{NFORMATION AND MODERNJZATION: A STUDY OF EASTERN NIGERIAN. FARMERS Thesis for the Degree of Ph. D. MECHIGAN STATE UNiVERSITY ROBERT F. KEITH 1968 I 1' t a l, [Ht-3‘9“? [RV/l“ ‘ :th [fighvxw u-y -,;---:-—-— _._ na- This is to certify that the thesis entitled .. 1 Information and Modernization: A Study of Eastern Nigerian Farmers presented by ' Robert F. Keith has been accepted towards fulfillment of the requirements for Ph.D. d . Communication egree 1n ____ Date May 1 7 ‘ ( F 0-169 ,rH—‘fi Luv-.Htfi We ,5: BINBING BY HDAG & saw 300x smnmv mc. LIBRARY amoms I""““xit?. iifltfll "is‘ an 3 10104 4 ‘Wwaee ” * ~ , M mo. W «fit RM @11 Wm mgr, #1 LIL! ""6911: ‘ WE: E Weds“ 3 My; cowl? W t W ABSTRACT INFORMATION AND MODERNIZATION: A STUDY AMONG EASTERN NIGERIAN FARMERS by Robert F. Keith ’ The present study explored the relationship be- tween levels of information about agricultural inno- vations and the relative earliness of adoption of such innovations. Information was viewed as a necessary condition for the adoption of innovations. The infor- mation level - innovativeness relationship was examined within the framework of a paradigm which considered: (1) characteristics of the adopters social system (the village); (2) the degree to which the individual #’::i;participates in his information environment; (3) per- ‘ sonal characteristics of the individual. Social systems were characterized by their capacity to acquire and provide information for ’1ndividua1 decision—makers. various indices of infor- _i Zgymation inputs and input facilities were constructed. , HiIndividual participation in one's information environ- rément was indexed by degree of exposure to various sources Robert F. Keith The present data were part of a study conducted in Eastern Nigeria. Data used to characterize social systems were derived from responses by village leaders in 18 'villages. Data used to characterize individual adopters (and non-adopters) were derived from responses by indivi- dual heads of households in the same 18 villages. The data were analyzed, first, by means of partial correlations on the relationship between each independent variable and (1) individual information level, and (2) innovativeness controlling for the effects of all other independent variables. Information level was positively related to (1) leader relay capacity; (2) variety of sources to which one is exposed; (3) cos- mopolite source exposure at awareness; (4) attitude toward-innovations. Innovativeness was positively related to (l) constant information inputs; (2)-variety of source to which one is exposed; (3) cosmopolite source exposure at awareness; (h) attitude toward innovations; and (5) information level. Multivariate analyses, namely, sequential interac- tion analyses, were performed to determine the explanatory sconfigurations of (1) information level; and (2) inno- vativeness. Those individuals with the highest infor- thation levels were characterized by very favorable fititudes toward innovations and exposure to several Robert F. Keith sources of information. Those with lower levels of in- formation were less favorable towards the innovations. Among this group, attributing awareness to cosmopolite rather than localite sources was associated with some- what higher levels of information. One hundred and nineteen respondents were unaware of any of the three innovations being studied. The most innovative respondents were typified by having been convinced of the innovations' utility, having favorable attitudes toward the innovations and being exposed to several sources of information. Those who were moderately innovative had been convinced of the innovations' usefulness, and either had (1) favorable attitudes toward the innovations and exposure to a few sources of information or, (2) less favorable attitudes toward innovations but exposed to several sources. The slightly innovative respondents were somewhat less favorable toward the innovations, were exposed to only a few sources and in some cases had not been convinced by others Of the innovations' utility. The least innovative or "laggards" were either unaware of the innovations or ~if aware exposed to few sources and unconvinced by any- one or anything of the innovations' utility. _ For the most part, characteristics of the social dxiuaystems did not feature in the explanatory configurations. is partial correlations reveal they account for only a Robert F. Keith f‘fqflmall portion of the variance. One of the problems in A ' measurement and the nature of the distributions of villages on these variables may account for these findings. The measure attitude toward innovations is also suspect. The partial correlation between attitude and information level .is .62 (the zero order correlation is .80). The extent and nature of the individual's participation in his in- formation environment was important in explaining innovativeness and to some extent information level. While some evidence is presented to warrant further consideration of the paradigm of information -and modernization, conceptual and operational clarity and further research, especially concerning attributes of social systems, is needed. at", 4' ." H". pr 1' f!“" .,$ \. l‘ I.\ I. a . J'gfi +1!“ " u' 1: ‘ ‘1‘- "5.. Accepted by the faculty of the Department of ;» 1cation, College of Communication Arts, Michigan 0 University, in partial fulfillment of the re- rec or 0 es s .9“ dance Committee: , Chairman Information and Modernization: A Study of Eastern Nigerian Farmers by ' \ Cw a I“ Robert FfiaKeith A THESIS Submitted to Michigan State university in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1968 ,, .. O - y . . ”3 .. _ g . A -U L". ‘ l; ' ‘ ‘ ..J.1'.__, , . . : .. n 4 .. {sf-'7 _ ,l . . ACKNOWLEDGMENTS 7 The author wishes to express his appreciation to his chairman, Dr. Everett M. Rogers for his considerable assistance and support throughout both the writer's dissertation and entire graduate program. Thanks is also extended to the other members of the author's committee; Drs. E. P. Bettinghaus, Eugene Jacobson, Hideya Kumata and V. C. Troldahl. The author is also indebted to the U. S. Agency for International Development and the Department of Communication, Michigan State University for permission to make use of data collected in Eastern Nigeria for the project, ”Diffusion of Innovations in Rural Societies." Many persons have substantially assisted the author during the course of the dissertation. The author is particularly indebted to Anant Saxena for his considerable help in the data analysis phase. Gerald D. Hursh, J. R. Ascroft and Hal W. Hepler contributed constructive thoughts and a manageable perspective to the dissertation process. To Mrs. Shirley Sherman, the typist, for her many hours of work, especially at the "eleventh hour", the writer ex- presses his thanks. Recognition must also be accorded the generous proposal of assistance from one four year old boy, who offered to assume the responsibility of writing the dissertation, if only the topic could be changed to hockey. To Dorinda - thank you. ii 5 TABLE OF CONTENTS CHAPTER Page I THE PROBLEM . . . . . . . . . . . . . . . . l The Setting . . . . . . . . . . . . . . . 7 II INFORMATION AND DECISION—MAKING . . . . . . 12 Information . . . . . . .'. . 12 Information and the Innovation Decision Process . . . . . 15 Information Levels and Thresholds . . . . 21 The Information Environment . . . . . . . 26 Source Orientations . . . . . . . . . 3h Attitudes toward Innovations . . . . . . #6 Innovativeness . . . . #9 Information Constraints, Information Levels and Behavior: A Multivariate Analysis . . . . . . . . . . . . . . . 50 III METHODOLOGY . . . . . . . . . . . . . . . . 55 Phase I Village and Respondent Selection Procedures . . . . . . . . 55 Village Selection . . . . . . . . . 55 Respondent Selection . . . . . . . . 56 Phase I Interview Schedules 57 Phase I Interviewer Selection and Training 58 Phase I Data Collection . . 