137' HOMOPHILY _ . IN THE DEFFUSEON 0F lNNOVATlONS IN BRAZILIAN COMMUNiTlES Thesis for the Degree of M. A MICHIGAN STATE UNIVERSITY YUNG'CHANG H0 1969 LIBRARY ' «mews Michigan State University . u [in um; {Hzllfljflflfll gm M! 1'!“ my, 11:, mug a ABSTRACT , HOMOPHILY IN THE DIFFUSION OF INNOVATIONS IN BRAZILIAN COMMUNITIES ‘ by Yung Chang Ho The present thesis related homophily with personal, socio- economic and demographic variables in interpersonal communication in the diffusion of technological innovations in 20 Brazilian communities. Homephily was defined as the degree to which pairs of interacting individuals are similar in certain characteristics such as mass media exposure, change_agent contact, cosmopoliteness, literacy, innovativeness, and opinionatedness. The objectives were (1) to define and measure general homophily in each of 20 communities, and test what variables are related to homo- phily, and (2) to relate community homophily to aggregate communication characteristics: mass media exposure, change agent contact, cosmopo- liteness, innovativeness, literacy, opinionatedness, and a non-communication characteristics, community economic development level. The present study examined relationships between two indicants of community homophily: community innovative and change agent homophily; and seven indices of modernity: community mass media exposure, change agent contact, cosmopoliteness, literacy, opinionatedness, innovativeness, and economic development. Since these are indicants of modernity, and traditional communities have a high degree of homophily, we predicted a negative relationship between the seven modernization variables and community homophily. The independent variables were mass media exposure, change agent contact, cosmopoliteness, literacy, innovativeness, opinionatedness, and community economic development. Dyadic analysis was used to investigate relationships among pairs of individuals. Zero-order Pearsonian correla- tion was used to test all hypotheses. The data analyzed in the present study were collected in a previous research project, "Diffusion of Innovations in Rural Societies." The project was funded by the United States Agency for International Develop- ment in cooperation with the Department of Communication, Michigan State University, under the direction of Dr. Everett M. Rogers in 196u—1968. There were 1,307 subjects interviewed in the Brazilian project, and 3,3uo dyads were found. Of these, 1,604 were friend dyads, and 1,736 were in- fluence dyads. None of the 28 theoretical hypotheses were confirmed. Two were statistically significantly different from zero, but in the direction opposite to that which was predicted. The remaining 26 were not statis- tically significant. There are at least two approaches which deserve attention: (1) In- vestigation of those variables negatively related to four types of homophily in the present study. These variables are movie exposure, TV exposure, newspaper and magazine readership, radio exPosure, change agent contact, opinionatedness, and innovativeness. (2) It might be fruitful for future researchers to use a combination of dyadic and individual analysis. Accepted by the faculty of the Department of Communication, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Master of Arts degree. £77174? Director of esis HOMOPHILY IN THE DIFFUSION OF INNOVATIONS IN BRAZILIAN COMMUNITIES BY Yung Chang Ho A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Communication 1969 (‘j ‘x a 7 "3g/z" 99/ l/Oc: {, 7 ACKNOWLEDGMENTS The author is most grateful to his advisor, Dr. Everett M. Rogers, for his advice, cooperation, and encouragement, without which this thesis would have been impossible. The author is also indebted to Dr. Gerald Miller, and Dr. Paul J. Hiniker, who not only served on the writer's M.A. Committee, but also gave their valuable suggestions and constructive criticism. Sincere appreciation goes to Dr. J. David Stanfield for his valuable counsel and guidance. The author is also indebted to the united States Agency for Inter- national Development because the data of the present study grew out of the research project, "Diffusion of Innovations in Rural Societies," which they sponsored, and which was conducted by the Department of Communication, Michigan State University, under the direction of Dr. Everett M. Rogers. Gratitude is also expressed to Michigan State University for its financial assistance and to the National Science Foundation for computer services. Special thanks are extended to Mrs. Anita Immele, the computer programmer, for her help, and to Betty Darlington for her assistance in punching IBM cards. A final word of appreciation to my wife for her understanding and to my son, Peter, who had to live with grandparents in Taiwan during the first year of my graduate studies. ii CHAPTER TABLE OF CONTENTS I INTRODUCTION .0.0.00.00.00.00.00000000000000000000000 Background and Reason for the Study ............... Limitations of Previous Interpersonal Communication Research .......................... Objectives of the Present Study ................... II RATIONALE AND THEORETICAL HYPOTHESES ................ ATheomtical Rationale OOOOOOOOOOOOOOOOOOOOOOOO0.. Homophily in Traditional Community vs. Heterophily in Modern Community ................. Theoretical Hypotheses and Modernization Variables. Community Mass Media Exposure ................... Community Change Agent Contact .................. Community Cosmopoliteness ....................... Community Literacy .............................. Community Opinionatedness ....................... Community Innovativeness ........................ Community Economic Development Level ............ summw 0.0000000000000000000000000000000.0.0.... III IETHOWLOGYhooooooooo00000000000000.0.000000000000000 The Data 0.0.0.0000...OOOOOOOOOOOOOOOOOO0......O... Criteria for Selecting Communities................. Criteria for Selecting Subjects ................... Method and Results of Data Gathering .............. The Measure of Homophily: Pearson Product-Moment Correlation Coefficient Analysis ................ Operationalization of Variables ................... Emirical Hypotheses 0......OOOOOOOOOOOOOOOOOOOOOOO Iv FINDINGS 0.0.0....OOOOOOOOOOOOOOOOOOOOOOOO0.0.00.0... Measummnt Of Homphily 0......OOOOOOOOOOOOCOOO... Pearson Product-Moment Correlation Coefficient Analysis....................................... Factor Analysis ................................. Correlates of Homophily ......................... Results of Intercorrelation between Homophily Indices and All Independent Variables.......... Test of Hypotheses Dealing with Homophily ....... iii PAGE l2 12 13 in l“ 15 16 l7 18 2O 20 20 21 22 23 2t: 29 31 31 31 32 38 38 H1 TABLE OF CONTENTS -- CON'T CHAPTER V SUMMARY, ADDITIONAL ANALYSES, DISCUSSION OF HYPOTHESES, AND SUGGESTIONS FOR FUTURE RESEARCH... Summary ........................................ Additional Analysis ............................ Discussion of the Hypotheses ................... Suggestions for Future Research ................ BIBLIOGMPHY O...OOOOOOOOOOOOOOO0..IOOOOOOOOOOOOOOOOCOOOOOO MPHDICES 0...0.0....IOIOOOOOOOOOOOOO.0.0.0....OOOOOOOOOOOOOO. iv PAGE 55 55 56 58 63 65 7O LI ST OF TABLES Results of dyad-gathering ....................... Pearsonian correlations of seekers' 20 variables to soughts' 20 variables in Sao Joao Nepomuceno community on the basis of friendship (modern community (N=109) ............................. Pearsonian correlations of seekers' 20 variables to soughtS' 20 variables in Sao Joao Nepomuceno community on the basis of influence (modern community N=l27) .............................. Pearsonian correlations of seekers' 20 variables to soughts' 20 variables in Corinto community on the basis of friendship (traditional community N=loo) OOOOOOOOOOO...OOOOOOOOOOOOOOO0.0.00.0... Pearsonian correlations of seekers' 20 variables to soughts' 20 varibles in Corinto community on the basis of influence (traditional community N=121) OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Varimax rotation: factor loadings for homophily iten‘s 00.....0...OOOOOOOOOOOCOOOOOOOOOOOOOOOOOO A comparison of Pearsonian correlations between innovative homophily on the basis of friendship or influence and modernization variables ...... A comparison of Pearsonian correlations between change agent contact homophily on the basis of friendship or influence and modernization variables ..................................... The zero-order Pearsonian correlations between innovative or change agent contact homophily and opinion leadership concentration or size of community .................................. PAGE 28 33 3a 35 36 37 39 no 59 LIST OF FIGURES Figure Page 1 Paradigm of information-seeking relationship between seeker and sought in different levels of competence in new technology and direction of information flow in modern social system .... ll 2 Paradigm of the relationships between modern— ization variables and homophily ................ 19 vi CHAPTER I INTRODUCTION Background and Reason for the Study Numerous research studies on the diffusion of innovations, i.e., the process by which information about an innovation spreads among members of a social system (Rogers, 1962, p. 13), have been completed by researchers in communication and rural sociology. A rich body of literature exists in the field of diffusion of innovation which relates behavioral change among selected individuals in certain social systems. However, researchers still know very little about: 1. How structural characteristics of a social system influence the diffusion of innovations within the social system. Katz (1963, p. 239) clearly indicated that there is a need for research designed to analyze how differential characteristics of the communication structure influence the diffusion of innovations within one social system as compared to others. 2. Determinants of the role of interpersonal communication in the diffusion of innovations. There is overwhelming evidence for the importance of interpersonal communication in the diffusion of innovations in less developed countries. Pye (1963, p. 27) emphasized the relative importance of interpersonal communication in less developed countries by saying that, "It is only necessary at this point to make it clear that the process of development is less dependent upon increased invest- ment in the modernized, urbanized mass media system than it is upon the adjusting of the informal, rural systems to each other and to the mass media system." 3. What variations in communication structure influence the level of develOpment in social systems. We need a better understanding of communication structures in comparative social systems with varying degrees of traditionalism and modernity. From a practical point of view, the present study is eSpecially important to change agents whose ultimate goal is to disseminate new ideas through a social system to all receivers in the shortest period of time with desired results and at least cost. Therefore, to be effective it is imperative for change agents to have an understanding of the social system's structural characteristics influencing the diffusion and adoption of innovations, particularly in less deve10ped countries. A model of homophily indicates why individuals tend to interact with others who have similar characteristics in the same social system. So also from a theoretical point of view, a model of homOphily is clearly needed in the field of diffusion of innovations. Rogers with Svenning (l969, p. 236) pointed out that, "In one sense, homophily acts as an invisible barrier to the flow of innovations within a social system. New ideas usually enter a village via higher status peasants, who tend to have higher mass media exposure and cosmOpolite relationships." In traditional communities, they further indicated, "These higher status peasants tend to Spread new ideas horizontally within a village to others of high status, therefore, innovations trickle down very slowly and indirectly to lower status peasants. Homophily can slow down the rate of diffusion of innovation in a system." An example of homophily as a barrier to the trickle down of mass media messages was given by Lerner (1958, p. 192) when he interviewed a lebanese villager who said, "TownSpeOple do listen, but I don't know them. Rich peOple have radios (but) none around me have radios." Research in measuring personal influence has been mainly concerned with identifying and classifying the opinion leaders and followers; little effort has been made to find relationships between individual seeker-sought dyads,* i.e. why some pairs of individuals tend to interact with each other. Limitations of Previous Interpersonal Communication Research We know much about the diffusion of innovation process in the United States. For instance, there are more than 1,500 diffusion publications available at the Michigan State University Diffusion Documents Center (DDC) as of January, 1969. New diffusion research papers come to the BBC at the rate of one every two days. Of these research papers, about 63 percent were done in the U.S. These past diffusion research studies have not focused on analyzing the characteristics of interpersonal communication networks in the *A dyad consists consists of a seeker and a sought. The "seeker is the one who seeks information from others who are either his friends or influ- entials. The "sought" is one who provides information when asked by some— one who seeks information. diffusion of innovations, but rather have focused on studying the inter- personal channels. Chou (1966, p. 7) found that none of the 37 percent of the diffusion studies completed outside of the United States investi- gated dyadic communication. A few researchers (Rogers and van Es, 196u; Lionberger, 