in part, at satisfaction with the work they do (Job Satisfaction). Correlates of job design and staff attitudes have been studied by Friedlander and Brown (1974). In the study of job design measures,factors influencing job design have been cited (Reppucci, Dean, & Saunders, 1975; Sarata, 1975). Sarata and Jeppesen (1977) researched the relation- ship between job design and satisfactions in a study of child serving agencies. Employees participating in this study included nurses, psychologists, remedial education teachers, social workers, and non-professional staff; all had direct service roles. A multi- ple regression analysis using job design variables (variety, task identity, feedback, autonomy, participation, learning, and 26 information) to predict level of job satisfaction revealed a signi- ficant relationship (R_= .34, p < .05). Further, simple correla- tional analysis indicated job satisfaction was significantly (p < .05) related to the specific job design characteristics of learning, variety, and information. Level of job satisfaction seems to be systematically related to job design variables. It should be recognized that these results came from employees of widely diversified agencies, thus correla- tions were subject to these variabilities. In turning to applied research in the social service organization of the AAA such investi- gations seemed potentially worthy of investigation since many of the job design variables seemed intuitively related to not only a variety of employee centered variables but also those associated with the information and referral service functions. Need for broadened scope in organizational research. All the organizational research reviewed above has studied the organi- zation and its members as an isolated unit. A relatively small number of researchers have recognized this as a research shortcoming and have initiated research which examined the organization as well as influences which impact on an organization from outside it. Sarason, Zitnay, and Grossman (1971) considered the problem of estalbishing work roles from this expanded perspective. This included concern for recognizing and utilizing the community in which the organization becomes established. And,rather than con- sidering these as tangential issues.these authors proposed 27 making them integral to staffing and goal setting in a community- based service. While this view, and that of Graen's group have to do with situations in which new relationships are established in an organization, one might ask what effects could be expected in an organization which changes rapidly as a result of interactions with situations external to its formal boundaries. As mentioned above, the AAA has been observed to have rapidly changing internal roles therefore such a question may have potential interest. Sarata and Reppucci (1975) addressed reciprocal relation- ships within an organization and did so from a perspective that placed the organization in an environmental context. Thus, factors outside an organization were also examined. This study investigated influences external to a rehabilitative organization which, in turn, influenced its internal functioning. In this service setting, an experimental re-entry program for adult offenders, nonprofessionals related directly to the resident adult offenders, over several months, in efforts to reestablish them as citizens of the community. Throughout this process 1) non-professionals rated resident aggressiveness, 2) residents responded to instruments indicating distrust of others on a modified Machiavellian scale (Christie and Geis, 1970), 3) independent observers recorded the nonprofessionals' behavior toward the residents, and 4) data on the frequency of residents returned to the facility for breaking program rules were recorded. This data was collected as part of a program evaluation effort and this circumstance made data available for an unintended 28 experimental design. What happened within one month of a seven month period was that Federal funding for the project became very questionable,(thus threatening the program and nonprofessional jobs) and also State corrections officials came to review this unorthodox program. The project began in February (Period 1); the forgoing outside events became salient in April (Period II); and later the funding was assured and official review terminated in August (Period III). Sarata and Reppucci analyzed the above mentioned data which had been recorded over this time span and made the following observations after the fact. 1. Staff rated residents as more aggressive during Period II than at other times; but there were no changes in ratings made on other dimensions. 2. Staff behaved less competently and supportively with residents during Period II than at other times. 3. The residents' scores on the Mach scale were highest during Period 11 but no systematic changes occurred in responses on other measures. 4. Residents exhibited more negative behavior during Period II than during either of the other periods. (p. 95) Data from the three periods consituted an A-B-A design. In the analyses, "indigenous staff" (nonprofessionals) and "other staff" (professionals) were compared to show consistency of effect. In a 2 X 3 ANOVA these two groups were not significantly different in their rating of resident aggressiveness and a time factor analysis showed ratings indicated significantly more (p < .05) aggressive behavior in Period II. Also, residents scores on the Mach Scale were significantly highest during Period 11 (ANOVA F significant p < .05). While no statistical analysis was offered for measures 2 and 3 above, reports of the observers 29 suggested nonprofessionals harangued residents, "expressing general frustrations" in Period II and records showed residents broke program rules most during this time (these were relatively negligi- ble during Periods I and III). This study illustrated that events external to an organiza- tion may have radiating impact on it. When staff are affected by external events, changes in their working attitudes and behaviors may subsequently impact on the services received by their clients. Summary and research direction. It is clear from the literature reviewed above that the variety of organizational theories and research have provided several clear notions of which dynamic characteristics of organizations are important to the work they do. From study of the "ideal type" organization proposed by Weber (1958) it becomes clear that the organization studied in this paper, the Area Four AAA, is substantially different from it. Conversations with and observation of AAA personnel repeatedly pointed out that while there was some hierarchy of authority the organizational units of the AAA were chiefly characterized as autonomous and informal. For these units,operating procedures, rather than being rule-bound, were clearly flexibly-determined by client needs and situation. Thus, Graen's (1973) proposal that the informal side of the organization should not be ignored seemed particularly appropriate as an area of emphasis in researching the AAA. Accordingly, the research discussed above addressed organization/worker issues related to the informal, inner workings of organizations. 30 In non-service settings, early organizational theory offered interesting findings regarding leadership styles, and participation of organizational members. Researchers examined leadership styles in small groups of boys (Lewin, Lippit, & White, 1939), worker participation in production changes (Coch & French, 1948) and parti- cipation and job satisfaction in clerical groups (Morse & Reimer, 1956). Although these studies suggest the importance of leadership approaches and member participation, weaknesses in experimental design and measurement cast doubt on their general importance across organizations. In particular, none were reported for social service settings such as the AAA. These factors may be operating in such settings but this investigationhas not been done. Exchange between leader and subordinate has been examined in more recent investigations. Homans (1958) pointed out the importance of looking at interpersonal exchanges. Then, in a more organizational context, Fiedler (1967) initiated a theory of leader- ship effectiveness based on leadership styles and favorableness of the situationto»the leader. Generally, it appeared that the leader's orientation determined effectiveness of those supervised. However, Graen, Alvares, and Orris (1970) point out that large numbers of studies using this approach have generally overstated the application of this theoretical model. Similarly, research employing the key factors of "Consideration" and "Initiation of Structure" by Shartle (1950) and others have also been found to have highly inconsistent or insignificant findings (Korman, 1966). 31 In the AAA situation, the outreach worker and director from theI and R center must necessarily have informational exchanges in order to serve the elderly; the essense of the work i§_information and referral. One component of this relationship is how the director represents the outreach worker within the organization. Perhaps the most useful work in the literature regarding the exchanges between leader and subordinate comes from Graen on role-making. In particular,the Dansereau, et a1. (1975) study would indicate that when the role relationship between supervisor and worker is in a process of change, how this relationship develops may effect the workers'attitude and behavior. This was particularly interesting, and potentially useful for researching the effect worker/director exchange may haveirlthe AAA since this exchange may subsequently influence how the elderly citizens are served. Finally, focusing on social service organizations Sarata and Jeppesen (1977) gave evidence that job design factors such as Opportunity for learning, variety, and information are systematically related to job satisfaction. In somewhat related findings (Sarata 8: Reppucci, l975),nonprofessionals in a community based service setting were found to be significantly affected in perception of and behavior toward clients by events which threatened their work. Thus. in researching the organizational processes Of the I and R centers Of the AAA, job design and factors effecting the jobs Of employees could be investigated for relationships with the satisfactions of the elderly. One could reason that for the elderly, who are Often 32 skeptical of social services, that which effects the working roles of outreachers may well effect satisfactions with outreach efforts. This body Of organizational work therefore suggests that the functional characteristics of an organization, (i.e. the way organ- izational members interact in doing their jobs) affect the organi- zation members,and may affect the product Of their work. In the case of social service agencies, the AAA organization among them, the “work product" was service provided. Therefore, in researching the AAA,the organizational features Of leadership, participation, exchange, job design, and job satisfaction form the context Of the organizational research measures discussed in the methods section. Beyond the research which has limited study Of the organi- zation bathe organization per se, Sarason, Zitnay, and Grossman (l97l),and also Sarata and Reppucci (1975) have recognized the