5 Phase II Village and ReSpondent Selection } Procedures . . 59 Village Selection . . . . . . . . . . 59 Respondent Selection . . . . . . . . 61 Phase II Interview Schedule . . . 62 Phase II Interviewer Selection and Training 63 L Phase II Data Collection . . . . . . 63 Operationalization of variables . . . . . 6h Dependent variables . . . . . . . 64 Individual Information Level . . . . 64 Innovativeness and Adoption . . . 66 Hypotheses and Independent variable Operationalizations . . . . 67 Information Inputs . . . . . . . . 67 Source Orientations . . . . . 69 Attitudes toward Innovations . . . 88 Data Analysis . . . . . . . . . . . 89 Tests of Hypotheses . 89 l Multivariate Analyses of Modernization 89 iii r“ . ER Page ~.Iv FINDINGS..................96 Information Levels . . . . . . . . . . . . 96 Descriptive Findings . . . . . . . . . . 96 Test of Hypotheses . . . . . . . . . .105 Adoption and Innovativeness. . . . . . . .112 Descriptive Findings . . . . . .112 Tests of Hypotheses Dealing with Innovativeness . . . . . . .116 Multivariate Analyses . . . . . . . .121 Information Level Typologies . . . . . .125 Innovativeness Typologies . . . . . . .130 V SUMMARY AND DISCUSSION . . . . . . . . . . .134 Summary . . . . . . . . . . . . . . . . .13h Discussion . .138 Levels of Information and Innovative- ness . . . .138 Bivariate Relationships: Information level .. ................... . .1h1 Bivariate'Relationships: Innovative- ness . . Multivariate Relationships: Information leve1.. .....1LL6 Multivariate Analyses: Innovativeness .150 Implications for Research . . . . . . .156 *7 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . .164 1‘} TABLE 10 ll 12 13 14 15 16 LIST OF TABLES Page A SUmmary of Past Research on Correlates of Knowledge of Innovations . . . . . . . . 37 Derivation of Innovativeness Scores for Three Innovations . . . . . . . . . . . . . 68 Rotated Factor Loadings for 18 Non-Specific Information Input Items on 71 Villages . . 72 Distribution of Intermittent Information Inputs in 18 Villages . . . . . . . . . . . 74 Distribution of Constant Information Inputs in 18 Villages . . . . . . . . . . . 75 Rotated Factor Loadings of Source Exposure 83 Hypothesized Relationships and Statistical Analysis . . . . 90 Distribution of Awareness of 14 Agricultural Innovations . . . . . . . . . . . . . . . . 98 Percentages of Awareness of 14 Agricultural Innovations (N = 1, 347) . . . . . . . . . 99 Distribution of Number of Innovations Correctly Known . . . . . . . . . . . . . . 100 Intercorrelations Among Awareness Scores for Agricultural Innovations (N = 1,347). . 101 Intercorrelations Among Correct Information Scores for Agricultural Innovations (N=1,347)101 Intercorrelations Among Individual Infor- mation Level Scores by Innovations (N=1,347) 102 Intercorrelations Among Adoption Scores for Agricultural Innovations (N=1,347) . . . . 102 Percentages of Awareness and Correct Information About Three Agricultural Innovations (N=1,347). . . . 103 Distributions of Individual Information Levels by Innovation (N=1,3 7) . . . . . . 104 V Page Intercorrelations Among Dependent and Independent variables . . . . . . . . . . . 106 Correlations Between Independent variables and Individual Information Level and Innovativeness (N=1,347) . . . . . . . . . 108 Percentages Adopting Agricultural Innovations . . . . . . . . . . . . 113 Percent of Respondents Adopting Each of 14 Agricultural Innovations . . . . . . . . 114 Adoption Distributions for Three Innovations in Eastern Nigeria: Fertilizer, Aldrin ' Dust and NS- 1 Maize (N=1, 347) . . . . . . 115 Intercorrelations on Usage of Three Innovations in 1966 (N=1, 347). . . . . 116 vi _ ,‘i LIST OF FIGURES Paradigm of Information and Modernization . Hypothesized Relationships of Constant- intermittent Input Combinations and Information Levels . . . . . . . . . . . . Configurational Analysis of Prediction of Individual Information Levels . . . . . Information Level Typologies . . . . . . . Configurational Analysis of Predictors of Individual Innovativeness . . . . . . . . . Typologies of Innovativeness . . . . . . vii .Page 74 123 127 128 132 I. APPENDIX A: r APPENDIX B: ; APPENDIX C: ."V. '. APPENDIX D: _'- ii APPENDIX E: LIST OF APPENDICES Page List of Innovations Studied in Phase I and II, Eastern Nigeria . . . 168 Selected Items from Phase II: Farm Family Schedule . . . . . . . 170 Rotated Factor Loadings on Information Input Items for 18 Villages 174 Items and Weighting Scheme for an Index of External Contact . . . . . . 176 Rotated Factor Loadings for Twelve Items on Village Affairs and Village Farming Opinion Leadership . . . . 178 V111 CHAPTER I THE PROBLEM The process of the diffusion of innovations1 is the transmission of innovation information. The process of adoption is the acquisition and processing of such in- formation, followed by certain behaviors. Information is an essential element, common to both processes. In spite of its importance, information is one of the least—studied concepts in diffusion and adoption research. For the most part, studies of diffusion and adoption have assumed the transmission of innovation information. Though such an assumption is not without foundation, the assumption appears "to have diverted attention from a systematic investigation of information and its dissemination and utilization. The intent of the present study is to focus upon innovation in- ‘formation in the innovation decision-making context. Schultz (1964) argued that innovative behavior can best be explained by economic factors associated with both 7 1An innovation is defined as "an idea perceived as {‘new by the individual" (Rogers, 1962, p. 13). 2 the innovation and the decision-maker.2 While economic variables are undoubtedly important, information about the innovation, economic and otherwise, must be diffused to members of the social system before innovation—decisions can be made. This is to say that the new idea must be 4communicable in such a way as to permit decision—makers to acquire sufficient information to make decisions. Adoption is not only dependent upon the diffusion of information. The information object must be perceived as having some utility (economic, social, cultural, etc.). It is the processes of diffusion and interpretation which to some ex- tent qualify the more limited view of economic explanation. With inadequate or irrelevant information, economic, social and cultural factors may fail to explain behavior patterns as anticipated. Thus, further understanding of modernizing behavior changes must include some consideration of the in- formation environment and those attributes of decision-makers which influence the acquisition and interpretation of in- formation about innovations. It is axiomatic that a person must have some infor- mation about a course of action before he can accept or reject that course of action. In other words, a course of 2Schultz stated, "Since differences in profitability areastrong explanatory variable it is not necessary to appeal to differences in personality, education and social environ- ment" (Schultz, 1964, p. 164). 