1963; Warland, 1963; Feldman, 1966; and Patel, 1966) studied dyadic relationships. A further review of diffusion studies shows that only six dealt with the role of interpersonal communication in the diffusion of new ideas. Of these studies, only three (Patel, 1966; Chou, 1966; and Yadav, 1967) investi- gated the relationship of interpersonal communication variables to the diffu- sion of innovations across several social systems. The other three dealt with diffusion of educational innovations in school systems in the United States (Eibler, 1965; Davis, 1965; Lin, 1966). Many limitations of these previous studies were pointed out by the researchers themselves. Yadav (1967, p. 15) said that his study was com- parative in nature; nevertheless, it was limited to the analysis of communication structure in only two Indian communities. Therefore, certain statistical limitations are inevitable. Such was also the case with Chou's thesis (1966, p. n), which was concerned with interaction patterns in only three Colombian villages in the process of the diffusion of innovations, using only three kinds of variables: Level of competence, communication contact, and social status. Patel (1966, p. 212) only compared the dyadic pairs in social status and competence. The associa- tion between interpersonal communication and such variables as education, mass media exposure, change agent contacts, opinionatedness, literacy, cosmOpoliteness, age, social status, and innovativeness of the receivers was not measured. Patel (1966) suggested that a complete sociometric survey of a system would give a more accurate picture of the flow of communication. Feldman (1966) was concerned with characteristics of the interpersonal communication dyad formed during the physician- selection decision process. Feldman (1966, p. 758), explored the dyadic relationships between the influences (infOrmation seeker) and the influencer (referent) in the process of selecting a physician, and then compared "age and social status characteristics of those influentials who advise in the selection of a physician, the purchasing of household goods, and the purchasing of fashion goods." Objectives of the Present Study The present study is concerned with personal, socio-economic, and demographic variables in interpersonal communication in the technological diffusion of innovations in 20 Brazilian communities. To avoid statistical and other limitations of previous studies, 20 purposively-matched com- munities in Brazil are included. The objectives of the present study are: 1. To define and measure general homophily in each of 20 communities, and test what variables are related to hom0phily. More Specifically, we want to define and measure community homophily. 2. To relate community homophily to such aggregate communication characteristics of 20 communities as opinionatedness, literacy, mass media exposure, change agent contacts, cosmopoliteness, and a non-communi- cation characteristic, community economic development level. CHAPTER II RATIONALE AND THEORETICAL HYPOTHESES A Theoretical Rationale In the present chapter, we shall present a theoretical background to support the relationship between the independent variables and the degree of homophily in dyadic communication in rural Brazilian communities. Also, we shall explain why certain types of communication structures exist in certain types of communities. General homophily is defined as the degree to which pairs of indivi- duals who interact with others with similar characteristics in the same social system (Lazarsfeld and Merton, 195M, p. 23). These characteristics in the present study are literacy, mass media exposure, opinionatedness, change agent contact, innovativeness and cosmopoliteness. If the two interacting individuals are alike in some characteristic, such as mass media exposure, we consider them to be homophilous. If two pairs of individuals who interact with each other are different in certain characteristics, then they are heterophilous, the opposite of homophily (Lazarsfeld and Merton, 195%, p. 23). Homophily in Traditional Communities vs. Heterophily in Modern Communities Previous studies in the United States and Netherlands (Lionberger, 1957 and van den Ban, 1963) showed that when there are more distinct differences in the degree of homophily between two individuals in a social system, interaction is less likely to occur_between them. Rogers and van Es (1964) computed an index of homophily* for the seeker-sought relationships in the Colombian modern and traditional villages. "This index is a measure of the degree to which seekers interact with adopter categories** quite different from themselves" (Rogers and van ES, 1968, p. us). The authors found that the homophily index was .80 for the two traditional Colombian villages, and .68 for the three more modern villages. Rogers and van Es (196n, p. #2) also found that in Colombian villages "In .general there was a marked tendency fbr 'seekers' to name 'soughts' who are in a similar adopter category. In fact, 117 of the 177 sociometric relationship (66 percent) depicted are to the same or an adjoining adopter category". According to these findings, it seems reasonable to conclude that: (l) Villagers in the two Colombian traditional villages tend to interact with others who are similar to themselves while villagers in the three modern villages tend to interact with those who are quite differ- ent from themselves, and (2) generally speaking, in all five Colombian villages, seekers tend to interact with others with similar characteristics. *Rogers and van Es (196%, p. #3) reported that "this index of resis- tence to communication flow was competed as the average length of the socio- metric arrows (the vertical distance in Figure u) between seekers and soughts. When the average sociometric choice is across fewer categories, the index is greater. For example, the index of resistence in the three modern veredas (villages), is .68, which indicates that the average sociometric choice is across more adapter categories (Figure a) than in the two traditional veredas, where the index is .80. " **Rogers (1962, pp. 168-171) defined adopter categories as innovators, early adopters, early majority, late majority, and laggards. Bose (1962) found a very marked degree of homophily among traditional Indian villagers on the basis of caste ranking, education, and size of farm. In the nearby city of Calcutta, there was a high degree of homo- phily based on income. Dasgupta (1968) found that in a highly structured societies in Uttar Pradesh, India--traditional communities--the mere presence of stratification in terms of caste retarded interpersonal communication between people. This restriction on whom can be talked to is an indicant of a high degree of homophily in these traditional communities. Harland (1963) reported that on the basis of his study of 28 socio- metric relationships among Iowa farmers individuals tended to interact with those who have similar attitudes, levels of technical competence, and socio-economic status. Katz and Lazarsfeld (19MB) and Merton (1957) supported the "likes-to-interact-with-likes" hypothesis, that individuals will interact with those who are similar to themselves. Chou (1966) examined the concept of homophily in the context of Colombian villages. She found little relationships between individuals' homophily based on competence, communication contact, or social status and their degree of interaction. Previous research studies indicate that a seeker of information in traditional communities will go to somebody similar to himself for most of the information he is looking for. This pattern of behavior is also made possible by less specialized and functionally differentiated roles in traditional communities. In short, there is a high degree of homophily among individuals in traditional communities. 0n the other hand, we expect that roles are more specialized and functionally differentiated in modern than in traditional communities (Rogers and van Es, 196u, p. 19). The more modern communities become, the more specialized the roles are likely to be. For example, as a traditional community becomes modern, an opinion leader may still be expert in farming but can not keep up with innovations in other fields. As change agents and mass media introduce more innovations, the process of modernization will accelerate. Very soon, the opinion leader may become highly specialized, since modernization has imposed complexity. Under these circumstances, the seeker of information can no longer go to one opinion leader like himself for answers to all of his questions. It is now beyond the latter's ability to answer all the questions because of specialization of roles in more modern community settings. Emery and Oeser (19n2) conducted studies of opinion leaders in four townships in Iowa, and concluded that opinion leaders are usually monomorphic. Rogers and Cartano (1962) suggested that opinion leaders are more likely to be polymorphic in traditional communities. They con- cluded that this was probably due to a greater separation of roles in more modern communities than in traditional communities. Katz and Lazarsfeld (1955) found that in the U.S., certain individuals are considered experts in one field, but not another. In other words, if an individual is an expert in fashion, then he can not be an expert in other fields such as movie-going, public affairs, and marketing at the same time. lO Merton (1949) concluded that opinion leadership is not a general characteristic of a person, but is limited to a particular issue. More importantly, Lionberger (1953, 1957) found that individuals who are sought for information are more competent than those who seek information. Rogers with Svenning (1969) observed "a general tendency for opinion seekers to obtain infOrmation and opinions from opinion leaders who are more com- petent than the opinion—seekers in technical knowledge, innovativeness, etc. Thus, when heterophily occurs, the opinion-seekers interact with opinion leaders who are more innovative but not too much so." Figure 1 shows a paradigm of information-seeking relationships between seeker and sought at different competence levels with technology and direction of information flow in modern social system. Therefore, when an individual seeks advice on a variety of topics in a modern community, he will likely find others different from himself in terms of, for example, technical competence, innovativeness. He will also find them different due to specialized roles in a modern community. In short, there is a high degree of heterophily in modern communities. High degree of competence in new technology Medium degree of competence in new technology Low degree of competence in new technology ll Technological information via external forces such as mass media, change agents 4; hAaél ‘)k b3 XJ’CCI3 \ ’(Dd D Villagers in Modern Social System Figure l. Paradigm of infbrmation-seeking relationship between seeker and sought in different levels of competence in new technology and direction of information flow in modern social system. A, B, C, D....N-l, N Stand for individual villagers who could be a, b, C, decon-l, 11 both seekers and soughts{ Stand for degree of competence in new technology, for example, Aa is slightly more competent than Bb who in turn is slightly more competent than Cc...etc. InfOrmation—seeking direction Information flow direction 12 Theoretical Hypotheses and Modernization Variables The first empirical problem is to determine the generality of homophily. For example, is a community that is homophilous on innova- tiveness also homophilous on literacy? Before we measure these variables to obtain community homophily indices, we shall examine the possible relationships between moderniza- tion variables and homophily. Modernization is the process by which individuals change from a traditional way of life to a more complex, technologically-advanced rapidly-changing style of life (Rogers with Svenning, 1969). The level of analysis of homophily in the present study is community rather than individual. Community Mass Media Exposure Theoretical Hypothesis I; Homophily i§_negatively related to_ community mass media exposure. Community mass media exposure is the relative proportion of members in a social system who are exposed to the outside world via radio, news- paper and magazine, TV and cinema. Lerner indicated (1967) that the mass media are bringing strange new worlds into traditional societies in the developing countries. Rogers (1965) found that mass media exposure leads peasants in Colombia down the road of modernization. MCNelly (196“) pointed out that "Much of the content in all of the media (is) designed to inform or persuade people about various kinds of modernization." Therefore, 13 it is clear that mass media exposure is an important indicant of modernity in developing countries. Furthermore, Rogers (1965) found that traditional communities have a lower level of mass media exposure than modern ones. Modern communities should have a higher degree of heterophily. Community Change Agent Contact Theoretical Hypothesis 3;; Homqphily i§_negatively related tg_community change_agent contact. Community change agent contact refers to the frequency with which proportion of members in a social system come into contact with certain extension service personnel whose mission is to diffuse new ideas into the social system. Villagers in the modern Colombian villages have more contacts with pereonnel of the National Agricultural ExtensiOn Service than those in traditional villages because there was no such service in the two traditional villages (Rogers and van Es, 196B, p. 13). Alves (1961), found that the Association for Credit and Rural Assitance (ACAR)* agents was one of the most important sources of new ideas for those villagers who participated in innovation programs; while others not in these programs received new ideas indirectly. Thus change agents represent external forces that bring new ideas into village social systems. These new ideas usually prove use- ful in accelerating the modernization of developing countries. The more contacts individuals have with change agents, the more new ideas they will likely have. Thus, change agent contact is an important force toward modernization. *Whiting and others (1968, p. 11) "A Summary of Innovation in Brazil: Success and Failure of Agricultural Programs in 76 Minas Gerais Communities," East Lansing; Department of Communication, Michigan State University (Diffusion of Innovations Report 7). Ill Community Cosmopoliteness Theoretical Hypothesis III: Homophily i§_negatively related to_community cosmopoliteness. Community cosmopoliteness is the degree to which the orientation of members in a social system are external to a particular social system (Rogers, 1960, p. 102). Merton (1957, pp. 387—u20) found that "the cosmopolite belongs to more organizations than the localite; is more willing to live elsewhere; makes friends with those with whom they can exchange ideas." Menzel (1960) found that the individual who is in touch with the outside world simply receives more information about new practices than those who do not. Opinion leaders not only have more contacts with the outside world (Rogers, 1955; Lionberger, 1953; and Menzel and Katz, 1955) but also tend to be seekers of technically accurate information sources. Thus, we may expect that social systems with a high degree of cosmopoliteness tend to be more modern than those without. In other words, cosmopoliteness is also an indicant of modernity. Community Literacy Theoretical Hypothesis 2!