ABSTRACT A STUDY OF DEVELOPMENT PREFERENCES AND SOCIOECONOMIC STRUCTURE IN THE RESOURCE CONSERVATION AND DEVELOPMENT PROGRAM: A COMPARATIVE ANALYSIS By David George Carvey The Resource Conservation and Development program depends on local participation and support. Without these the program could not function. The objective of the program is to help improve economic and environmental conditions in rural areas by offering local residents the opportunity to identify problems, evaluate needs, propose remedial ac- tions, set priorities, and initiate actions. The program's openendedness allows them to change their minds and plans. In view of the program‘s expansion over the years, encompassing over one—third of the counties in the U.S., and in view of the variation in program success in widely different physical, institutional, social, and economic settings across the Nation, research of program response and factors affecting decision making by local volunteer participants was undertaken. Comparative analyses were used to test the following hypotheses: 1. Direct relationships exist between program response or actions and local development proposals, suggesting one indicator of program effectiveness; David George Carvey 2. In specifying development preferences, the behavior of local decision makers is closely associated with socio- economic attributes of the RC&D projects in which they live; 3. Relationships between development preferences of local decision makers and socioeconomic attributes of their respective project areas will differ over a range of development alternatives. A system for classifying development proposals and actions was constructed using mutually exclusive categories based on primary deve— lopment intentions. Records of 48 RC&D projects were examined. Propo- sals made and actions taken between 1963 and 1970 were classified. Statistically significant association was found between rankings of prOposals and actions. This suggests that the RC&D program seems to be consistent in reflecting locally specified development preferences. Factor and discriminant analysis techniques were used to examine important linkages between shifts in local development preferences, as actions were taken on proposals, and socioeconomic structure as represented by 76 socioeconomic variables. The conclusion is that man's views, as represented in his group decision making, seem to be distinctly influenced by his surroundings as defined by socioeconomic structure. Analyses also showed that the relationships between basic elements of socioeconomic structure and local development preferences varied considerably for a wide range of development activities. Major predictors of shifts in development preferences were found to be those aspects of structure concerning health and education finances, the David George Carvey minority aspect of other rural-farm population, banking deposits, non- resident workforce, education specialty, government debt compared to revenue, and measures of wholesale efficiency. Results of this research suggest that consideration be given to use of the classification system developed for categorizing RC&D activities; that additional effort beyond measuring program consistency be given to developing an efficient measure of client satisfaction for evaluating program effectiveness; and that the comparative analytical approach presented in this study be considered for application to the RC&D program for use in program management, evaluation, and planning. A STUDY OF DEVELOPMENT PREFERENCES AND SOCIOECONOMIC STRUCTURE IN THE RESOURCE CONSERVATION AND DEVELOPMENT PROGRAM: A COMPARATIVE ANALYSIS By David George Carvey A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Resource Development 1975 ACKNOWLEDGMENTS I thank Professor Milton H. Steinmueller for long years of coun— sel and encouragement. I hold his work in great esteem. I also thank the Economic Research Service, Natural Resource Economics Divi— sion, USDA for their support of this research and other asPects of my education. Though they may not fully realize their contribution, my children, Roger, Terri, Christine, and Cynthia, must be credited with cooperation and understanding. Their mother, my wife Nancy, must be credited for her great love and I love her. ii TABLE OF CONTENTS CHAPTER PAGE I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . 1 Problem Setting . . . . . . . . . . . . . . . . . . . 1 Purpose of Study . . . . . . . . . . . . . . . . . . . 5 Objectives ... . . . . . . . . . . . . . . . . . . . . 7 Hypotheses . . . . . . . . . . . . . . . . . . . . . . 8 Assumptions and Limitation , , , , . . . . . . . . . . 8 11. REVIEW OF LITERATURE , . . . . . . . . . . . . . . . . . 11 The Community: A Social Action Arena . . . . . . . . 12 Social Structure and Social Action: An Analytical Approach . . . . . . . . . . . . . . . . . . . . . . 14 Research Methods . . . . . . . . . . . . . . . . . . . 18 Literature Review Conclusions . . . . . . . . . . . . 24 III. RESEARCH PROCEDURES . . . . . . . . . . . . . . . . . . 25 Data Requirements . . . . . . . . . . . . . . . . . . 26 Statistical Methods . . . . . . . . . . . . . . . . . 34 IV. COMPARATIVE ANALYSIS . . . . . . . . . . . . . . . . . . 39 Resource Development Preferences . . . . . . . . . . . 39 A Measure of Program Effectiveness . . . . . . . . . . 41 Project Analysis . . . . . . . . . . . . . . . . . . . 46 Development Preferences Analysis . . . . . . . . . . . 54 Comparative Analysis in Retrospect . . . . . . . . . . 62 V. SUMMARY, CONCLUSIONS, IMPLICATIONS AND RECOMMENDATIONS . 66 Summary . . . . . . . . . . . . . . . . . . . . . . . 66 Conclusions . . . . . . . . . . . . . . . . . . . . . 68 Implications . . . . . . . . . . . . . . . . . . . . . 70 Recommendations . . . . . . . . . . . . . . . . . . . 75 Summary of Recommendations . . . . . . . . . . . . . . 77 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . 78 APPENDIX A: METHODS . . . . . . . . . . . . . . . . . . . . . . 36 APPENDIX B: TABLES . . . . . . . . . . . . . . . . . . . . . . 92 10. ll. 12. 13. 14. LIST OF TABLES Classification System for Cateogorizing Development Activities . . . . . . . . . . . . . . . . . . . . . . . Socioeconomic Variables Selected for Structural Analysis . Resource Conservation and Development Projects . . . . . . Overall Development Preferences by Percentage and Rank . . Distribution of RC&D Project Development Emphases . . Rank Correlations for Proposals and Actions Within RC&D PIOj ects O O O O O O O O O O O O C O O O O O O O O C O 0 Listing of Socioeconomic Factors, Major Components, and Loadings I O I I O O I O O O O O O O O O O O O O O I O 0 Primary Development Classification Determinants and Their Rankings 0 O I O O O O O O O O O O 0 O O O O O O O O O 0 Development Classification Determinants and Their Rankin 88 O O O O O O I O O C 0 I C O O C O O O O O O O 0 First Entry of Each Socioeconomic Factor in a Discriminant PrOblem O O O O O O O I O O O O O O O O I O O O O O O O Discriminant Function Coefficients Showing Relationships Between General Development Directions and Socioeconomic Structure 0 O O O O O O O O O O O O O O O I O O O O O I Discriminant Function Coefficients Showing Relationships Between Human Resource Preferences and Socioeconomic Structure 0 C I C O O O O O O O O O O O O O O O I O O O Discriminant Function Coefficients Showing Relationships Between Natural Resource Preferences and Socioeconomic Structure 0 O O O O O O I O O O O O O O O O O O O RC&D Projects Factor Score Matrix . . . . . . . iv PAGE 30 32 4O 42 43 45 48 58 62 64 92 93 95 97 FIGURES PAGE 1. Primary Soil Conservation Service and Local Client Relationships in an RC&D Area . . . . . . . . . . . . . 4 2. A Profile of RC&D Measure Plans and Actions for the Northwest Michigan RC&D Project . . . . . . . . . . . . 72 3. Factor Score Profiles for Selected RC&D Projects . . . . . 73 CHAPTER I INTRODUCTION Problem Setting In recent decades, several efforts have been made to stimulate community improvement in rural areas. Several versions of community- oriented development programs have been tried which incorporated various levels and degrees of community involvement and responsibility.1 Some critical questions facing administrators of those program adaptations concerned the desirability and consequences of local involvement and the strategies and policies for insuring a level of local participation consistent with program needs.2 This concept of development policy based on the precept that community involvement must be contingent on program needs is one indicator of why a national policy of intervention has conceived and given birth to relatively short run programs. Feder- al program designers seem to have stressed community adjustment to program requirements. This is a major failing of an intervention poli- cy. While an overall policy to intervene to change the deterioration rate of rural areas is worthy, the objective of community development requires a coordinated, functional approach including a process of 1James L. Sundquist and David W. Davis, Making Federalism Work, (The Brookings Institution, N.W. washington, D.C.: 1969), Chapter 5. 2Kenneth P. Wilkinson, "Special Agency Program Accomplishment and Community Action Styles: The Case of watershed Development," Rural Sociology, XXXIV, (March,1969), p. 29. comprehensive planning and action embracing a wide range of community shortcomings; mobilization of resources of many agencies, public, pri— vate, federal, state, and local; and vigorous leadership with more extensive citizen participation.1 Successful programs, i.e., specific guidelines for community adjustment, involvement, and adherence, must at least allow collaboration and cooperation by the communities and program administrators. Program needs should not be weighted as heavily in development efforts as the community viewpoint. A key to effective, community-oriented, development programs is a concern for a level of local involvement consistent with community needs and objectives. Community involvement forms the basis of a current rural develop- ment program under the auspices of the Department of Agriculture's Soil Conservation Service (SCS). The Resource Conservation and Deve- lopment (RC&D) program was suggested in the Agriculture Act of 1962 and later defined and authorized in November of that year.2 An RC&D project has been defined as: a locally initiated, sponsored, and directed project designed to carry out a program of land conservation, land utilization, acce- lerated economic development, and reduction of chronic unemployment in an area where these activities are needed to foster a local economy.3 Thus, local people participating in the RC&D program are described as main ingredients for successful analysis, planning, and action processes vital to a well-rounded program of community improvement. 1Sundquist and Davis, op. cit., p. 131. Secretary's Memorandum Number 1515, U.S. Department of Agricul- ture, November 2, 1962. 3U.S. Department of Agriculture, Soil Conservation Service, Resource Conservation and Development Projects: RC&D Handbook, Washing- ton, D.C., 1972, Sec. 100.2b. In searching out means for improvement of economic and environ- mental problems, project sponsors and other citizens voluntarily participating in the program are faced with responsibility for the decision making process. This includes identifying local problems, specifying solutions by formulating and submitting proposals for deve- lopment measures, and setting local development goals and general priorities. Local citizens are also required to make final decisions as to specific priorities for seeking action on proposed development measures. A generalized view of SCS and local responsibilities is shown in Figure 1. Each RC&D project is required to complete a project plan summar- izing proposals for development actions to alleviate local problems. Development goals are established in this manner. These development proposals represent one measure of development preferences of people participating in the program. A second measure of local preferences is suggested by program response through recorded actions. Annual progress reports for each project list all development proposals for which actions have been taken, but not necessarily completed. Comparison of proposals and actions should reflect one measure of program effective- ness, i.e., consistency, in working toward the general development goals of the RC&D clientele which is the local people and their communities. The RC&D program has been operating since 1963. Federal policy is that the program be extended where needed, given the local leadership to effectively plan and implement activities necessary to achieve the goals of the program.1 While the USDA encourages local volunteer leadership to take an orderly, coordinated, natural resource-oriented 1Ibid., Sec. 100.2c3. mmfioamwm wofiumummooo I “concavuooo uumnoua I manoma Hwoofi I mmaufiuowum w mmfiofiaom onwahmumv I mamow hmaomam I mamanoum mMHucouw I mmmooum coauo< mwouufiaaou mmmuuwaaoo + wswummum mousommm umwmaomam monommme coauom I mmawon umsd0wum>ummcoo uofiuumwa ooHuu< now coamwomo Hmswm ofiapsm manfiwwam I fil. “H + muomcoam PI woumafiwuooo uoomoum Qwom mowuauowua w .mamow .mamanoua .muamaumnxm .:0Huma umficoaum>ummcoo mmu< Iuomcw :« mmwcmzo ou umsfium I soaumauomca soc mumafiawmmm I mofiuwuofium a .mamom .mamapoum uooawmu I mmum moammo mom mumum mmHoHHoa umuaumuca I w mmwuasoo awom mo AmouuHESoo wcwummum mcowumasooa I mmum uov Hwocooo awom mamuamfiao muw>umm coaum>ummaoo HHom AoneHHumm cowum>ummaou Hwom humeum .H oudwfim approach to improving economic and evironmental conditions, there is no provision for comprehensively judging the merits of the program in relation to its contribution toward facilitating and effecting a full range of local development preferences. There is also no way in which to judge the effects of variation in the socioeconomic makeup of RC&D projects and the influence that this variation might have on local deve— lopment preferences. Finally, there are no means for evaluating the implications of structural and preference variation on the administra— tive and operational aspects of the RC&D program. A basic problem underlying this research is that of a growing program to encourage development in primarily rural areas without appro- priate methods of program analysis and guidelines for program growth. Appropriate information concerning development response to local preferences through the RC&D mechanism, is not readily available and that which exists is slanted toward natural resource aspects of RC&D activities. Many resource development activities being reflected in locally conceived project plans deal directly with improvements of the human condition or social well-being but are not visible in the report- ing format currently in use inthe RC&D program. In this situation, there is no base for properly analyzing and evaluating either the federal policy for aiding qualified rural areas or the development mechanism for implementing such a policy. Purpose of Study The purpose to which this research is directed is that of formu— lating a systematic approach to the analysis and evaluation of the RC&D program in relation to program response to local preferences for development and also in relation to determinants of local development preferences. Major aspects of the research problem can be specified by these questions: 1. Is the RC&D program reflecting development preferences held by local citizens participating in planning and other decision making? 