3 action must be known before it can have some probability of being accepted or rejected. Thus, information constitutes a necessary condition for adoption of an innovation. An intriguing and problematical aSpect of the informa— tion-decision making relationship is how much information an individual requires before he makes the decision to adopt or reject some course of action. various diffusion studies find a positive relationship between knowledge and adoption. Knowledge, for the most part, has been indexed as awareness of a new idea. Few attempts have been made to discern how levels of information are distributed among members of some population and how such variation is related to innovative behavior, information-seeking activity, attitudes toward innovations and characteristics of the information environment. The notion of an optimal level of information suggests a behavioral threshold with respect to levels of information. At some point on the continuum of information the decision maker perceives himself to have adequate information; or he decides that the cost of acquiring additional information is in excess of the utility of that additional information. Circumstances may also be such that the decision maker per— "3 ceives himself to be in a "forced-choice situation and 3For a discussion of knowledge types see Johnson and others (1961). thus, may adopt with less than a desired level of informa- ( tion. As part of the present study, the analysis will seek to determine how differential levels of information about innovations are related to decisions to adopt. To more fully discuss the information—behavior re- lationship, one must consider a variety of possible con- straints upon both information and behavior. Constraints f ' are those factors which are related to, and thus affect information levels and adoption, whose variation the present analysis attempts to explain. Three classes of such con— straints are considered: (1) the information environment; I (2) source orientations; and (3) attitudes toward innovations. 5 The information environment is defined as the extent F ‘ to which a social system is exposed to or possesses informa— tion about some idea or object. Individual information levels are necessarily constrained by what it is possible to know. It is suggested that social systems vary on this attribute of information levels. The more isolated and traditional systems experience fewer information inputs than more modern systems. Information inputs, in a sense, con- stitute the "objective information environment", that which could possibly be known. Three forms of information inputs are considered in the present analysis: (1) non-specific information input W" facilities; (2) specific information input facilities; and (3) leaders' information relay capacity. "Non-specific in- formation input facilities" are those facilities and mechanisms in a social system by which or through which information can pass. Specific information input facilities refer to diffusion agencies which disseminate innovation information. Their pur- pose is to transmit specific information to decision makers. The message content of these inputs is almost entirely in- strumental and occupationally relevant. The non—specific inputs may be more diverse in terms of kinds of content. Leaders in a social system may affect the innovation de- cisions of others in the system. To the extent the leaders' opinions about innovations are sought and, in addition, have contact with information systems external to their immediate social system, they operate as "gatekeepers."u Thus, the leaders can affect the levels of information of members of their social system. Subjective information environments or individual in- formation levels may differ from the objective environment “Village leaders may be looked upon as gatekeepers and perform information relay functions, as described by Westley and MacLean (1957). The "C role," that of the communicator, in the ABXC model, would seem to approximate the information relay function of village leaders. In essence, the leaders are extending the information environment of other members of the social system. for many reasons. One reason could be the sources utilized by decision makers to acquire innovation information. The diversity of kinds of sources to which one is exposed and the degree of exposure to these sources varies among in- dividuals. The present analysis seeks to ascertain how source exposure is related to information levels. In essence, we are concaned with the individual‘s degree of participation in the information environment, acknowledging that such environments may vary. Evaluative predispositions of the decision maker are another type of constraint upon information levels and be- havior. The relationship between attitudinal states and voluntary exposure to information is unclear at the present time.5 Assuming the decision maker is not in a forced- choice situation, we would expect those with positive attitudes toward innovations to be more inclined to engage in information seeking activity and, as a result, to be more knowledgeable. The present analysis seeks to ascertain how favorability toward innovations is related to information leVels. In addition, the attitude-behavior relationship will be explored to ascertain how attitudes affect decisions 5Hovland (1959) suggested that while attitudes may limit or constrain information seeking activities, the acquisition of new information could alter one's attitudes, which could alter information seeking behavior. Sears and Friedman (1967), i to adopt innovations. In summary, the present analysis is an attempt to determine how variation in information level affects decisions to adopt innovations. Three classes of constraints upon in— ) formation level will be investigated. Information inputs to the social system, source orientations and attitudes toward innovations are considered as potential constraints upon what one knows and, hence, upon adoption behavior. In addition to determining the nature of the bivariate relationships between these three classes of variables on the one hand, and infor— F mation levels and behavior on the other, the present analysis ) will seek to determine the conjunctive effects of these variables upon levels of information and innovative behavior. r The Setting The present analysis derives from a larger investigation I in Eastern Nigeria of the diffusion of agricultural infor- [ mation and the acceptance of agricultural innovations.6 Con- , in reviewing the research on selective exposure and prior attitudes, find equivocal evidence for the proposition of attitude—consonant information seeking. In addition to prior attitudes, they argue for consideration of the utility of l information and situational requirements, as possible deter- minants of voluntary exposure to supportive and non-supportive information. This study, conducted by the Department of Communication, Michigan State University, and sponsored by the U.S. Agency for International Development, was carried out concurrently in Brazil, India and Nigeria. In Nigeria the project was affiliated with the Economic Development Institute, University of Nigeria, Enugu Campus, and the Ministry of Agriculture. A fifwr sideration was given to social, psychological, cultural, economic and geographical factors in attempting to describe and explain the processes of diffusion and adoption. The present study utilizes some of the data collected by the re- search team in Nigeria. The author was involved in the initial research planning phase prior to the conduct of the study. All data were returned to Michigan State University for analysis. The author has used some of the data as obtained, while some data were rescored and combined into various indices as required by the concepts being investigated in the present study. The Nigerian study is composed of two phases. In Phase I the unit of analysis is the village. The intent of Phase I is to characterize villages in terms of their success in programs of planned change. The villages were indexed on such attributes as: level of leader modernization; presence of village improvement groups; change agency activities; in- stitutional development; social differentiation; communi- cation behavior. Attempts will be made to determine which factors best predict village modernization. In Phase II the unit of analysis is the individual head of a farm household. This phase sought to determine how the individual's social characteristics, family structure, farm business, communication behavior and selected social— psychological attributes (indicative of states of modernity), are related to his innovativeness, knowledge, attitudes toward change and opinion leadership. The present analysis utilizes data from both Phase I ( and II in Nigeria. The Phase II data provide the basis for ascertaining levels of information, source orientations, attitudes toward innovations and innovativeness. Phase I data are used to derive measures of the social systems' characteristics, notably information inputs. Respondents in Phase I were mainly village leaders. They were selected on the basis of frequency of sociometric mentions in specific leadership roles.7 Leadership roles were defined as administrative, religious, educational and civic, with the added distinction of each being modern or traditional. In addition, the information roles of village affairs opinion leadership and farm opinion leadership were utilized. Phase I respondents also included a small number of progressive farmers or innovators (17 percent of the total sample). Almost all (99 percent) of the Phase I respondents listed farming as either their primary or secondary occupation. Of the total sample, 65 percent inherited their land, nearly 60 percent were 49 years of age or less, and 40 percent had never attended school. 