} Homophily i§_negatively related to community literacy. Community literacy is the degree to which proportion of members in a social system possess mastery over symbols in their written form. Lerner (1963, p. 3H1), also pointed out that "Literacy is indeed the basic personal skill that underlies the whole modernizing sequence... the vary act of achieving distance and control over a formal language 15 _gives people access to the world of vicarious experience." Rogers and Herzog (1966, p. 192) indicated that literacy is a "crucial element in the modernization process." The authors pointed out an interdependent relationship between literacy and modernization by saying that "...literacy is necessary for modernization, but modernization, as it develops, also impels literacy forward. Thus, the arrows of relationship are mutual and reciprocal: at Literacy 1 Modernization..." \ They concluded that literacy was important not only for individual modernization but also for community modernization in less developed coun- tries. It is also a key antecedent concept from which numerous modern- ization consequences flow (Rogers and Herzog, 1966, p. 202). We may expect that as the level of literacy in the community improves, so does people's ability to learn new ideas. Therefore, communities with higher levels of literacy tend to be more modern than those which are not. In other words, literacy is an indicant of modernity. Community,Opinionatedness Theoretical Hypothesis V; Homophily i§_negatively related to community Opinionatedness. Community Opinionatedness is the degree to which members in a social system have opinions about things of concern to the people in that social system. 16 Lerner (1958) reported that as people become more opinionated, the community is more likely to become modern. When people become more opinionated they tell others about whatever ideas they learned through such external forces as mass media or change agents. Chances are that both the talker and the listener will eventually benefit as a result of this exchange of ideas. Understandably, the former is seeking psycholog- ical support from the latter (Festinger, 1957). On the other hand, the latter benefits by_gaining useful scientific ideas from the former. As these actions and reactions take place quite often in the community, the community is bound to show progress. Therefore, opinionatedness is also an important indicant of modernity. Communityfilnnovativeness Theoretical Hypothesis 2;: Homophily i§_negatively related to_ innovativeness. Community innovativeness is the degree to which proportion of members in a social system are relatively earlier in adopting new ideas than the other members of his social system (Rogers, 1962, p. 19). Rogers (1962) found that innovative members of a social system are relatively earlier in adopting new ideas than the other members of his social system. The more innovative individuals a social system has, the more new ideas this social system might adopt. Therefore, these social systems tend to be relatively more modern since new ideas usually help modernize social system. Thus, innovativeness is another indicant of modernity. 17 Community Economic Development Theoretical Hypothesis VII: Homophily i§_negativeiy_related to community economic development. According to Stanfield (1968, p. 3): "Community economic develop- ment is the degree of the general physical well-being of a community's inhabitants, which is reflected by their possession of various comforts of life and an income sufficient to purchase these comforts." Deutschmann (1963) found that the best single economic index is an Andean village in Colombia was size of farm, followed by radio set owner- ship. Stanfield (1968) reported that items such as running water, bath in house, electricity and radio constitute a measure of economic develop— ment in the state of Minas Gerais, Brazil. According to a United Nations report,* community development'u.. designates the utilization under one single programme of approaches and techniques which rely upon local communities as units of action and which attempt to combine outside assistance with organized local self- determination and efforts, and which correspondingly seek to stimulate local initiative and leadership as the primary instrument of change..." As communities in less developed countries achieve a higher degree of economic development, so will the social structure become more complicated due to clearer separation of roles in society. In other words, in the communities with high level of economic development, indivi- duals depend heavily upon others for their daily necessities, either mat- erially or non-materially. *United Nations Document E/CN 5/291, Programme of Concerted Action in the Social Field of the United Nations and Specialized Agencies. 18 People have to interact with all types of individuals in order to satisfy their own needs regardless of the latter's different characteristics. Thus, we may assume that the higher level of economic development of a social system is an important and necessary charac- teristic of modernity. Summary As a result of previous discussions and research findings, we may conclude that literacy, opinionatedness, innovativeness, change .agent contact, mass media exposure, innovativeness, cosmopoliteness, and community economic development are important indicants of modernity of social systems in less developed countries. Modern communities have a higher degree of heterOphily. Out of necessity, if not willingness, people in these communities have to interact with all types of other individuals regardless of their relative characteristics. Therefore, we may expect that the higher the degree of literacy, opinionatedness, cosmopoliteness, mass media exposure, innovativeness, change agent contact, and community economic development, the more modern communities will be. In other words, a low degree of homophily would be found in modern communities. Figure 2 summarizes the expected rela- tionships of modernization variables and homophily. 19 mammooeon mom moaomHnm> cowuonwonoooe . coozvon mowgmcowumaon on“ mo swarmsom .m onswwm mirage: T 539mm. Howoom mo common + Hocowpwomme Homemoaoboo oweocooo hpwcoaeoo momnoqu monsoopmcowcwoo mason mmoco>fivm>oooH A11: momma zooxuoowmpso one 3.2.3020: Taovmhm Hmaoom mmooouqoooEwoo . . . amom . . . mo common u cameo 09 over. mpoouooo poems omomno Chin—“NH mane . onsmooxo memos who: sawaaosoe assumsm Hmaoom spas amamaoapmamm are: magmaoaomaom moaamanms eowumuaenoaoz ommwmoz monsom CHAPTER I I I METHODOLOGY The Data The data for this present study grew out of the project, "Diffusion of Innovations in Rural Societies" originally including three countries: India, Nigeria, and Brazil. There were 1,307 subjects interviewed in the Brazil Phase II Diffusion Project. Criteria for Selecting Communities Community sampling of the present study was based on a purposive selection of 18 communities chosen from the 76 communities studied in Phase I of the Diffusion Project. All of these 18 communities were situated in the state of Minas Gerais. The communities were matched on size, highway access, distance from urban center, approximate level of literacy, and availability of mass media. Specifically, criteria for selection were: 1. They must be equally suitable for four treatments (plus a control) in Phase III of the Diffusion Project. These treat- ments consisted of two sets: Set A included literacy train— ing and a community development program known as "animation." 20 21 Set B consisted of radio farm forums and community newspapers. 2. They must be within the range of a single broadcasting station so that any of 18 communities can be a part of radio forum. 3. There must be a meeting place in all communities for community members to use in literacy training and radio forums. u. All communities must be relatively easily accessible to facili- tate visits. Besides these original 18 communities, we incorporated one "success" and one "non-success" community selected from the Association for Credit and Rural Assistance local offices outside our Phase II-sample.* The procedure for collecting data from these two communities was identical to those used in the original 18. Criteria for Selecting Subjects Subjects selected for the present study are on the basis of the following criteria: 1. Subject must work on the farm, in other words, he is not an absentee landowner. 2. Subject should be the major decision maker for his farm. 3. Subject should own at least part of the land they worked. If the subject can not meet any of the above criteria, he might not either care about innovations or be reached by communications directed to people in the community. *Whiting and others, op: cit. 22 Method and Results of Data—Gathering An interview schedule was designed and pre-tested by the Diffusion Project staff in Brazil in 1966. Given interviewer training prior to hiring, university students were sent out under supervision to interview farmers in 20 communities in the state of Minas Gerais. They were instructed to interview all farmers who owned at least part of the land they operated and made major decisions for their farm. A total of 1,307 interviews were held in 20 rural communities during July and August, 1966. The data were coded and mailed to Michigan State University for card punching and computer analysis. The dyadic relationship between the seekers and the soughts is the unit of analysis in the present study. The "seekers" are not completely distinguished from the "soughts" since they make exchange roles. The seeker is the one who seeks information from others who are either his friends or influentials. Pertaining to friendship, reSpondents were asked the question "Who are three friends with whom you talk most frequently?" Influentials were located by asking the question "Who are the three persons in this community who are most listened to or imitated when it comes to farm- ing and cattle raising in general?" The people mentioned in response to these questions are the soughts of the seeker. Excluded from the present analysis are those who did not seek information from anybody or who sought someone outside of the community. for information. The interaction dyads were classified into two types according to the nature of their relationships: 23 1. Friend Dyad -- the dyad in which the sought was designated by the seeker as his friend. 2. Influence Dyad -- the dyad in which the sought was deisgnated by the seeker as a person who has influence on him. IBM dyadic relationship cards (see Table l for results of dyad- gathering) were prepared which included community identification num- ber, seeker's identification number, codes recording his personal char- acteristics, i.e., age, education, cosmopoliteness, innovativeness, etc., and same information on the person designated as a sought (see Appendix B). The Measure of Hom0phily: Pearson Product-Moment Correlation Pearson product—moment correlation is the statistical method used in the present study to measure homophily of dyads, where one var- iable of the seeker is related to the same variable of the sought. A high positive correlation coefficient indicates high homophily. Pearson product—moment correlation coefficient analysis was used to relate the characteristics of seekers and soughts because: 1. The independent variables used in the present study are continuous rather than discrete. Had the variables been discrete, then Coleman's (1958) homophily index would have been used. 2. Highest degree of homophily is indicated by correlation coef- ficient of +1. 3. Lowest degree of homophily is indicated by a correlation coef- ficient of -1. 