2. To what extent is variation in the socioeconomic structure of RC&D projects associated with variation in local development pre- ferences? Of prime concern in this research is the development of a syste- matic capability to consistently categorize local development preferences suggested by RC&D proposals and resultant actions. A second target concern involves examining relationships between inherent, socioeconomic attributes of RC&D projects and local patterns of development prefer— ences specified within the context of the RC&D program. Development of empirical models to classify RC&D projects with respect to changes in development preferences is desired. A final concern deals with the interpretation of research findings for the purpose of encouraging a more comprehensive analysis and evaluation of the RC&D program. Further, it includes the identification of useful analytical approaches and tools so as to allow their consideration for future applications with regard to the RC&D program or in any feasible situation where it is useful to compare variation in socioeconomic structure to variation in some dependent variable. This research should not be construed as an attempt to evaluate RC&D effectiveness as might be accomplished by analysis of capital investment or employment efficiency stemming from development activities. The only relevant measure of effectiveness pertinent to this study involves the concept of comparing program response to clientele preferences as revealed by development proposals and actions within the RC&D context, i.e., response consistency. Objectives Meaningful analysis of RC&D activities must take into account an accurate classification of RC&D activities, their variation between projects, and their variation over time, i.e., variable preferences. The following objectives were formulated to strengthen the data base describing RC&D activities and to examine the role of socioeconomic structure in relation to these activities. The primary motive is to enhance the understanding of the program thereby facilitating improve- ment in overall program administration (management, analysis and evalua- tion, policy, and program planning) as well as in coordination of and participation in activities at the local level. These objectives are: 1. To develop a consistent system for classifying local resource development preferences suggested by local proposals for deve- lopment and resultant actions. 2. To determine general resource development directions (human vs. natural resource) and specific development emphases of participating local citizens and to analyze the consistency of program response to development preferences. 3. To develop an analytical approach for examining socioeconomic structure and identifying those influences which seem to determine shifts in development preferences. 4. To examine implications of findings of this study for overall program administration (including planning and evaluation), coordination, and participation at the project level. Hypotheses The following working hypotheses have been formulated to assist in the accomplishment of objectives established for this study: 1. Direct relationships exist between program response and local development proposals, which suggest one measure of program effectiveness, i.e., consistency. 2. In specifying development preferences, the behavior of local decision makers is closely associated with socioeconomic attri— butes of the RC&D projects in which they live. 3. Relationships between development preferences of local deci- sion makers and socioeconomic attributes of their respective areas will differ over a range of development alternatives. Assumptions and Limitations Limitations of this research and resultant findings can be better understood by examining the assumptions which facilitated the analysis of certain aspects of development preferences within the RC&D program and the exploration of their linkages with socioeconomic characteristics of RC&D projects. A primary assumption involves the propriety of using only deve- -lopment proposals included in project plans as baseline measures of local preferences, as opposed to looking at all proposals made through- out the operation of the project. It is assumed that the research should try to examine shifts in development preferences over time. This is achieved by comparing initial preferences (proposals) with cumulative actions over all years of project operation. This approach does not allow a periodic examination of the influence of time which could act as a dummy variable to express the effect of continued local involvement and the knowledge and experience gained by citizen volunteers in for- mulating new proposals and initiating actions. Although this is a serious weakness in the research the primary support for the assumption is the analysis of measures of initial and cumulative preferences. Cumulative data collection on proposals would entail a much longer data collection period involving indepth contact with all RC&D projects selected for study. Another assumption worth noting concerns the degree of importance placed on the measurement of cumulative actions. The actions reported in progress reports for RC&D projects do not constitute completed actions but only those which were initiated. An action can be dropped from a later progress report and would not be identified as such in this research. The justification for the assumption involves a need for a measurement of cumulative action preferences at the latest cutoff date, July 1970. In all probability there are many weaknesses in the proposed sys- tem for classifying resource development activities. There is an entire set of assumptions concerning the types of categories to be included in the system and the decisions necessary to insure consistent classifica- tions. In support of the general assumption that such decisions can be made so as to accurately quantify development preferences at two points in time, it can be stated that some dependence on mutually exclusive categories is necessary. In essence the classification system stands as the cornerstone to this research. Another general assumption concerns socioeconomic structure. Out of a multitude of variables which could be included in the exploration of socioeconomic structure only a relatively few are to be chosen. The 10 analyses which depend on and reflect the socioeconomic parameters in this study will be limited by the original variables chosen. The final selection of variables is to be based on the best current knowledge and experience available in the literature and on the judgment of the researcher. A final set of assumptions concerns the theoretical relationships of socioeconomic structure and with what it may or may not be linked. There is a movement toward general theories of the relation of community socioeconomic structure and decision-making.1 Recently, researchers have been proposing the hypothesis that socioeconomic structure can be related to specific "issue areas."2 In accordance with logical positi— vism in social science, it is suggested that logical consequences or conclusions derived from assumptions of a theory are subject to inde- pendent, empirical verification although the general theory may not be verified. Thus, the assumption is made that linkages between different "issue areas" in resource development can be found in and empirically described by socioeconomic structure. 1Terry N. Clark, Community Structure and Decision-Making: Compar- ative Analyses, (San Francisco: Chandler Publishing Company, 1968). 21bid., p. 67. CHAPTER II REVIEW OF LITERATURE An important aspect of this analysis of local development prefer- ences concerns the involvement of local citizens participating in a rural development program-—RC&D. Local participation is not perfunctory, it is central and essential to any accomplishments forthcoming through the program. Local citizens who reside within the established or proposed multi-county RC&D projects and wish to participate in decision making must first become volunteers. After accepting this commitment, they must then assume the responsibility for identifying conditions or problems which might be improved through the program, evaluating local needs, proposing means for achieving improvement or solution to identi- fied problems, and finally stimulating actions for implementation of their proposals. This sequence of grassroots involvement implies an ordering of local development preferences within the context of multi- group decision making. The process of social choice or decision making is therefore an integral element in the social action arena of the RC&D program. The concept of social action or collective response, in search of an acceptable mix of resource development activities to meet locally defined needs, establishes a relevant basis for examining literature concerning the community. Although RC&D projects often include many counties, the concept of community is still very relevant, as the intent ll 12 and purpose of social actions forms the basis for "community." A community has been defined as: a collective response to conditions of life in a given territory formed by the establishment of social action paradigms necessary to meet common needs of residence, sustenance, and other societal functions. In addition, the overwhelming importance of social choice or local decision making in improving conditions of life and surroundings added a second major element to the literature review-—decision making, not the process itself, but the outcomes. The third important element in this literature search stems from the RC&D program and how it has spread to embrace a wide range of types of geo-political areas characterized by variant attributes and variant resource development preferences. To gain more understanding of two variant entities, RC&D projects and development preferences, it is logical to pursue a more complete understanding of their individual variations and then search for meaningful associations between the variations. Thus, the review in its final stages turns toward litera- ture concerning analysis of structural variation in geo-political areas and the relationship between socioeconomic variation and variation in decision outcomes. The Community: A Social Action Arena There are many diverse views as to what the concept of community means. Reiss' view of community incorporates collective response in a social action territory based on common needs. A similar view is shared by Wilkinson who has outlined a definition of community which 1Albert J. Reiss, Jr., "The Sociological Study of Communities," Rural Sociology, XXIV, (March, 1959), p. 118. 13 he suggests as an aid in the process of clarification, examination, and understanding characteristic aspects of community. Wilkinson desires that community be recognized as a field of study and that it be clearly defined. His review of several sciences has provided a definition of field which he uses to define community. The attributes of Wilkinson's field or community are: a holistic, interaction nexus, i.e., internally interactional in relation to causes and consequences; unbounded but distinguishable; dynamic, i.e., contin- uously changing; and emergent, i.e., resulting from its own interactions.1 Both Wilkinson and Reiss recognize the validity of social interaction as a basis for community definition. Social action to achieve change is a basic tenet of the RC&D pro- gram. A crucial relationship between social action and change is reflected in Wilkinson's statement that: . . . the eternal fact of change in human societies is to be found in the gap between what people expect and wish to happen, and what actually transpires when they behave and interact with one another! Local residents must band together for the purpose of making decisions which will influence the magnitude and composition of the changes pre- ferred by these residents. Barkley and Seckler, in relating economic development to environmental decay, have testified to the significance of man's decisions or choices in effecting preferred changes. They recognize the complexity of the societies within which man must act and they suggest the consequential nature of mankind's choices as determi- nants of his environment and vice versa. They state: 1Wilkinson, op. cit., p. 31. 2Ibid., p. 32. 14 In sum, the environment of the human organism is a complex system of physical, biological, and social mechanisms that must contin- ually adapt to the consequences of man's choices. While man is unique in that he can significantly determine his environment, he is similar to other organisms in that his behavior at any point in time is highly constrained by the environment he has created. Choice not only determines man's immediate welfare, it also deter- mines the various options open to him in the future. The consequential relevance of mankind's choices within the con- text of his environment parallels the concept of consequential decision making within the context of the community. Social Structure and Social Action: An Analytical Approach The works of the aforementioned writers and researchers have been used to establish the important link between community, social action, and decision outcomes. In so doing, the process of resource development, through change based on interactions of community residents, has been reviewed. Other questions remain to be answered. Given the importance of the community and social action, i.e., decision making, what is known about their relationship? Do communities provide clues as to preferred decision outcomes? If so, do such clues exist for resource development preferences? Such questions suggest the importance of establishing a suitable framework for community-oriented research in relation to the process of development. In proposing a framework for community oriented research, Wilkin— son specified that comprehensive development and change requires coor— dination and social structure differentiation. Differences of values, ideas, and desires within a community must be viewed in some logical, organized manner, achieving an equitable degree of coordination. Such 1Paul W. Barkley and David W. Seckler, Economic Growth and Envi— ronmental Decay: The Solution Becomes the Problem, (New York: Harcourt Brace Jovanovich, Inc., 1972), p. 6. 15 coordination aids in the social process of community decision making. The ideal is movement toward a final decision or set of decisions spe- cifying a preferred resource use or allocation. This decision would perhaps yield some optimum level of social welfare or community bene- fits. In other words, differentiation and coordination within the context of social decision making and action facilitates desired changes within a society, i.e., community. When community decision making is directed exogenously, the thrust of change is most likely to be through intervention. This approach says that local values and desires must be altered for the sake of the program. Endogenous choice direction is more likely to be collaborative in nature, allowing a closer coordination between local planning and choice processes. Goodenough recognizes the value of local inputs as he writes that "the best customers for community deve- lopment are those with a need they are themselves aware of."1 To extend this toward Wilkinson's view, the degree of community involvement in decision making is a function of the importance of the need for material and human resources to obtain community development goals. It is also a function of the need to legitimize a development program or thrust. Community involvement hinges on the need to achieve a congruence with the values of democratic society. Although Wilkinson uses resources, legitimation, and value congruence as means to justify local participation in decision making within the context of a govern- mental program, the very same conceptual framework would hold for any community oriented development or improvement effort.2 That is, if 1Ward Goodenough, Cooperation and Change, (New York: Russell Sage Foundation, 1963), p. 309. 2Wilkinson, Rural Sociology, XXXIV, No. 1, op. cit., p. 35. 16 some local project is needed, e.g., library or sewer extension, local participation in decision making through bonding or millage votes or through public meetings would lend to the probability of success for the project. With regard to research of local participation in the local choice or decision making process Wilkinson says: Among the many variables to be considered as a state or federal agency plans a special interest, development program within a given local society, the one which has received perhaps the least atten- tion in research is the extent of local participation in decision making to be encouraged or permitted in the program. This statement should not be limited to a state or federal agency, but should include even locally oriented and initiated plans for improvement as per the examples of the library and sewer extension used above. However, while identifying the extent of local participation in the choice or decision making process as inherently important, Wilkinson fails to give recognition to the importance of community influences or attributes on community preferences. A. J. Reiss has written that community research generally fails to apply the scientific comparative approach and techniques of multi- variate analysis in their design and execution.2 Furthermore, he believes that there is no systematic approach to the study of diverse community problems, community attributes of these problems, and their community variation. He calls for the characterization of communities in terms lIbid., p. 35. 