importance Of an expanded organizational context. Especially for community based service organizations, the service community appeared to have great relevance. Since the AAA was an organization which, in coordinating services for the elderly also interacted with local service organizations the importance Of these interactions could not be ignored. For this reason the following discussion provides a review of literature on interorganizational theory and research. Interorganizational Theory and Research Interestin the study Of organizations within the context of the environments Of which they are part has developed in the 33 relatively recent past. In the General Systems theory of Ludwig von Bertalauffy (1956) animate systems (such as work organizations) were defined by their need to be open to and have exchanges with a surrounding environment in order to survive. Organizational systems were viewed to be not unlike such systems. Thompson (1967) noted that Old conceptions of organizations such as "scientific manage- ment" (Taylor, 1911), "administrative management" (Gulick and Urwick, 1937) and "bureaucracies" (Weber, 1947) did not account for this need for environmental exchange. Bennis (1969) and also Burns and Stalker (1961) have made similar Observations. The need for organizations tO act and react with systems outside the organization was termed "the problem Of adaptability" by Bennis (1969). While most Of the theory on how and why organizations inter- act has come from the field Of economics, the most pertinent liter- ature has come from writings in psychology and sociology. William Evan (1966) recognized that organizations outside a given organiza- tion could be considered the enviornment's relevant elements in terms Of the common exchange of 1) information, 2) products or services, and 3) personnel. Further, this exchange was said to take place through the actual members Of the organization which were party to an exchange. Emery and Trist (1965) even went so far as to classify these exchanges by origin, receipt, and directional flow. Beyond the identification of the components of exchange. several investigators have written on the significance exchanges 34 with the extraorganizational environment have for today's organiza- tions. Howard Aldrich (1971) pointed out that organizations in social service systems can only exist when their management recog- nizes interactions between the social service agency and such environmental elements as other service organizations and govern- mental entities are matters just as important as are internal concerns. Looking to the future Shirley Terreberry (1968), sub- mitted evidence that organizations are likely to face environments characterized by "an accelerating rate and complexity of interactive effects (which) exceeds the component's systems capacities for prediction and hence, control Of the compounding consequences Of their actions (p. 393)." Moreover, she argued that organizational change is increasingly externally induced and an organization's ability to make needed changes will depend uponits ability to act on what it knows Of the changing environment. Drucher (1964), Hood (1962) and Gardner (1963) support these conclusions. Learning from the extraorganizational environment has become vital to organizations. In dealing with changes in environment,organizations may be seen as being directed by the cross product Of purposes and environment (Hood, 1962) or by goal-setting groups which necessarily interact with the environment (Thompson, 1958). Study of the process Of interactions between organizations was suggested as an important focus Of research, particularly where community organizations were concerned (Warren, 1972). 35 Environmental characteristics. Osborn and Hunt (1974) postulated that environmental complexity includes 1) broad factors facing all organizations, 2) conditions unique tO a system, and 3) interorganizational characteristics. Similarly, Perrow (1967) Offered the observation that the nature of the work (e.g. task, social structure, goals) determines how the organization is struc- tured. A small amount of literature, specific to the AAA as an organization, came from the 27th Annual Meeting of the Gerontologi- cal Society in Louisville, Kentucky. Jones and Jones (1975) reported that in surveying 30 AAAs in Alabama it was generally found that they were becoming more complex. Don Hull (1975) suggested that with the growing proportion of the elderly in the population, types and numbers Of services needed increases and community ser- vices goals will increasingly be more dependent on feedback between agencies to maintain a common goal direction. These positions seem tenable but the fact that the above literature only attempted to describe the situation for the AAAs revealed that very little empirical research has been done with these agencies. Research from other sources have, however, Offered findings which suggest direction for research of social service organizations. The first research Of interest came from studies Of effec— tive organizations in private industries. Lawrence and Lorsch (1967) selected highly effective organizations from the plastics, food packaging, and container industries to investigate what effects their respective types Of extraorganizational environment had on 36 their management of internal matters. Thepflastics industry was selec- ted because management must typically react to the characteristically diverse and dynamic business environment of this field. The food packaging industry was characterized by less diversity and change- ability and for the container industry these influences were relatively stable. Evidence from this study suggested that the plastics industry, while having the conflicting characteristics of high differentiation gpg_high integration, was able to maintain high effectivenss through employees designated to resolve conflicts on an ongoing basis. (In this way highly specialized chemists could effectively interact with those in charge of highly diversified production personnel.) Measurement of the characteristics Of all three types of successful companies showed that they all had roughly equal organizational integration but respective industry types seemed to have a degree Of differentiation consistent with the diversity of the commerical market in which they competed. Accord- ing to Lawrence and Lorsch these findings indicated that the characteristics of the environment which are relevant to the organ- ization should determine how they structure internal matters. As in most developing fields Of investigation, the current preoccupation of researchers appeared tO be the description Of organizations relative to the extra organizational matters in which they are involved. Representing interest in this new area of research Warren (1972) suggested that for community organizations, in particular, the interorganizational field of investigation may be useful in guiding the "Optimization Of mix of community values 37 values and goals (p. 305).“ Wohlhill (1970) noted that the emerging discipline of environmental psychology has started to address these issues. Exchange between organizations. In research Of health organ- izations Levine, White, and Paul (1961) reported that exchange was an important conceptual framework for studying interorganizational relationships. They observed that l) referrals, 2) labor services, and 3) non-labor services (such as commodities and information) were common elements Of exchange. Levine, White, and Paul (1963) administered a brief survey to 147 health and welfare agencies in two cities and found that of 377 respondents 113 wanted more information on other agencies' services, and 96 wanted more mutual planning or coordination. These findings began to identify sources of exchange between individual organizations and other organiza- tions situated in the relevant community services environment. Aiken and Hage (1968) chose to focus on number of joint programs as an index of organizational interdependence. Their research Of 16 social welfare and health organizations in a mid- western city investigated the relationships between structural characteristics of organizations and the number of joint programs they had with others. They hypothesized that high degree of complex- ity, program innovation, and internal communication would be posi- tively related to number Of joint programs, and high degree Of centralization and formalization would be negatively related to number of joint programs. Results Of this study came from a total 38 Of 520 program management personnel in the above organizations. Findingscnithe relatedness of the organizational determinants to the organizational interdependence index, "number Of joint programs" showed a consistent positive relationship, "complexity" was pre- dominantly positively related, and "internal communication" showed some positive relationship but results were mixed. Those deter- minants expected to be negatively related to organizational inter- dependence generally were not: "centralization" results were mixed and "formalization" results were only weakly negative. While "centralization" had mixed results,detailed analysis of component measures revealed that number of joint programs did have a weak positive relationship to degree Of staff participation. In summary, only program innovation and complexity Of the organization seemed to maintain the expected relationship to the interdependence or connectedness Of the service organizations. Propositions about the correlates of organizational inter- dependence as stated by Aiken and Hage (1968) have been tested by Steven Paulson (1974). Following the assumptions given by Hage,and in review of the research findings, Paulson discovered that the partial correlation matrix Of the specified organizational deter- minants contained relatively few significant relationships; only 16 out Of 45 such correlations were significantly (p < .15) differ- ent from zero. TO further investigate these relationships for health and welfare agencies Paulson collected responses from the 138 administrators Of these agencies. Then, using assumptions from Aiken and Hage (1968) variables were organized into a recursive 39 assymetrical model. The variable sequence had the following order: 1) structural variables: complexity, stratification, centralization, and formalization, 2) linkage variable: communication, 3) perfor- mance variables: efficiency, job satisfaction, effectiveness, and innovativeness, and 4) the focus variable: interorganizational relations (number Of joint programs). Eight regression equations were constructed using each variable in the sequential ordering as a dependent variable with all variables preceding it in the causal ordering as independent variables. Thus organized, path analysis was used to investigate the causal network among the variables. Paulson used a path analysis procedure to test the axiomatic model implied by Hage (1966). Results from this test revealed nonsignificant path coefficients which represented nonsignificant relationships between variables within the network. Thus, of the 44 pathway coefficients only 18 were found to be significantly different from zero (a = .15). Then, prior tO the second analysis, this original model was improved by setting to zero (i.e. eliminating from the analysis) all nonsignificant path