2” For example, if the seeker's age is highly correlated with the sought's age, there is homophily on age. On the basis of this correlation for each community measured separately, community homophily was calcu- lated. Homophily is defined for of such independent variables as mass media exposure, change agent contact, opinionatedness, literacy, innova— tiveness and cosmopoliteness. As stated in the rationale, there is a negative relationship between homophily and these modernization variables in traditional communities. The coefficient of multiple determination, r2, indicates the common variance of two variables. In this data, the two variables are always the same characteristic of the seeker and that of the sought. This r2 is the basic measure of homophily for each community. To avoid negative correlations we decided to add .50 to every score. Therefore, the formula for homophily in the present study is: Homophily = r2 + .50 Operationalization of Variables Independable Variables 1. Individual mass media exposure is measured by Question No. 8* which asks each respondent "Have you read (or has anybody read for you) newspapers or magazines lately? (If yes) How many times a month." No. 8 also indicates that the individual mass media score ranges from 0 to 99 * Questions used in this thesis can be fOund in Appendix A, n. 25 times a month. Community mass media exposure in the present study is the percent of members in the social system who read a newSpaper or maga- zine one or more times per month. 2. Individual change agent contact is measured by Question No. 25 which asks each respondent "How many times have you talked to the Association for Credit and Rural Assistance agent (ACAR agent) last year?" Question No. 25 also indicates that the individual score may range from 0 to 99 contacts. Community change agent contact is the percent of members in the social system who have had one or more contacts with ACAR agent in the last year. 3. Individual cosmopoliteness is measured by Question No. 4 which asks "Did you visit a large city last year?" A large city is one with more than 40,000 inhabitants. Question No. 4 also indicates individual score ranging from 0 to 99 times. Therefore, community cosmopoliteness is the percent of members in the social system having made five visits to a large city with population more than 40,000 in the past year. 4. Individual literacy is measured by Question No. 22, which asks each respondent to read a 50-word reading card: "He who cannot read is like a blind man who has to be guided according to other peOple's wishes; or then he will stumble his way. The illiterate man is not altOgether free; he is a slave of his ignorance. Never stOp reading something every day and keep learning." Item No. 22 indicates how many words reSpondent can not read. Individual functional literacy score range from 00 to 50 words. Community literacy is the percent of members in the social system who can read 40 to more words in the 50-word passage. 26 5. Individual opinionatedness is measured by Question 38 which asks each reSpondent "How many of ten questions on which have you expressed Opinion?" The range of scores is from 0 to 9. The higher the score, the greater the individual Opinionatedness. Community opinionatedness is the percent of members in the social system who expressed Opinions on 9 items of opinion questions. 6. Individual innovativenss is measured by a series of five questions. The first question, No. 16, ascertains how many agricultural practices have ever been adopted, while the second question No. 17, indicates year of adoption by practice. The last question, No. 18 shows continuing use of practices by adopters. The agricultural practices are: (1) Correct spacing for corn; (2) Hybrid corn (3) Chemical fertilizer - corn (4) Chemical fertilizer - other crops (5) Contour plowing (6) Septic tank (7) Reforestation (8) Lime on acid soil (9) . Ant control (10) Irrigation (ll) Termite control (12) Forage grass plot 27 Community innovativeness is the percent of members in the social system who adopt one or more practices in the average year of adoption for 12 selected agricultural practices. 7. Community economic development is a measure of home and farm equipment and improvements possessed by each respondent in the social system. This home and farm equipment or improvements included water filter, wood or tile floor in the house, refrigerator, stove with chimney, plumbing for running water,:bathroom inside the house, elec- tric lighting, radio, relevision set, motorized vehicle, property, house in town and agricultural machines. The range of scores is 0 to 26. Question No. 26 indicates that community economic develOpment is the percent of members who possessed 12 or more items mentioned. 28 The distribution of friend, influence and general dyads in 20 Brazilian communities are shown in Table 1. Table 1. Results of dyad gathering. Names of Influence General Communities Friend dyads dyads dyads 10 Tres Pontas 71 79 150 ll Tres Coracoes 68 70 138 20 Sao Joao Nepomuceno 109 127 236 21 Sao Joao del Rei 83 75 158 22 Santos D 58 49 107 23 Bicas 8O 95 175 24 Rio Voro 85 97 182 30 Parao-Poba 95 130 225 31 Sete Ligoas 78 102 180 32 Podro Leopaldo 85 63 148 34 Corinto 100 121 221 35 Cordisburgo 99 109 208 42 Itauna 99 61 160 43 Divin0polis 88 83 171 51 Formiga 52 78 130 70 Uba 72 71 143 71 Catagueses 77 85 162 72 Tocantins 74 95 169 80 Fonta 40 48 88 82 Alvinopolis 91 98 189 Total 1604 + 1736 = 3340 29 Empirical Hypotheses In the present study, the level of homOphily analysis is the community. Theoretical Hypothesis if Homephily i§_negativeiy_related to_ community mass media exposure. Empirical Hypothesis la; Homophilv i§_negativeiy related to community_radio exposure. Empirical Hypothesis £25 Homophily i§_negatively related to_ community T!_exposure. Empirical Hypothesis is} Homophily i§_negatively realted to_ community movie exposure. Empirical Hypothesis id; Homophily i§_negative1yrelated to_ community newspaper and magazine readership. Theoretical Hypothesis 2;; Homophily ia negatively related to community change agent contact. Empirical Hypothesis 1;; Homophily i§_negatively_related to community ACAR agent contact. Theoretical Hypothesis III: Homophily i§_negatively related to community cosmOpoliteness. Empirical Hypothesis III: Homophily i§_negatively related to_ community number o£_trips to_city in_the last year. Theoretical Hypothesis 1!; Homophily i§.negativeiy_related to. communityyliteracy, 30 Empirical Hypothesis 2!} Homophily i§_negatively related to_ community literacy. Theoretical Hypothesis X; Homophily i§_negativeiy related t3 community opinionatedness. Empirical Hypothesis 2} Homophily i§_negatively related to. community Opinionatedness. Theoretical Hypothesis Xi; Homophily i3 negatively related to_ community innovativeness. Empirical Hypothesis Xi} Homophily i§_negative1y related to_ community average years of adoption of innovations. Theoretical Hypothesis VII: Homophily i§_negativeiy related to community_economic development. Empirical Hypothesis VII: Homophily i§_negatively related to_ community economic development. CHAPTER IV FINDINGS Measurement of Homophily Pearson Product-Moment Correlation Coefficient Analysis Most correlations between seekers' and soughts' 20 variables across 20 Brazilian communities were found to be very low and positive, but not significant. For example, only in the case of friend dyadic relationships in Sao Joao Nepomuceno community, Pearsonian correlations on home and farm improvement and equipments, income, and most influential source A were .402, .325, and .306. They were significantly different from zero. In the case of influence dyadic relationships in Sao Joao Nepomuceno community, none of the correlations was significant. Similar results are found in other 19 communities as well. Therefore, we are unable to say that seekers in Brazil who are homophilous on one variable should interact with soughts who are also homophilous on the same variable. frhe expectation that there is a high degree of homophily in Brazilian communities was not fulfilled. Findings are consistent with Chou's findings (1966) in three Colombian villages when she reported that homophily was also very low there. 31 32 Tables 2, 3, 4, and 5 show Pearsonian correlations between variables of seekers' and soughts' in two representative communities: Sao Joao Nepomuceno (relative modern community) and Corinto (relative traditional community) on the basis of friendship and influence. FactorlAnalysis Given the low correlations between seekers' and soughts' charac- teristics in Brazil, we tried to find what variables, if any, constitute . general homophily. No main factor indicating general homophily was found. Using the criterion that loadings should be .40 or larger on one factor and less than .40 on other factors, four groups of homophily emerged from our factor analysis: (See Table 6 for factor loadings of homophily items). (1) Cosmopolite group homophily consists of: education, visits to the city, political knowledge, and opinionatedness. (2) Information and source group homophily consists of: radio listening, total number of sources utilized and opinion leadership. (3) Adoption group homophily consists of: ACAR contacts, social participation, and innovativeness. (4) Economic group homophily consists of: information source, trust in neighbors, home and farm equipment and improvements, and income. 33 Table 2. Pearsonian correlations of seekers' 20 variables to soughts' 20 variables in Sao Joao Nepomuceno community on the basis of friendship (Modern community N = 109). Seekers', Soughts' 20 variables in Sao Joao Nepomuceno Pearsonian Correlations Community _Age . 091 Education -.074 Visit to city . -.001 Newspaper 6 Magazine Readership .000 Radio Listening -.040 Cinema Attendance .436+ Total Sources Utilized -.017 Source A .263+ Most Influential Source A .306+ Trust in Neighbor .151 Functional Literacy .142 ACAR Contact -.052 Home and Farm Improvement 8 Equipment .402+ Political Knowledge .185 Income .325+ Number of cows owned .135 Opinion Leadership -.104 Social Participation .193 Opinion Leadership Total .007 Innovativeness .087 *Significantly different from zero at the 5 per cent level. +Significantly different from zero at the l per cent level. 34 Table 3. Pearsonian correlations of seekers' 20 variables to soughts' 20 variables in Sao Joao Nepomuceno community on the basis of influence (Modenzcommunity N = 127). Seekers‘, Soughts' 20 Variables in Sao Joao Nepomuceno Pearsonian Correlations Community ,Age .030 Education .025 Visit to city .023 Newspaper 8 Magazine Readership -.l20 Radio Listening .000 Cinema Attendance .122 Total Sources Utilized -.065 Source A .128 Most Influential Source A .064 Trust in Neighbor -.049 Functional Literacy .168 ACAR Contact -.055 Home and Farm Improvement 8 Equipment .105 Political Knowledge .034 Income .058 Number of cows owned -.020 Opinion Leadership -.071 Social Participation .020 Opinion Leadership Total -.051 Innovativeness .062 *Significantly different from zero at the 5 per cent level. +Significantly different from zero at the l per cent level. 35 Table 4. Pearsonian correlations of seekers' 20 variables to soughts' 20 variables in Corinto community on the basis of friendship (Traditional community N = 100). Seekers', Soughts' 20 Variables in Sao Joao Nepomuceno Pearsonian Correlations Community Age .178 Education .062 Visit to city -.075 Newspaper 8 Magazine Readership_ -.029 Radio Listening .362+ Cinema Attendance -.003 Total Sources Utilized .319+ Source A -.018 Most Influential Source A E -.162 Trust in Neighbor -.014 Functional Literacy -.015 ACAR Contact .161 Home and Farm Improvement 6 Equipment .139 Political Knowledge .103 Income .162 Number of cows owned .196* Opinion Leadership .062 Social Participation .210 Opinion Leadership Total .197* Innovativeness .147 *Significantly different from zero at the 5 per cent level. +Significantly different from zero at the 1 per cent level. 36 Table 5. Pearsonian correlations of seekers' 20 variables to soughts' 20 variables in Corinto community on the basis of influence (Traditional community N = 121). Seekers', Soughts' 20 Variables in Sao Joao Nepomuceno Pearsonian Correlations Community Age -.046 Education -.082 Visit to city .030 Newspaper 8 Magazine Readership -.056 Radio Listening .069 Cinema Attendance -.049 Total Sources Utilized .225* Source A -.009 Most Influential Source A -.l42 Trust in Neighbor .167 Functional Literacy -.019 ACAR Contact .071 Home and Farm Improvement 6 Equipment .108 Political Knowledge -.022 Income .003 Number of cows owned -.037 Opinion Leadership -.091 Social Participation .039 Opinion Leadership Total .027 Innovativeness .057 *Significantly different from zero at the 5 per cent level. +Significantly different from zero at the l per cent level. 