2Albert J. Reiss, Jr., "The Sociological Study of Communities," Rural Sociology, XXLV, No. 2, (March 1959), p. 126. 17 of their attributes and for comparative analyses to show how such attributes affect decision making in the community.1 Peter Rossi substantially agrees with Reiss with regard to the relevance of studies of the social environment of decision makers and the decision making process. However, he contends that it is also relevant to examine the characteristics of the decision makers them- selves and their relation to decision outcomes.2 He supports Reiss' comparative analysis position and argues that understanding of particu- lar decisions should recefve less emphasis than the understanding of tendencies within classes or types of decisions. He suggests compara- tive research of: decision makers of different types; different community and institutional settings; and a range of issues.3 The comparative, community research approach is also supported by Summers, Clark, and Seiler who recognize that we know a great deal about communities, but what we know does not add up to a coherent, systematic body of propositions, concepts, and explanations which can be recog- nized as a sociological theory of community.4 Comparative analysis seems to be a reasonable approach to understanding the community and inherent influences on social choice. 1Ibid., p. 129. 2Peter H. Rossi, " Community Decision Making," Administrative Science Quarterly, 1, No. 4, (March 1957), p. 415. 3Ibid., pp. 438-39. 4Gene F. Summers, John P. Clark, and Lauren H. Seiler, "The Renewal of Community Sociology," Rural Sociology, XXXV, No. 2, (June 1970), p. 218. 18 Research Methods Given a feasible framework for comparative research on community attributes and issues the question remaining is, what is relevant to know? Relevance, in this case, is anything that will aid the under- standing of the influence of community attributes on the choices, decisions, or tradeoffs communities must make to satisfy their demands for improvement through change. In response to the need for a scientific comparative approach and the use of more meaningful multivariate analytical techniques, the advent of the 1960's brought a flurry of comparative, structural studies directed toward multi-county, geo-political areas. These com— parative studies facilitated the identification of attributes which seemed to be most relevant to area differentiation. The basis of this new comparative thrust can be traced back to 1941 work by Hagood, Danilevsky, and Beum.1 In this work, factor analysis, a relatively new analytical technique in sociology, was used to group gee-political areas. Factor analysis was seen to be a valuable tool for exploring socioeconomic structure by reducing exceedingly complex relationships within a set of variables to more understandable proportions. This was similar to its role in psychology. Daniel Price recognized the value of factor analysis for compara- tive structural studies and in 1942 he published results of factor analysis of characteristics of 93 American cities with populations of 1‘Margaret J. Hagood, Nadia Danilevsky, and Merlin O. Beum, "An Examination of the Use of Factor Analysis in the Problem of Subregional Delineation," Rural Sociology, 6. (September 1941), pp. 216-233. 19 100,000 as of 1930.1 One purpose of his writing was to further expose the academic world of sociology to the concept and value of factor ana- lysis. Although he used static measures of metropolitan population and area characteristics, he recommended the use of measurements of changes in such characteristics. He suggested that such an approach would be meaningful in explaining and predicting social change. Although factor analysis continued to be used in comparative stu- dies in education and psychology in the next two decades, little pro- gress was made in advancing comparative community studies. Then in 1961, Johassen and Peres published factor analytic research which sought to simplify the complex structure of communities. Eighty-two census measurements (1950 data) for 88 Ohio counties were analyzed and reduced to seven basic elements characterizing differences between counties.2 Also in the early 1960's, Hadden and Borgatta published their par- allel factor analyses of census measurements on 644 American cities. These cities, with populations of 25,000 or more as of 1960, were grouped into eight combinations: all cities; four sets of cities grouped accord- ing to size; and three sets of cities grouped according to a location rule.3 Sixty-five census measurements were reduced to fourteen basic factors or elements of urban structure. The eight parallel factor analyses allowed comparisons of structural differences. 1Daniel 0. Price, "Factor Analysis in the Study of Metropolitan Centers," Social Forces, XX, No. 4, (May 1942), pp. 449-455. 2Christen T. Jonassen, "Functional Unities in Eighty-eight Com- munity Systems," American Sociological Review, XXVI, No. 3., (June 1961), pp. 399—407. 3Jeffrey K. Hadden and Edgar F. Borgatta, "The Factor Analytic Structure of American Cities," American Cities: Their Social Charac— teristics, (Chicago: Rand McNally and Co., 1965). 20 In a follow-up to the Jonassen and Peres work, Munson, in 1965, presented the results of a second factor analytic study of 88 Ohio counties, using 113 1960 census measurements.1 Munson, as did Jonassen and Peres, found seven basic elements of community structure, four closely paralleling the earlier findings: urbanism, socioeconomic level, population growth, and governmental expenditures. Munson tentatively suggested these may represent the most fundamental elements or dimen- sions of the community. In the late 1960's Bonjean, Browning, and Carter responded to the well documented need for comparative community research with their factor analytic study of all counties in the 48 contiguous states.2 They chose 79 census measures, 46 of which were identical or similar to those used by Hadden and Borgatta. They searched for refinements in the lists of variables forming the basic dimensions or elements of community structure. In their analysis, they found considerable parallelism with results of Hadden and Borgatta and Jonassen and Peres. They found 15 basic community dimensions, twelve of which compared similarly to those in the other studies. The thrust and progress in comparative research in community structure in the decade of the 1960's set the stage for a melding of two concepts--community structure and what it can say about community action. Green and Mayo, in their research of actions of community groups in the early 1950's, recognized that structural studies were lByron E. Munson, "Structural Analysis of the Community," Rural Sociology, XXXLLL, No. 4, (December 1968), pp. 450—459. 2Charles M. Bonjean, Harley L. Browning, and Lewis F. Carter, "Toward Comparative Community Research: A Factor Analysis of United States Counties," Sociological Quarterly, X, No. 2, (Spring, 1969), pp. 157-176. 21 fundamentally important but generally had not been very fruitful for predicting actions of organized social groups.1 Although predictive structural analysis was found useful in classifying individual behavior and its determinants very early in psychology; only in the late 1960's was comparative community research through factor analysis recognized as an integral aspect of predictive studies of social choice. In 1968, Kevin Cox published research relating the geography of political party preference and participation to various characteristics of the population of metropolitan London.2 Factor analysis was used to define socioeconomic dimensions used in the development of causal models of political affiliation and participation. Cox's work is particularly relevant because voter behavior is an essential element of policy formation. The advances in comparative research at the geo-political or community level urged political scientists into proposing new hypotheses concerning public policies and political system characteristics. Tradi- tional variables in widely different political systems, e.g., electoral and institutional circumstances, did not explain much of the variation in public policy. In 1969, Sharkansky and Hofferbert published com- parative research on the dimensions of state politics, economics, and public policy using factor analysis. They provided a basic statement upon which much current research is founded. They stated: 1James W. Green and Selz C. Mayo, "A Framework for Research in the Actions of Community Groups," Social Forces, XXXI, No. 4, (May, 1953), p. 320. 2Kevin R. Cox, "Suburbia and Voting Behavior in the London Metropolitan Area," Annals, Association of American Geographers, LXIII, (March, 1968), pp. 111—127. 22 . . our findings show that different social and economic charac- teristics have different relevance for policies, and their rele- vance varies between substantive areas of policy. This position recalls the Rossi position of the late 1950's that the understanding of tendencies within types or classes of decisions may be the key to understanding social action.2 Adelman and Morris' work in the mid-1960's preceeded the Sharkansky and Hofferbert research.3 It dealt with social, political, and economic relationships. They sought to understand developmental processes in under-developed nations by means of a factor analytic, comparative approach. Although their main objective concerned dimensions of economic development, they added encour- agement for further, important, comparative research at the are level. In 1967, P. T. Cox published comparative research findings from a study of small watershed developments in Oklahoma.4 This consisted of the usual exploration of a large socioeconomic data set and its reduc- tion to a small number of dimensions accounting for most of the variance in the original set. Cox, however, pushed further with his comparative studies, breaching the gap between techniques in community studies and psychological and educational techniques. He employed the results of factor analysis in the classification of watersheds by discriminant analysis. The combination of the two techniques, 1Ira Sharkansky and Richard I. Hofferbert, "Dimensions of State Politics, Economics, and Public Policy," American Political Science Review, LXIII, No. 3, (September, 1969), p. 867. 2Rossi, 0p. cit., p. 415. 3Irma Adelman and C. T. Morris, Society, Politics and Economic Deve- lopment: A Quantitative Approach, (Baltimore: John Hopkins Press, 1967). 4P. Thomas Cox, "A Sociological Analysis of Upstream Watershed Development in Oklahoma," (unpublished Ph.D. dissertation, Graduate College, Oklahoma State University, 1967), 141 pages. 23 factor and discriminant analyses, allowed the derivation of models which would predict classification probabilities for small watershed develop- ment based on socioeconomic data. In the late 1960's another Rossi position gained some support. Rossi firmly believed that the comparative research approach should not be limited to community characteristics. He asked that the decision makers, the members of social action groups, be examined in light of their decisions.1 In 1968, Kivlin and Fliegel published comparative research of Pennsylvania farmers suggesting that the way in which a farmer relates to his business may be at least as important as percep- tions and stimuli in accounting for their behavior in the adoption of agricultural technology.2 This work represents a part of the break- through into comparative research on decision makers. Then in 1972, Smith and Martin analyzed the association between socioeconomic attributes and the behavior of cattle ranchers.3 As in P. T. Cox's work, they applied factor and discriminant techniques. Findings included classification probabilities showing the degree of accuracy of the classification of ranchers based on their socioeconomic characteristics and views. 1Rossi, op. cit., p. 415. 2Joseph E. Kivlin and Frederick C. Fliegel, "Orientations to Agriculture: A Factor Analysis of Farmers' Perceptions of New Prac— tices," Rural Sociology, XXXIII, No. 2, (June, 1968) pp. 127-140. 3Arthur H. Smith and William E. Martin, "Socioeconomic Behavior of Cattle Ranchers, with Implications for Rural Community Development in the West," American Journal of Agricultural Economics, LIV, No. 2, (May, 1972), pp. 217-225. 24 Literature Review Conclusions Over time, the value of comparative research in community and group structure has been borne out. The technique of factor analysis, developed for description and classification in psychology, has proved invaluable in the study of structure in the community as well as group context. The comparative research possible with factor analysis, when combined with the classification capabilities of discriminant analysis overcomes many of the problems of deriving empirical models for explain— ing the behavior of social groups. Given this review, one conclusion is that comparative structural research should be performed on socioeconomic attributes of selected RC&D projects. Factor analysis should be used to reduce a large socio- economic data set to a set of basic dimensions summarizing or account— ing for most of the variance in the original data. This would identify similarities and differences between projects. A second conclusion is that empirical models should be derived which depict relationships between shifts in resource development preferences, i.e., decision maker behavior, and socioeconomic structure of RC&D projects. This can be done by discriminant analysis. The procedure is to group the RC&D projects according to their known shifts in resource develop— ment preferences and use this as a dependent variable in conjunction with each projects' measurements on elements of socioeconomic structure to find those equations which best reproduce the actual, known groupings. Resultant equations will depict relevant functional relationships between structure and various classes or categories of development emphasis. Such equations can be used to predict shifts in development preferences in new or proposed RC&D projects. CHAPTER III RESEARCH PROCEDURES This chapter presents an overview of data requirements and sta- tistical techniques necessary to carry out a comparative analysis of community socioeconomic structure and the identification of linkages between structure and tendencies for development preference shifts within certain types or classes of development decisions. In the review of literature, it was noted that concerted social action or collective response by representatives of a particular geo- political area defines that area as a community. Thus by definition a community is an arena for social action. The RC&D program with its emphasis of local, collective decision making fits well this definition of community. The literature review also followed the development of a comparative analytical procedure for examining socioeconomic structure of geo—political areas, i.e., communities, and suggested the need for this type of research in view of an apparent lack of solid theory of community.1 The review discussed many examples of comparative analyses and emphasized the relevance of factor analysis for exploring and defining important aspects of community structure. Relevant findings concerning the relation of variation in community structure to variation in sub- stantive areas of policy, (see Sharkansky and Hofferbert, et. al.), lSummers, Clark, and Seiler, op. pip., p. 218. 