coefficients. Job satisfaction, as measured by turnover, was one of the most crucial variables thus deleted. Equations were solved again for estimates Of remaining coefficients. Using this procedure, (see Duncan, 1966; Mulford, Klonglan, Warren, & Schmitz, 1972) examination of the dir- ect effects on "interorganizational relations" showed "complexity" followed by "efficiency" had greatest effects. The most important total effects (the sum of direct and indirect effects) were, in 40 order of strength, "centralization," "efficiency," and “complexity." Several indirect effects included "communication" as the linkage variable. In the revised mode1,the sign (negative or positive) Of the relationships between some variables changed from the model implied by Hage (1966). The new relationships of the model seemed tO center around "formalization," "effectiveness" and "innovation." Findings were as follows: 1) "formalization" and "centralization" were inversely related to "complexity," but "centralization" was inversely related to "formalization”; 2) "effectiveness" was inversely related to "centralization" and "stratifcation"; and 3) "innovation'was inversely related to "effectiveness" but directly related to ”cen- tralization," "efficiency," and "stratification." In summary, three points deserve comment. First, previous positions held that formalization and centralization were responsi- ble for an efficient and effective bureaucracy. Data in this article suggested that they were responsible for an effective but not necessarily efficient bureaucracy. It seems that innovative organi- zations have low formalization and high complexity, stratification and centralization. Second, interorganizational relations seem to be a function Of organizational characteristics, but this accounts for only 34 percent Of the variance (from structural and performance characteristics). Third, even though relationships in the revised model were all statistically significant, the overall magnitude of the effects were low. Paulson recapitulated that "these findings suggest that the revised model is relatively weak in terms Of 41 accounting for organization behavior generally, and interorganiza- tional relations specifically (p. 330)." Thus, for the small or- ganizational units of the AAA, the I and R centers, a clear research question seemed to be "How effective are these apparently informal organizational units?". Also, one might ask "How do the interor- ganizational relations between I and R centers and key service agencies relate to the organizational determinants Of the work situation for I and R center staff?". Research of the AAA attempted to measure these variables and relationships. This research there- fore was directed toward the need for more research of these questions. Summary and research direction. For the AAA, whose function it was to coordinate existing services to the elderly, the level at which the organizational/interorganizational systems were typically measured seemed inapprOpriate. These most recent studies (Aiken & Hage, 1968; Lawrence & Lorsch, 1967; Paulson, 1974), in using the gpganization as the unit of analysis Trad neglected the importance of the organizational members which actually carry out interorganizational exchanges. Focusing on the organizational member seemed particularly appropriate in researching the processes Of the AAA since component I and R staffs appeared to be highly autonomous, individualized, and small. Thus, the research in this paper measures staff perception of environmental characteristics. For the concept Of exchange between organizations, Paulson's (1974) findings regarding communication, efficiency, job satisfac- tion, effectiveness, and innovation, as organizational referants 42 Of interorganizational exchange, were found to be Of mixed impor- tance. The similarity in content Of these variables tO those mentioned earlier under organizational theory and research raised the argument that they might best be examined using organizational members as the unit of analysis. Such an approach also benefited from an increased sample size; instead Of looking at the AAA (n = 1) or the I & R centers (n = 7), using all staff Of the I and R centers (n = 20) allowed for more sources Of data. Moreover, the primary function Of the I & R centers, informing and referring seniors, took place on a one to one basis, therefore job characteristics which seemed logically related to this function warranted investi- gation: Feedback, information, learning, and participation are among them (Sarata & Jeppesen, 1977). Adding interorganizational components to these member-based organization measures also seemed to increase their meaning for the AAA context. Research Rationale Researching the AAA appropriately involved extracting data at three levels Of content: 1) the AAA organization, 2) its rela- tions with other organizations, and 3) the elderly clients served. Thus, this community-based research was based on a multilevel approach. Several researchers have noted that multilevel research provides a stronger information base. McGuire (1973) pointed out that to get at the complexities Of naturally occuring events, such as those in community settings, a multiple focus seems best able to get at major influences. Consistent with this Observation, 43 White (1974) suggested that organizational and interorganizational areas have some overlapping of important variables and that research design could be improved by measuring both Of these content areas. Therefore, this research of the AAA included measures of the AAA organization and of two other agencies (the Department of Social Services and the Social Security Administration). (In addition to the AAA,these two agencies gave service to the elderly clients in this study.) The individual responses Of AAA staff regarding the characteristics of their work in their component I and R centers and their responses about certain key elements Of exchange with the two other service agencies provided the data for the first two content levels above. The AAA research provided focus on the AAA and two other key service organizations but, importantly, it also provided focus on the perspectives of the elderly clients selected for this study. In this way, relationships Of the organizational process were compared to service outcomes for the elderly clients. Study Of these important relationships were central to the research of the AAA. In previous research Goodwin (1971) identified that social service system research should include data from both service sys- tem (organization) gpg_public served (clients). Also, from a general systems research perspective, Burgess (1975) had recognized the importance Of examining the internal Operations gpg outcomes of organizational efforts. Examination of the AAA organization and the elderly clients served by it seemed especially salient since the poor or disabled elderly, as members Of a subpopluation with 44 low socioeconomic economic status, are Often grossly underrepresented in the decision-making processes which impact on community based served organizations (Grain and Rosenthal, 1967). Thus the AAA research included measures of the service received by the elderly clients and the perceptions of these services held by the elderly clients. Therefore, the objective of this research was to examine the characteristics of the component I and R centers Of the AAA, the interaction OfI and R staff with DSS and SSA agencies, and how these characteristics were related to service received by elderly clients. Organizational and interorganizational data came from I and R center personnel, the directors and outreach workers. Client data come from the elderly clients served by the I and R centers. The variables used for each content level are discussed, by level, below. First, measurement of organizational variables involved the I and R center staff perceptions Ofsevenalcharacteristics Of their jobs. These variables were staff autonomy, social contact, feedback, information, learnipg, task identity, varieiy_(Reppucci, Dean, & Saunders, 1975; Sarata, 1975; Sarata & Jeppesen, 1977), and participation in decision-making (Coch & French, 1948; Lewin, Lippet & White, 1939; Morse & Reimer, 1956; Sarata & Jeppesen, 1977). Degree Of agreement between director and outreach worker regarding feedback, information, and participation in the outreach‘worker's jOb situation was also measured. These indices of director/outreach worker agreements provided data specific tO director/outreach worker 45 exchange. Further, pegotiation latitude, the extent to which a superior was willing to consider requests from a subordinate con- cerning role development, was used as a specific measure of leader/ subordinate relations (Dansereau, Graen, & Hage, 1975; Fiedler, 1967; Halpin & Winer, 1957). Finally, job satisfaction (Sarata & Jeppesen, 1977) was also measured. Second, interorganizational variables (relations of the AAAvnth the other two organizations) were measured separately for both the Department Of Social Services (DSS) and the Social Security Administration (SSA). Degree Of interorganizational exchange (Aiken & Hage, 1968; Lawrence & Lorsch, 1967; Levine, White, & Paul, 1961; Paulson, 1974) was measured via several separate variables (Tornatsky, 1964; Tornatsky & Lounsbury, 1974). These variables were the amount of funding received, amount known about the agency, amount of referrals. importance Of their services to elderly clients, number Of ways of communication with them, similariiy_of goals, and similarity Of tasks and services. Third, data on elderly client variables were collected. The demographic variables were age, amount of education, contact with own children, client's ability to read, client's memory/attention, life satifaction, whether or not living alone, number Of correct social services identifications, Observed client health, race, response level, sex, social contact, social service knowledge, and type Of assistance. Attitudes toward each Of three agencies (AAA, DSS, and SSA) were also measured. Thus, satisfaction, frequengy Of use and intended use of these agencies were measured for 46 individual elderly clients. And last, whether or not a client asked for and received service from the I and R center was also collected. Examination of these variables involved analysis of the relationship between variables within each context level. Then, relationships between the organizational/interorganizational variables and the client variables were analyzed. CHAPTER III METHODS Setting The Area Four Area Agency on Aging was selected as the main organization to be studied by virtue of its special contractual mandate with the local Michigan Department of Social Services Office. Under the special Title XX contract the AAA provided in- formation and referral to local elderly persons. In functional terms this meant that the AAA organized a system of staff members in its three county area to provide personal contacts with elderly persons. This contact effort, termed "outreach" was managed from seven Information and Referral (I and R) centers in the three counties. Berrien county was served by five I and R centers; Cass and Van Buren counties each had one I and R center. Since these centers received 75% of their funding locally, service boundaries had little overlap. I and R centers were typically located in existing structures such as store fronts, and church basements. Staff at these centers included a director, a secretary, and one or more outreach workers. The numbers and organizational relationships of the composite staff to the Area Four AAA are detailed in Figure 2. This organizational system provided the framework for service coordination with key social service agencies in the area. 47 48 FIGURE 2 INFORMATION AND REFERRAL SYSTEM I 8 R SUPERVISOR. AREA AGENCY ON AGING 7K I 8 R I & R I 8 R I & R I & R I X R DIRECTOR DIRECTOR DIRECTOR DIRECTOR DIRECTOR DIRECTOR AREA 1 AREA 2 AREA 3 AREA 9 AREA 5 AREA 7 River Niles Berrien Benton Saint Cassopolis Hartford Valley Springs Harbor Joseph \\\E Berrien County Cass Van County Buren County KEY: 1) ON = Outreach Worker 2) Areas are numbered here as they were by the AAA. 3) Shaded areas indicate participant attrition. 