37 Table 6. Varimax rotation: factor loadings for homophily items. Item No. Homophily Factor Item Name 1 2 3 4 1 _Age -.0578 -.3080* -.1011 -.0891 2 Education .6905* .3481 -.1708 -.1780 3 Visits to City .7396* -.503u .0110 - .1818 l u Newspaper and 1 Magazine Readership -.1580* .1039 -.O288 .0617 5 Radio Listening -.0221 .7665* -.3017 .0677 6 Cinema Attendance -.2577 -.0162 .u093 .7881* 7 Total Sources Utilized .1333 .7990* .0072 -.2767 8 Source A .1611 -.1467 -.4788 .5370* 9 Most Influential -.1388 .0163 .0772 .56u9* 10 Trust in Neighbor .3319 -.0903 .1546 .6011* 11 Functional Literacy .u950* -.1372 -.17HO .9777 12 ACAR Contact -.0751 -.0672 .8582* -.ouus 13 Home and Farm Improve- ment and Equipment .1235 .9551 -.0755 .6137* 14 Political Knowledge .5967* .2669 .H3H0 .3156 15 Income .2366 -.0597 .0898 .7838 16 Numbers of cows owned -.0217 .u766* .9535 -.ou12 17 Opinion Leadership .8636* .1226 .1227 .1057 18 Social Participation .1269 -.0u82 .7379* .12us 19 Opinion Leadership Total-.1557 .6910* .0877 .0319 20 Innovativeness .0900 .0692 .7807* .2453 Proportion of Variance Explained by the Factor .1395 .1351 .1458 .158# *Largest factor loading 38 -Corre1ates of Homophily To find relationships between types of homophily and indepen- dent variables used in the present study, it was decided to select innovative and change contact homophily on the basis of friendship and influence. There are several reasons for selecting these four types of homophily. Innovativeness is a key variable in the modernization process. Rogers with Svenning (1969) pointed out that "the best single indicator of (an individual's) degree of modernization is innovativeness, indicating a behavioral rather than a cognitive or attitudinal change." Change agent contact is an important external force in bringing new ideas into a social system. It is also an important indicant of modernity, that is, there is a higher degree of change agent contact in the modern than in the traditional community. Furthermore, it is not known what the relationships are between these four types of homophily and the independable variables: mass media exposure, change agent contact, cosmopoliteness, literacy, opinionatedness, innovativeness, and community economic development. Results of Intercorrelation between Homophily Indices and All Independent Variables Pearson product-moment correlation coefficient was used to analyze relationship between homophily indices and all independent variables. Tables 7 and 8 show results of intercorrelation between homophily indices and all independent variables. 39 Table 7. A comparison of Pearsonian correlations between innovative homophily on the basis of friendship or influence and modernization variables. Innovative Homophily on the basis of Friendship Innovative Homophily on the basis of Influence Radio Exposure TV exposure Movie exposure Newspaper and magazine readership Change agent contact Cosmopoliteness Literacy Opinionatedness Innovativeness Community Economic Development .109 .303 -.048 -.063 -.033 .256 .292 -.026 -0189 .046 .uo7* .169 .159 .055 -.036 .130 .183 -.073 .099 .170 *Significantly different from zero at the 5 per cent level. '40 Table 8. A comparison of Pearsonian correlations between change agent contact homophily on the basis of friendship or influence and modernization variables. Change agent contact Change agent contact homophily on the homophily with basis basis of friendship of influence Radio exposure .oug -.107 TV exposure .22“ .222 Movie exposure .035 -.052 Newspaper and Magazine readership .118 .008 Change agent contact -.231 -.l83 Cosmopoliteness .963* .631 Literacy .180 .109 Opinionatedness .223 .113 Community Economic Development . 10 7 . 018 *Significantly different from zero at the 5 per cent level. 1&1 Test of Hypotheses Dealing with Homophily In the present study, none of the hypotheses were confirmed. Of the 28 hypotheses, 26 did not reach statistical significance; Pearsonian correlations of .360 and .u92 are necessary at the five and one per cent levesl respectively. The remaining two hypotheses were statistically significant, one at 5 per cent level and the other at l per cent, but in the direction opposite from that predicted. Theoretical Hypothesis I Theoretical Hypothesis 5} Innovative homophily gn_the basis g£_ friendship i§_negative1y related :2_community mass media exposure. Egpirical Hypothesis la; Innovative homophily en the basis g£_ friendship i§_nggative1y related £2_commmnity radio exposure. The zero-order Pearsonian correlation is .10“, which is not significantly different from zero. Empirical Hypothesis Ia is not confirmed. Empirical Hypothesis 32; Innovative homophily gn_the basis g§_ friendship i§_negativelyrelated £2_community I!_e§posure. The zero- order Pearsonian correlation is .303, which is not significantly different from zero. Empirical Hypothesis Ib is not confirmed. Empirical Hypothesis is} Innovative homophily gn_the basie g§_ friendship i§_negatively related £2_community_movie exposure. The zero-order Pearsonian correlation is -.0u8, which is not significantly different from zero. Empirical Hypothesis Ic is not confirmed. Empirical Hypothesis E: Innovative homophily 93 the basis 2: friendship i§_nggatively related 32 community newspaper and magazine 1+2 readership. The zero-order Pearsonian correlation is -.063, which is not significantly different from zero. Empirical Hypothesis Id is not confirmed. Since none of four Empirical Hypotheses is confirmed, Theoretical Hypothesis I is not confirmed. Theoretical Hypothesis II Theoretical Hypothesis ii} Innovative hompphiiyypp_the basis p§_ influence i§_pggativeiy related pp_community mass media exposure. Empirical Hypothesis IIa: Innovative homophiy pp the basis 93 influence i§_negativeiy related £p_commmnity radio exposure. The zero- order Pearsonian correlation is .uo7, which is significantly different from zero but not in the direction predicted. Empirical Hypothesis 11a is not confirmed. Empirical Hypothesis IIb: Innovative homophiy 33 the basis 93 influence i§_npgatively related Ep_community Iz_exposure. The zero- order Pearsonian correlation is .169, which is not significantly different from zero. Empirical Hypothesis 11b is not confirmed. Empirical Hypothesis IIc: Innovative homophiiy gp_the basis 9: influence i§_negatively_related £p_community_movie exposure. The zero- order Pearsonian correlation is .159 which is not significantly different from zero. Empirical Hypothesis IIc is not confirmed. Empirical Hypothesis IId: Innovative homophiiy pp_the basis pf_ influence i§_pggatively_re1ated Ep_community_newspaper and magazine readership. The zero-order Pearsonian correlation is .055, which is not 1+3 significantly different from zero. Empirical Hypothesis IId is not confirmed. Since none of four Empirical Hypotheses is confirmed, Theoretical Hypothesis II is not confirmed. Theoretical Hypothesis III Theoretical Hypothesis III: Change agent contact homophily pp_ the basis pf_friendship i§_negatively related Ep_community mass media e' osure. Empirical Hypothesis IIIa: Change agpnt contact homophily pp_the The zero-order Pearsonian correlation is .089, which is not significantly different from zero. Empirical Hypothesis IIIa is not confirmed. Empirical Hypothesis IIIb: Change agent contact homophily pp_the basis pf_friendship i§_negative1y related pp community I!_e§posure. The zero-order Pearsonian correlation is .22“ which is not significantly different from zero. Empirical Hypothesis IIIb is not confirmed. Empirical Hypothesis IIIc: Change agent contact homophily pp_the basis p§_friendship i§_pggatively related Epgcommunipy movie exposure. The zero-order Pearsonian correlation is .035, which is not significantly different from zero. Empirical Hypothesis IIIc is not confirmed. Empirical Hypothesis IIId: Change agent contact homophily pp_the basis pf friendship ip_negatively related Ep_community movie readership. The zero-order Pearsonian correlation is .118, which is not significantly different from zero. Empirical Hypothesis IIId is not confirmed. uu Since none of four Empirical Hypotheses is confirmed, Theoretical Hypothesis III is not confirmed. Theoretical Hypothesis IV Theoretical Hypothesis 1!; Change agent contact homophily gp_the Empirical Hypothesis IVa: Changp ggent contact homophily pp_the basis pf_influence i§_negatively related Ep_communityradio egppsure. The zero-order Pearsonian correlation is —.107, which is not significantly different from zero. Empirical Hypothesis IVa is not confirmed. Empirical Hypothesis IVb: Changg_ggent contact homophily pp_the basis of influence is negatively related Ep_community I! e§posure. The zero-order Pearsonian correlation is .222, which is not significantly different from zero. Empirical Hypothesis IVb is not confirmed. Empirical Hypothesis IVc: Changg agent contact homophiiy_pp_the basis p£_influence ip negatively related pp_community_movie exposure. The zero-order Pearsonian correlation is -.052, which is not significantly different from zero. Empirical Hypothesis IVc is not confirmed. Empirical Hypothesis IVd: Changg_agent contact homophily pp_the basis pf influence i§_negativelyre1ated £p_communipy_new§paper and magazine egposure. The zero-order Pearsonian correlation is .008, which is not significantly different from zero. Empirical Hypothesis IVd is not confirmed. 1+5 Since none of four Empirical Hypotheses is confirmed, Theoretical Hypotheses IV is not confirmed. Theoretical Hypothesis V Theoretical Hypothesis X: Innovative homophily pp_the basis p£_ friendship i§_negatively related pp_community change agent contact. Empirical Hypothesis H} Innovative homophily_pp_the basis pf_ friendship i§_ppgatively_related pp_community ACAR change ggent contact. The zero-order Pearsonian correlation is -.033, which is not significantly different from zero. Empirical Hypothesis V is not confirmed. Since Empirical Hypothesis V is not confirmed, Theoretical Hypothesis V is not confirmed. Theoretical Hypothesis VI Theoretical Hypothesis Vi: Innovative homophiiy 93 the basis 21: influence i§_negatively related £p_community change agent contact. Empirical Hypothesis 2;: Innovative homophily 93 the basis 25. influence i§_negatively related pp_community ACAR changp agent contact. The zero-order Pearsonian correlation is -.036, which is not significantly different from zero. Empirical Hypothesis V1 is not confirmed. Since Empirical Hypothesis VI is not confirmed, Theoretical Hypothesis VI is not confirmed. Theoretical Hypothesis VII Theoretical Hypothesis VII: Changp_agent contact homophily pp_the basis pf friendship i§_pggatively_related pp_community changg_agent contact. 1+6 Empirical Hypothesis VII: Change agent contact homophily pp_the basis pf friendship ip negatively related £p_commmnipy_ACAR change agent contact. The zero-order Pearsonian correlation is -.231, which is not significantly different from zero. Empirical Hypothesis VII is not confirmed. Since Empirical Hypothesis VII is not confirmed, Theoretical Hypothesis VII is not confirmed. Theoretical Hypothesis VIII Theoretical Hypothesis VIII: Change agent contact homophiiy_pp_the basis p£_inf1uence i§_negativeiy related £p_communipy change agent contact. Empirical Hypothesis VIII: Change agent contact homephily_pp_the basis pf_inf1uence i§_negatively related Ep_commmnity_ACAR change agent contact. The zero-order Pearsonian correlation is -.l83, which is not significant different from zero. Empirical Hypothesis VIII is not con- firmed. Since Empirical Hypothesis VIII is not confirmed, Theoretical Hypothesis VIII is not confirmed. Theoretical Hypothesis IX Theoretical Hypothesis EH; Innovative homophily pp_the basis 2: friendship i§_ppgatively related Ep_pgmmuni:y_cosmopoliteness. Empirical Hypothesis EH} Innovative homophilypp_the basis of friendship i§_negatively_related pp_communipynumber p£_visits pp_1arg§ '47 aiEy_ip_paa£_yaa£3 The zero—order Pearsonian correlation is .256, which is not significantly different from zero. Empirical Hypothesis IX is not'confirmed. Since Empirical Hypothesis IX is not confirmed, Theoretical Hypothesis IX is not confirmed. Theoretical Hypothesis X Theoretical Hypothesis H3 Innovative homophily pp_the basis a: influence ia_negativeiy related pa_community cosmopoliteness. Empirical Hypothesis H: Innovative homophily 93 3133 922.13 3: influence ia_negatively_related pp_community number p§_visits pp_1arge gi£y_ip_pa§5_yaap, The zero-order Pearsonian correlation is .130, which is not significantly different from zero. Empirical Hypothesis X is not confirmed. . Since Empirical Hypothesis X is not confirmed, Theoretical Hypothesis X is not confirmed. Theoretical Hypothesis XI Hypothesis Hi: Change agent contact homophily all the basis 21: friendship ia_negative1y related Ep_community cosmopoliteness. Empirical Hypothesis Hi; Change agent contact homophily pp_the £2.iEEEE.EiEZ;iE.EE§E.XEEE: The zero-order Pearsonian correlation is .463, which is significantly different from zero, but not in the direction predicted. Empirical Hypothesis XI is not confirmed. Since Empirical Hypothesis XI is not confirmed, Theoretical Hypothesis XI is not confirmed. #8 Theoretical Hypothesis XII Theoretical Hypothesis XII: Chaage agent contact homophily pa_the basis pf_inf1uence ia_negativeiy related £p_community cosmopoliteness. Empirical Hypothesis XII: Change agent contact homophilypp_the basis pf_influence ia_negatively_related £p_community number pf visits £p_large city ip_past year. The zero-order Pearsonian correlation is .631, which is significantly different from zero, but not in the direc- tion predicted. Empirical Hypothesis XII is not confirmed. Since Empirical Hypothesis XII is not confirmed, Theoretical Hypothesis XII is not confirmed. Theoretical Hypothesis XIII Theoretical Hypothesis XIII: Innovative homophily pa the basis a: friendship ia negatively related Ep_communipyliteracy. Empirical Hypothesis XIII: Innovative homophily pa_the basis af_ friendship ia_negative1y related £p_community literacy. The zero-order Pearsonian correlation is .292, which is not significantly different from zero. Empirical Hypothesis XIII is not confirmed. Since Empirical Hypothesis XIII is not confirmed, Theoretical Hypothesis XIII is not confirmed. Theoretical Hypothesis XIV Theoretical Hypothesis XIV: Innovative homophily pp the basis p£_ influence ia_negatively related Ep_community literacy. Empirical Hypothesis XIV: Innovative homqphily pp_the basis p£_ influence ia_negatively related £p_commmni£y literacy. The zero-order 49 Pearsonian correlation is .183, which is not significantly different from zero. Empirical Hypothesis XIV is not confirmed. Since Empirical Hypothesis XIV is not confirmed, Theoretical Hypothesis XIV is not confirmed. Theoretical Hypothesis XV Ipeoretical Hypothesis HZ} Changa_agent contact homophily pp_the basis pf friendship ia_negatively related 32 community literaay. Empirical Hypothesis HZ: Change agent contact homophilypa the basis g£_friendship ia_negative1y related :a_community literacy. The zero-order Pearsonian correlation is .140, which is not significantly different from zero. Empirical Hypothesis XV is not confirmed. Since Empirical Hypothesis XV is not confirmed, Theoretical Hypothesis XV is not confirmed. Theoretical Hypothesis XVI Theoretical Hypothesis XVI: Change agent contact homophily pa_ EH9 basis p£_inf1uence ia_negatively related Ea community literacy. Empirical Hypothesis XVI: Change agent contact homophily pp_the pasis p£_influence ia_negatively related £p_communipy_literacy. The zero-order Pearsonian correlation is .109, which is not significantly different from zero. Empirical Hypothesis XVI is not confirmed. Since Empirical Hypothesis XVI is not confirmed, Theoretical Hypothesis XVI is not confirmed. 