25 26 issue areas (see Clark), and tendencies within classes of decisions (see Rossi) are presented. Another statistical technique, discriminant analysis, was identified as a key to pinpointing the linkages between structure and development preference shifts in classes of decisions or issue areas. Two necessary types of data are essential to the comparative research proposed for this study. First, a set of data which clearly defines substantive issue areas or classes of decisions occuring within the RC&D program is needed. This set of data is developed by review of records of RC&D development preferences. Numbers and types of development proposals are categorized and counted. The same is done for actions initiated. The second set of data consists of an array of socioeconomic measurements of community structure. Selection of varia- bles to include in this set is based on the review of literature. Data Requirements To realistically evaluate program response to community needs as indicated by deve10pment preferences and to specify important relation- ships between socioeconomic structure and shifts in preferences, a consistent classification system is needed with which to categorize these preferences. The development classification scheme used by the Soil Conservation Service, USDA, in the RC&D program does not allow for clear, concise consideration of a wide range of community preferences, thus making the evaluative process uncertain. Activities covered by the SCS system are closely aligned with the natural resource-oriented program offerings. In addition, many development proposals offered and actions desired by local citizens do not fit into mutually exclusive categories and thus cannot be considered for analysis and evaluation 27 in terms of initial development preferences versus resultant program actions. Meeting the objectives of this research requires two basic types of data. First, data describing local resource deve10pment pre- ferences across 48 RC&D projects are necessary. Secondly, data describ— ing the structure or socioeconomic makeup of these project areas must be examined. Development Preference Data Local development preferences were obtained by reviewing project plans which specify development proposals and progress reports which specify resultant actions. Preferences determined in a context of local decision making have been found to embrace a wide range of activities including such things as further detailed studies of various proposals, requests for assistance from various agencies, planning and technical assistance, and cost sharing. It is assumed that an accounting of development proposals and resultant actions can provide enough data to evaluate the consistency with which the RC&D program responds to locally determined needs. Quantification of local development preferences requires a classi— fication system for consistently categorizing proposals and actions. Review of the SCS system revealed several major weaknesses which hindered attempts to adequately evaluate RC&D response in view of the implied comprehensive rural development mission of the program. The system used by SCS for classifying RC&D activities is shown below: Accelerated resource developments ‘Agricultural water management developments Recreation developments Wildlife developments Watershed projects (under Public Law 566) Water developments other than P.L. 566 28 Land and critical area stabilization Special resource studies and inventories Highways, scenic highways, trails, and roads Range improvement groups and associations Agricultural and wood using processing and marketing industries Other industries Public service facilities (hospitals, schools, sewage systems, etc.) Industrial parks Rural water lines Rural sewer systems Beautification Education measures Other measures not classified Accelerated soil surveys Accelerated conservation planning Accelerated land treatment Accelerated land conversion: cropland to grass and woodland One major problem presented by this type of framework is that some categories overlap. Examples are Accelerated land treatment and Land and critical area stabilization. Both deal with land and its treatment. Secondly, some categories are too general as exemplified by Accelerated resource developments and Special resource studies and inventories. Measures grouped by these categories would have no unity of intent. A third problem is that some categories are too specific. Rural water lines and Rural sewer systems are good examples. These types of pro— blems present serious dilemmas for program analysis which requires unity of intent in each category entering into the analysis. One attempt to achieve unity of intent through mutually exclusive categories is discussed below. A major study of 48 RC&D work plans was undertaken in recognition of the serious problems in the SCS classification system. The objective was to build a framework that would serve as a reliable classification instrument for any type of development proposal. To solve the problem of ensuring mutually exclusive categories, the basic intent or concern of each proposal was used as the primary decision criterion for classi- fication purposes. 29 Study of the work plans revealed two major areas of concern. Proposals which were directed toward improvement of the human condition were grouped together as human resource measures. Those proposals primarily directed toward improvement in the natural condition were grouped as natural resource measures. These two major groups are defined to represent resource development directions. Further categor- izations were made within the human and natural resource groupings using mutual exclusiveness and basic intent as the decision rule. The end result was seven categories in each of the two major groupings. Together these 14 categories provide the basis for evaluation of pro- gram effectiveness and for examining the relationships between socioeconomic structure and shifts in development preferences. The entire classification system is presented in Table l. A listing of basic concerns is provided for each category. These acted as the cri- teria for classifying proposals and actions. In summary, the 14 category classification system is designed for use in quantifying local development preferences as indicated by basic intent of proposals and as reflected by actions occurring within the RC&D context. In classifying a proposal or action, two questions were asked. Is this an attempt to improve a human or natural condition? Assuming a satisfactory decision on this, what is the basic intent of the proposal? This final decision serves to properly classify the proposal or action. Socioeconomic Structure Data Meaningful analysis and evaluation of the RC&D program will have to take into account variation between RC&D projects with regard to preferences and program response. One means of so doing is to relate 30 TABLE 1 Classification System for Categorizing Development Activities Human Resource Related 1. Education elementary and secondary college adult vocational Health and Medical Services medical personnel medical facilities medical programs Industry lack of management personnel lack of development Employment low wages lack of job training seasonal work lack of industry and business Transportation highways and roads harbors and channels rail facilities air facilities Housing shortage dilapidation presence of vacation housing housing development controls Community Facilities and Services water supply and distribution systems police and fire service urban improvements business services historical and cultural improve- ments sewage and treatment and disposal systems Source: Natural Resource Related 1. Environment air pollution loss of natural beauty changing land use Land erosion lack of soils data land development Water pollution flooding drainage Agriculture management farm size and ownership land use and treatment marketing Forestry management timber quality and species marketing land ownership Recreation management land and water use conflicts public access underdevelopment overdevelopment financing Planning and Development comprehensive planning . land use planning development controls or guidelines Developed for this study from a review of 48 RC&D project plans. 31 variation in the makeup of RC&D projects to variation in shifts in local development preferences. Area analysis is implied. Insight into patterns of deve10pment which reflect local needs requires research of the involved areas. Some researchers stress the need for examination of variation in community attributes which affect human behavior and decision making while others call for indepth research into the charac- teristics of decision makers, their social environment, and the decision process itself.1 Most researchers, however, recognize the importance of relevant community attributes. This study focuses on such attributes and their association with resource development preferences within the context of the RC&D program. Review of literature pertaining to voter preferences, community and regional structure, and decision making suggested a wide range of variables that could be valuable in examining the association between development preferences and socioeconomic structure. In all, 76 varia- bles were chosen for the structural analysis of the 48 RC&D projects in the study. All are displayed in Table 2. These county-level census measurements were collected for all of the 297 counties included in the selected RC&D projects. All variables were transformed to represent multi—county attributes in accordance with project boundaries. Percen- tages, rates, and averages were used whenever possible to partial out any dramatic influences of size of raw data figures. This procedure was followed as past studies have shown that the amount of variance of a variable may be a direct function of its size. While rates and averages may have little or no relationship to size, they are useful 1Reiss, op. cit., p. 119. Also see Peter H. Rossi, "Community Decision-Making," Administrative Science Quarterly, 1, No. 4 (March, 1957). Number #WNH H OOCDVO‘U‘I ll. 12. 13. 14. 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 32 TABLE 2. Socioeconomic Variables Selected for Structural Analysis Description total population in 1960 Z population change 1950 to 1960 Z population change due to migration, 1950 to 1960 Z population change due to natural increase 1950 to 1960 population density, 1964 change in number of families, 1950 to 1960 NNNNNNNNNNNNNN population voted in 1960 urban population, 1960 population population population population population population population population population population population rural-farm, 1960 rural-farm white, 1960 rural-farm negro, 1960 rural-farm other, 1960 in group quarters, 1959 minority, 1960 foreign born, 1960 foreign stock, 1960 under 5 years, 1960 over 65 years, 1960 21—65 years, 1960 median age of population, 1960 change in median age, 1950 to 1960 Z population of voting age, 1960 per capita income, 1959 median family income, 1959 Z change in family income, 1949 to 1959 Z family incomes of $3,000 or less, 1959 Z family incomes of $10,000 or more, 1959 number of cars per capita, 1960 Z NNNN NNNNNNNNN population 5-34 years old in elementary school, 1960 population population population population 5-34 years old in high school, 1960 5-34 years old in college, 1960 completed 5 grades or less, 1960 25 years old or more completed high school, 1960 median years of education, 1960 population 21-65 years old in labor force, 1960 civilian labor force male, 1960 civilian labor force female, 1960 labor force white collar, 1960 of employed employed 13 labor force labor force working outside home county, 1960 weeks or less, 1959 employed in agriculture, 1960 employed in manufacturing, 1960 labor force employed in construction, 1960 Data Type population H resident type I! N H H H ethnicity H II age H H II I! H income H H H H education H labor force H Number 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. ' 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 33 TABLE 2. (continued) Description Z labor force employed in retail and wholesale trade, 1960 Z labor force employed in finance, insurance, and real estate, 1960 Z labor force employed in educational services, 1960 Z labor force employed in public administration, 1960 property tax per capita, 1962 general expenditure per capita excluding capital outlay, 1962 Z general expenditures for education, 1962 Z general expenditures for highways, 1962 Z general expenditures for public health and hospitals, 1962 Z debt of government revenue, 1962 Z revenue for education, 1962 Z revenue for highways, 1962 Z revenue for public health and hospitals, 1962 manufacturing productivity per employee, 1963 wholesale sales per employee, 1963 Z capital expenditure of value added in manufacutur— ing, 1963 retail sales per employee, 1963 selected services sales per employee, 1963 occupied houses with washer, 1960 occupied houses with freezer, 1960 occupied houses with air conditioning, 1960 occupied houses with television, 1960 occupied houses with telephone, 1960 occupied houses with car, 1960 commercial farms with sales of $10,000 or more, 1964 parttime farms of commercial farms, 1964 farm tenancy, 1964 change in farm size, 1959-1964 farm operator households with non-farm income, 1964 farm family living index, 1959 Z time deposits of total deposits, 1964 Z demand deposits of total deposits, 1964 Z change in bank deposits, 1960-1964 NNNNNNN NNNN Data Type labor force H revenue and expenditures productivity II I! housing ll agriculture H II banking " 34 in that they may measure somewhat less obvious but perhaps more relevant portions of differences.1 Statistical Methods Development Preference Consistency The product of resource development decision making is viewed at two points in time. First, development proposals were classified and counted. The same was done for development actions resulting from the proposals. In order to evaluate the effectiveness of the RC&D program relative to program response to local proposals, comparisons of propos— als and actions were made. For each project, a set of rankings was established for proposals and another for actions. These rankings were based on a 14 unit scale corresponding to the 14 development categories of the classification system. The association between proposal rank- ings and action rankings was tested by Spearman's rank correlation statistic.2 With this statistic, the degree of association of rankings is represented by the magnitude of the correlation coefficient. The direction of the relationship between rankings is indicated by the sign of the coefficient. Socioeconomic Structure and Development Preferences In the past, many studies have focused on the social and economic structures of various types of geo-political areas. Numerous studies of characteristics of decision makers and their decisions have also been completed. However, little research has been directed toward deriving 1Hadden and Borgatta, op. cit., p. 34. 2Sidney Seigel, Nonparametric Statistics for the Behavioral Sciences, McGraw Hill Series in Psychology, (New York: McGraw-Hill Book Company, 1956). 35 models which may aid our understanding of the influences and importance of socioeconomic structure in the outcome of local resource development decision making. A basic question is, what are the primary elements of socioeconomic structure of these essentially rural RC&D projects? Another is, to what degree do various elements of socioeconomic struc— ture seem to be associated with shifts in various resource development preferences? The literature review has identified one analytical method which has proved fruitful in examining socioeconomic structure of geo- political areas--factor analysis.1 Factor analysis is the generic name for a variety of procedures developed for analysis of intercorrelations within a set of variables, and for facilitating the discovery of regu- larity, order, and patterns of variation present in many observations on many variables. Principal component analysis is a useful factor technique for determining the minimum number of linear, independent dimensions (factors) needed to account for most of the variance in the original set of observations and is used in this study. This particular technique not only reveals how several measures (socioeconomic variables) can be combined to produce maximum differentiation among cases along a single socioeconomic factor, but also often reveals that several independent factors are required to adequately define the domain or socioeconomic structure under investigation. Factor analysis can be used and has been in this research, to: untangle linear relationships into separate patterns with each pattern appearing as a factor delineating a distinct cluster of interrelated data; reduce a mass of information to its essential meaning; discover 1R. J. Rummel, Applied Factor Analysis, (Evanston: Northwestern University Press, 1970). 