7The sampling procedures in both Phase I and Phase II are discussed in more detail in Chapter III. A --.-...— fw -——- 10 Data for Phase II were obtained from 1,347 male heads of households in 18 Eastern Nigerian villages. Nearly four- fifths (78 percent) of the respondents cited farming as their primary occupation. Fifty—two percent of the sample are from 30 to 49 years of age. Levels of education are low. Fifty-six percent have no formal education and thirty-three percent have an incomplete primary education. Thirty-seven percent of the sample have never participated in any of the several agricultural programs. The study considered 18 agricultural innovations (programs)8 in Phase I and 14 innovations in Phase II (see Appendix A for the list of innovations). The 14 programs studied in Phase II were also studied in Phase I. In Phase I, from 17 percent to 81 percent of the respondents knew at least one item of correct information for the various innovations. The range among the innovations in Phase II is very similar; 11 percent to 79 percent. In Phase I, 53 percent had not adopted any of the innovations. The most widely adopted innovation was fertilizer (22 percent). In Phase II, 37 percent had not adopted any of the innovations. 8The Ministry of Agriculture refers to each specific innovation it promotes and the efforts related to its promo- tion, as a "program". This differs somewhat from North American usage, where the entire extension program involving several innovations or projects is referred to as a "program". Throughout the present discussion, the term "innovation" will be used. v v‘ _ V -r———wvv V . V . 11 “$1Wifd adopted fertilizer, aldrin dust and NS—l maize. “ ' In summary then, the respondents in both phases are _:bxpr1marily farmers and are typified by low levels of education, gg',lbw levels of information and generally little participation 4 in programs of planned change. CHAPTER II INFORMATION AND DECISION MAKING Information A review of definitions of the concept of informa— tion reveals considerable variation. Information is fre- quently defined by use of a paraphrase. Smith and others (1956) define the concept of "informational support” as: ...the knowledge an individual possesses relevant to his attitude. It merely identifies the amount'ofhinformation a person is capable ofbringingto bear in appraising the topic of an attitude. The terms "information” and "cognitions" have been interchangeably used. Katz (1960) refers to the cognitive dimension of attitudes as that which constitutes one's non— evaluative description of an attitude object and its re- lationships to other objects. A somewhat different approach to the notion of in- formation was taken by Ackoff (1958), who suggested that a message has three potential functions with respect to a message receiver in a purposeful state: (1) to inform; (2) to instruct; and (3) to motivate. 12 13 A message is said to inform if it alters a receiver's perception of the number of courses of action available to him. This is analogous to the process of creating awareness of an innovation. A message is said to instruct when it alters a receiver's perceived efficiency of a course of action for an outcome. Messages or message segments which advise a receiver how to perform certain behaviors are in- structional in nature. Demonstrations, trials, skills, training, etc., are instructional messages. Messages which have outcomes as their referents and which are concerned with outcome values or advocate change in outcome assessments, are said to be motivational. Persuasive messages which emphasize the utility of increased yields, improved health, higher status, etc., are motivational messages. Any one message may perform more than one of these functions. Furthermore, a message may serve one function for one person and a different function for another person. All three types of messages contain "information" for the receiver in the sense that they alter the probabilities with which courses of action will be selected. As Ackoff (1958) indicates, the ”determinate state” where the probability of one course of action being selected is 1.0, and 0.0 for all the other courses of action, is the state of "maximum possible information". Where all courses of action are 14 equally likely to be selected, an indeterminate state of 3 minimum information exists. In the latter instance there is no basis of choice for selecting among alternatives. The role of a message source with reSpect to a potential adopter is to attempt to produce a sufficient degree of determinateness in the receiver to induce the receiver to select a particular course of action. In an information- seeking framework, the seeker is attempting to acquire inno- vation-relevant information which will produce that state of determinateness which is sufficient to permit him to select among alternative courses of action. Ackoff's formulation derives, in part, from the un- certainty reduction or information theory point of view. Shannon and Weaver (1949) use the term information to refer to knowledge that one does not have about a code, i.e., which symbol is coming next in a sequence of symbols. Where all events are equally likely, i.e., complete uncertainty (entropy), the state of maximum information (receptivity) exists. Once a choice has been made, that state no longer exists. Degrees of freedom have been lost, uncertainty is reduced, hence the assumption, information has been trans- mitted. Information is characterized as "bits" and the number of 2335 in a situation is defined as the number of choices (amount of information) to be made until only one alternative remains. 15 This cursory resume of some of the uses of the concept of information serves to accentuate variations in use of the concept. Present use of the term information follows that of Smith and others (1956). Information level about innovations is viewed as the degree to which the respondent can correctly identify certain aspects or characteristics of the innovation. Information and the Innovation Decision Process The present discussion focuses upon information and innovation decision-making. Thus, it is relevant at this point to discuss the decision—making process and, in so doing, attempt to discern the functions of information in that process. Various formulations of the "adoption process" are to be found in the literature. Among the most widely used is the fiveestage model (Rogers, 1962; NCRS Subcommittee, 1955). The five stages are: (l) awareness; (2) interest; (3) evaluation; (4) trial; and (5) adoption. Johnson and others (1956) provide a six component model of deciSion— making as follows: (1) problem definition; (2) observation; (3) analysis; (4) decision making; (5) action; (6) reSponsi— bility bearing. Brim and others (1962) also posit six elements in a model of decision—making: (1) problem identification; (2) information gathering; (3) determining problem solutions; (4) evaluating solutions; (5) selecting strategy; and (6) per- forming, learning and revising. 