49 From the state level survey findings on problems Of elderly people discussed in Part One Of this paper the reader will recall income and health problems were among the three most frequently mentioned by the elderly. For this reason the social service agencies having large programs designed to address these needs were included as focal organizations in this study. These agencies were important to the I and R centers in terms Of service coordina- tion functions,and to the elderly in terms Of assistance received from these major programs. The local Department Of Social Services (DSS) was providing the Old Age Assistance/Medical ("M") program and the regional Social Security Administration (SSA) Office was providing the Supplemental Security Income ("551") program. The clients who participated in this study were selected by virtue of their enrollment in one of these two programs. Logistics This study was largely made possible through the COOperation and support Of both the Michigan Office on Services to the Aging (OSA) and its regional agency, the Area Four Area Agency on Aging. At the time this study was proposed there was keen interest in detailed research Of AAA functions, and, in particular, these agencies wanted to research the Information and Referral function Of the I and R centers in Area Four. OSA was interested in comparing different methods Of initiating outreach efforts (the substance Of an experimental study Of information diffusion which was completed concurrent with the research described in this paper) and were 50 willing to extend this to the correlational study Of the AAA. The correlational study of the AAA was coordinated with the experimental information diffusion study such that elderly clients from only two Of the experimental conditions (those initially contacted by telephone or by personal contact) were included. Also, all measures used in the correlational study were exclusive to this research but two measures (The Client Survey and Outreach Response Sheet) were used in both studies. Only part Of the data in these shared measures were used in the correlational study therefore for these measures discussion is limited to the selected variables. A Memorandum Of Agreement between the author, another graduate student,and the Area Four AAA was drawn and signed (see Appendix A). This document contained agreements on the conduct and reporting of the research. While these agreements were being negotiated I and R center directors were consulted and verbal agreements were secured. Participants Three participant groups were required for the research: elderly clients, I and R outreach workers and I and R directors. While staff from all seven I and R centers agreed to participate in the research, data on clients in the Benton Harbor I and R center area, to have been collected by outreach workers, was not collected. Correlational analysis between staff and client data was precluded for this I and R center and the Benton Harbor I and R center was therefore dropped from the data analysis. Figure 2 above shows 51 the participant attrition for Benton Harbor as well as for two other outreach workers (from Cassopolis and Hartford centers respectively). Discussion Of participants does however include these people since they were initially part Of the research. Elderly_Clients From lists Obtained from the Department Of Social Services a sample of 350 senior citizens which were registered as receiving Supplemental Security Income (SSI or program "A") or Old Age Assistance/Medical (program "M") were identified for the information dissemination study mentioned above. From these raw listings, I and R center directors were asked to eliminate those names of people already contacted in information and referral efforts. This reduced list was then used for the random selection Of equal numbers of elderly persons in each program for each Of the seven centers. While the information diffusion study conducted concurrent to this study required a more extensive sample this study only required a subset Of it. The research of this paper included the elderly clients who were contacted by the I and R center either by phone or in person. The purpose Of these contacts was to inform the person Of the available services in the area and possibly to make referrals. These two types Of contacts were selected as those which were most direct and therefore, if there‘was to be any effect from a contact,it would presumably occur subsequent to these approaches. The resulting sample for this correlational study was initially 20 per I and R center or 140 total. Thirty-nine of the 52 140 elderly clients could not be reached during the initial contacts by phone or in person. Then, of the 101 then in the sample only 74 allowed the follow-up data collection which provided the bulk Of elderly client data. Thus, the final sample included 74 elderly (flients. Of these, roughly half were in the SSI program, and half were in the M program. The elderly clients selected for this study were receiving service from either DSS or SSA; contact by the I and R center was to serve the purpose of linking the elderly client with any services needed beyond assistance provided by $51 and program M monies. Outreach Workers In the seven I and R centers in Area Four there were from one to three outreach workers in each center. All were females (23 to 66 years old) who did door to door canvassing tO locate senior citizens needing local services. Their task was one Of linking the needy senior with appropriate services. Thus, the information and referral function of the centers was largely conducted through out- reach worker efforts. As shown in Figure 2 above, 13 outreach workers were initially included in the study. Four outreach workers left their I and R centers just as the research was beginning. The final sample, therefore, included nine outreach workers. I and R Directors Directors from theI and R centers participated in this study. In the original sample all the directors were females (27 to 50 years old). The basic tasks required of directors included 53 supervision of outreach efforts and administration Of center Operations. This often involved negotiating with county commission- ers and other service agency personnel. Directors were coordinated by the main office of the Area Four Area Agency on Aging. Again, Figure 2 shows these supervisory relationships, and the attrition Of one I and R Director. £251.91 This research was designed to investigate, through correla- tional methods, the variables attributed to characteristics of the I and R center-based AAA organization, the interoganizational inter- action between I and R centers and DSS and SSA Offices, and the elderly clients served by these organizations. Basic correlational analyses were completed on the variables 1) within the organizational/ interorganizational data set, 2) within the client data set and 3) between the organizational/interorganizational and client data sets. Thus, for the first set Of analyses, staff autonomy, social contact, feedback, information, job satisfaction, learnipg, negatia- tion latitude, participation, task identity, mdvariety (organizational variables) were examined separately,and the interorganizational variables Of amount of funding received, amount known about the about the agencies, amount of referrals, importance Of their services, number Of ways of communicating, similarity of ggals, and similarity of tasks and services were also examined separately. 54 Then, correlations between organizational and interorganizational variables were computed. Correlational analysis of the second set Of data, the client data, involved examination Of elderly client demographics (age, education, contact with own children, abilityito read, memory/attention, life satisfaction, whether or not livipg alone, number of correct social services identifications, Observed health, race, response level, sex, social contact, social service knowledge, and type of assitance), client attitudes about the AAA, DSS and SSA organizations (in terms Of satisfaction, fregueney Of use, and intended use), and whether or not a client asked for and received service from thel and R center. Client demographics and attitudes were first studied separately, then together with the "service received" variable. Finally, the fundamental question Of what relationship organizational/interorganization variables might haveixiactual service outcomes for the elderly clients was analyzed. Thus, organizational/interorganizational variables, as well as client variables, were considered the independent variables and service received by elderly clients was the dependent variable. More detail on the actual procedures used to do these analyses follows an overview of all measures and their administration in the following section. 55 Procedures Overview of Measures and Their Administration Several measures were employed for the organizational, interorganizational, and elderly client data sets. Summary infor- mation, including instrument name, type Of measure and item content provided in Table l clarifies the source of data on variables men- tioned above. Four measures were used for the organizational data set (the Job Design interview, the "Faces" Satisfaction Measure, the Negotiation Latitude index, and the Director/Outreacher Exchange questionnaire), one for the interorganizational data (the Inter- organizational Interaction Questionnaire), and five for the elderly client data set (the Senior Citizen's Opinions questionnaire, the Outreach Response check sheet, the Follow-up Survey, the Phone Call Format Sheet, and the Outreach Visit Format Sheet). Administration of Organizational and Interorganizational measures. The author administered all organizational and inter- organizational measures to the seven I and R directors and 11 outreach workers (two outreach workers left prior to this data collection). Since the I and R centers were geographically remote from one another (spread across three counties) appointments were made on a center by center basis. When the author arrived at the center, the director and outreach workers were seen privately, one at a time, according to a schedule derived by the director. Before administering any instruments to an individual, the author explained that responses to the measures were to be held confidential and that 56 TABLE 1 MEASURES Measure Type of Measure Item Content “Faces” Satisfaction Measure (F. S. M.) Negotiation Latitude (N. L-) Job Design Interview (J. 0.) Director/ Outreacner