36 the basic structure of a given domain; develop an empirical typology for classification or description; transform data to meet the assump— tions of other analytical techniques; and explore.1 The literature also suggested a technique for analyzing linear relationships between discovered elements of socioeconomic structure of RC&D projects and shifts in local development preferences-- discriminant analysis.2 This technique is used to find linear combi- nations of variables that maximize the ratio of among—groups to within—group variability. It produces an optimum discriminant function for a two-group situation that includes a linear combination of varia- bles capable of discriminating between two groups better than any other linear combination.3 The probabilities of each case having come from each group are computed and used for evaluating the classification of an area in a given group. Multi-group discrimination is possible but was not feasible for this research. Discriminant analysis has the general capability to: test for significant differences among average score profiles of two or more a priori defined groups, assuming multi-normal distributions and equal dispersions; determine which variables account for most of the inter- group differences in average profiles; find linear combinations of variables which allow the representation of groups by maximizing among— groups relative to within-group separation; and establish models for lIbid, p. 449. 2Maurice M. Tatsuoka, Multivariate Analysis: Techniques for Educational and Psychological Research, (New York: Wiley, 1971). 3 and David V. Tiedeman, "Discriminant Analysis," Review of Educational Research, XXIV, No. 5, (December, 1954), p. 402. 37 assigning new individuals whose profiles, but not group identity, are assumed to be from one of the a priori defined groups.1 For each RC&D project development proposals and resultant actions were compared for each of the 14 development categories. For any given category, if a project's percentage share of development actions was found to be greater than the corresponding share of proposals, that project was said to have emphasized that particular type of development. Thus, for each development category it is possible to obtain two groups of projects-~one group that has emphasized and one that has not. Discriminant analysis uses measurements on each socioeconomic fac- tor discovered through factor analysis to discriminate between the two groups of projects for each development category. The result is a best, linear equation for the group emphasizing the activity and 3 correspond- ing equation for that group not emphasizing the given activity. These equations are derived for each specific development activity. They are also derived for the two group situation formed by grouping the projects with respect to their emphasis of human versus natural resource develop- ment. As in the 14 categories, the term "emphasis" is used to indicate a percentage share increase in actions as compared to the corresponding share of proposals. The overall concept behind the discriminant analyses in this research is the discovery of the magnitude and direc- tion of relationships between socioeconomic structure and shifts in development preferences quantified by the use of the 14 category system for classifying resource development activities. 1Paul Green and Donald Tull, Research for Marketing Decisions, 2nd ed., (Englewood Cliffs, New Jersey: Prentice—Hall, Inc., 1970), p. 368. 38 In summary, two analytical techniques, factor and discriminant analysis, are combined to produce equations which serve to identify seemingly functional relationships between socioeconomic structure and development preferences as measured by changes in emphases of develop- ment activities in RC&D projects across the Nation. A major advantage these techniques offer is the capability of assessing and predicting qualitative dependent variates, i.e., Yes or No emphasis groupings representing tendencies toward development preference changes, from a set of quantitative independent variates representing socioeconomic structure of geo—political areas. CHAPTER IV COMPARATIVE ANALYSIS Resource Development Preferences Examination of local resource development preferences began with the review and analysis of development proposals and actions. Proposals are presented by the local people in the major planning documents of their respective RC&D projects and resultant actions are recorded in corresponding cumulative progress reports. Forty-eight RC&D projects across the Nation were selected for study. These were Operational for a period of years between 1963, when the first ten RC&D projects were authorized, and 1970, the cutoff date for the analyses in this study. The projects selected for study are shown in Table 3. Resource Development Directions Study of 48 RC&D planning documents resulted in the classification of 8,341 development proposals. Corresponding progress reports con- tained records of 6,590 measures acted on through the intiiative of local participants and their respective project coordinators. These data revealed a strong, natural resource preference in development. Overall, two-thirds of all proposals and actions were classified in natural resource-related categories. In each of the projects studied, natural resource-related proposals and actions outnumbered those primar— ily concerned with human resource conditions. However, when the percen- tage shares of human and natural resource proposals and actions of each 39 40 TABLE 3. Resource Conservation and Development Projects Selected for Study 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. RC&D Areas East Connecticut St. John - Aroostook North Country East Central Vermont South Central New York Seneca Trail Penn Soil Endless Mountains Shawnee Lincoln Hills Northwest Michigan Buckeye Hills Pri Ru Ta Lumberjack Sunflower Top of the Ozarks South West Missouri West Central Minnesota Onanegozie Randall Black Hills North Central Piedmont Low Country Crossroads Little Kanawa Mountain Dominion Coosa Valley Wiregrass Tradewater River Southeast Delta Northeast Mississippi Hull - York Lakeland Arkansas River Valley Ozark Foothills Trail Blazer Cherokee Hills Southeast Texas Eastern Hill Country Western Wyoming Box Elder North Idaho Upper Willamette Northern Rio-Grande South West New Mexico Central Nevada North California Sangre De Cristo Bitter Root State Connecticut Maine New Hampshire Vermont New York New York Pennsylvania Pennsylvania Illinois Indiana Michigan Ohio Wisconsin Wisconsin Kansas Missouri Missouri Minnesota Minnesota South Dakota South Dakota — Wyoming North Carolina South Carolina South Carolina West Virginia West Virginia — Virginia Alabama Alabama Kentucky Mississippi Mississippi Tennessee Arkansas Arkansas Louisiana Oklahoma Texas Texas Wyoming - Idaho Utah - Idaho Idaho - Washington Oregon New Mexico New Mexico Nevada Nevada - California Colorado Montana Total Included Counties -Number— H I_.I \IOOVWWO‘O‘O‘QWbWOU‘NQOLflWb-L‘mbw\l-L‘wML») I..I I_I h‘h‘k) ¢~U1C>h‘c> NUWMU‘Uval-‘UTWDUJHW l N \D \l 41 project were considered, directional shifts toward human resource pre- ferences were found in twenty-seven or 56 percent of the projects. Resource Development Emphases Overall development emphases were determined by comparing percen— tage share of all proposals for each development category with the cor- responding share of actions.r These percentages are shown in Table 4. Increased shares, indicating slight shifts in development preferences, were found for Education, Industry, Housing, Land, Agriculture, and Recreation. In the process of converting proposals into actions, over half the studied areas were found to have increased their emphasis in five development categories: Education, Housing, Environment, Land, and Recreation. The distribution of RC&D projects' emphasis shifts is provided in Table 5. A Measure of Program Effectiveness One measure of the effectiveness of a rural development program such as RC&D is the identification and quantification of measurable economic impacts of development-related activities, i.e., capital invest- ment and job creation. This is the traditional type of measure used to judge program success. This approach is beyond the scope of this study. This research is directed toward the local context. The measure of effectiveness chosen concerns the degree to which program response through initiated actions corresponds to local development proposals drawn up in accordance with locally identified problems and needs. The assumption has been made that the intense process of problem and need specification provides one rather accurate picture of local development preferences. A second, time-lapse picture of such preferences is 42 0H m o.~ o.m " ucoaaon>mo vow mcHsamHm H N m.e~ m.- " COHummuomm w n H.q m.q u ummuom m m N.¢ m.m u musuHsoHuw< N H m.wH o.- " umumz q q H.0H «.0 " ncmH HH OH H.N ¢.N “ ucmecouH>sm " mouaommm Hmusumz m m n.5H H.mH ” mmoH>umm mam mmHuHHHomm %uHsoaBoo NH mH a.H ~.H ” wsHmsom o m H.m N.o " sOHumuuoamamuH «H «H N.o m.o " uamE%OHmBm m o e.m o.¢ u anomaocH MH NH m.H m.H " mmUH>umm Hmovaz was nuHmom a HH n.~ ¢.~ “ sOHumooom " mousommm cmesm msoHuu< mHmmoaoum maOHuo< m mmoooum " mmaHxsmm oumnm mmmwcmoumm mommnmam usmEQOHm>mQ xcmm mam ammusmoumm an moosmummmum usmEQOHm>on HHmum>o .q mHmmn a wchcmHm oq mH 00 mm H soHummuomm no Hm mm NH H anamouom No on mm mH ” musuHsoHuw< mm mm as ON H swans HN oa an an m same we NN cm 0N . ucmEcouH>sm H mousomom Honoumz qm ow sq mm H mmua>umm a mmauaaaumm sufiasesoo we mm Nm mm H wchsom mm mm we mm H GOHumuuommsmuH cm as ea m H uamasoaaam co am os SH H suumaeaH mm mm Ne ON H mmofi>umm HmUHemz a euammz an as me cm W coaumusem . muusommm smasm usmuumm " mmmu< " usmuumm " mmmu< " mo umnasz mo umnaaz . mmHuowoumu MWNMHUOQ mmNNHUQH mHmmnaam untQOHm>oo mommamam uamEmoHo>ma mmu< Qwom mo sOHuanuumHo .n mHm¢H 44 provided by final decisions by local people to seek action on proposals and actually have them initiated. These two pictures provide the only measurement of the consistency with which local development preferences are adhered to through program response. This approach to evaluation is justified by the argument that improvements in human and natural conditions cannot always be realisti- cally or accurately measured in terms of dollars or jobs. Local view- points, attitudes, leadership, and cohesiveness may all be important considerations. Furthermore, the type of developments or improvements desired may not require significant investment, employment, or resource reallocations and therefore resultant effects would not be identified. For the purpose of evaluating program consistency, development categories have been given overall ranks according to their shares of total proposals and actions. In the two resulting rankings shown in Table 4, the lower rankings indicate larger shares. The null hypothesis is that there is no difference between the set of rankings. A Spearman rank correlation statistic of .96, statistically significant at alpha = .05, was found, indicating close agreement between rankings of pro- posals and actions. The same test procedure was used for each RC&D area. The hypothesis of no difference between rankings for proposals and actions was rejected for only three projects. Correlation coeffi- cients are shown in Table 6. Local participants in the RC&D programlunnaan open—ended opportu- nity for identifying problems, evaluating needs, proposing remedial courses of action, setting priorities for actions, initiating actions, and changing their views. Given the findings above, the only possible conclusion is that program response seems to be effectively mirroring 45 TABLE 6. Rank Correlations for Proposals and Actions Within RC&D Projects Project Correlation Project Correlation Coefficient Coefficient 1 .7165 25 .5308* 2 -8473 26 .4737* 3 -8435 27 .5253* 4 '6770 28 .5594 S .8506 29 .8077 6 .7913 30 .7429 7 ~7325 31 .8935 8 -8770 32 .7737 9 -7638 33 .7869 10 .9066 34 .5429 11 .7011 35 .8924 '2 -3649 36 .9055 13 .9451 37 .8594 14 .6990 38 .7110 15 .8440 39 .7539 16 .8847 40 .8957 17 .9462 41 .7913 18 .9055 42 .6385 19 .6292 43 .7506 20 -7935 44 .7044 21 .8624 45 .8242 22 -8044 46 .7374 23 -7605 47 .8880 2“ ~7759 48 .6935 * Not significant at alpha = .05 46 development preferences of local participants. The degree to which local peeple are satisfied with program response can only be measured through survey techniques beyond the scope of this study. Project Analysis Thus far, comparative analysis has provided measures of locally specified preferences for improvements in human and natural conditions in rural areas and measures of the consistency with which the RC&D pro— gram has responded to these choices. In keeping with the objectives of this study, comparative analysis of RC&D projects was undertaken for the purpose of identifying basic elements of their collective socio- economic structure. This process began with the selection and collection of 76 county- 1eve1 census measurements descriptive of socioeconomic structure for each of the 297 counties comprising the selected projects. The 76 variables were transformed such that measurements on 297 counties would represent the socioeconomic structure of the 48 RC&D projects. The first data matrix of 76 x 297 was reduced to 76.x 48 in this manner. Application of factor analysis to this matrix further reduced it to 20 x 48 by gathering highly intercorrelated variables together in 20 groups, i.e., factors, which explained 94 percent of the variation in the 76 x 48 matrix. These factors represent the basic elements of socioeconomic structure of the studied project areas as limited by the 76 original variables selected for study. Factors derived in this manner are important in that they lend themselves to indepth interpre— tation of socioeconomic relationships within areas. Factors also allow for the computation of weighted scores for each project on each factor. Such scores allow for structural comparisons between projects 47 and also serve as the basis for analyses identifying relationships between socioeconomic structure and development preferences. Relationships between variables and the factors in which they have primary importance are used in factor interpretation and definition. In factor definition, major component variable loadings are considered as are their signs which indicate positive or negative association with the factor. Generally, the higher a variable's loading is, the greater is its association with the factor and the more descriptive the variable is concerning relationships within the factor. Major factors discovered in this analysis are shown in Table 7. Structural Elements The largest factor found was Socioeconomic Status. This factor represents strong, positive influences of income levels and distribution, population change, residence, education, age, labor force, and employ- ment. Minority Population was the second element found. A strong minor- ity aspect in this factor is signified by strong, positive minority component loadings. It is reinforced by contrasting, strong, negative loadings on the white rural-farm population component and on level-of- 1iving and political participation variables. These variable loadings are often associated with minority circumstances. In the Health and Education Finances factor, a distinct bipolar relationship exists between education and health revenue and expendi- tures components. Highly negative loadings on public health and hospi- tals components suggest low levels of revenue and expenditures for health systems are important in explaining variation in structural make-up of rural areas. The education components are inversely related 48 NHHm. oocm. comm. mHNm. Hmmm.I