16 The similarities between these models are apparent and inevitable; the various models all deal with the decision process. All three models imply some degree of temporal order, though their authors would undoubtedly maintain that some of the elements or subprocesses may occur simultaneously or perhaps in reversed order. Mason (1964) demonstrated that the sequential order of the stages in the adoption process may vary by innovation. He also showed that the process was not terminated by adoption (or rejection). Evidence of post- adoption information seeking was noted. As Mason indicated, such a finding is consistent with Festinger's (1957) theory of cognitive dissonance. Such evidence suggests that the adoption process may be alternatively viewed as an interrelated set of processes (information—seeking; attitude formation or change, formu- lating alternatives; etc.) as contrasted to a single process view. From such a vantage point, Rogers with Shoemaker (1968) has re-cast the adoption process into four "functions" or sub- processes and labelled the overall process the "innovation decision process." Rogers defines the innovation decision process as "the mental process through which an individual passes from first knowledge of an innovation until adoption or rejection of that innovation and confirmation of the decision." The 1? four functions of the innovation decision process are: (1) knowledge; (2) persuasion; (3) decision; and (4) confir— mation. The model implies a temporal sequence. The knowledge function is defined as the subprocess in which ...the individual is exposed to the new ideas, but lacks complete information about it. He gains knowledge of the innovation but is not yet motivated to consciously seek additional information. The primary effect of the knowledge subprocess is that of "initiating a sequence of cognitive activities" which result in either activation of the remaining subprocesses or else in rejection. The persuasion function is that process by which the individual acquires additional information and develops an evaluative disposition toward the innovation. In other words, this function is the attitude formation/change process. Having developed an adequate evaluative disposition, the individual subsequently passes to the decision phase. Rogers describes the decision function as the activities associated with the selection of a particular course of action from some set of alternative courses of action. Presumably an innovation decision consists of at least two alternatives: (1) adopt the innovation; or (2) not adopt the innovations. In many instances innovations may be perceived as divisible i -..l. l8 phenomena, thus permitting choice among some set of alter- natives greater than two. Likewise, the decision maker may have available several innovations among which he may make a decision. The confirmation function is a post—decision process involving information acquisition (usually sought or solicited). Rogers explains information seeking at this phase of the process on the basis of dissonance reduction. The proposition is that decision makers will solicit or expose themselves to information consistent with their choice, thus confirming that choice. Of course, it is possible for the reverse to occur. Doubting the wisdom of his choice, an individual may expose himself to contradictory or behavior-discrepant in- formation, thus disconfirming his choice and providing a basis for subsequent disadoption. The explication of the innovation decision process elaborates the role of information in somewhat greater detail than do previous models. Knowledge and awareness are equated in the knowledge function. Information and evalua- tion formation/change are linked in the persuasion function. Information and reinforcement are associated in the confir- mation function. 19 The four function model implies a temporal sequence. To some extent this would seem to be a valid claim. Con— firmation of a decision must necessarily follow the decision. Selection of an alternative must necessarily follow cog- nizance of that course of action. However, the temporality of the scheme seems ambiguous with respect to the first two functions, knowledge and per- suasion. Both involve information acquisition and are thus not temporally distinct on this dimension. The primary distinction between these two functions is the development of an evaluative disposition, an attitude. Past researchers equate the knowledge function with being aware. It is difficult to entertain an evaluation-free knowledge state. Upon encountering an innovation, a potential adopter must make, among other decisions, a decision of relevance. To. the extent that an innovation is deemed relevant, the inno- vation is related to existing behaviors, associated with which are various evaluative states. If only by associational relevance, the knowledge function is evaluative as well as informative. If in addition to the decision of relevance, the potential adopter draws tentative conclusions about the innovation's potential utility, then the evaluative or attitudinal process is operative. '4 -__-_-_ _ r :‘-1 51.1: ‘1 '1’ 20 If one takes the position that evaluative activity takes place in both the knowledge and persuasion subprocesses, and that information acquisition occurs in both these sub— ! processes, the temporal distinction between these two functions becomes more difficult to maintain. This writer would argue ; that both are temporally concommitant processes, although } conceptually distinct. ( The suggestion then is that information acquisition and evaluation formation/change are concommitant subprocesses which interdependently affect each other to produce that state of determinateness which will permit or induce the individual to select a particular course of action. How these two processes affect each other is largely unknown at this time. Hovland (1959) hypothesized that though attitudes may result in selective information acquisition, information may subsequently modify the attitudes in question, which may in turn result in further information seeking, and so on. What are the implications of the suggested concommitant subprocess view? One implication is that information about innovations should be looked upon from a quantitative stand— point. Implicit in the concommitant subprocess view is the notion that the individual will attempt to acquire an adequate amount of information, which in conjunction with the evaluation of that information with respect to the perceived outcomes, i '1' 1'. 21 will permit or induce him to select a course of action. In effect then, we are concerned with how much information is needed in order to make a decision. Information Levels and Thresholds The discussion of innovation-relevant information will focus on three aspects of individual information states: (1)degree of awareness of fourteen agricultural innovations; (2) degree to which individuals have some correct information about these fourteen innovations; and (3) degree of correct information about each of three selected innovations. The first two aspects of information states serve the purposes of providing some basis for discerning the applicability of innovations to individual enterprises and of contextually orienting the discussion of three selected innovations for which more in-depth information states were obtained. To more meaningfully progress from a discussion of levels of information alone, to its relationships to other elements of the diffusion and adoption process, it is necessary to consider how information about innovations is distributed among members of some population. Fourteen agricultural innovations were investigated. In settings like Eastern Nigeria it may be presumptuous to anticipate widespread knowledge of all programs, let alone acceptance of such 22 innovations. Consideration of