50 Theoretical Hypothesis XVII Theoretical Hypothesis XVII: Innovative homophily pp_the basia a: friendship_ia negatively paiated_pp_communityopinionatedness. Empirical Hypothesis XVII: Innovative prophily pp the basis p£_ friendship_ia negatively paiapaa_pp_gommunity opinionatedness. Zero- order Pearsonian correlation is -.026, which is not significantly different from zero. Empirical Hypothesis XVII is not confirmed. Since Empirical Hypothesis XVII is not confirmed, Theoretical Hypothesis XVII is not confirmed. Theoretical Hypothesis XVIII Theoretical Hypothesis XVIII: Innovative homophily pa the basis a: m m -~_ influence ia negatively related pp coummnity opinionatedness. Empirical Hypothesis XVIII: Innovative homophiiyaa_the basis 2: influence ia_negative1y palated £a_community opinionatedness. The zero- order Pearsonian correlation is -.073, which is not significantly differ— ent from zero. Empirical Hypothesis XVIII is not confirmed. Since Empirical Hypothesis XVIII is not confirmed, Theoretical Hypothesis XVIII is not confirmed. Theoretical Hypothesis XIX Theoretical Hypothesis XIX: Change agent contac£_homophily pp the basis p£_friendship ia_negatively related pp_community opinionatedness. Empiricai Hypothesis XIX: Change agen:_contact prophily pp the basis of friendship_ia_negative1y related £p_community Opinionatedness. MM“.- The zero-order Pearsonian correlation is .223, which is not significantly different from zero. Empirical Hypothesis XIX is not confirmed. 51 Since Empirical Hypothesis XIX is not confirmed, Theoretical Hypothesis XIX is not mnfirmed. Theoretical Hypothesis XX Theoretical Hypothesis HH: Change agent contact homophily ap_the basis p§_inf1uence ia_negatively_related £a_community opinionatedness. Empirical.Hypothesis HH; Change agent contact homophily pa_the basis p£_influence ia_negatively related pp_commmnity_opinionatedness. The zero-order Pearsonian correlation is .113, which is not significantly different from zero. Empirical Hypothesis XX is not confirmed. Since Empirical Hypothesis XX is not confirmed, Theoretical Hypothesis XX is not confirmed. Theoretical Hypothesis XXI Theoretical Hypothesis XXI: Innovative hbmophily_pp the basis a: friendship ia_negativelyrelated Ep_community innovativeness. Ejpirica Hypothesis XXI: Innovative homophiiy 2’. the basis pi friendship ia_negatively related pp averaga years a: agoption p£_ ~J innovations ip_the community. The zero-order Pearsonian correlation is -.189, which is not significantly different from zero. Empirical Hypothesis XXI is not confirmed. Since Empirical Hypothesis XXI is not confirmed, Theoretical Hypothesis XXI is not confirmed. Theoretical Hypothesis XXII Theoretical Hypothesis XXII: Innovative homophily_pp_the basis p£_ influence ia_negative1y related pa_community innovativeness. 52 Empirical Hypothesis XXII: Innovative homophi1y_pa the basis of influence ia_negatively related Ep_community average_years of adoption a£_innovations. The zero-order Pearsonian correlation is .049, which is not significantly different from zero. Empirical Hypothesis XXII is not confirmed. Since Empirical Hypothesis XXII is not confirmed, Theoretical i Hypothesis XXII is not confirmed. 7 Theoretical Hypothesis XXIII Theoretical Hypothesis XXIII: Change agent contact homophily pp_the basis friendship ia_pagatively related Ep_community innovativeness. Empirical Hypothesis XXIII: Changa_agent contact homophily pp the basis p£_friendship ia negatively related pa_community average_years p£_ adoption p£_innovation ip_the community. The zero-order Pearsonian correlation is -.080, which is not significantly different from zero. Empirical Hypothesis XXIII is not confirmed. Since Empirical Hypothesis XXIII is not confirmed, Theoretical Hypothesis XXIII is not confirmed. Theoretical Hypothesis XXIV Theoretical Hypothesis XXIV: Change agent contact homophily aa_the basis p£_inf1uence ia_negativeiy related Ep_community innovativeness. Empirical Hypothesis XXIV: Change agent contact homophily_pp_the basis a: influence ia_negative1y related pa_average years p£_adoption p§_ innovations in the community, The zero-order Pearsonian correlation is -.ll7, which is not significantly different from zero. Empirical Hypothesis XXIV is not confirmed. 53 Since Empirical Hypothesis XXIV is not confirmed, Theoretical Hypothesis XXIV is not confirmed. Theoretical Hypothesis XXV Theoretical Hypothesis XXV: Innovative homophily pp the basis a£_ iiiendship ia_negative1y related Ep_commmnityeconomic development. Empirical Hypothesis XXV: Innovative homophily pp_the basis p£_ friendship ia_negativeiy_related Ep_community home and farm improvements and equipment. The zero-order Pearsonian correlation XXV is .046, which is not significantly different from zero. Empirical Hypothesis XXV is not confirmed. Since Empirical Hypothesis XXV is not confirmed, Theoretical Hypothesis XXV is not confirmed. Theoretical Hypothesis XXVI Theoretical Hypothesis XXVI: Innovative homophily pp_the basis 3:. influence ia_negatively related Ep_commnnipy_economic development. Empirical Hypothesis XXVI: Innovative homophily pp the basis a: influence ia_negatively related pp community home and farm improvements and equipments. The zero-order Pearsonian correlation for Empirical Hypothesis XXVI is .170, which is not significantly different from zero. Empirical Hypothesis XXVI is not confirmed. Since Empirical Hypothesis XXVI is not confirmed, Theoretical Hypothesis XXVI is not confirmed. 54 Theoretical Hypothesis XXVII Theoretical Hypothesis XXVII: Changa agent contact homophily pp- the basis friendship ia_negatively related pp_community economic develop- uent. Empirical Hypothesis XXVII: Change agent contact homophily ap_the basis of friendship is negatively related to community home and farm improvements and equipment. The zero-order Pearsonian correlation is .107, which is not significantly different from zero. Empirical Hypothesis XXVII is not confirmed. Since Empirical Hypothesis XXVII is not confirmed, Theoretical Hypothesis XXVII is not confirmed. Theoretical Hypothesis XXVIII Theoretical Hypothesis XXVIII: Change agent contact homophily pp_ the basis pf_inf1uence ia negatively related 32 community economic development. Empirical Hypothesis XXVIII: Changa_agent contact homophily ap_the basis p§_inf1uence ia_negative1y related 52 community home and farm improvements and equipments. The zero-order Pearsonian correlation is .018, which is not significantly different from zero. Empirical Hypothesis XXVIII is not confirmed. Since Empirical Hypothesis XXVIII is not confirmed, Theoretical Hypothesis XXVIII is not confirmed. CHAPTER V SUMMARY, ADDITIONAL ANALYSES, DISCUSSION OF HYPOTHESES, AND SUGGESTIONS FOR FUTURE RESEARCH Summary The present thesis was concerned with personal, socio-economic and demographic variables in interpersonal communication in the diffusion of technoIOgical innovations in 20 Brazilian communities. Homophily was defined as the degree to which pairs of individuals who interact are similar in certain characteristics such as mass media exposure, change agent contact, literacy, cosm0politeness, innovativeness, and Opinionatedness. Dyadic analysis is a method to study relation- ships among pairs of individuals. The objectives were (1) to define and measure general homophily in each of 20 communities, and test what variables are related to homophily. More Specifically, we wanted to define and measure community homoPhily, and (2) to relate community homophily to such aggregate communication characteristics as opinionatedness, literacy, mass media exposure, change agent contact, cosmopoliteness, innovativeness, and a non-communication characteristics, community economic develOpment level. The data used in the present study were collected during 1964-68 in the research project, "Diffusion of Innovations in Rural Societies." The project was funded by the United States Agency for International 55 'LL‘.). ..-- . _ . . 56 Development, and conducted by the Department of Communication, Michigan State University, under the direction of Dr. Everett M. Rogers. There were 1,307 subjects interviewed in the Brazilian Phase II project, and 3,340 dyads were found. Of these, 1,604 were friend dyads, and 1,736 were influence dyads. The present study examined relationships between innovative or change agent contact hom0phily on the basis of friendship and influence and community mass media exposure, community change agent contact, cosmopoliteness, literacy, opinionatedness, innovativeness, and economic develoPment. There were 28 theoretical hypotheses in the present study, none of which were confirmed. However, two theoretical hypotheses were signi- ficantly different from zero in the direction opposite to that which was predicted. The remaining 26 were not statistically significant. Of all 40 empirical hypotheses tested, Pearsonian correlations of 13 were negative as predicted, with the remaining 24 positive. All except 3 Pearsonian correlations were very low and not significantly different from zero. Three empirical hypotheses were statistically different from zero, but not in the direction predicted. Additional Analyses Two additional variables, opinion leadership concentration* and size of community were correlated with innovative and change agent contact homophily. The purpose was to see whether any relationships *Opinion leadership concentration is measured by Question No. 36 which asks each reSpondent, "Who do you think would be the best person to organ- ize this (CooPerative community project) in the community? Community Opinion leadership concentration is giniratio for each community. 57 between these two variables and the four types of homophily could clarify the previous results. Reasons for relating opinion leadership concentration and size of community to homophily are: 1. As we know that Opinion leaders are characteristically different from seekers of information. In a community with a low degree of Opin- ion leadership concentration, it might be likely that opinion leaders and seekers are relatively more similar to each other than their counterparts in a high degree of Opinion leadership concentration. Since there are more Opinion leaders in the former case, opinion leaders might be more similar to seekers in terms of social status, literacy level, income, etc., except the specific knowledge seekers are looking for. On the other hand, in a community with only a handful of opinion leaders, we might also find less interacting dyads than we could in a community with a larger number of opinion leaders because seekers tend to seek others like themselves. Thus, we assume that there is a higher degree of homoPhily in a community with a lower degree of opinion leadership concentration. 2. In a community with larger numbers of individuals, we might find relative more Opinion leaders as well as seekers of information. As a result, we might find not only more individuals with similar char- acteristics but also a higher degree of interaction among individuals. However, intervening variables such as physical locations of individuals in the community, geOgraphical size of the community, etc., must be taken , .U 58 into consideration. In the present study, failure to take these varia- bles into consideration might contribute to the unexpected results. We found negative Pearsonian correlations that are not signifi- cantly different from zero. Pearsonian correlations between size of community and change agent contact homophily on the basis of friendship was negative as predicted, but not significant. Correlations between size of community and other three types of homophily were positive and not significant. Table 9 shows the zero—order correlations between innovative or change agent contact homOphily and Opinion leadership concentration or size of community. Therefore, we conclude that there are no significant relationships between Opinion leadership concentration or size of community and fOur types of hom0phily. Discussion of Hypotheses The failure to obtain support in the present study for any of the 28 hypotheses might be due to the following reasons: 1. One of the criteria for selecting the 20 Brazilian communities studies was having a place where villagers could attend literacy training classes. Under these circumstances, two possible results could take place. a. Those villagers who go to the literacy training program might very well be similar to one another in terms of level of literacy; other- wise, they would not be assigned to the same program. Besides their level of literacy, they probably have other similar characteristics since 59 Table 9. The zero-order Pearsonian correlations between innovative or change agent contact homophily and opinion leadership concentration or size of community Types of Homophily Opinion Leadership Concentration Variables Size of Community Innovative homophily on the basis of friendship Innovative homophily on the basis of influence Change agent contact homophily on the basis of friendship Change agent contact homophily on the basis of influence -.048 -.022 -0235 '0161 .082 .050 -001“ .079 *Significantly different at the 5 per cent level. ‘“”"u--r -_..._ “Waxy 60 level of individual literacy in rural areas is related to "measure of power and social status (such as income, control of land, number of employees) and to mass media usage" (Whiting and others, 1968, p. 139). An old saying, "Birds of a feather flock together," may be proved true among these Brazilian villagers. b- Those who are not in the program may use the training center as a gathering place. Thus on the basis of unexpected results* we might speculate that despite the presence of a higher degree of mass media eXposure, more frequent change agent contacts, etc,, Brazilian villagers still interact with others who have similar degrees of innovativeness and change agent contacts. 