wmwm. wmoo. quc. mane. emec. ammo. ammo.I mmmo. mnmo. mmMN. News. oqmw. wmmm. meow. meow. NNom. Noam. Nmmw. Nomm. coqm. coqm. mmsvaOH ooaH .ousuHsoHuwm sH vaOqum wou0m noomH N momH .mmNOHaam nun NuH>Huo=voua wsHusuommscma coma .6Ho 6666s meIHN coaumaaaoa N oomH .muu0m noan :H vHo mumoh moIHN aOHustnoa oomH .aummIHmusu GOHumHsaoa mesa .smauao HmuHamo wsHvsHuxm muHamu “on musqusomxm Hmumsow ooaH .oOHmH>mHmu nuHs momson oqusooo N NooH .muHamo you woo zuuoqoua oomH .umo :uHs momson vaasouo N oomH .:0HumHsaoa amen: N «00H .muoa no ooo.0Hw mo mmHmm :uHB maumm HmHouossoo N oomH .mmmH no mmomuw n voumHaaoo GOHumHsaoa N oooH .muHamU use mumu mo popes: oomH .umHHoo muHss wou0m Hoan oomH .msonmmHmu nqu momson vamsooo ooaa .mumumm How» was .mosmch .mosmusmsH sH oohonam mouom noan oomHIommH .moHHHawm mo Hogans sH swamps oomHIomaH .mmm amHoma sH mwcmso ooAHIOmmH .mmameu coasmaaaoa N oomH .Hoonom :an omuMHaaoo oHo mumoz mm GOHuMHsaom N oomH .mummm Hoosom schoa oomHIommH .aOHumuwHe cu was mmcmso sOHumHsaoa N mmaa .mmma to coo.mm mmaooaa sesame N mmmH .maoocH AHHSmm amvaa mmmH .muoa no ooo.OHm oaoocH NHHamm omoH .maoocH oanmo pom N N 6°N NN mucoaomaoo msumum oHEocoomOHoom muouomm mwusmoq mam .muoosoaaoo Mano: .muouomm UHBonoomOHoom mo wcHumHH .N NHmmu oomH .vHo mumom no um>o GOHumHsaoa oomHIommH .mmmwuocH Housums ou one owcmnu soHumHsaoa oomH .vHo mums» m nova: soHumHsaom §£§£§2§2§9 $8§9§€ §2§2§8 oomH soHumHsaoa mo mwm amvaE ooaH .mmm wcHuo> wo sOHumHsmom. oomH .COHumuumHsHaom oHHpna :H voNOHaam wou0m noan oomH .GOHuosuumaoo aH noaoHnam mouow uopmH oooH .umnuo summIHmusu :OHuMHsuoa NomH .sOHumosoo now ossm>mu NomH .soHumoswo you mouauHosmaxm Hmumcmm NomH .mHmuHamo: was suHmm: oHHnsa “cm mququcmaxm Hmumcmw NomH .mHmuHamon was nuHmon UHHnaa How msso>mu somHImmmH .mNHm snow sH mwcmnu $2 §2§2§£ §8§8 momH .mmNOHmam pom mmHmm mmoH>umm vmuomHmm oomH .msHsOHancoo uHm nuH3 mouse: umHasooo oomH .muuow uoan smHHH>Ho cH memamw oomH .mouom woan GMHHH>H0 sH mmHma oomH aH vmuo> aOHumHsaom oomH .umnmmz :uHa momson vaasuoo ooaH .uuHuoaHa aOHumHsmom oomH .ouwo: BumMIHmusu aOHumHsaoa Hanan oooH .oana aumeHmusu GOHuMHsuoa Hausa N §Q§2§Q§£&£&Q&£ mcaxcmm .N moosmsHm mmBLme .0 6we .m aOHumHsmom BumeHmusm umnuo .q mmUGmCHh COHumUDfim fifim SuHmmm .m :oHumHsaom muHuost .N Hemaafiucoov .N mamHH NHHamw Show NomH .ossm>mu unmesum>ow mo unmo N momH .mmNOHmEo you mmHmm HHmumu oomH .mwmuu mHmmoHoza was HHmumu :H nmNOHmEo mUHOm noan eomH .maumm HmHoumEEoo mo maumw mSHquumm momH .wcHusuommssma sH coves oaHm> mo ouaqucmnxo HmuHamo mmmH .mmoH no mxmms MH umNOHaao ocmH .mooH>uom HmsOHumoswm aH uohoHnam wou0u noan oomH .mmeHoo :H vHo mumm> emIm QOHumHsaoa oomH .Nucsoo mac: omeuso waquos voNOHaao mo N N NNN N N muHmooQ cam Hm>oH oOHumHsaom .om mosmHonmm onmmHonz .oH mwamno muHmome mstamm .wH mwcmnu meousH NHH8mm .NH ousuosuum :OHumosom .oH musuHsoHuw< HmHouanoo .mH msousH annulcoz .eH xmosH wsH>HH mHHamm Show .MH osco>mm\unma unassum>oo .NH evoke mHmmmHoanHHmumm .HH mocoHUHmwm unmaumo>cH wcHusu06w=smz .oH huHmHooam GOHumosvm .m monomxuoz ucmvaouIcoz .m HomscHucoov .N mHmma e wcHacmHm mm o HA A+e maouum moocmaam coaumuaem can :uHmom m " coaumouowm Ne eH me A+e waouum muuom xuos unmeHmouIaoz m “ Nuummuom ow mH oe HIV xmms mousmsHm cOHumusem ecu suHmmm m " mHSuH30Huw< an OH ee HIV Mama mooomsHm cOHumosem eam nuHmom m u umumz OOH m mm HIV xmos GOHumHsaom anomIHmusm umsuo e " eGMH He mH me H+V wcouum mousmaHm sOHumosem eco nuHmmm m " ucmssouH>sm " mousommm Hmusumz He NH mm HIV xmms mouom xuoa unmeHmmuIcoz w u mooH>pmm " e mmHuHHHumm muHs58800 om eH me A+v wsouum muaoHonmm QHMmoHop3 mH “ msHmsom om mH we A+v wsouum huHmHommm aOHumosem m " cOHumuuoamsmuH em OH me HIV xmms wstsmm N “ uamENOHaam em mH me A+v wsouum ouuom xuoz ucmeHmouIsoz w “ muumsecH me NH Ne HIV xmms mouom xuos usoeHmouIaoz m “mmoH>uom HmUHeoZ e :uHmmm me HH mm HIV xmos xmesH mssm>mm\uema unmasum>ou NH " sOHumosem “ mousommm amaze me 0H oe A+v wcouum mousmsHm coHumusem ecm :uHmmm m “ mousommm Hmusumz .m> amaze N homusoooo HmsHm ”.02 HmuoH H NumeHum H H H H muqmsHanmumo ucmSQOHo>mn humaHpm mumusoo< GOHumoHMHmmmHo eam mucmcHaumuon uamBQOHo>mo humEHum .m mHmmH aoHumHsmom .ON m.OH e mH e . NocmHonem mHmmmHonz .OH m.OH NH OH I “ mwcmno muHmoOme mcchmm .eH OH eH OH e " owcmso maoosH NHHEmm .NH mH e ON I " musuosuum sOHumosem .eH mH mH mH I " ousuHsoHumm HmHoumaaoo .mH mH OH NH O ” moaoocH BummIcoz .eH ON NH NH I u xmesH wsH>HH NHHame Show .mH m N m.m I H xmecH ossm>mulueoe Hmusmesum>ou .NH m.eH wH mH N " NuHmHomam memuu mHmmmHonaIHHmumm .HH m.eH mH eH I " NusoHoneo unmaumm>cH wcHusuomwsamz .OH O m mH w ” NuHmHommm aOHumosem .O m.N m.N m.N N " ounce xuoa unmeHmmulsoz .e w HH O I " onxcmm .N m.e m m m " musuHeammxm e mssm>mu mNmsanm .e m.e OH N I ” mw¢ .m m.N H e I " coHumHsaon same Hanan umnuo .e H m.N H H " woodman GOHumusem w nuHmmm .m NH ON m.m m m aOHumHnaom NuHuost .N eH eH HH I . msumum oHsocoomOHoom .H HHmum>O " Hmusumz “ amesm ” Hmuaumz .m> smasm mHmmsaEm . cOHuomuHQ muouomm untQOHo>oa mousomom mwcHxsmM uHmLH eaw mucmsHaumuoa GOHumoHMHmmmHO unmamon>oQ .O MHm¢H 63 In the latter case, the factor rank is the same as the degree of impor- tance in the problem. The human and natural resource development rankings were statisti- cally tested to determine the degree to which they are similar. The re- sulting rank correlation coefficient of .41 is significant at alpha = .05. This finding implies that there is a significant association between the values of a factor in the classification of either human or natural resource development emphases. Although for some factors there are wide differ- ences in ranks for human and natural resource-related emphases, notably Minority Population, Education Structure, and Wholesale Efficiency, each of the 20 factors was found to have been one of thetxn>five determinants of classification for one or more development categories. Table 10 shows, as previously indicated, that only seven different factors acted as pri- mary determinants of increased development tendencies. Only these and four more ever acted as secondary determinants; these eleven and three different factors ever acted as tertiary determinants; etc. The overall importance of each factor as a determinant of development emphasis is shown by the overall ranking of Table 9. The most important factor in the 14 classification problems concerning specific development emphases was Health and Education Finances while the least important was Farm Family Living Index. Comparative Analysis in Retrospect The RC&D program offers project areas and included communities an open-ended opportunity to identify problems, evaluate needs, prOpose reme— dial courses of actions, set priorities for action, initiate actions when possible, and perhaps most importantly, local participants are allowed to change their minds relative to problems, needs, priorities, and preferences for action. This research represents an attempt to quantify local 64 acuomm x " NuHmcoe e Ho>oH GOHumHsaom .ON N " NusmHonmm onmmHon3 .OH ” omamno muHmonoe wcHxsmm .wH u mwsmnu maoocH NHHamh .NH x " ousuosuum coHumusem .eH x " ousuHsoHuwm HmHouosEoo .mH x " moaoosH sumeIooz .eH x u x85 9:3: SEE Ems .2 x " xmeaH msso>mqueme Housmaaum>oo .NH x ” NuHmHoomm memuu mHmmmHonzIHHmumm .HH N “ .NoamHonmm usoeumm>GH waHuauommsamz .OH N u muHmHumnm nOHumosem .m x " mouom xuoa unmeHmmuIsoz .e x “ 335m .N x u musuHecoOxo e msam>ou mNmzanm .e x n «we .m x “ cOHumHsmoa sham Hmusu umnuo .e x " ousuHecmnxo e msam>ou aOHumoseo a nuHmmm .m x u sOHuMquoa NuHuoaHz .N x " msumum UHaosoomOHoom .H u e u m u N u H m Nausm umuHm mo Hm>mH EmHnoum unmaHaHuomHn m CH scuomm oHEosoumOHuom comm mo Nuuam umuHm .OH MHmMn e oszzzm mMUH>me e mmHHHHHUmMm HoH aOHumHsmom NuamHonmo mHmmoHonz owsmso muHmoame waHMamm Own—.050 080.35" .AHHn—mm ............. I ousuusuum aoEmosem A: ousuHsoHuww HmHoumsaoo osoosH Summsoz xoecH wsH>HH NHHamm aumm ,msso>ou\upme unmaswo>ou meow» onmoHonsIHHmumm NosoHonmm unmeumm>cH wsHusuommssmz NuHmHooam soHumosem mouomxuos usoeHmmusoz wcHxsmm w mmmsanm mw< sOHumHsaoa same Hausa umnuo w aOHumoseo e zuHmmm GOHumHsmon aumu Hound Northwest Michigan Project, Michigan -——~——~- Shawnee Project, maumum oHaosouoOHuom Illinois 1.5'—- _ _ .5 .U 9“ L 74 given variables in these factors, thus providing clues for program adjustments in given projects. Determinants of Development Directions and Emphases The relevance of analysis of socioeconomic factors is supported by the derivation of the empirical models used to successfully classify projects in terms of shifts in development preferences. The modeling approach to understanding factors affecting decision making can be extended for purposes of predicting likely development directions and emphases for any proposed RC&D project for which appropriate secondary data have been collected. Another interesting extension of this approach consists of the derivation of empirical models based on the classification of existing RC&D projects in accordance with their ranking of proposals and actions. Such rankings could be determined by the proportions of measures in each development category or by survey methods. Different classifica— tion models, again based on scores on socioeconomic factors, would then classify projects according to priorities in planning or action. This sort of analysis could also be used for predicting proposals and actions for potential RC&D projects, given the appropriate secondary data. This method would define relative relationships between different types of development whereas the models derived in this research define only the direction of changing preferences within a given category of development. The use of both methods would present more detailed inputs for the evaluation of program planning and response. Such inputs could serve as additional criteria for RC&D project selection and could be important at the RC&D administrative level. Improvement in policy formulation, program planning, management, and evaluation require a continual quest 75 for improvement of indicators of program success or progress. Increased intelligence concerning socioeconomic influences and their relationship to various aspects of the RC&D effort might improve the possibility of discovering other perhaps more meaningful empirical models which could serve well in various aspects of program administration. Recommendations RC&D Classification System The classification system developed and used in this study embraces a wide range of development activities and intentions. It provides an additional data base for quantifying RC&D planning and pro- gress. Consideration should be given to the use of this system to monitor local resource development preferences indicated by project proposals and resultant actions. This would facilitate analysis of planning and action trends on project and national levels. It should also be considered for use as a guide to problem identification and formulation of planning proposals in existing and potential RC&D projects as well as for evalution of applications with respect to the total development concept of the RC&D program. Analysis of Socioeconomic Structure The exploration of socioeconomic structure of projects is useful for defining elements of considerable variation in their makeup. Pro- files can be built which pinpoint socioeconomic differences and similar- ities in projects. Along with providing insight into the makeup of RC&D projects, the analyses provide the data input (factor scores) necessary for relating structural variation to changing development preferences. 76 Determinants of Changes in Development Preferences Improvement in policy formulation, program planning, management, and evaluation require improvement in indicators of program activities and reSponse. The success of the empirical socioeconomic models in classifying projects according to shifts in resource development pre- ferences points out the relevance and importance of analyzing socio- economic data as a means of gaining insight into what factors influence decision-making in the RC&D program. Consideration should be given to extending the modeling approach to include the prediction of likely development tendencies for proposed RC&D projects for which appropriate data can be collected. This approach could also be extended to the prediction of priorities in planning and action given some additional data concerning local rankings of priorities. Such analyses could help in establishing firm guidelines for planning and action in proposed RC&D projects. Further Research Additional research is needed to insure that resource development programs deal effectively with problems of people while assuring socially acceptable impacts on the resources involved. As the RC&D program grows and as other programs related to resource development grow, care must be given to the task of developing improved approaches to effective resource development. Careful identification and considera- tion of development preferences in conjunction with scientific analysis and evaluation of results of development activities can help achieve this objective. To this end, consideration should be given to encour- aging the type of comparative analysis suggested by this study. 77 Summaryyof Recommendations l. The development classification system, consisting of 14 development categories and 53 development purposes, should be considered as a basis for monitoring, analyzing, and evaluating the broad range of development possible within the concept of the RC&D program. 2. Consideration should be given to the search for relevant socioeconomic dimensions or factors characterizing project simi- larities and differences (including those RC&D projects still in the application stage). 3. Further effort should be directed toward the development and use of empirical models specifying functional relationships between socioeconomic influences and shifts in development direc- tions and emphases and priorities for all RC&D projects (including those in the application stage). 4. Consideration should be given to the concepts and techniques employed in this research project in terms of their potential value and usefulness in RC&D and other development programs where additional knowledge of geo—political areas and development ten- dencies is important and where some level of citizen decision- making is required to insure socially acceptable resource development and use. LIST OF REFERENCES BIBLIOGRAPHY Books Adelman, Irma and Morris, Cynthia Taft. Society, Politics, and Economic Development: A Quantitative Approach. Baltimore: The John Hopkins Press, 1967. Barkley, Paul W. and Seckler, David W. Economic Growth and Environ— mental Decay: The Solution Becomes the Problem. New York: Harcourt Brace Jovanovich, Inc., 1972. Dixon, W. J. (ed.) Biomedical Computer Programs. Berkley: University of California Press, 1970. Goodenough, Ward. Cooperation and Change. New York: Russell Sage Foundation, 1963. Green, Paul and Tull, Donald. Research for Marketing Decisions. 2nd ed. Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1970. Hadden, Jeffrey K. and Borgatta, Edgar F., "The Factor Analytic Struc- ture of American Cities," American Cities: Their Social Characteristics, Chicago: Rand McNally and Company, 1965, pp. 32-66. Rummel, R. J. Applied Factor Analysis. Evanston, Illinois: North- western University Press, 1970. Seigel, Sidney. Nonparametric Statistics for the Behavioral Sciences. New York: McGraw—Hill Book Company, 1956. Sundquist, James L. and Davis, David W. Making Federalism Work. The Brookings Institution, N.W. Washington, D.C., 1969. Tatsuoka, Maurice M. Multivariate Analysis: Techniques for Educational and Psychological Research. New York: Wiley, 1971. Articles Bonjean, Charles M.; Browning, Harley L.; and Carter, Lewis F., "Toward Comparative Community Research: A Factor Analysis of United States Counties," Sociological Quarterly, X, No. 2, (Spring, 1969), pp. 157-176. 