Exchange Questionnaire (D./O- E.) Interorganizational Interaction Questionnaire (l. I- 0.) Staff Questionnaire Staff Questionnaire Staff Interview Staff Questionnaire Staff Questionnaire satisfaction with job in general satisfaction with field Of social services satisfaction with actual work satisfaction with own agency satisfaction with coworkers satisfaction with types of clients worked with satisfaction with supervision satisfaction with pay satisfaction with promotion satisfaction with client progress satisfaction with l and R center services satisfaction with DSS services satisfaction with SSA services Supervisor flexibility in negotiating changes in one's job likelyhood of supervisor using influence to help employee number of separate tasks in the job rate of client contact number of coworkers interacted with hours of face-tO-face contact per day number of different individuals from other agencies seen in a month number of times individuals from other agencies seen in a week sOurce of job clarificatiOn degree to which job is perceived as integral autonomy in scheduling work time autonomy in actual work frequency of general feedback from supervisor frequency of periodic feedback from supervisor frequency of feedback on individual clients aMOunt of feedback on client referral degree of participation in decisions about clients degree of participation in decisions about program and policy amount of information received about AAA amount of infornotion received about other agencies amOOnt of general informatiOn received frequency of optional learning experiences frequency of learning about other agencies outreach worker's actual participation in work decisions outreach worker's needed participation in work decisions outreach worker's actual participation in client decisions outreach worker's needed participation in client decisions outreach worker's actual feedback on work with elderly clients Outreach wOrker's needed feedback on work with elderly clients outreach worker's actual infornotion on the l and R center outreach worker‘s needed infornotion on the l and R center Outreach worker's actual information on outside agencies Outreach worker's needed information on outside agencies water of ways of comunicating with DSS number of ways of cannunicating with SSA amount of referrals from DSS amount of referrals from SSA amount of funding from DSS amount of funding from SSA similarity of goals to those of DSS similarity of goals to those of SSA similarity of tasks and services to those of DSS similarity of tasks and services to those of SSA importance of DSS services to seniors importance fO SSA services to seniors amount known abOut DSS amount known abOut SSA 57 TABLE I-Continued Measure Type of Measure Item Content Phone Call Format Sheet (P. C. F. S.) and Outreach Visit Format Sheet (0. V. F. S.) Outreach ReSponse Sheet (0. R. S.) Follow-Up Survey (F. S.) Senior Citizen‘s Opinions Questionnaire (S. C. 0 Client Contact Check Sheet Client Response Check Sheet Client Interview Client Questionnaire response level OUCPQGCM response type Of assistance age sex race living alone observed health memory ability to read social contact contact with own children education social service knowledge life satisfaction satisfaction with AAA satisfaction with DSS satisfaction with SSA frequency of use of AAA frequency of use of DSS frequency of use of SSA intended use Of AAA intended use Of DSS intended use of SSA 58 any results reported would be in aggregate form. Further, staff members were requested not to confer about the measures with I and R staff members in their own center or other centers in the region until all had taken them. To further prevent collaboration about responses, the author also administered measures on the same day tO staff in centers known to communicate frequently. Thus, all I and R staff were interviewed within two weeks. Each administration period with a staff member required approximately one half hour. During this time the author adminis- tered the measures each time, in the same order, beginning with the "Faces" Satisfaction Measure (the FSM). For the FSM the researcher gave the staff member a copy of the instrument, commented briefly on its format and directed that for each item (regarding an aspect of work satisfaction) the respondent was to place a check (V) on the iage_which best represented how they felt about the aspect of work to which the item referred. A continuum Of happy to sad "faces" under each item provided eleven response Options. Attached to the back Of the FSM, the two-item Negotiation Latitude index (the NL) was next in the battery of measures. When finished with the FSM,the author directed the staff member to complete responses to these two, self explanatory, multiple choice questions. The staff member selected a response Option by placing a check (V) by the appropriate choice. The Job Design interview (the JD) followed collection of the FSM/NL packet. The author explained that the JD employed an interview format in which the researcher was to ask a series Of 59 questions about several aspects Of the respondent's job and then the author would simply mark down each response directly on the instrument. Thus, the instrument acted as a protocol of interview items for the author. In order to subsequently compute an inter- rater reliability estimate of this method the author trained a graudate student in several practice sessions and arranged for him to be present and corate six out of the 15 interviews across all centers. Each interview contained two types of interview items. First, for the items referring to "variety" and "social contact" the researcher asked the question suggested for each item, and wrote down longhand the staff member's response. Then, the counting number associated with the response was recorded. Second, for all remaining items on the JD, the instrument listed a continuum of response-rating categories. For each item the researcher(s) simply circled the rating judged most appropriate. When the co- researcher corated on this measure, he was present only during the JD interview part Of the time spent with the staff member. Up to this point,measure administration was identical for both director and outreach worker. For the Director/Outreacher Exchange questionnaire (the D/OE) two forms were used: outreach worker's received the outreachers form, directors received the directors form for each outreach worker at their center. Outreach workers answered the items with regard to themselves and directors answered the items with regard to each outreach worker at the center, one at a time. Each form given to the director had the name of the outreach worker to be referred to written in a blank space at 60 the top. For both forms the questions asked about the degree to which the outreach worker's job a) provided the characteristic named, and b) how much of the named characteristic was needed. These two questions were listed for the characteristics of participation, feedback, and information. The respondent was directed to place a check (v/) in the box corresponding to their judgment. Last in the battery of organizational/interorganizational measures was the single interorganizational measure, the Interorgan- izational Interaction Questionnaire ( [10). Director and outreach worker received the same instrument and instructions. Respondents were instructed to indicate the response selected by placing a check (v’) next to it. Each question asked for a response--first regarding the Department of Social Services, then regarding the Social Security Administration. These responses provided quantifi- cation Of several suggested types of exchange I and R staff had with DSS and SSA. All organizational/interorganizational measures were administered beginning with the week after outreach workers com- pleted the initial contacts to all selected elderly clients. This time was selected since it was the earliest Open time after the listed clients had been contacted. Administration Of elderiy client measures. Five instruments were used to collect information about the elderly clients in the study. Two of the instruments were used during the initial contacts 61 with the elderly clients, all the others were administered afier. I and R center outreach workers had completed all the initial con- tacts. All measurements Of the elderly client component of the research were directed toward a) outreach worker contacts with elderly persons not previously contacted and b) a series of possible responses to this initial contact. The first measures were used by outreach workers during each contact. These measures were actually protocols of standard information on the services available to the elderly to be stated directly tO the elderly client contacted. When the elderly client was on the list to be contacted byitelephone, the outreach worker used the Phone Call Format Sheet (the PCFS). If, however, the client's name was on the list to be contacted ip_ parapp, the outreach worker used the Outreach Visit Format Sheet. In either case,the outreach worker received a lecture on reasons for standard information in each contact and was given training in practicing use of both format sheets prior to the initial contacts. As the content Of each section on the format sheet was completed during the course Of the contact, the outreach worker recorded having related that section to the elderly client by placing a check (v’) by that section, thus the format sheets acted as measures of the degree to which the client was content to listen to the in- formation. In addition to this, if the elderly client permitted the outreach worker to enroll him/her with the I and R center, the outreach worker recorded the enrollment as an index of a response to the contact which could be categorized as more positive than merely 62 listening to all information given in the protocol. All format sheets were collected from the I and R centers when all telephone and personal contacts were completed. Immediately after all initial client contacts were completed, the author arranged a meeting with I and R directors and outreach workers. During this meeting I and R center staff were trained in the use of the Outreach Response Sheet (the ORS). They were instructed that in using the ORS they would be recording information about any contacts made by the elderly clients initially contacted for this study. The author handed out c0pies of the ORS to each director and talked through several examples of how to enter record information. Also, the author related several hypothetical contact situations, asked for I and R staff to practice making record entries,and then checked them. For each contact the staff member was to locate the person's name already listed on the instrument and then record type and date of the contact as well as which type Of service the center then provided. I and R staff were asked to please place this ORS form next to the telephone at their center so that it would be easily accessible. Once each week, for the four weeks the ORS was used at the centers, the author stopped at the center to talk with the staff and monitor the continued use Of the ORS. At the end of these four weeks the ORS was collected from all I and R centers. This took place just prior to the follow up phase of the project. During the follow-up phase the elderly clients were again contacted, this time to gather important information directly from 63 them. Outreach workers administered the measures during these follow- up contacts. Just prior tO these contacts outreach workers were assembled at a meeting during which each worker read through each instrument and practiced asking the questions. Following this, the directors were again given lists Of the elderly clients to be contacted and plenty of blank instruments. The first instrument used in the follow-up phase was the Follow-Up Survey (the FS). This instrument was primarily used in the concurrent research already described, however, the demographic data from each survey was extracted for use in the correlational research described in this paper. Outreach workers visited each elderly client's residence, used the FS as an interview protocol, and after asking each question recorded the client's response directly on the FS form. When the client had finished answering each question given in the FS, the outreach worker gave the elderly client a copy of the Senior Citizen's Opinions questionnaire (the SCO) and explained its general format and content. While this questionnaire was basically self-explanatory, the outreach worker explained that they were being asked to share their Opinions about the AAA, DOS and SSA organizations by selecting the appropriate answers. If the person could not read or had poor eyesight, the outreach worker helped by reading the question to them and then marked down their response. When outreach workers completed all follow-up contacts, all FS and SCO measures were collected. 