2. The present study was exploratory in nature. This is the first study of community homephily using the village as a unit of analysis. There was no past research to guide its design. Ways and means to improve its design are open to investigation by future researchers. 3. Since we did not find a high degree of general homoPhily among 20 Brazilian communities, we decided to use innovative and change agent contact homOphily as indicants of community hom0phily. If we *We expected negative relationship between four types of homophily, i.e., innovative and change agent contact homophily on the basis of friendship or influence and seven modernization variables such as mass media exposure, change agent contact, cos- mOpoliteness, etc., but we found correlations between them are low and positive and not significant. 61 had used four groups of homophi1y*, i.e., cosmopolite, information and source, adoption and economic group homophily, found in the present study as indicants of community homophily, we might have been able to _ get more expected results. The reasons for selecting innovative and change agent contact homophily are: a. In traditional communities, change agent contact is one of the best indicators of villagers' external contacts with new ideas vital to modernization. b. In traditional communities, innovativeness is the crucial fact of villagers which determines how many new ideas they might adopt. However, we found that innovative and change agent contact homo- phily are only two of the three elements in adoption group homophily.** In other words, we suspect that adoption group homophily might be a better indicant of community homophily than either innovative or change agent contact homophily alone. 4. Based on the same Brazilian data, Guimaraes (1968, p. 50) inferred that "the (relative) modern community exhibits relatively more integration in its communication system*** than does the (relative) traditional community (in Brazil)." Whiting and others (1968, p. 31) also reported that "The (Brazilian) success cOmmunities (i.e., more modern) *See p. 32 of this thesis for a clear description of these four groups of homophily. ' **See p. 32 of this thesis for different elements in different . groups of homophily. ***Guimaraes defined communication integration as "a communication system which maybe a subsystem of a social system (in one case, a community) embraces such communication subsystem as cliques, subgropps, chaina, and dyads." (1968, p. 44) Unpublished M.A. thesis. East Lansing, Department of Communication, Michigan State University. 62 tend to have fewer cliques."* Therefore, we might speculate that there are many cliques in Brazilian communities. Based on our findings again, despite a high degree of mass media exposure, more frequent change agent contacts, etc., villagers still tend to interact with others in the same cliques. 5. The data were not originally collected for the present study. For example, of all 1,307 subjects, we found only 3,340 pairs of general dyads consisting of friend and influence dyads when we asked respondents, "Who are three friends with whom you talk most frequently?", and "Who are the three persons in this community who are more listened to or imitated when it comes to farming and cattle raising in general?" We might better have gone beyond the sought to find out from whom did he sought information, because he might seek information from someone not included in 1,307 subjects. Rogers and Jain (1968, o. 28) suggested that "past diffusion research has over emphasized investigation in which the individual, rather than communication relationship, is the unit of analysis." If we had utilized pairing, chain, or snowball sampling techniques, we might have gotten a better measure of general homophily in Brazil as a result. 6. Katz and Lazarsfuld (1955) indicated that gregariousness is an important characteristic of opinion leaders. Perhaps in the information- seeking interaction, seekers in Brazilian communities rely more upon the *Guimaraes, pp, cit., p. 31, defined a clique as a "subsystem (of the communication system) in which at least three members mutually interact." 63 sought's social ability i.e., gregariousness, rather than his knowledge of innovations as an information-seeking condition. We may assume that popular villagers are similar to seekers in different attributes; other- wise, they can hardly be popular among those who are quite different from themselves. On the basis of our unexpected results, we might say that despite a high degree of mass media exposure, more frequent change agent contacts, etc., seekers of information in Brazilian communities still tend to seek information from other who are similar to themselves on the basis of innovative or change agent contact homophily. Suggestions for Future Research Importance of homophily in the field of diffusion of innovations was suggested by the present study. However, the study of homophily, particularly at the community level, is only in its infant stage. This present study was exploratory in nature; therefore, exploration of new variables such as interpersonal, psychological, and other communication variables which might be related to homophily indices is very important. There are at least two approaches which deserve attention: 1. Investigation of those variables negatively related to four types of homophily in the present study. These variables are movie ex- posure, TV exposure, newspaper and magazine readership, radio exposure, change agent contact, opinionatedness, and innovativeness. We found only weak evidence that the higher the degree of these variables, the less likely farmers will interact with others who have similar degree of change agent contact and innovative homophily. 64 On the other hand, it does not mean that those variables positively related to homophily should be discarded because they tend to increase degree of homophily in the community. These variables are visits to city, literacy, and community economic development. These results could be due to special circumstances in Brazil as we pointed out in the first part of this chapter. Interested researchers are encouraged to look into these variables as well. In addition to modernization variables used in the present study, special attention should be paid to interpersonal, psychological, and other communication variables. These variables such as personal attractiveness, personality, and motivation, could be relevant to homophily. 2. We emphasized that dyadic analysis, i.e., using pairs of in- dividuals as the unit of analysis, is better than conventional analysis, i.e., using the individual as the unit of analysis. However, we feel that it might be fruitful for future researchers to use a combination of dyadic and individual analyses. For example, future researchers should first investigate how homophilous each individual is in the community. Then they could get a composite homophily score for all dyads in the community and investigate relationships between their scores and modern- ization variables. BIBLIOGRAPHY Alves, Eliseu Roberto Andrade, Adacao de Practica: Area Atinigida pelo Escritorio Local de Viscosa, Minas Gerais, The Association for Credit and Rural Assistance in Brazil, 1961. Bose, Santi Priya, Dasgupta S., Rudra, B. and Biswas, T. K. Eadpur, A West Bengal Village, Calcutta: Department of Agriculture, Government of West Bengal, 1963. Chou, Teresa K. M. Hom0phily in Interaction Patterns in the Diffusion of Innovation in Colombian Villages, M. A. thesis, Michigan State University, East Lansing, 1966. Coleman, James S. "Relational Analysis: The Study of Social Organi- zations with Survey Methods," Human Organization 14:28-36, 1955. Dasgupta, Satadal, Caste Structure and Agricultural Development: A Case Study of Two Villages in Uttar Pradesh, India, Paper presented at the Annual Meeting of the Rural Sociological Society, Boston, Mass., August 24-26, 1968. Davis, Richard H. Personal and Organizational Variables Related to the Adoption of Educational Innovations in Two Liberal Arts Colleges, Unpublished Ph. D. thesis, University of Chicago 1965: Deutschmann, Paul J. "The Mass Media in an UnderdeveloPed'Village"‘ Journalism Quarterly, 1963, Vol. 40, p. 35. Eibler, John H. A Comparison of the Relationship between Certain A8pects of the Structure of High School Faculty and the Amount of Curriculum Innovation, Unpublished Ph. D. thesis, University of Michigan, 1965. Emery, F. E. and Oeser, O. A. Information, Decision and Action: A Study of Psychological Determinants of Changes in Farmimg_ Techniques, New York: Cambridge University Press, 1962. Emery, F. E. and Oeser, O. A. and Ryan, Bryce, Social and Ecological Patterns in the Farm Leadership of Four Iowa Townshipa, Ames Iowa Agricultural Experiment Station Bulletin 306, 1942. Feldman, Sidney P. "Some Dyadic Relationships Associated with Consumer Choice," in Hans, Raymond M. (ed.) Science, Technology and Marketing, Chicago: American Marketing Association, 1966. Festinger, Leon, A Theory of Cognitive Dissonance, Row, Peterson, and Company, Evanston, Ill., 1957. 65 66 Guimaraes, Lytton L. Matrix Multiplication in the Study of Inter- personal Communication, M.A. thesis, Michigan State University, East Lansing, 1968. Herzog, William A., Whiting, G. C. and Standield, J. D. Patterns of Diffusion in Rural Brazil: Correlates of Innovativeness and Opinion Leadership in 20 Minas Gerais Communities, Michigan State University, East Lansing, 1968. Katz, Elihu, and Lazarsfeld, Paul F. Personal Influence: The Part Played by People in the Field of Mass Communication, New York, The Free Press of Glencoe, 1955. Katz, Elihu, "Traditions of Research on the Diffusion of Innovation," American Sociological Review, 28:237-252, 1963. Lazarsfeld, Paul F. and Merton, Robert K. "Friendship and Social Processes: A Substantive and Methodological Analysis," in Berger, M., Abel, T., and Page, C. H. (eds.), Freedom and Control in Modern Society. New York: Octogon Books, Inc., 1954. Lerner, Daniel, The Passing of Traditional Society: Modernizing the Middle East, New York: Free Press, 1958. Lerner, Daniel, "Toward a Communication Theory of Modernization" in Lucian W. Pye (ed.) Communications and Political Development, Princeton; Princeton University Press, 1963, 311. Lin, Nan and Leu, Donald J., Rogers, Everett M., and Schwartz, Donald F. The Diffusion of an Innovation in Three Michigan High Schools: Institution Buildinnghrough Change, Institute for International Studies in Education and Department of Communication, Michigan State University, 1966 . Lionberger, Herbert F. and Campbell, Rex R. The Potential of Interpersonal Communicative Networks for Message Transfer from Outside Informa- tion Sources: A Study of Two Missouri Communities, Columbia: Missoufi Agricultural Experiment Station Research Bulletin 842, 1963. Lionberger, Herbert F. and Coughenour, M.C. Social Structure and Diffusion of InfOrmation, Columbia, Missouri: Agricultural Experiment Station Research Bulletin 631, 1957. Lionberger, Herbert F. "Some Characteristics of Fram Operators Sought as Sources of Farm Information in a Missouri Community" Rural Sociology, 18:237-338, 1953. 67 McNelley, John T. Mass Communication and the Climate for Moder- nization in Latin America, Paper presented at the Association for Education in Journalism, Austin, Texas, 1964.' Menzel, Herbert, and Katz, Elihu, "Social Relations and Innovation in the Medical Profession: The Epidemiology of the New Drug," Public Opinion Quarterly, 19:337-352, 1955. Merton, Robert K. "Patterns of Influence." in Lazarsfeld, Paul F. and Stantion, Frank N., (eds) Communication Research. New York: Harper and Brothers, 1949. Patel, Nasi, Status Determinants of Interpersonal Communication: A Dyadic Anaiysis, Ph. D. thesis, Lexington: University of Kentucky, 1966. Pye, Lucian W. "Models of Traditional, Transitional and Modern Communication Systems" in Pye, Lucian W. (ed.), Communication and Political Development, Princeton, N.J.: Princeton University Press, 1953. Rogers, Everett M. Diffusion of Innovations, New York: The Free Press of Glencoe, 1962. Rogers, Everett M. and Herzog, William A. "Functional Literacy among Colombian Peasants," Economic Development and Cultural Change, 14:190-192, 1966. Rogers, Everett M. "Mass Media EXposure and Modernization Among Colombian Peasants," Public Opinion Quartemiy, 29:614-625, 1965. Rogers, Everett M. and Cartano, David G. ”Method of Measuring Opinion Leadership," Public Opinion Quarteriy, 26:435-437, 1962. Rogers, Everett M. with Lynne Svenning. Modernization Among Peasants: The Impact of Communication, New York: Holt, Winston and Rinehart, 1969. Rogers, Everett M. and Jain, Nemi C., "Needed Research on Diffusion within Educational Organizations," Paper presented at the National Conference on the Diffusion of Educational Ideas, East Lansing, Michigan, 1968. Schram, Wilbur, "Communication Development and the Development Process," in Pye, Lucian W. (ed.) Communication and Political Development, Princeton, N.J.: Princeton University Press, 1963. _, AL” A“ i I 68 Stanfield, J. David, The Correlates of Community Economic Development in Brazil, Paper presented at the Rural Sociological Society Convention, Boston, August 26, 1968. United Nations Document E/CN5/291, Programme of Concerted Action in the Social Field of the United Nations and Specialized Agency, New York. Van den Ben, Anne W., De Communicatie van Niewe Landbonwmethodeu, Asseu, Netherlands: Royal Vansgorsum, 1963. Yadav, Dharam P. Communication Structure and Innovation Diffusion in Two Indian Villages, Technical Report 2, Project on the DiffusiOn of Innovations in Rural Societies, East Lansing, Department of Communication, Michigan State University, 1967. Warland, Rex H. Personal Influence: The Degree of Similarity of Those Who Interact, M.S. thesis, Iowa State University, Ames, Iowa, 1963. Whiting, Gordon C., Herzog, William A., Quesada, Gustavo M., Stanfield, J.D., and Guimaraes, L. Innovation in Brazil: Success and Failure of Agricultural Programs in 76 Minas Gerais Communities, Diffusion of Innovations Research Report 7, Diffusion of Innovations in Rural Societies, Department of Communication, Michigan State University, East Lansing, 1968. _ar 0" APPENDIX A THE QUESTIONNAIRE 70 DO NOT WRITE IN THIS SPACE Project 712 Phase 02 Subject Number Reitoria da Universidade Federal de Minas Gerais and Michigan State University September 1, 1968 Date: Interviewer: Name of Community: Type of Community: (INTERVIEWER: INDICATE BETTER OR POOR COMMUNITY) 1. low old are you? (INTERVIEWER: IF THE RESPONDENT DOES NOT KNOW HIS OWN AGE ASK YEAR OF BIRTH) 2. How many years did you attend school? 