a diverse set of innovations is suggestive of the decisionrmaking range or context. When one knows how information and behavior with respect to several diverse practices are distributed, statements about selected innovations take on additional meaning as expressed in the relativity of their attributes to those of the other innovations. In particular, it becomes possible to speak relatively about variation in information levels and behavior patterns. Three innovations were selected for investigation in greater depth. The criteria for selection included applicability to individual respondents throughout the study area. The criterion of applicability was largely derived from reputational data, i.e., change agency personnel testified as to the appropriateness of the innovations. Reputational evidence is used since reSpondents who are unaware of innovations cannot be expected to make decisions about innovation applicability. The three innovations, fertilizer, aldrin dust (an insecticide) and NS-l maize——the only new variety of maize are reputed to be widely applicable for the population studied. This is a population of Eastern Nigerian villagers whose primary occu- pation is farming, and who are relatively early in the transi— tional phase from subsistence to modern commercial agricul- ture. ‘- 23 One of the primary objectives of the present analysis is to discern the relationship between (1) levels of informa- tion about innovations and (2) behavior. One might ask at this point why there ought to be a direct relationship be- tween information possessed about some course of action and novel behavior. Perhaps the important aspect of the innovation— decision situation among peasants in a developing country is that of limited modern alternatives. Whereas in modern society the variety of instances of a given type of object are often numerous, in the Eastern Nigerian context there is but one fertilizer, that provided by the government distribution system. There is but one appropriate insecticide, aldrin. The only instance of an improved variety of maize is the variety NS-l. In more developed nations we often note the existence of numerous instances of virtually the same innovation, i.e., several formulations of fertilizer and insecticide and many hybrid varieties. In such cases decision makers may acquire similar amounts of information for each of the alternatives; thus, information levels would not serve to discriminate adopters from non-adopters. More likely, attitudinal variables would distinguish these two categories. One may possess similar amounts of information about two objects, but differ markedly on the evaluative dimension. 24 In the context of the present analysis, however, only one instance of the attitude object, the innovation, is available. This does not negate the influence of attitudinal states upon behavior. However, the bases of comparison are the traditiOnal, less productive techniques rather than similar, modern and equally productive alternatives. Like- wise, there are few competing and contradictory messages from external diffusion agencies. Most negative information arises from within the social system. The effects of such information upon attitudinal and behavbral states cannot be ignored; analysis of the relationships between levels of information and innovative behavior must also account for the effects of favorability toward the innovations. However, it is suggested that in the one—modern-alternative situation, with individuals whose values appear to be strongly oriented to community and self bettermentg, the level of information about an innovation will be related to innovative behavior and will serve to dis- tinguish between adopter categories. Consideration of the relationship between the degree of information one has about a course of action and one's behavior with respect to that course of action, carries with 9 Uchendu (1965) testified to the strong desire on the part of Eastern Nigerian villagers for both self and community betterment, often expressed in the phrase ”getting up" 25 it the implication of both a continuum of information and of various thresholds. Two thresholds, in particular, are relevant. The first is that of being aware of an innovation. The ability to report that one knows of an innovation or to distinguish it from other objects, is usually taken as evidence of awareness. There is yet another important threshold. It was sug- gested previously that individuals will acquire information until that state of determinateness is produced which enables the individual to select among alternative courses of action. The postulation of states of determinateness and indeterminate— ness suggests a threshold, prior to which and after which some behaviors will differ. In the present analysis, the problem is one of attempting to ascertain those levels of information which are associated with decisions to adopt the innovations. In other words, we are asking how the continuum of information can be partitioned to maximize the degree to which variation in behavior can be explained. The partition points un— doubtedly vary among individuals and particular categories of individuals. Evidence of acceptance of innovations under conditions of low information inputs and non-acceptance under conditions of high inputs suggests the existence of variation in both levels of information and the optimal partition point. The discussion will now focus upon various factors which may be associated with such variation. at“ 26 The Information Environment It was suggested previously that individual information levels are subject to a variety of constraints, some of which can be typified as attributes of the social system. One im- portant constraint is the system of information inputs. In- dividuals in information-acquiring and decision—making roles do so within some socio-cultural and physical environment. The potency of such factors to affect information-acquiring behavior, decision-making activities and acceptance-rejection behaviors, demands their consideration. In considering such elements, two functions are performed. The first is that of accounting for observed individual differences which otherwise cannot be explained within the micro-level framework of the analysis. It is conceivable that members of various social systems may be sufficiently similar on some of the personal attributes being examined, such that these attributes may not account for variation in a criterion variable. Explanation of such variation may, in some instances, be possible by examining the manner in which social system characteristics relate to the observed differences. The second important function is that of a systematic ordering of social systems on those variables which have some 27 relationship to the criterion variable. Known differences exist between the socialsyetemiuuis (villages) in the present analysis. Villages were selected to represent the range of variation in institutional development and external accessibility. Villages that are more remote and that lack various modern facilities seldom experience the intensity of information inputs of less remote, more developed communities. Of particular concern to the present analysis is variation in the capacity of villages to receive and/or relay external information to members of the social system. Since virtually all modernizing innovations are developed externally to the adopters' social systems and are promoted by externally based diffusion agencies, the notion of information inputs as a social system or cultural variable would seem to merit attention. It is postulated that villages vary with reSpect to their capacity to be exposed to information. Furthermore, it is hypothesized that as this capacity increases, individual levels of information and adoption will increase. Two im- portant notions are implicit in this viewpoint. The first is that of some "objectively" defined in- formation environment. The second is