64 The Measures and Their Development All measures in the study were subjected to the same data reduction procedures. First, within each measure, items having 80 percent or morelrfthe responses in one category were omitted from further analysis. Second, where applicable, items were combined into scales. This scaling process employed both empirical (Cronbach, 1970) and rational (Jackson, 1970) standards. Generally, within each measure Of the organizational, interorganizational and elderly client data sets scales were rationally (Jackson, 1970) created then, using corrected item-total correlations and coeffi- cient alpha (Cronbach, 1970), scales were tested and revised until an Optimal rational fit and coefficient alpha were achieved. Items from each measure not excluded for poor frequency distribution, but not selected for component scales were retained as one-item, “singlet," variables. Below, the procedures and results leading to the scaling of data in each measure is discussed. Organizational and Interorganiza- tional measure development is described first and is followed by the same type Of information for the Elderly Client measures. Organizational and interorganizational data. The reader will remember from the section on research rationale that the primary focus of the research was to determine the relationship between the organizationallinterorganizational characteristics Of the I and R centers and service received by elderly clients. Five separate instruments were used to collect the organizationall 65 interorganizational data. In the order presented below, Table l of of Appendix 8 summarizes scaling for each instrument. Also, an overall scaling list, by instrument, is given in a table following discussion Of the organizational/interorganizational measures below. 1. "Faces" Satisfaction Measure, FSM. This questionnaire was designed to index the satisfaction of directors and outreach workers toward the general context of their jobs. Using a "faces" technique developed by Kunin (1955) thirteen items were included in the FSM. Items for the FSM had three sources. The context Of six items came from the Job Descriptive Index, the JDI, (Smith, Kendall, a Hulin, 1969): the job in general, the work itself, coworkers, supervision, pay, and prpmotion policies. Four items used in a study of social service employees by Sarata and Jeppesen (1977) (working in this field, working in the particular agency, the type of clients worked with, and the progress made by_the clients worked with) were also included. Finally, since the clients in the current study had contact with three primary service organizations, the last three items addressed the services of the local I and R center, the local Department of Social Services and the local Social Security Administration office. The final form of the FSM contained one item for each of the thirteen content areas above. For all "faces" items subjects responded to a series Of eleven schematic faces ranging in expression from a pronounced smile to a pronounced frown, thus representing degrees Of satisfaction. Each subject placed a check (8’) on the face indicating his/her degree Of satisfaction. 66 The validity Of using the "faces" technique for measuring job characteristics was examined in the Smith, Kendall, and Hulin (1969) study mentioned above. Direct scoring of the JDI, when compared to the same content using the “faces" technique, correlated an average of r = 0.54. Thus, the technique seemed appropriate for meaasuring the areas of job satisfaction covered in the FSM. The first attempt tO reduce the thirteen items of the FSM into a single scale revealed that an eleven-item scale achieved the highest coefficient alpha (a = 0.89). This scale was identified as the general job satisfaction scale. Of the two remaining items one (satisfaction with working in the field Of senior citizen services) was omitted according to the high endorsement criterion. The item which was retained but omitted from the job satisfaction scale, item 3 (satisfactionwith the actual work) had a conspicuously low correlation (r = 0.13, p < .32) with the scale. Table l Of Appendix B details this data reduction solution. 2. Negotiation Latitude questionnaire, NL. In the battery of measures following the "faces" measures a two item scale devised by Dansereau, et al., (1975) was included. These questionnaire items asked about a) the perceived flexibility of one's sgpervisor in negotiating changes in one's_job and b) the perceived likelihood _pf one's supervisor using his influence in heiping the employee. Both questions referred directly to the relations between the worker and the supervisor. For the outreach worker the supervisor was the I and R center director, and for the director it was the coordinator 67 of I and R services for the AAA. Each question had a Likert-like scale having four categories. In the Dansereau, et al. (1975) paper these component itmes were found to correlate an average of r = .68 over a series of repeated measures. Aside from the rational judgment that these items were similar, in the current study they were in fact highly corre- lated ( r = 0.86, p < .001). Thus these two items were combined as the negotiation latitude scale, NL. Table l, of Appendix 8, also provides this information. 3. Job Design interview, JD. The Job Design interview was used to measure job characteristics of the I and R staff. In previous research (Sarata and Jeppesen, 1978) this measure demonstrated its usefulness in studying the jobs of those in the field of social services. The job design content areas were: variety (two items) task identity_(two items), feedback (four items), autonomy (two items) (Hackman & Lawler, 1971), participation (two items) (Lawler & Hackman, 1969), learning (two items) Sarason, et al., 1971) and information (three items) Sarata, 1972). Also, an additional context area of social contact (four items) was included. The original instrument (21 items) was used with only minor modifi- cation to items. Basically, the items were changes to the extent that they referred to the I and R setting. For the six interviews which had independent ratings by the author and coresearcher, an interrater exact agreement of 74 percent was achieved. (The interrater reliability coefficient was r = .87. 68 These reliability estimates were generally comparable to the 87 per- cent exact agreement found in a previous study (Sarata and Jeppesen, 1977). Scale construction for the JD was suggested by the correla- tion matrix of all 21 items minus three items (source of task identity, autonomy in ordering work tasks, and feedback on client referrals) which met the high endorsement criterion. Two scales were formed. The most highly intercorrelated items from the content areas of Feedback, Participation, and Information were combined into a single scale, intraoffice communication (01 = 0.82). All Participa- tion and Information items and two items of the Feedback content area were included in this scale. Thus, the final scale included seven items. The second scale, social contact, comprised of two Contact items was formed on the basis of high interitem correla- tion (r = 0.70, p < .002). After these scale constructions seven items remained. Since no consistent patterns Of intercorrelations were observed, these items were retained as singlets. Table l of Appendix B, details the above data reduction information on the Job Design measure. 4. Director/Outreacher Exchange questionnaires, D/OE. The D/OE was basically designed to measure the difference between the I and R director's perceptions and the outreach worker's perceptions Of three functional job component areas: feedback, participation, and information. Levels Of these job components seemed logically tied to their salience as components of exchange between the director and outreach worker. Thus the D/OE questionnaires were 69 designed to index the degree to which the director and outreach worker agreed on the importance Of feedback, participation and information exchange. For this questionnaire director and outreach worker(s) at each I and R center were asked to answer questions about a) how much participation, feedback, and information existed for the out- reach worker in exchanges with the director and b) how much Of each Of these types of exchange were needed by the worker. Two forms Of this questionnaire were devised. The outreach worker form inquired about the amount Of each type of exchange the worker had, and needed while the director's forms asked for ratings Of how much of each type of exchange the individual outreach worker had and how much the director thought the outreach worker needed to dO the job. Thus, while outreach workers only filled out one form, regarding themselves, the director completed a form regarding each outreach worker at the center. From the ratings of participation, feedback, and information by both outreach worker and director the absolute value of the difference between outreach worker and director the absolute value of the difference between outreach worker and director on each question was computed. Thus, an absolute value of the difference was considered an index of agreement between director and outreach worker. Then, with these computed indices of agreement a scale com- ,biliing them was devised. First, all ten items were combined in a single scale. Items 1 and 2 had considerably lower corrected item- total correlations than the other items inthe scale, and, when 70 deleted improved the coefficient alpha from 0.88 to 0.92. The eight-item scale, agreement about intraoffice communication, was established as the final version. Subsequently, the two deleted items (1. agreement on actual outreach worker participation in work decisions and 2. agreement on needed outreach worker participation in work decisions) were retained as singlets since they were not significantly correlated with each other (r = 0.37, p < .085). Table 1, of Appendix B, provides a summary. 