0) None or doesn't apply 1) Part of primary 2) Primary completed 3) Secondary-high school 4) University 3. Have you ever lived away from this community? 0) Never lived outside community 1) Lived outside but not in a large city; doesn't know; no reSponse 2) Lived in a large city “if ‘ ‘u-lL-p-ejm 9. 71 Did you visit a large city in past year? (One with more than 40,000 inhabitants) 1) Did not visit a large city in past year 1) One visit in past year to a large city 2) Two visits in past year to a large city 3) Three 1: n n I") H H H H H) H H H H 99) 99 H H H Are you a member of the cooperative? O) No 1) Doesn't know; no response 2) Yes To how many other organizations, clubs, societies, charities (conferencias), etc. do you belOng? 0) Zero organization as a member of 1) One organization as a member of H) H N H 9) Nine organizations as a member of Can you read (or someone in family can read) a newspaper? 0) No one in family can read a newspaper 1) He cannot, but someone in family can, doesn't know; no response 2) Re3pondent can read a newspaper Have you read (or has anybody read for you) newspapers or magazines lately? (IF YES) How many times a month? 0) Zero time per month exposed to newspaper or a magazine H) " H 99) 99 times " Do you have a radio at home? 0) No 1) Doesn't know; no response 2) Yes lo. 11. 12. 13. 14. 72 How often do you listen to the radio? Never (SKIP TO QUESTION 28) Almost never; doesn't know; no reSponse Sometimes More or less than an hour per day What program do you like best? Doesn't listen Musical Sports Others, doesn't know; no response Agricultural Do you watch TV? 0) 1) 2) Never sees it Sometimes sees it, in some other place; doesn't know; no response More or less regularly Do you go to the cinema? 0) Doesn't go 1) Once per year attended cinema 2) Twice per year attended cinema 3) Three times per year attended cinema H) H H ) N 99) " Do you usually receive news about agriculture through: a. Radio 0) No 1) Doesn't know; no response 2) Yes b. Television 0) No 1) Doesn't know; no reSponse 2) Yes C. e. 15. 0) 1) 2) 3) 4) 5) 6) 7) 16. a. 0) 1) 2) b. 0) l) 2) 0) 1) 2) Has practice 73 Newspaper O) NO 1) Doesn't know; no response 2) Yes Magazine 0) No 1) Doesn't know; no response 2) Yes ACAR Bulletin 0) No 1) Doesn't know; no response 2) Yes How many sources you utilized Zero sources utilized One source utilized Two sources utilized Three sources utilized Four sources utilized Five sources utilized Six sources utilized Seven sources utilized ever been used? Practice A No Doesn't know; no reSponse Yes Practice B No Doesn't know; no response Yes Practice C No Doesn't know; no response Yes Practice D O) No 1) Doesn't 2) Yes Practice E 0) No 1) Doesn't 2) Yes Practice F O) No 1) Doesn't 2) Yes Practice G 0) No 1) Doesn't 2) Yes Practice H O) No 1) Doesn't 2) Yes Practice I 0) No 1) Doesn't 2) Yes Practice J 0) No 1) Doesn't 2) Yes Practice K O) No 1) Doesn't 2) Yes know; know; know; know; know; know; know; know; no no no no no no no no 74 response reSponse reSponse response response response reSponse response 17. How many years since you used it for the first 75 1. Practice L 0) No 1) Doesn't know; no response 2) Yes time? a. Practice A b. C. d. e. f. 00) Year of adoption 66) Practice B 00) Year of adoption 66) Practice C 00) Year of adoption 66) Practice D 00) Year of adoption 66) Practice E 00) Year of adoption 66) Practice F 00) Year of adoption 66) of practice of practice of practice of practice of practice of practice 18. 'g. Practice G 00) Year of adoption 66) h. Practice H 00) Year of adoption 66) i. Practice I 00) Year of adOption 66) j. Practice J 00) Year of adoption 66) k. Practice K 00) Year of adOption 66) l . Practice L 00) Year of adoption 66) What practices are you 0) None practice 1) One practice H) I! H) H I!) H 12) Twelve practices 76 of practice of practice of practice of practice of practice of practiceL still using? 19. 20. 21. 22. 23. 77 From whom did you first hear about? 0) Neighbor 1) Person from outside community 2) Doesn't know; no response 3) Agronomist 4) Radio or Newspaper Who convinced you to use it? 0) Neighbor 1) Stranger 2) Doesn't know, no reSponse 3) Agronomist 4) Radio or newsPaper What do you consider best when it comes to dealing with your neighbors: trust, or "trusting them but at the same time doubting them?" 0) "Trust ... untrustingly" 1) Doesn't know; no reSponse 2) Trust Can you read this card for me? "He who cannot read is like a blind man who has to be guided according to other people's wishes; or then ’he will stumble his way. The illiterate man is not altogether free; he is a slave of his ignorance. Never stop reading something every day and keep learning." 0) Cannot 1) Slow and stammering 2) More or less satisfactorily CODER: SUBTRACT THE NUMBER OF WORDS MISSED FROM .50 AND CODE THE RESULT 0) 00-04 -— Functionally illiterate 1) 05-09 2) 10-14 3) 15-19 N) 20-2u 5) 25-29 6) ao-au 7) 35-39 8) 40-44 9) 45-50 -- Functionally literate 24. 25. 26. 78 All right, from what you have just read can you tell me now in what sense "the illiterate is not free?" 0) Didn't understand the reading 1) Understood vaguely 2) Understood well How many times have you talked to the ACAR Agent last year? 0) Zero times last year reSpondent talked to ACAR agent. 1) One time last year reSpondent talked to ACAR agent 2) Two times last year respondent talked to ACAR agent 99) Ninety-nine times last year respondent talked to ACAR agent Do you own ... a. Water filter 0) No 1) Yes b. Wood or tile floor in your house 0) No 1) Yes c. Refrigerator 0) No 1) Yes d. Stove with Chimney 0) No 1) Yes e. Plumbing for running water 0) No 1) Yes 27. 79 Bathroom inside the house 0) No 1) Yes Electric lighting 0) No 1) Yes Radio 0) No 1) Yes Television Set 0) No 1) Yes Motorized vehicle 0) No 1) Yes Any property 0) No 1) Yes House in town 0) No 1) Yes Agricultural machines 0) No 1) Yes Who is the present governor of Minas? To what American country has Brazil sent troops last year? Who was the President of Brazil who was deposed two years ago? What is the name of one of the parties created after the 1964 revolution? Which country of Latin America became communist a few years ago? 28. 29. 30. 31. 80 0) Zero right answers 1) One right answer 2) Two right answers 3) Three right answers 4) Four right answers 5) Five right answers All sources considered, what was approximately the total cash income of your family last year? (salaries, revenues, etc.) 1) Less than Cr$500,000 2) From Cr$500,000 to Cr$,000,000 -3) From Cr$l,000,000 to Cr$2,000,000 4) From Cr$2,000,000 to Cr$3,000,000 5) From Cr$3,000,000 to Cr$4,000,000 6) From Cr$4,000,000 to Cr$6,000,000 7) From Cr$6,000,000 to Cr$10,000,000 8) More than Cr$10,000,000 Conversion rate at the time of data collection: Cr$2,210 = US $1.00 What is the total area in pasture? 0) Zero hectares of pasture land 1) One hectare of pasture land 99) Nine- How many milk cows do you have? 0) 0 Milk cows 1) One milk cow 99) Ninety-nine milk cows What is the total area of your property? 0) Zero hectares of respondent's property 1) One hectare of reSpondent's property 99) Ninety-nine hectares of respondent's pr0perty 81 32. Who are three friends with whom you talk most frequently? 0) 1) «36) Zero times reSpondent was mentioned by others as being a best friend One time reSpondent was mentioned by others as being a best friend Ninety-nine times reSpondent was mentioned by others as being a best friend 33. Who are the three persons in this community who are more listened to or more imitated when it comes to farming and cattle raising in general? 0) 1) 99) Zero times respondent was mentioned by others as being influential in regard to agriculture in general One time respondent was mentioned by others as being influential in regard to agriculture in general Ninety-nine times respondent was mentioned 34. In this community who is more listened to in regard to Practice A? 0) 1) 99) Zero times reSpondentwwas mentioned by others as being influential in regard to practice A One time respondent was mentioned by others as being influential in regard to practice A Ninety-nine times respondent was mentioned by others as being influential in regard to practice A 35. SCORE: Polymorphic Opinion leadership on three practices 0) 1) 2) Zero practices for which respondent received any nominations in question 41 One practice for which reSpondent received any nominations in question 41 Two practices fOr which respondent received any nomincations in question 41 82 36. If you need to borrow money, who would you ask for it? 0) Zero times reSpondent was mentioned by others as a potential source for a loan 1) One time respondent was mentioned by others as a potential source for a loan 99) Ninety-nine times respondent was mentioned by others as a potential source for a loan 37. Who do you think would be the best person to organize this (COOperative community project) in the community? 0) Zero times reSpondent was mentioned by others as a potential leader for a cooperative community project 1) One time reSpondent was mentioned by others as a potential leader for a cooPerative community project 99) Ninety-nine times respondent was mentioned by others as a potential leader for a COOperative community project. 38. SCORE: Opinionatedness How many questions on which did you express opinions? 0) Zero opinion questions on which respondent expressed any opinion at all 1) One opinion question on which respondent eXpressed any Opinion at all 9) Nine opinion questions on which respondent expressed any Opinion at all APPENDIX B Friendship Dyadic Relationship Card and Friendship and Dyadic Card Format Appendix B Friendship Dyadic Relationship Card* Column no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 ... 36 37 38 48 49 50 ... 80 O 1 . p l l l 2 l I ' l 3 I u l I 5 I l ' ' 6 i l 7 Q I 8 9 l Column Contents 1,2,3 Project number; 712 4.5 Community number; 11 6,7,8 Seeker's number 1,2,3 9,10 Deck number; 50 .11,12 Card.number; 2 13,14 ‘Age; 50 95.37.38 Idfluence I.D. number; 4,5,6 ”3. Type of sought; 4 or 5 or 6 (Influential) 49,50 Age; 59 *Influence dyadic relationship card is same as that of friendship's. 84 Card Column 85 Friendship and Dyadic Card Format Contents 1 1—3 4-5 11-12 13-14 15 16-17 18-19 20 21-22 23 24 25 26 27-28 29-30 31-32 33 34 35-36 Project number. Community identification number. Respondent number and Reproduce all 6—8. Deck number (50) Dyadic Deck. Card.number. (01) Age of seeker. Seeker's education. Seeker's visits to city in the past year. Newspapers and Magazines seeker reads Seeker's radio listening. Seeker's cinema attendance. Total number of sources seeker utilized. Seeker's source for first knowledge of practice "A". Seeker's most influential source in making decision to adopt practice "A". Seeker's trust in neighbors. Seeker's functional literacy. Seeker's contact with ACAR supervisor in past year Seeker's home and farm equipment and improvements. Seeker's political knowledge. Seeker's income. Seeker's ownership of cows. Card Column 86 Content 37 38-39 40-41 42-44 45-47 48 49-50 51 52-53 54-55 56 57-58 59 6O 61 62 63-64 65-66 Seeker's opinionatedness. Seeker's social participation. Opinion leadership total Seeker's innovativeness. Source's I.D. No. lst. friend 1 2nd. friend 2 3rd. friend 3 lst. Influence 1 2nd. Influence 2 3rd. Influence 3 1, 2, or 3 for lst, 2nd, or 3rd. Source for friend 4, 5, or 6, for lst, 2nd, or 3rd. Choice for influence. Sought's age. Sought's education. Sought's visits to a large city in past year. Newspapers and magazines sought reads. Sought's radio listening. Sought's cinema attendance. Total number of sources sought utilized. Sought's source for first knowledge of practice "A". Sought's most influential source in making decision to adopt practice "A". Sought's trust in neighbors. Sought's functional literacy. Sought's contact with ACAR supervisor in past year. Card Column 87 Content 67-68 69 70 71-72 73 74-75 76-77 78-80 Sought's home and farm equipment and improvements. Sought's political knowledge. Sought's income. Sought's income. Sought's opinionatedness. Sought's social participation. Opinion leadership total. Sought's innovativeness. "IT'Eiifljili'ifllifllifliITS