a "subjective” in- formation environment. By objective is meant that body of information about some object or idea which could be known to the decision maker. By subjective is meant the body of 28 information actually known to the individual. For any number of reasons the subjective information environment will differ from the objective environment. One reason may be the nature of the individual's source orientationlo, a concept of concern to the present analysis. But, regardless of the variation that exists in the subjective environments, for any social system there may be said to exist some common potential population of items of information which places a limit upon that which individuals can know. Thus, it is essential to recognize that information acquisition and attitude formation/ change must occur within the confines of the objective in- formation environment. It is hypothesized that as the incidence of information inputs into some social system increases, the frequency of decisions of innovation acceptance will increase. However, there will still be instances of acceptance in social systems with relatively low levels of inputs. The question then becomes: ”What factors precipitate and explain the acceptance of innovations under conditions of low information inputs?" One possible factor is the complexity of the in- formation environment. Where inputs are relatively few the decision maker may be faced with correspondingly fewer behavioral alternatives. Under conditons such as these the complexity and ambiguity of the judgment situation is less than it might be for systems with a high level of both fre- 10Source orientation is the configurationCs) of sources to which respondents attend. This concept is discussed further in the present chapter and in Chapter III. 29 quency and diversity of inputs. Jacobson (1967, p. 10), ' characterized in- in discussing "communication cultures,‘ dustrialized, literate societies as experiencing "increased ambiguity of choices about behavior opportunities that accompanies increased complexity of society." The reverse of this point suggests that members of social systems with relatively few information inputs and, hence limited social— behavioral controls as a function of such inputs, experience less ambiguity about the alternatives at their disposal. If, coexistent with less ambiguity, there exists some level of awareness of more modern life styles and an accompanying desire to acquire the amenities of modernity, then one can expect to encounter instances of modernizing behavior change under conditions of relatively low levels of information in- put. In addition to multiple inputs themselves, the degree of uncertainty and complexity may be, in part, a function of the available alternatives. The population studied is characterized as confronted with one-modern-alternative situ- ations. These respondents seldom have to make decisions among several equally modern innovations. Thus, one would expect less perceived complexity and subsequent uncertainty. In desoribing the Igbo of Eastern Nigeria, Uchendu (1965) accords the values of "beneficial reciprocity" and "equal 30 opportunityftn‘achievement” high priority in the value structure. These values seem to manifest themselves in the concept of "getting up", a characteristic which appears to influence considerably the modern motives of many Eastern Nigerian communities. Thus, it is suggested that the per- vasiveness of such values in conjunction with their considerable importance will be reflected in attempts at both the village and individual level to acquire and enact more modern ways of life. Since the communities to be investigated in the present analysis are not without some contact with the ex- ternal, more modern world, one would expect to find instances of modernizing behavior even where the level of information inputs is comparatively low. When one attempts to define information inputs, the aforementioned definitional problems associated with the concept of information become salient. The definition of information used in the present analysis, as well as the nature of the data collected, preclude a direct measure of the amount or frequency of information which is available to a social system. Thus it becomes necessary to approach the measurement problem indirectly, i.e., one must utilize some related attribute whose variation can be said to parallel that of the unmeasured variable. For purposes of the present 31 analysis it is assumed that level of information inputs varies directly with the number of input and relay mechanisms which exist in a social system. By input and relay mechanisms is meant those institutions and roles which facilitate entry of information into a social system. These entities may act as carriers and/or relayers of information to individuals internal to the social system in question. In the context of the present analysis three types of information input mechanisms are recognized: (1) non- specific information input facilities; (2) specific infor- mation input facilities; and (3) leaders' relay capacity. Non—specific information input facilities are those in- stituions and roles by which innovation information may pass into the community but whose essential purpose is not 11 the diffusion of innovation information. Examples of this llMarkets include both daily and weekly market places. News vendor and agents are the area and local distributors of neWSpapers. Lorry and taxi facilities refer to transportation depots. Sons abroad organizations are groups of persons from a village now living somewhere other than the village and who contribute in some manner to the community's development. University students also constitute a means by which modern information is available to a village. Advertising vans are vehicles equipped to promote new ideas and manned usually by two or more persons. Commercial advertising is often con- ducted in this manner. The Ministry of Agriculture also has used such units. Media circulation refers to the number of families in the village receiving a neWSpaper. Permanent strangers are persons who live in the village but who were not born there. 32 type include markets, news vendors and agents, post offices, inter-village meetings, lorry and taxi facilities, sons abroad organizations, students at university, advertising vans, media circulation and the presence of ”permanent strangers." These mechanisms all possess the potential to carry infor— mation from external sources to the community. Specific information input facilities are the formal diffusion system such as the agricultural extension agency. The majority of the agricultural officers' messages are innovation-oriented. The strategies used to disseminate such information include the use of demonstrations, village meetings, tours of government farms, posters, films, news- letters and pamphlets. It is hypothesized that as the in- cidence of the extension agent's activity increases, individual information levels and adoption behavior increase. Village leaders constitute yet another type of input mechanisms. The capacity of leaders to act as relayers of information is a function of two characteristics. The first is village leaders' degree of Opinion leadership, in essence the extent to which they act as givers of opinions. The second characteristic is their degree of contact with ex- ternal sources of information. To be maximally effective as information input mechanisms, leaders must have both high 33 telling activity and high contact with external sources. As either of these decreases, the leaders effectiveness as a relayer of information decreases. For purposes of the present analysis leaders' diffusion activity is indexed as the degree of village farming opinion leadership.l2 External contact is based upon contact with the extension agent, exposure to radio and the print media and cosmopoliteness, indexed as external travel behavior. Thus, it is hypothesized that the more externally oriented and the more active the village leaders are as opinion leaders, the greater will be individual levels of information and adoption in such villages. The hypotheses presented in the foregoing discussion involve two units of analysis: (1) the individual respondent; and (2) the village. In both cases the unit of resEonse is the individual. Leaders' responses are aggregated to produce 13 village-level indices. 