5. Interorganizational Interaction Questionnaire, 110. The basic format of this questionnaire, was devised by Louis Tornatsky (1974) and was used by Tornatsky and Lounsbury (1974) in a study of local geriatric and juvenile delinquency agencies. In their study, perceptions Of communication and organizational similarities were found to be important correlates of actual interagency interaction. This measure was included in the current study in order to examine the bearing I and R staff perceptions Of interorganizational inter- action with the Department Of Social Services and the Social Security Administration might have on service to the elderly clients. (The reader is reminded that elderly clients were selected from the ser- vice rolls Of these two agencies.) The questionnaire was modified for use in the current study in two ways: a) only selected questions were used and b) the list of agencies to be considered for each question was changed to include only the local DSS and SSA Offices. The basic question format was retained; the instrument was modified to index inter- actions between the staff of the I and R center and these two 71 agencies. Items asked about ways in which communication took place (items 1 and 2), degree to which referrals were provided (items 3 and 4), degree to which funding was provided (items 5 and 6), degree of goal similarity (items 7 and 8), degree Of task and service similarity (items 9 and 10), importance Of other agency's services (items 11 and 12), and knowledge Of other agency' 5 programs (items 13 and 14). Using a Likert-like scale checklist format, respondents indicated, for the local DSS unit, then for the local SSA unit, the levels of each of the seven elements listed above. Thus, for each of the seven elements two distinct answers were given. The general data reduction strategies previously outlined when used on the 110 data resulted in four distinct scales. Since each item had two answers, (one referring to DSS, the other to SSA) an attempt was made to generate a set of parallel scales, thus retaining distinctions between DSS and SSA. However, since the DSS and SSA items were highly intercorrelated,the final scales incorporated both answers to each question. The first scale interorganizational exchanges, which included responses to items 1, 2, 3, and 4, had a coefficient alpha Of .84. Items 11, 12, 13, and 14 comprised a second scale interorganizational awareness with a coefficient alpha of .79. The last two scales each contained two items. Items 7 and 8 simply became the similarity Of gpals scale (r = .88, p < .001) and items 9 and 10 resulted in the similarity Of tasks and services scale (r = .75, p < .001). Finally, item 5 (funding from DSS) was retained as a singlet, and item 6 (funding 72 from SSA) was omitted on the basis Of the high endorsement criter- ion. Table l, Of Appendix B, provides the compiled scaling solutions for the 110 measure. TO summarize, scales derived from the FSM, NL, JD, D/OE, and 110 measures, along with the remaining singlets, comprised the final elements of the organizational/interorganizational data set. Table 2 indicates all the final scales and items resulting from data reduction of this data set. Thus, nine scales and thirteen singlets resulted from data reduction. In the following section reduction of the elderly client data set is discussed. Elderly client data. Data on the elderly clients in this research came from five instruments. The first three to be dis- cussed did not require scaling since data from each Of them provided one summary variable. Information on how data was derived from these three measures is covered first. An overall scaling table such as the one presented for organizational/interorganizational data above is given in the summary following discussion of the elderly client measures. 1. Phone Call Format Sheet, (PCFS) and 2) Outreach Visit FOrmat Sheet, (OVFS). The PCFS and OVFS acted as scripts for the I and R staff member during the initial contact but they also pro- vided data as to how much information the elderly client received and whether or not the client enrolled with the I and R center. Thus, the highest rating was given to an initial response by the 73 TABLEZ FINAL DATA REDUCTION SOLUTIONS FOR EACH MEASURE OF THE ORGANIZATIONAL/INTERORGANIZATIONAL DATA Measure Scale Name Scale Items Singlets 'Faces‘ Satisfaction Measure (F. S. M.) Negotiation Latitude (N. L.) Job Design Interview (J. 0.) Director/ Outreacher Exchange Questionnaire (D./O. E.) InterOrganizational InteractiOn Questionnaire (l. I. 0.) job satisfaction negotiation latitude intraoffice conmwication social contact agreement abOut intraoffice co-nunication inter- Organizational exchanges inter- organizational awareness similarity of goals similarity of tasks and services satisfaction with Job in general satisfaction with own agency satisfaction with coworkers satisfaction with types of clients worked with satisfaction with supervision satisfaction wdth pay satisfaction with promotion satisfaction with client progress satisfaction with I and R center services satisfaction with DSS services satisfaction with SSA services supervisor flexibility in negotiating changes in one's Job likelynood Of supervisor using influence to help employee frequency of periodic feedback from supervisOr frequency of feedback on individual clients degree of participation in decisions about clients degree of participation in decisions about program and policy amount of information received about AAA annunt of information received about other agencies ampunt of general information received number of different individuals from other agencies seen in a month number Of times individuals from other agencies seen in a week agreement on actual participation in client decisions agreement on needed participation in client—EEETsions agreement on actual feedback on work with elderly clients agreement on needed feedback on work with elderly clients agreement on actual information on the l and R center agreement on needed information on the l-ZTETTT center agreement on actual information on outsidETSEEncies agreement on needed information on Outsidi agencies number of ways of conlnunicating with DSS nunber of ways of cannunicating with SSA amount of referrals from DSS amOOnt of referrals from SSA importance of 055 services to seniors iaportance of SSA services to seniors amount known about DSS amount known about SSA similarity of goals to those Of OS similagity of goals to those of S similarity of tasks and services to those of DSS similarity of tasks and services to those of SSA satisfaction with the actual work frequency of general feedback from supervisor number of coworkers interacted with hOurs of face—tO-face contact per day number of separate tasks in the job rate of client contact degree to which job is perceived as integral autonomy in scheduling work time frequency of optional learning experiences frequency of learning about other agencies agreement on actual participation in work BeCiSions agreement on needed participation in work EEETETOns amOunt of funding from DSS 74 client including listening to the entire script and enrolling with the center. The variable was therefore an index of response level. 3. Outreach Response Sheet, ORS. When elderly clients contacted their I and R center,the staff simply recorded the date on the con- tact, and type of service they gave the client directly on the Out- reach Response Sheet, the ORS. If a client asked for and received service from the I and R center during the project period, it was recorded. Thus, the ORS provided a single dichotomous variable, outreach response indicating whether or not client initiated contact and service was given subsequent to the initial contact and prior to the project end (approximately one month). 4. Follow—Up Survey, FS. Demographic data on elderly clients contacted during the last part of the project was extracted from the Follow-Up Survey. Of the thirteen items taken from the FS eight appeared to be related to a general profile of self-help determinants for elderly clients. This rationally-formed scale was tested for internal consistency. Thus, the scale comprised of the items living alone, observed health, memgry, ability to read, social contact, contact with own children, education, and life satisfaction, yielded a coefficient alpha of .49. Corrected item- total correlations indicated an increase in the coefficient when living alone and contact with own children were deleted from the scale. When these two items were deleted, the resulting six-item scale achieved a coefficient alpha of .66. Since all these client characteritsics seemed to collectively be identifiable as assets which would influence ability to obtain needed services, the scale 75 was termed client ability, Other demographic items (type of assis- tance, age, sex, race, social service knowledge, along with the previously deleted livigg_alone and contact with own children) were retained as singlets since a consistent pattern of intercorrelations was not observed. Table l of Appendix B summarizes the above scale characteristics. 5. Senior Citizen's Opinions questionnaire, SCO. The final measure administered in the study, the SCO. was the only question- naire used for elderly client data. The SCO was administered to the elderly client so that in addition to whether or not they received service, their opinions (attitudes) about the key services agencies in thisresearch (the AAA, DSS, and SSA) could broaden the base of data on them. Accordingly the SCO asked the elderly client about the following: satisfaction with the AAA satisfaction with the DSS satisfaction with the SSA frequency of use of the AAA freguengy of use of the DSS frquengy of use Of the SSA intended use ofthe AAA intended use of the DSS intended use of the SSA OmflmthN—f e 0 e e e 0 e e a Thus, the instrument had nine items, three for each agency. The author wished to discriminate these items into two groups: 1) opinions about the AAA and 2) opinions about DSS and SSA agencies. (In this way the combination of DSS and SSA items in the SCO could be correlated with those items in the Interorganiza- tional Interaction Questionnaire (110), which similarly combined DSS and SSA items. Thus, the two items referring to the AAA (l and 76 7) were placed in one scale and items referring to the DSS and SSA agencies (2, 3, 8, and 9) were combined for a second scale. This rational approach to scaling provided two scales with fair internal consistency. The first scale, senior citizen's opinions about the .44 AAA_(with two AAA items), had interitem correlations of r (p < .OOl) and the second scale, senior citizen's opinions about DSS and SSA (with two DSS items and two SSA items), had alpha equal to .50. As a strategy to improve the statistically defined internal consistency of these scales the "frequency of use" items (4, 5, and 6) were deleted from the two scales. Subsequently, the three "frequency of use" items were retained as singlets, but were not reduced any further since separate scales for frequency of use of AAA and frequency of use of the other two agencies were desired but not supported by inter-item correlations. Table l of Appendix B details the scale characteristics. Thus, the elderly client data from the PCFS, OVFS, ORS, FS, and SCO measures reduced to three scales and twelve singlets. Table 3 shows these scales and items listed by measure. With these data reductions on the elderly client data and those previously discussed for the organizationallinterorganizational data the final analyses were completed. These findings follow in the Results section. 