78 79 Cox, Kevin R., "Suburbia and Voting Behavior in the London Metropolitan Area," Annals, Association of American Geographers, LXIII, No. 2, Green, James W. and Mayo, Selz C., "A Framework for Research in the Actions of Community Groups," Social Forces, XXXI, No. 4, (May, 1953), pp. 320-327. Jonassen, Christen T., "Functional Unities in Eighty-eight Community Systems," American Sociological Review, XXVI, No. 3, (June, 1961), pp. 399-407. Kivlin, Joseph E. and Fligel, Frederick C., "Orientations to Agricul- ture: A Factor Analysis of Farmers' Perceptions of New Practices,‘ Rural Sociology, XXXIII, No. 2, (June, 1968), pp. 127-140. Munson, Byron E., "Structural Analysis of the Community," Rural Socio- logy, XXXIII, No. 4, (December, 1968), pp. 450-459. Price, Daniel 0., "Factor Analysis in the Study of Metropolitan Centers," Social Forces, XX, No. 4, (May, 1942), pp. 449-455. Reiss, Albert J., "Some Logical and Methodological Problems in Commu- nity Research," Social Forces, XXXIII, No. 1, (October, 1954), pp. 51-57. _________, "The Sociological Study of Communities," Rural Sociology, XXIV, No. 1, (March, 1959), pp. 118-130. ‘ Rossi, Peter H., "Community Decision Making," Administrative Science Qparterly, I, No. 4, (March, 1957), pp. 415-443. Rummel, R. J., "Understanding Factor Analysis," Journal of Conflict Resolution, XI, No. 4, (December, 1967), pp. 444-480. Sharkansky, Ira and Hofferbert, Richard 1., "Dimensions of State Poli- tics, Economics, and Public Policy," American Political Science Review, LXIII, No. 3, (September, 1969), pp. 867-879. Smith, Arthur H. and Martin, William E., "Socioeconomic Behavior of Cattle Ranchers, with Implications for Rural Community Develop- ment in the west," American Journal of Agpicultural Economics, LIV, No. 2, (May, 1972), pp. 217-225. Summers, Gene F.; Clark, John P.; and Seiler, Lauren H., "The Renewal of Community Sociology," Rural Sociology, XXXV, No. 2, (June, 1970), pp. 218-230. Tatsuoka, Maurice M. and Teideman, David V., "Discriminant Analysis," Review of Educational Research, XXIV, No. 5, (December, 1954), pp. 402-420. 80 Wilkinson, Kenneth P., "Special Agency Program Accomplishment and Commu- nity Action Styles: The Case of Watershed Development," Rural Sociology, XXXIV, No. 1, (March, 1969), pp. 29-42. Other Materials Cox, Thomas P., "A Sociological Analysis of Upstream Watershed Develop- ment in Oklahoma," (unpublished Ph.D. dissertation, Graduate College, Oklahoma State University, 1967), 141 pp. U.S. Department of Agriculture, Soil Conservation Service, Resource Conservation and Development Projects: RC&D Handbook, washington, D.C., 1972. . GENERAL REFERENCES Books Adcock, C. J. Factor Analysis for Non-mathematicians. New York: Cambridge University Press, 1954. Cattell, Raymond B. Factor Analysis: An Introduction and Manual for the Psycholggist and Social Scientist. New York: Harper and Brothers, 1952. Clark, Terry N. (ed.) Community Structure and Decision Making: Comparative Analyses. San Francisco: Chandler Publishing Com- pany, 1968. Cooley, William W. and Lohnes, Paul R. Multivariate Procedures for the Behavioral Sciences. New York: John Wiley and Sons, Inc., 1962. Cooper, Joseph D. The Art of Decision Making. Garden City, New York: Doubleday, 1961. Fruchter, Benjamin. Introduction to Factor Analysis. Princeton, New Jersey: D. Van Nostrand Company, Inc., 1954. Hare, Paul A. Handbook of Small Group Research. New York: The Free Press of Glencoe, 1962. Harman, Harry H. Modern Factor Analysis. 2nd ed., Revised. Chicago: University of Chicago Press, 1967. Munson, Byron E. Changing_Community Dimensions: The Interrelationships of Social and Economic Variables. Columbus, Ohio: Ohio State University Press, 1968. Orcutt, Guy H.; Greensberger, Martin; Korbel, John; and Rivlin, Alice M. Microanalysis of Socioeconomic Systems: A Simulation Study. New York: Harper and Row, 1961. Shaffer, Albert and Woodruff, Ruth. A Study of Community Decision Making. Chapel Hill, North Carolina: University of North Caro- lina Press, 1970. Wallace, Luther T.; Hobbs, Daryl; and Vlasin, Raymond D. (ed.) Selected Perspectives for Community Resource Development. Raleigh, North Carolina: Agricultural Policy Institute, School of Agriculture and Life Sciences, North Carolina State University, 1969. 81 82 Wasserman, Paul and Silander, Fred. Decision Making: An Annotated Bibliography. Ithaca, New York: Graduate School of Business and Public Administration, Cornell University, 1958. Articles Brooks, Michael P., "The Community Action Program as a Setting for Applied Research," Journal of Social Issues, XXI, No. 1, (January, 1965), pp. 29-40. Brunn, Stanley D.; Hoffman, Wayne L; and Romsa, Gerald H., "The Youngs- town School Levies: A Geographical Analysis in Voting Behavior," Urban Education, (April, 1970), pp. 20-51. Eber, Herbert W., "Multivariate Analysis of a Vocational Rehabilitation System," Multivariate Behavioral Research Monographs, No. 66-1, (1966), pp. 1-52. Fessler, Donald R., "The Development of a Scale for Measuring Community Solidarity," Rural Sociology, XVII, No. 2, (June, 1952), pp. 144-152. Finely, James R., "Farm Practice Adoption: A Predictive Model," Rural Sociology, XXXIII, No. 1, (March, 1968), pp. 6-18. Folkman, William 8., "Board Members as Decision Makers in Farmers' Cooperatives," Rural Sociology, XXIII, No. 3, (September, 1958), pp. 239-251. Freeman, Charles and Mayo, Selz C., "Decision Makers in Rural Community Action," Social Forces, XXXV, No. 4, (May, 1957), pp. 319-322. Freeman, Howard E.; Novak, Edwin; and Reeder, Leo C., "Correlates of Membership in Voluntary Associations," American Sociological Review, XXII, No. 5, (October, 1957), pp. 583-593. Freeman, Howard E. and Sherwood, Clarence C., "Research in Large Scale Intervention Programs," Journal of Social Issues, XXI, No. 1, (January, 1954), pp. 11-28. Hanson, Robert C., "Predicting a Community Decision: A Test of the Miller-Form Theory," American Sociological Review, XXIV, No. 4, (August, 1959), pp. 662-671 Hoffer, Charles R., "Social Action in Community Development," Rural Sociology, XXIII, No. 1, (March, 1958), pp. 43-51. Matthews, Joseph L. and Holland, Linnea 8., "Procedures and Methods for Community and Resource Development," Review of Educational Research, XXXV, No. 3, (June, 1965), pp. 224-230. 83 Miller, Paul A., "The Process of Decision Making Within the Context of Community Organization," Rural Sociology, XVII, No. 2, (June, 1952), pp. 153-161. Sower, Christopher and Freeman, Walter, "Community Involvement in Com- munity Development Programs," Rural Sociology, XXIII, No. 1, (March, 1958), pp. 25-33. Sutton, Willis A. and Kolaja, Jiri, "Elements of Community Actions," Social Forces, XXXVIII, No. 4, (May, 1960), pp. 325-331. Young, James N. and Mayo, Selz C., "Manifest and Latent Participators in a Rural Community Action Program," Social Forces, XXXVIII, No. 2, (December, 1959), pp. 140-145. APPENDIX A METHODS APPENDIX A METHODS Factor analysis is the generic term for a variety of procedures developed for analysis of intercorrelations within a set of variables. Such techniques facilitate the discovery of regularity, order, and patterns within sets of observations on many variables. Principal com- ponent analysis (component factor analysis) is a useful factor technique for determining the minimum number of independent dimensions needed to account for most of the variance in the original set of variables. It not only reveals how several measures of a given domain can be combined to produce maximum discrimination among cases along a single dimension, but also often reveals that several independent dimensions are required to adequately define the domain under investigation. This technique is described below. The generalized linear factor model is:1 = + o o o F o o zji aleli asz21 + + ajp p1 + aquJu where zji a standard score on test j for individual i, j = 1, 2, . . . m measurements, i = l, 2, . . . n cases, p = l, 2, . . . p common factors 1R.J. Rummel .,App1ied Factor Analysis, (Evanston, Northwestern University Press, 1970), pp. 107-108. Also see pp. 101-155. 84 85 ajp = factor loading for the pth factor on the jth variable, Fp1 = the factor score for area i on the pth factor and a U = a unique term (including the coefficient a and the factor ju ju ju score U u) describing the specific and random error variance :1 in 1 measurements on the jth variable. The following model displays the factor model for the elements of vector 2 for n cases: J = O O O + zlj ajlfll + aj2f12 + + ajpflp ajuflu - + o o o 2j aj1f21 + aj2f22 + ajprp + ajuf2u nj — ajlfnl + ajanZ + . . . + ajpfnp + ajufnu th th where f p 8 p factor score for the first case on the j variable and h = pt factor loading for the first case on the jth variable. a JP When all factors (common and unique) are considered the sum of the squared factor loadings for a given row is equal to one: " 2 Z aju = 1.00 k=l where k = any factor. In the case of principal component analysis, no differentiation is made for unique variance representing both specific and random error variance in measurements on variables. The unique terms are not included in the generalized linear factor model or in the zj vector model presented above. The correlation matrix would be fac- tored with unities in the diagonal yielding p common factors explaining most of the variance in the data. Thus in principal component analysis, the sum of the squared factor loadings for a given row (variable) is equal to: 86 hj = 1.00 - (specific + error variance) 2 2 2 or h = a + a + . . . + a ’ j 31 32 JP where hJ.2 = the observed communality of variable j when p factors are used, 2 ajp = the proportion of a variable's total variance accounted for by factor p. 2 J variance accounted for by all p factors. The proportion of total The communality h represents the proportion of a variable's total variance in all variables explained by factor p is: m 2 vp = Z a : (trace of correlation matrix) 1P i=1 where the trace a sum of diagonal elements or m. The following definitions are offered for purposes of review and clarification. A factor loading is a weight for each factor dimension measuring the variance contribution the factor makes to the data vec- tor. Each variable has a loading on every factor. Loadings can be interpreted generally like correlation coefficients, that is their values vary from -l.00 to +1.00 with the signs indicating that the variable varies inversely or directly with the factor. Loadings are crucial as they form the basis for factor interpretation. For a given variable, the sum of the squared loadings on each factor equals its communality, or the proportion of a variable's total variation that is included in the factors. Use of the closed factor model, factoring with unities in the diagonal of the correlation matrix, allows computation of factor scores according to: 87 F11 = allz11 + a21221 + . . . + aplzpi where F11 8 score on factor 1 for case i, a = loading on factor 1 for case 1, 11 z = standard data score on test 1 for case i. 11 Each variable is weighted proportionally to its involvement in a pattern or factor; the more involved, the higher the weight. To deter- mine a factor score, F1, for a case on a pattern, the case's data, zji’ on each variable is multiplied by the pattern weight, for that aji’ variable. The sum of the weight-times-data products for all variables for a given case equals the factor score for that case on that factor. Multiple factor analysis involves two basic steps. First a tech- nique, principal components analysis for example, is used to derive an initial set of reference dimensions. Then a rotational technique is used to convert the reference or principal factor pattern to a pattern of simple structure. Rotation causes a shift from factors maximizing total variance to factors delineating separate groups of highly inter- correlated variables. The basic requirements that simple structure should satisfy are:1 1. Each variable should have at least one zero loading in the factor matrix. 2. For a factor matrix of p factors, each column of factor load- ings should have at least p variables with zero loadings. 3. For each pair of columns of loadings (factors), several varia- bles should have zero loadings in one column but not in the other. 1Ibid., p. 380. 88 4. For each pair of columns of loadings (factors), a large pro- portion of the variables should have zero loadings in both columns. 5. For each pair of columns of loadings (factors), only a small proportion of variables should have non-zero loadings in both columns. In this study, rotation was restricted to orthogonality, meaning that the resulting factors are mutually orthogonal. Orthogonality ensures that factors will delineate statistically independent variation and are amenable to subsequent mathematical manipulation and analysis. One primary characteristic of interest is that factor scores obtained from orthogonal factors are linearly independent and uncorrelated. Such factor scores were derived and used in a discriminant analysis technique. The Varimax criterion was used to obtain an orthogonal rotation. This procedure maximizes the sum of the variances of squared factor loadings in the columns of the factor loading matrix. The Varimax criterion is defined as: V=mg I; [211314-21 ['3 fjii-]2=max i=1 j=l j i=1 j=1 h. where V = variance of normalized factors, J aji 8 factor loading of variable xj on factor Fp, hj2 = communality of variable xj and j = 1, 2, . . . m variables i = 1, 2, . . . n cases Discriminant Analysis Discriminant analysis is a technique used to find linear combina- tions of variables that maximize the ratio of among-groups to within-group 89 variability. The optimum discriminant function for the two-group situa- tion is that function yielding a linear combination of variables which would discriminate between two groups better than any other linear combination.1 This optimum function, Fisher's, is described by the following matrix equation: Wv = dk where W square matrix whose elements are the sums-of-squares and the sums-of-cross products within the two groups, of the p ori- ginal variables; d = column vector of the differences between the group-means on the p variables; k = arbitrary constant; and v = column vector of weights which satisfy the equation and yield an optimum linear combination. The two-group discriminant criterion can be defined as:2 ssb (Y) = v'Bv ssw (Y) v'Wv where SSb (Y) SSw (Y) between groups sums-of-squares of Y; within groups sums-of-squares of Y; and B = between groups SSCP matrix, and W within groups SSCP matrix. 1Maurice M. Tatsuoka and David V. Tiedeman,"Discriminant Analy- sis," Review of Educational Research, XXIV, No. 5, (December, 1954), p. 402. zMaurice M. Tatsuoka , Multivariate Analysis: Techniques for Educational and Psychological Research, (New York: Wiley, 1971), p. 159. 9O Mahalanobis' D2 statistic is used to measure the "distance" between two groups assuming the populations are multivariate normal with equal dispersions (variances and covariances). Upon failure to reject the hypothesis of no difference between groups, the discrimi- nating functions are calculated according to:1 = + FLMK Z zmkj cmj cmo where FLMK = mth discriminant value for case K in group L; zmkj = observation'(factor score) for each variable (factor); cmj = mth classification function coefficient for variable j; = mth constant; mo L, M-= two groups k = l, 2, . . . t for each L; and j = 1, 2, . . . p factors (variables). Next, the posterior probability of case k in group L having come from group m is computed according to:2 LMK) 151Exp(FL1) where i = l, 2, . . . g functions. Basically, group differences are determined by means of the Mahalanobis D2 statistic and discriminant function values and posterior probabilities are computed and used to classify cases into groups. 1w,J, Dixon, fled.), Biomedical Computer Programs, Berkeley, University of California Press, 1970, p. 214k. 2Ibid. 91 Analytical Objectives The main advantage of factor and discriminant analyses is the capability of assessing and predicting a qualitative dependent variate from a set of quantitative independent variates. Factor analysis techniques may be used to: (l) untangle linear relationships into separate patterns with each pattern appearing as a factor delineating a distinct cluster of interrelated data, (2) reduce a mass of information to its essential meaning, (3) discover the basic structure of a given domain, (4) deve10p an . empirical typology for classification or description; (5) transform data to meet the assump- tions of other analytical techniques and (6) explore.1 At various stages of this research, most of these capabilities were used advanta- geously. Discriminant analysis has the capability to (1) test for signifi- cant differences among average score profiles of two or more a priori defined groups, assuming multinormal distributions and equal disper- sions, (2) determine which variables account most for such intergroup differences in average profiles, (3) find linear combinations of varia- bles which allow the representation of groups by maximizing among-group relative to within-group separation, and (4) establish models for assigning new individuals whose profiles, but not group identity, are assumed to be from one of the a priori defined groups. lR.J. Rummel, "Understanding Factor Analysis," Journal of Conflict Resolution, XI, No. 4, (December, 1967), pp. 449-451. Paul Green and Donald Tull, Research for Marketing Decisions, 2nd ed., (Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1970), p. 368. APPENDIX B TABLES 92 mucoHonmwoo sOHuossm Ne.I mN.I " usmumsoo HH. wO.I " NuHmsoe a Ho>oH sOHumHsnom .ON mm.I NN. " NosoHonmo oHomoHonz .OH II II " owsmno muHmonme waHxamm .OH eN. ON.I " mmcmno maoosH NHHamm .NH II II " eunuosuum aoHumosem .eH II II n eunuHsoHuwm HmoHumaaoo .mH wH. eH.I “ mmaooaH summIaoz .eH I II " x35 wan/S 328m 52 .2 II II " xoesH osso>ou\unoe Housmaauo>ou .NH NN.I NH. u NuHmHooam memuu onmoHogaIHHmuom .HH II II “ NoamHonwo usoaumo>aH waHusuomNssmz .OH ON. mH.I " NuHmHoomm cOHuousem .O mm. me.I u mouom Mao: unmeHmouIcoz .w II II " wstamm .N ee. em.I u unauHeconxo e ossw>mu mhmaanm .e II II " ow< .m II II ” cOHuMHsnoO shoe Hausa nonuo .e Ne.I Nm. ” musuHesmoxo e msco>ou GOHumusem e nuHmmm .m NN. HN.I “ GOHuMHsnon NuHuost .N II II n maumum 0Haoaooo0Hoom .H muusonom Hausumz muusomom sass: " sOHuOHuomon noeasz " muouomm musuosuum UHaosouooHoom one msOHuoouHa ustOOHo>oO kumcmo coosuom manm:0HumHom mafiaonm muaoHonmuou sOHuossh usmsHaHuumHO .HH MHm¢H 93 emssHusoo oHemu mo ecu um ouosuoom mom mo.l mH.eI me.l ee.HI mm.l eN.I mo.HI mm.l u ucmumcoo o~.I «N.H em.I mm. mm.I ee. Nm.I NN. “ Nuamame a H6>oa coaumasaoa .ON II II II II II II Hm. Om.I “ NosoHoHewm onmoHonz .OH II II oe. He.I ma. m~.I we. a~.I " owcmeu muamoame museums .ma No. m~.I mm. mm.I II II II II " «mango 6866:“ Nassau .NH II II eN. Oe.I II II II II " ousuuauum sOHumusem .eH mH.I a~.H II II II II ee.I as. " unauazoaumm Hmuaumano .ma II II mm. Hm.I an. em.I me.I NN. " 66566:“ aumeIcoz .eH II II II II om.I on. II II " xmeaa maa>aa Hanan“ same .ma II II eH.I mm. mm.I as. om.HI wk. “ xonaa maaw>ou\unme Hmucoacu6>ou .NH II II me. mo.I II II ow. we.I " Nuamaumam menus mammmaoeaIHamumm .HH eH.I em.H II II II II II II “ NusmHoHumm u usuaumm>sH msHusuuwmscmz .OH II II II II eH.I mN. II II “ NuHmHooOm sOHumosem .O II II Ne.I mm.a NN.I wo.H mm.I em. " muuom egos aamoammuIaoz .m NN. em.HI II II II II II II " maaxamm .N no.I me. an. mm.I mo. mH.I II II " manuaecoaxm a osco>ou msmsnmax .0 mo. MN.I Ne. No.HI NH. NH.I II II n mw< .m OH. me.HI we. mo.HI mH. NN.I mm. Ne.I " GOHumHamon Show Hanan Hmnuo .e eH.I em.H II II ee. 00.! II II " QHSqucmmxm “ a ossm>ou GOHumoseo e nuHmmm .m HH. NN.I mm. oo.I NN. mm. NN.I NH. ” aofiumasaoa Nuauocaz .N II II mO. wO.I II II OH.I HH. “ unusum 0HaocooooHoom .H 62 H mm» H oz H mm» H oz H mow H oz H mm» H . H . H . MI . H GOHuOHuomon nonasz usmBNOHoam H NuumsecH H HooHeoz .nuHmom H aOHumusem H . . . H mucuomm GOHumusmHuo mousomom amaze . musuusuum UHaoaoooOHoom ecu moocmuomoum sass: sowsuom mOHnmaOHumHom wsHaonm muamHonmooo sOHuossm usasHaHuomHn .NH mHnda mouaomom .ommu ousmmoa some you Hoeoa o>Hu0HeouO m MOM mHmmn one mau0m ecm menu ucmsHEHuumHe oumuwnmm mo muHsmou musmmounou msBSHoo «0 “Hum some .mcoHuow cucH emuum>coo mums memomouO sm£3 sOHusmuum emmmouosH eo>Hmoou max» musmmoa so>Hw m uoc no assumes moumoHesH me .oz 94 H ON.I Ne.I ee.I em.I ee.I eN.I " acoumcoo eO.I OH. ON.I OH. eN. ON.I u NuHmcoe e Ho>oH aOHDMHsOom .ON II II He. eN.I eH. eH.I u NosoHonmo onmmHonz .OH II II NH.I eH. eN. eN.I " swamgo muHmoaoe wstamm .eH II II II II em. He.I " mwamnu maousH NHHamm .NH II II II II mm.I em. " enouusuum cOHuwosem .eH II II eH.I NH. mm.I mm. " musuHsoHuwm HMHuuoaaoo .mH NH. ON.I II II eH.I NH. " moEoosH anamIaoz .eH II II mm.I om. NM.I mm. " amuse waa>aa Nassau 868m .mH mm.I Ne. NH. eH.I HN.I mN. u amesH oscm>ouxueoe Hmuooasuo>ou .NH NN. eN.I NO. NO. HH.I NH. " NuHmHumnm oemuu onmoHogaIHHmuom .HH NN.I NN. eN. Ne.I OO.I OH. "NosoHonmo usoaumo>aH maHusuommscmz .OH NN.I mm. II II eN.I Ne. " NuHoHooam GOHumosvm .O me.I He. II II ON.I HN. u ounce xuoa unmeHmmuIaoz .e eN.I em. MN. HN.I w~.I on. " museums .N II II ee.I On. me.I me. " mousuHesmmxo e osao>ou mNmaanm .e ee. em.I eH.I eH. Nm. mm.I “ ow< .m II II ee. Oe.I II II " sOHumHsaoO snow Hausa uonuo .e eN. ON.I mm. HN.I Ne.I me. " unauHeamOxo " e msao>ou aOHumuseo e :uHmmm .m NN.I Nm. II II Ne.I Hm. " GOHuMHsaon NuHuoaHz .N NH. eH.I eN.I eN. on. em.I " usumum OHBocOUOOHoom .H 62 u was H oz H mm» H oz H mm» H . H . H . H sOHuOHuummn noessz muHuHHHowh NuHasano H waHmsom H =0Humuuoamswua H . . H muouomm aoHumuaoHuO mousommm amass H Asmaaauaoov .NH money 95 aOHuouamHuO uousommm Hmuaumz eosaHuaoo oHemu no use us ouocuOOe mom NN.I me.I OH.I ON.I Nm.mI NN.I Ne.I eN.I " unsumaoo HN.I em. HN.I ee. Hm.HI Om. eN.I Om. " NuHmsme e Hm>mH soHumHsaom .ON ON.I we. Om.I Ne. eN.HI mm. en.I On. “ NosoHUHmeo onmoHonz .OH ON. em.I eH. mN.I ON.H eN.I OH. OO.I “ mwamno muHmonme wachmm .eH NN.I mm. II II Ne.H NN.I II II " owamno oaoocH NHHEwm .NH NO. eO. ON I me. Nm.HI em. me. He.I u ousuosuum GOHumusem .eH eH. HN.I em Oe.I II II HN.I eN. " eunuHsoHuwo HmHouoaBoo .mH II II eH OH.I II II OH. eO.I " moaooaH shamIsoz .eH OH. NH.I OH I NN. II II NN. eH.I " xmeaH waH>HH NHHamm Bumm .NH eO.I OH. eH I ON. Ne.HI me. NN. NN.I “ xoecH msso>mu\unoe Housoaaum>oo .NH mH. eN.I II II II II HN. OH.I " NuHMHomOm oemuu onmmHonaIHkumm .HH eN.I ee. II II II II eN.I NN. u NoaoHu ” IHmmo uaoaumo>cH waHusuowmssmz .OH eN.I He. NO OH.I II II Ne. oe.I " muHmHomam GOHumosem .O eN.I oe. II II II II Om.I Ne. “ ounce xuoa acoeHmouIaoz .m mm. me.I II II em. NN.I eH.I NH. " wcHxsmm .N NN. ON.I HN ON.I Hm.H ON.I NH.I OH. " eunuHeaomxo e osao>ou ohmsanm .e mm. Oe.I II II II II we. He.I " mw< .m me. mN.I II II OH.m He.I mm. Om.I “ sOHumHsaoa spam Housu noeuo .e Ne.I NH.H Ne I am. II II oe. He.I “ unauHeaonxu " a osso>ou aOHumoseo e :uHmom .m OH. NN.I II II II II ON. NH.I " GOHumHsnom NuHuost .N mH. eN.I II II II II eN.I ON. " unusum UHaoaoooOHoom .H 02 m mu» m oz m mow m oz m mm? m oz m no» N . H . H . H . H oOHuOHuommO nonasz ousuHsUHuw< H umum3 H ean H unmaaouH>am H . . . H neuumm ouauusuum UHBosoooOHoom ecu moosoummonm usuaOon>mO mousomom .NH mHmHoomu onhu museums ao>Hw m uoc no nonuoss mouooHesH mom .oz .mOOu whammoa some now Hoeoa o>HuoHeouO w you mHmme onu mauoe .m=0Huuo eusH eouum>soU mums H eN.I ee.I ee.I eN.I eH.I ms.I m sausages Ne.I ON. II II HH.I HN. H NuHmsme e Ho>mH cOHumHsOom .ON NH.I eN. Oe. ee.I NH.I mm. H OosoHonmm onmoHO£3 .OH NH. NN.I II II eH. eN.I H mesons muHmoaoe NcHxssm .eH OH.I ON. II II eO. NO.I H mesons oaooaH OHHamm .NH OH.I eH. mm. eO.I eH.I Om. H ousuusuum coHumosem .eH ON. eO.I II II II II N musuHsoHuwm HmUHuanoo .mH NN. eO.I II II II II N moaooaH BummIcoz .eH Na.I SN. II II ma.I mu. m enema waH>HH Naaaum auam .mH Oe.I me. II II eH.I ON. . xoeaH asao>ou\unoe Hmuauasuo>ou .NH ON. HN.I II II OH.I OH. H OuHsHoonm oemuu onmoHosaIHHmumm .HH II II eO.I mm. OH.I eN. ”NoaOHonmo usuaumo>sH wsHuauummsamz .OH eH. HN.I II II ON.I mm. H NuHmHooum sOHumosem .O ee.I HN. ON.I Hm. ee.I Oe. H ounce Hues usoeHmmuIsoz .w ea. NN.I II II om. Am.I H museums .N eN. NO.I II II II II N musuHesmnxm e osco>mu ohmsanm .e II II II II mm. ee.I H 6w< .m mm. ON.I mm. eN.I eH. NN.I H sOHumHsOom snow Hausa nonuo .e II II mo.a HN.I NH.I an. m manuaeaoaxm . a osso>ou :OHumoseo e :uHmmm .m eH. eN.I II II II II N sOHuMHsmoa OuHuost .N NH.I eN. II II NH. HN.I H msumum 0Haos0000Hoom .H oz m mow H oz H mow H oz N mm» H . H . H . H sOHuOHuummO nonasz unmaOOHo>oO .wcHssmHmu :OHummuuom H Ouumouom H . . H muouumm soHumucmHuO oousommm Housumz H asuaaauaouO .ma mam00 0600000060 .00 00.0 00.0 00.0 00.0 00.0I 00.0 00.0I 00.0 00.0I 00.0I 00.0I” 0600060000 000000603 .00 00.0 00.0 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0 00.0I 00.0 0 000000 00006000 0000000 .00 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0 M 000000 006600 000000 .00 00.0 00.0 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0 00.0 0 000060000 060006000 .00 00.0I 00.0I 00.0I 00.0 00.0I 00.0I 00.0I 00.0 00.0 00.0 00.0 H 00000060000 0006000060 .00 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 0 006600 0000I0oz .00 00.0 00.0 00.0I 00.0I 00.0I 00.0 00.0 00.0 00.0 00.0I 00.0 M 600000 000>00 000000 0000 .00 00.0 00.0 00.0I 00.0I 00.0I 00.0 00.0I 00.0 00.0 00.0 00.0Iu 00000 000060000000 0000006660 .00 00.0 00.0I 00.0 00.0I 00.0 00.0I 00.0I 00.0I 00.0 00.0 00.0 M 000006000 00000 000000603I000000 .00 00.0I 00.0I 00.0I 00.0 00.0 00.0 00.0 00.0 00.0I 00.0 00.0 0 0600060000 0000000>00 000000600000: .00 00.0I 00.0 00.0I 00.0 00.0 00.0 00.0 00.0I 00.0I 00.0 00.0 H 000006000 060006000 .0 00.0I 00.0 00.0 00.0 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0 0 06060006: 00000000062 .0 00.0 00.0I 00.0I 00.0I 00.0 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I“ 0000000 .0 00.0 00.0 00.0I 00.0I 00.0I 00.0I 00.0I 00.0 00.0I 00.0I 00.0I" 00600000 00030000 .0 00.0I 00.0 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I“ 000 .0 00.0 00.0 00.0 00.0I 00.0I 00.0 00.0 00.0I 00.0I 00.0I 00.0 0 0600000060 0000 00000 00000 .0 00.0I 00.0I 00.0I 00.0 00.0 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I” 00600000 060006000 0 000000 .0. 00.0I 00.0I 00.0 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I 00.0I" 0600000060 0000600: .0 00.0I 00.0I 00.0I 00.0 00.0 00.0I 00.0 00.0I 00.0I 00.0 00.0 M 000000 6006066660660 .0 HH OH O m N e n e m N H m oawz ens nonasz 000000 00000002 06000600000000 0000 0000 " x0000: ououm heuomm uoofioum .eH mHn00 0600000060 .00 00.0 00.0- 00.0- 00.0- 00.0- 00.0- 00.0 00.0- 00.0- 00.0- M 0600060000 000000603 .00 00.0 00.0 00.0 00.0 00.0- 00.0 00.0: 00.0- 00.0- 00.0- 0 000000 00006000 0000000 .00 00.0- 00.0 00.0 00.0 00.0- 00.0- 00.0- 00.0 00.0 00.0 M 000000 006600 000000 .00 00.0 00.0- 00.0- 00.0- 00.0- 00.0- 00.0- 00.0 00.0- 00.0 0 000060000 060006000 .00 00.0 00.0 00.0- 00.0- 00.0- 00.0- 00.0 00.0- 00.0- 00.0- M 00000060000 0006000060 .00 00.0- 00.0- 00.0- 00.0- 00.0 00.0- 00.0- 00.0 00.0 00.0 0 006600 0000-062 .00 00.0 00.0 00.0- 00.0- 00.0- 00.0 00.0- 00.0- 00.0- 00.0- M 06000 000000 000000 0000 .00 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 0 00000 000000000000 0000000060 .00 00.0 00.0- 00.0- 00.0- 00.0 00.0- 00.0 00.0- 00.0- 00.0 M 000006000 60000 000000603-000000 .00 00.0 00.0- 00.0- 00.0- 00.0 00.0- 00.0- 00.0- 00.0- 00.0- 0 0600060000 0000000000 0000006000002 .00 00.0 00.0- 00.0- 00.0 00.0- 00.0- 00.0 00.0 00.0- 00.0 N 000006000 060006000 .0 00.0- 00.0- 00.0- 00.0- 00.0 00.0- 00.0 00.0 00.0- 00.0 0 060600063 00000000060 .0 00.0- 00.0 00.0 00.0- 00.0 00.0 00.0 00.0 00.0 00.0- M 0000000 .0 00.0 00.0 00.0- 00.0 00.0- 00.0- 00.0 00.0 00.0 00.0- 0 00600000 0030000 .0 00.0- 00.0 00.0- 00.0- 00.0- 00.0- 00.0- 00.0 00.0- 00.0- M 000 .0 00.0- 00.0- 00.0- 00.0- 00.0- 00.0- 00.0- 0000- 00.0- 00.0 0 0600000060 0000 00000 00000 .0 00.0 00.0- 00.0- 00.0- 00.0 00.0 00.0- 00.0- 00.0- 00.0- M 00600000 060006000 0 000000 .0 00.0- 00.0- 00.0- 00.0- 00.0 00.0 00.0 00.0- 00.0- 00.0- 0 0600000060 00006000 .0 00.00 00.00- 00.0- 00.0- 00.0: 00.0- 00.0 00.0- 00.00- 00.0- H 000000 6006066060660 .0 00 00 00 00 00 00 00 00 00 00 m 0000 000 000000 060600 Hmumnaflz QOHumUflwfiUH—QVH mmh< Qwum 0 00000000660 .00 00000 99 «o.a N¢.H mq.H- ~m.o wm.o mH.o NH.H- HH.o -.Hu om.H " %ufimamn a Hm>mg coaumfiaaom .ON mc.H wo.Hu mm.o- q~.H m~.H H¢.o- oo.~ ao.~ mm.H Nq.o M muamfiuammm mammmaosz .mH mm.ou ew.o oo.o qq.Hu a~.~u mq.ou mm.o ¢~.ou mq.H Ho.on " unsung mufimoamn wafixcmm .wH qq.ou mh.on om.o- ma.ou «H.o ~m.~ q~.o- oc.H o¢.ou o~.o M «mango «sooaH “Hflamm .NH eH.ou mH.o ca.H wo.o Ne.H mo.o ofl.ou mo.H oq.H ¢~.~ " ousuuauum aoflumuaom .OH -.o w~.~ 00.0 NH.o ~m.ou HH.o Ne.H- «m.o mm.o mm.H- M auguH=UHum< Hmfiuumaaoo .mH No.~- om.¢- qo.a- Ne.H- No.o mm.o Ho.H “w.o ow.oa «H.ou " meouaH anam-aoz .«H om.H- wq.Hu Hm.H- Hm.ou o~.o- ¢N.o m~.H mm.Hu oo.au «m.on H xmuaH mafi>fiq mafiamm aumm .MH eN.HI mh.ou wH.H- oq.H- oq.~: No.o mq.H- Ho.H mo.H «o.Hu " xmuaH wsam>mm\upoa unmacuo>oo .NH mm.ou ~m.ou H¢.H No.o wH.o HH.~ mH.o «m.Hu Ne.H oq.o M huamfluugm «cape mamaoaonzuaflmumm .HH «H.ou mm.ou eq.H oo.o no.o mq.o- qo.au m~.o mm.o mH.~ " “camHUfimmm unmaumm>cH mafluauummsamz .OH qq.¢u eo.m- a~.o- H~.m- o~.H ma.o om.~ m~.ou mq.o oc.~ M zuamfiooam aofiumuaum .m ma.ou o¢.o ~o.ou m~.o co.H sq.~ «H.H «H.H Nm.o qo.o “ muuowxuoz gamvfimaucoz .w NH.ou «H.Hu am.ou eH.o- eN.Hu mo.ou Ne.H- Hm.o om.ou mH.o M wafixcmm .5 m~.H Nm.o Hm.Hu Hm.H mm.ou mo.H- ca.Hu HH.H- qm.au -.o- " mmocmcfim “msnwfim .0 qm.~ «H.q mq.ou oc.~ om.~ mm.m ~m.o Hm.m No.o ao.o M mm< .m “H.ou Ne.H- NH.o qm.ou oq.o NH.H- -.o oN.H- oa.o mo.o " coaumaanom sum» Hausa aunuo .¢ -.o ma.o NH.H ow.o o~.o q~.o mq.ou mH.o mw.o mm.o M mmuamaflm aofiumusnm a suammm .m no.m No.m mN.~ nm.m cm.H Hm.H Ho.o nfl.¢ H~.¢ me.~ “ cofiumaaaom muauocfiz .N mw.hu N~.~Hu om.m oo.n- om.~ Na.~- mq.e Hm.ou m¢.~ mH.o M maumum afiaoaoumoaoom .H an on mm mm “N on mm «N mm mm M memz can umnaaz uouomm Hmuwnfinz GOHHQUHMfiuflwUH wmu< Qwum u Auoacaucouv .eH mammq coaumasaom .ON H~.Nn mm.o no.0: mm.H om.o mm.o mo.o m¢.o ww.H «5.0 H moamfiuammm mammmHosz .¢H um.ou Nq.on mH.H qq.H N<.H c~.m wo.o mm.ou Ho.OI «H.ou “ mwcmso mufimoamo wcaxcmm .ma mm.ou mo.~n no.0: mm.o mo.o 00.0 ow.H mm.o: m~.ou om.o M mmcmnu «aouaH hawamm .NH no.0: cm.ou mm.o wN.Hu m~.Hn Hm.ou mm.ou mH.o «H.H| Hm.H " mucuoauum aoflumusvm .eH mm.au q~.N| mm.o n~.Ha mm.H NH.H Nc.o ~m.~ Hm.o ¢H.o M manuasofiuw< Hmofiumaaoo .ma om.H mc.Hu No.o mm.ou Ho.H mn.0| Ho.oa am.a| mq.o Hm.~| " maoucH Bummucoz .qa «H.H «H.Hn ow.o mm.o mo.o om.ou mm.Hn mm.ou mm.o ow.o M xmwcH wcfi>wq hafiamm Bump .MH om.o mq.Hn m~.o 0m.o qm.m| om.H oo.~| mo.ou $0.0 eN.Hu " xmwaH maam>om\unwa unmacum>ou .NH mH.N mm.on mH.o mm.o «H.o co.H o~.o -.H| o¢.H mm.a| M %uamfiomam mummy mammmaonzlaaaumm .HH oo.o mm.o om.H cm.on mm.Hn n~.ou an.o cm.ou oo.on Hw.ou ” hocmfioammm unmaumm>cH wcwuauommaamz .OH mm.~ mo.ou mm.m «o.au mo.m n~.ou NN.HI cm.mn w~.~| No.m| H huamuooam coaumusvm .m mq.Hu -.o oq.o Hm.au -.o om.H -.H o~.ou om.o wq.o " muuowxuo3 ucmvfimouaoz .w oo.H om.Hu Ne.H mo.H mm.o ~¢.H| oe.ou oq.o mc.H o~.ou M waaxcmm .m mo.on mw.a ww.o mm.~ <~.H| Hm.~ mo.o mm.Hu mm.o on.H " mmuamcam mzmsnwfim .o mn.Hu mq.on mn.o ¢o.~| om.o mo.Hu mH.H mm.~ oo.o oq.~ M mw< .m eN.OI m~.o mm.ou cm.ou qm.o -.m mH.ou mw.on ma.on m~.on " aofiumaanom aumm Hausa umnuo .o No.o m~.on nm.a mH.H wn.o co.ou mm.o nH.on Nw.o ~m.o H mmoamaam aowumosvm m nuammm .m ¢n.H| wm.au Ho.mn mm.o on.o «H.o m~.c oo.a am.~ NH.~ " coaumasaom mufiuocfiz .N ow.q ma.q mm.w «H.H mm.o qa.mu oo.~ mm.wn om.~| ma.n| M msumum oaaoaoumowoom .H He cc ”muwnsmm scammu«Mfimmme MMH< owwm mm mm H mamz can umnaaz uouumm AvosaauaouV .eH mgmmq cofiumflsaom .ON mm.ou mm.H om.Hu Ho.Hu Hq.on eq.o wq.Hu zoawflOHmmm mammmaopz .mH w~.o mo.H oo.o ~o.au mn.H- m¢.ou mm.H mwcmnu mugmoamo wafixcmm .wH qq.as mH.o m~.o no.0 H~.N «w.ou oo.ou «mango maouaH »Hfiamm .NH om.o Hq.o mH.o qo.a NN.ou oo.m- HH.o muauusuum aofiumosum .eH oo.~- 00.0 mm.ou Ne.H co.H «m.ou ~¢.ou ma:~a:uauw< Hmfiuuoaaou .mH mm.o mq.a «o.H «o.a «m.H cm.H- ao.~ msoocH summucoz .qfl ~m.o mm.o- mo.¢ mH.on Nm.ou mo.H mq.o noucH mcfi>fig xafiamm sump .MH ~m.o -.ou mm.N- m~.o ON.o mm.o mq.ou xmwaH oaao>om\unmo ucmamum>oo .NH oq.o mm.a mw.o Ho.Hu mm.ou H~.o ¢~.ou xuamflomam «vane mammoflonzuafimuom .HH mm.~u on.H mo.m H5.ou mm.o om.o mH.Hu mocmHUfimmm unmaumo>aH mafiuauummscmz .OH No.ou on.m NH.m «H.m oa.o ~q.o mn.H muaaaooam aofiumosum .m w¢.H- m~.o- Na.o- mq.an -.H- sq.o «q.ou monomxuoz ucmvflmmuaoz .m m~.o mm.H- ~q.ou me.H- mH.~ No.o N¢.o maflxamm .5 co.H mm.~- mm.o- «o.a- mH.Hu mo.ou m~.H mmoamaflm mumsnwfix .o mo.au qo.o no.m- «N.qn so.m m~.~ mq.on mm< .m Hm.ou -.c «m.H mm.m mm.o- Hm.~ om.ou aofiumasaom aumm Hmuam “mayo .q w~.ou mm.au o~.o wo.o ~c.ou q~.~ o~.o mmuamcfim aoflumusnm a auammm .m mm.Hu 0N.N- mm.~- Ho.Hu oo.ou mm.o- eN.H- cofiumasqom zufiuoafiz .N mH.oa «m.~ o~.cH mw.~H en.H1 m~.~ 0H.N maumum afiaocoomofluom .H ma Ne Hmumnamm coflquMWfiuamcH wwu< awommq Nq mamz can umaaaz uouomm Auuaafiuaouv .eH mgm