77 Pp cowuummmpuom mew— mmnmpzocx mow>com —mmuom copuouauo moo; uumucou "apoom xmm new; on xu.pwnu was xgosms Au__wnm A.m .mv mucaamvmma co max» zupoog um>cmmno acme—u xu>c=m aanxo_pom Aa_aaac_aaa A.m .a .ov oaagm mmcoamwc gummcpao Ampamu*~aaa uocv uocv onconmoz gouacuao cm .a .> .8 uomgm unseen upmv> guowcuao new 3 .a .u .3 Ampaou._anm uoogm unseen ~m>mp mmcoamoc Ampaau_paaa uocv uocv —pou mecca mumpocvm m5¢u~ opoum was: «Faun menace: 4mmo4m MI» mo manmcmm one pmcmmucw mxmmp mo xumcmpwewm am.- we um>wwucoa we con mmucmwcmnxw mcwccmmp we» sows: op magnet . m¢.- pmcowuao mo zucoaomcm Hm. mmo seem mcwucsc mo “cacao mm. sup: umuumcmucw pmcmmucp mcoxcozou mo guess: . mm. mm um>mmocma mp non mmmcosuxo we» sows: on mocmmo om.- chowum~wcmmcocmuc_ mm. now on» cw mxmou wumcmamm co Logan: . mo.- uuoucou pmwuom em.- new and cm momma mumcmamm mo conga: em. comuuommwumm now . mm. ouaumuap cowumwuoam: me. man: mpnuwcm> c mam: mpnmwcm> c «an: apnowce> L mcopumuwcassoo mummmomeucp meowumuvcassou mowcmoucucw ozone acmemmcmm cowuumwmwumm non mopnoeeu> aux mm4m 4ux wmmzh mam mzo~p<4mmmou zommPmucma my now ecu gave: on mmcmmu . ¢~.- m:o_pmowczseou new use =_ mxmma mowmmoccucw . mu. aumucou —wwuom . Hm. uomucou pmwuom . RN. mpmcmamm mo conga: . em.i mam: mpnmwca> c mam: m_amwcm> c mac: opamwcm> c «an: wpamwcm> c <<< on“ mo mmosu op mmg soc» mmmcocmzm momcmcuxm mmo mo mpmom mo xuwcmpwawm mcwvcac mo oesoEm pm:o_um~w:mmcocmucw _m:o_um~_cmmcocmu:_ mm—no_co> chopuu~vcomcocmucm Neuzv mm4m<~m<> zummmo moa oz< 4cmnzm now on» c_ mama» scum xumnummw mamgmamm mo Logan: - es. panacea mo zucoacmcw . mm.- mam: m_nowcm> c mam: «Pamwca> c mmmcmcmzm —mco_am~_ceacoemucw mmmcmsuxm pmcowumwpcmmcocmuc_ mmpnmwcm> Po:o_um~wcmmcocmucm Maize mm4m<~m<> zoummo moa az< 4 cmpmmu non mmpamwcm> =m_mmc non mm—nmwcm> :mwmmu no. e mwmmxcoz summeuso c m.cmxco3 zomocuzo c m.cmxcoz gummcuzo <<< on» we omega on Pocowumwwcmmcocmucm m.couumcwo Mouzv mmgm<~m<> zwmmmo mow m.mmxmoz zu 4 L can: mpnmwcm> c mam: opnmpcm> c wmcoqmmc condense pw>mp mucoamwc <<< we» «soon meowcpao m.cm~wuwo cowcwm mmpnowcm> xmx .1 Aeeuzv ma4m m w4m4mmoem >mx ummzh «on mzo~h<4mmmou zommeEec neueza eeem ceue— e um ve>esec neuez+ 96 ooo. o_ memmo.oe —oN~v. moomm. Foo—p.p obese; cewue_ecceo ouceewmwcowm .u.o eeoeem-wco mxpwz peewceceo eope>ceowm mmm. omomo. _Nomm.oo ooo. _o-v. moNPo. mega anew—o mom. omeoo. Fpeoe.oo ooo. o_-¢. Nemmo. moeepzecg eep>cem paweem ooo. oooom. mmemv.oo ooo. ummfiv. ¢_o~o. xuo Leo ueeuceo eueulep-eeeu we mcee: owe. PMMNm. mm~_~.oo ooo. .oMNe. ommpw. mmo we on: me xeceeeecu ooo.— oweoo.- comeo.mn ooo. ooeue. ¢m_oo. m:e_mweeo xcex co :ewume'epucoo eeeeez co aceEeeco_o moo. eomoo.e ooseo.o~ ooo. Nooom. oeono.— eeauoueo cewuewueoez oo_. moooo.p o_~mo.mo ooo. ~o_~m. moooo. pecoeuco ma ve>_eecea mo goo news: eu eecoeo ooo. mmomo.~ _m~oo.eo ooo. ~_o~m. mkeoe._ mucoumvmm< we oexp opp. “mmmv.~ oomo_._o ooo. _ooem. emo-.p cewueeocesseo eewmweecuco ueeo< aceseeco< mmo. oo_mm.e ~ooe~.wm ooo. ooomm. ammov.~ eo< aceppo FNo. owopm.m mv~P~.em ooo. «comm. mmoom.~ mcewm_eoo gee: c. :evueo_uwuceo veueez :e aceseeco<+ moo. coopo.o e_~oo.oe ooo. ommom. ooooo.e <<< ezu usene m:e_:woo m.:e~wuro cewcem moo. Noomo.o o~ooo.oe ooo. omxmo. moc~o.m xce: —o:ue< ego no.3 ce_uee»m_uem ooo. Peoe~.ep o_m_o.~m ooo. mmwoo. o_oo~.pp gov: emuecceuco mcexcezeo we geese: ooo. osmon.mp memo~.~p ooo. oomoo. osmo~.~p <<< we em: we xeceeeecu .owm > meem > memo .owm cease; e>esoc ce me_aaoce> cw emcage mx_ez twee“ 0e a umZOQmmm zo u:» «on mmmoum > mm¢zz=m a m4m=eowu eee. _em~e. eemmo.me e~_. cemNe. «em—o. new use =. momae ueacuaem co cease: ope. c.4mo. «Name.me mmo. emmNe. emewe. ee< one._e Na“. Aweeo. e_eem.~e ooo. mmeme. mmmme. «mm co an: ea Aegeaeecm Fee. ema_~. Amoem.~e ooo. meeme. emao_. new: eeeeaceeco meaxcozoe co cease: _mm. moomm. ,mmme.~e ~me. Peame. mepe_. xom eee._e mmm. .eeaw. mqamm.~¢ _~o. oemme. e_mm_. aqua—zoex ouoscem .aruom «Rm. mem_m. eem_P.~e ope. «meme. cepe_. co.eaeec=esoe advocate»:— eme. emwpe. memaa._e moo. memme. e_m_m. xe._,n< ecu._e ace. mmmmm. me~m_.Pq ooo. __eme. Noeaw. <<< eaoa< meeo=.ae m.=e~ee_e ce.=em MMN. wmmpe.p _emee.oe moo. Keene. mmeae. u=o_< m=.>_4 m.N. eaepm.p omaa~.mm Noe. wwaae. mmemw. mmo me as: me Auguseaca _p~. memem.F mm-~.~m Noe. A_eme. mao_e. mmececaz< _a=o_ea~_=aecoc~eco amp. eoqmm.~ m_¢ep.em Poo. memee. mom—v._ etc: .a:ee< age gee: ecceuacm.eam Pap. mmeo~._ mmomm.mm .oo. «meme. emmee.. _acmae=_ m< ea>_oecaa a. ace 28.5: 0» outage amm. oooo¢._ Nemmp.mm Foe. aepoe. _mmme. mmo sate a=.e==a co eeaoe< mac. om~e_.m NeN-.om ooo. ~oo0a. «ae~_.~ em< oce._e _~P. m_~e¢.~ Neemm.em ooo. e_m-. N~_me.o moeucem< canoe eaon< m=_=caae co xecmsemca mmo. ~m~e~.m me~m_.m~ ooo. eo_e~. m~o55.~ <<< co am: co xucaaeaca mwo. Nmam_.m __mmm._~ ooo. Necks. em_me.¢ cutep_;e use see: euaeeoe moo. ee_P~.~ maam~.e_ .oo. compo. weea~.e (mm es. mmo eaon< «co.eeao m.=e~ee.e co_=am ooo. ecemm.e oaemm.e moo. __eee. eaemm.e maecoccenxu e=.=caee ~a=o_eae co seceseece .orm > menu > meom .owm moose; e>eeec ce wepnaoce> c, unease mx__z cause ea c ‘14.]! I144} om>mo mmzoommm men zoohuzom hz mzh «Om mmmoum > mo<¢ oz< aceowm . . . memo uce__o emu. oommo. oNoNo.oo ooo semnv mmmmo eo< aceppo om“. oommp. mmoum.o~ ooo. oNMNe. oeeeo. eoeuvuao cepuewuemez oso. oueNP. mmmmo.os ooo. e_e~e. omeoo. euceumpmmm ce weak mom. ommom. moofio.o~ ooo. moexe. _ommp. :evumu.::EEeo.eepeoeececo com. ookmo.p emmpmflmm ooo” moose” mmooen Au¢_,e< u:e__o omm. meoeon momm~.o~ ooo. opouv. oseum. mcewmweeo gee: co :e_ueo.u?uceo eeeeez co aceEeeco< emo. _mooo m oekom “A ooo mmmme mmmp_ N esph gee: ocwpeoegum co xseceee< om—. oeom~.~ ooo-.- ooo. Fmooe. eoNoo. _e>eo emceemex ~o_. oeoom.~ _eo~o.eo ooo. omuom. m~_m_.p e e c m me e ounce e c e: Nee. 33: 238.3 ooo. N33. 3:: n A. 5 H e. h e «we? mums.“ ono. “Romp.m om__o.eo ooo. nmomm. m-_m._ cem _e_eem «mo. Nopko.e o-_~.~e ooo. eomoo. vm~o~.~ m e m c eP me e a a: e esee o Se. $8: 23%? 80. $80. 8854 c _ 68 e. o: S c .e o 3.. a _ e < coxwm 2L.“ moo. oopom.o omom_.mm ooo” moNNoH Posmhuo eecuppgo czo goo: ueoecoo s—o. emmoo.m oommm.m~ ooo. moons. ~m_vm.e Lemw>ceesm seem xecoeeeu Pmceceo we xeceaoecu Poo ummoo op emooo mp ooo comm“ Npmmm o <<< me em: ue xeceaeecu woo. -~o_.m -~o_.o moo. mofioo. usmop.o .o_m > menu > menu .o_m cease; e>eEe¢ Le me_ae—Le> E 855 3:; .85 3 o EEQWTIT mp m4m

mzh «Om mucoum > mom<12=m 102 Eight of the variables entered achieved the .05 level of signifi- cance and were therefore included in the description of the function. They were: frequency of use of AAA (client variable) frequencyof;general feedback from sppervisor (staff variable) contact with'own_Ehildr n (client variable) ss5_of client (client variable) agreement on actual participation in work decisions (staff variable)’ social service knowledge (client variable) perceived amount of funding from DSS (staff variable) frpquency 6f_use of SSA (client variable) I071 monoch- Then, using the initials above to designate each variable, the following equation was defined: ? = -.28(A) + .59(B) + .55(C) - .56(0) - l.0l(E) -.36(F) + .l3(G) - .60(H). (standardized scores and coefficients were used) Tabled means for the high group and low group on senior citizen's opinions about the AAA were used to determine the directionality of the eight significant independent variables. This function suggested that elderly clients had more positive Opinions about the AAA when they had more frequently used the I and R center and Social Security Administration, knew more about social services, were female, and had fewer contacts with their own children. Further, the function showed that when the I and R staff member who contacted them had less frequent general feedback from her super- visor, tended to be more in agreement about actual participation in work decisions,and perceived less funding coming from DSS,these client opinions were more positive. lO3 Prediction results for this equation showed 86.5 percent correctly classified. The Chi-squared statistic comparing actual group membership to that predicted is shown in Table 14 below. For this equation the improvement over chance prediction (52.3% correctly classified) was 34.2 percent, an amount roughly equal to the two previous prediction equations. TABLE 14 CAPABILITY OF THE DISCRIMINANT FUNCTION T0 PREDICT CLIENT OPINION OF THE AAA Actual Group Predicted Group Membership Membership 1 2 Low 1 38 7 45 Opinion (84.4%) (15.6%) High 2 3 26 29 Opinion (10.3%) (89.7%) 41 33 74 x = 39.405 0 < .000 104 Prediction of senior citizen's opinions about DSS and SSA. The final discriminant function analysis involved the opinions senior citizens had toward DSS and SSA agencies: This variable was dichotomized by recoding the original coded values to either a positive or negative group depending on which side of the median value the individual's response fell. The same independent variable list used for discrimination just described was employed for this analysis. Results are reported in Table 15. The first two variables listed in Table l5 achieved the .05 level of signi- ficance and therefore were selected as entries for the prediction equation. In order of selection they were: A. freqpency of learning about other agencies (staff variable) 8. client ability_(client variable) Thus, the discriminant function, employing the letter symbols above, was as follows: if: -.27(A) + .46(B). (standardized scores and coefficients were used) The tabled means for the high group and low group on senior citizen:s_ opinions about DSS and SSA were used to determine the directionality of the two significant independent variables. Thus, from this examination, a client was more likely to have positive opinions about DSS and SSA agencies if that client was less able to provide self-help and if they were contacted by an I and R staff member who had more frequent learning experiences with other agencies. eeum mpg» an ee>eeec "eueza neon copep e as eo>eeec Heuez+ 1(35 wee. mo amoem.a~ cameo. momma. eNmem. cease; :ewuepecceo wucoucwcmmm .ud 0.5035120 9.2.: 33:23 mapa>cmmvm Nme. eeame. Pmeem.mm Nae. cameo. o_N_o. cemosceaam sate xaaneeea Facecae co xecaaeace 0mm. momma. mem_m.mm omo. cameo. Nee—o. mmo me «a: co zuemaeacu «as. eeeoo. maoma.am mmo. memae. meame. woea_zo=x weescem _a_uom ooo.. moose.- Peeam.am mmo. “cove. mmoeo. mmoeacaz< .ocoeeaaeeaacocoe=_. wee. e~_.~. momam.om one. eeeee. emeop. ae.p ace: e=__=eagum co xsocoea< .me. ~p_m~. meme..mm mwe. aeaee. «mom—. «on: eeeo_e wow. mammm. ~a~mm.mm a.e. «cove. «weep. new we» no memap oeacaaam co cease: mam. eeeem. neemm.mm ooo. .e.me. eeee_. co.eeacm_eam nee NMN. emome.F eo~e~.mm moo. ammme. me_ee. <<< me am: co xecuaeoce Non. eeome. _meme.em ooo. ~e_ee. eveae. .aucaa no gee £0.53 a» autumn Awe. ~moom.e o.omm.mp Noe. -_am. em~_m.¢ »e__ee< ecu__e moo. mmoee.m Nmoee.e coo. _m~oe. ~moe0.e meeecam< canoe eaoe< ee_=taoo co Auceaeeca .orm > menu > menu .owm cease; e>esec ce wepaeoce> c. manage we.e: twee“ cu m 4mm oz< mmo - rDom< mzoozoeo m.zMN~»~u moozwm «cu zoohuzog hz oz» «Om mwaoom > mo1<£2=m m— m4m