12Village farming opinion leadership is the extent to which a person is sought fOr or gives opinions about farming topics and issues. The opinion leader's relay capacity is seen here as the natire of his gatekeeping activity, or the manner in which he controls the flow of opinions at one step of the multi-step diffusion process. 3The non-specific input facilitts are not indexed as aggregates. The project interviewers obtained these data in the course of their study in the village and recorded such information on the village description schedule. So these data were originally obtained in an aggregated form. 34 In summary, one of the constraints thought to be operating on information levels is the nature of the infor- mation inputs into the social system. Three types of inputs have been discussed: (1) non-specific input facilities; (2) specific input facilities; (3) leaders' relay capacity. All three types possess the capacity to carry and/or relay information to members of a social system. The following hypotheses will be tested. H1 Individual information level varies directly with the village level of non-specific input facilities. Individual degree of innovativeness varies directly with the village level of non-Specific input facilities. Individual information level varies directly with the village level of specific input facilities. Individual degree of innovativeness varies directly with the village level of specific input facilities. Individual information level varies directly with the village level of leaders' relay capacity. Source Orientations The second major classification of constraints upon information levels and behavior is the configurations of 35 sources1 used by decision makers. Decisions, either by active consideration or by habit, to attend to particular means of acquiring information represent constraints upon that which a person could potentially know. The following discussion considers both recognizable types of individuals and the mass media as sources of information. The recognizable sources refer to several particular sources. The agricultural officer in his institutionalized role as an agent of diffusion and influence is one such source. Permanent strangers in the villages are another interpersonal source of information. Friends, neighbors and kin are also important interpersonal sources. The mass media include radio, newspapers, films, extension pamphlets, newsletters and books. In the innovation context, most past research on in- formation acquisition behavior sought to relate such behavior to innovative behavior. The intent has been to ascertain the extent to which the various media and personal sources had an impact on the innovation decisions of their audience. Little attention has been given to the manner in which ex- posure to various sources affects the information level of decision makers. Information level, for the most part, has been indexed solely as awareness. Even here, few relationships 1“While several of the sources referred to are media, the term "source" will be used throughout the discussion to refer to both media and sources. 36 15 are reported. A review of the Diffusion Documents Center yielded the data in Table l. The evidence in Table l is ambiguous. First, there is insufficient evidence to derivevwfljrsupported generaliza- tions. Second, the data express relationships between communi- cation behavior and being aware of, or knowing about, inno- vations. It is, therefore, tenuous to extrapolate these findings when discussing levels of information beyond the point of awareness. A third source of ambiguity arises from the Ink of empirically-defined attributes of the environment in which the foregoing relationships were found. One might legitimately ask why three of the nine relationships involving mass media exposure were not statistically significant. Not- withstanding measurement error, it may be the case that the content of the media was inappropriate or there was little variability in mass media exposure (perhaps due to: (1) low levels of literacy where the print media are concerned; or (2) the general unavailability of mass media). Social systems or "cultures" must necessarily be indexed on such attributes 15The Diffusion Documents Center, located in the Department of Communication, Michigan State University, con- tains 1,300 documents on the diffusion and adoption of innovations. 37 Table 1. A Summary of Past Research on Correlates of Knowledge of Innovations. Type of Relationship Found with Knowledge of Innovations variable Correlated with Knowledge of No Con— Innovation Positive Relationship Negative ditional* 1. Mass media ex- posure 5 3 — l 2. Interpersonal communication 7 - - - 3. Cosmopolite sources (mass media and interpersonal) l 2 — - 4. Interpersonal cos- mopolite sources (e.g., extension agents, public health agents, etc.) 12 4 - - 5. Commercial sources 1 — - _ *A conditional relationship is one that depends on the degree to which a third variable is present. 38 as levels of literacy, availability of the media, and relevance of content if differences such as these are to be meaningfully explained. Much source and channel research has related communi— cation behavior to innovative behavior. Most of this research has been conducted in North America and the other more developed areas of the world. Rogers (1966) reports only three studies among peasant villagers and six among urbanites (in develOping countries) in which the impact of the mass media were studied. Seldom has knowledge of innovations been indexed as a continuum. Neurath (1960), in studying the effects on individual knowledge levels of radio farm forum members, utilized an eighteen-item scale. Spector gt El. (1963) indexed "Specific knowledge" of various inno- vation campaigns studied, as the ability of the reSpondent to report any information contained in the campaigns (dealing with latrines, ovens, marmalade and vaccination). Thus, it would seem useful to discern how source orientations affect individual levels of information. In recent years, the role of communication in the process of modernization has been the subject of considerable theoretical attention and a growing amount of research. Lerner (1958) classified systems as oral-traditional, transitional and modern-mass. He accorded the mass media a crucial role in the creation of "psychic mobility" or empathy, 39 the capacity to put oneself in modern roles. Pye (1961) asserted that the transformation of societies from traditional to transitional to modern is, in part, a function of the en- croachment of modern communication systems upon traditional cultures. Both Lerner and Pye based their positions, to a large extent, on the impact of the mass media. Rogers and Bebermeyer (1966) suggested that communication strategies combining the efficacies of the mass media and the potency of interpersonalcnnmunication be utilized to effect a more rapid rate of social and economic development. It is this focus which the present analysis takes. Since the unit of analysis is the individual and since it has been shown that he relies on a variety of sources, both interpersonal and mass media, it is necessary to explore the conjunctive re- lationships between the various kinds of sources as they affect information levels. Research on communication behavior in the innovation decision process illustrates the complementary functions of the mass media and interpersonal communication. From North American research, two generalizations emerge (Rogers, 1962). The first is that mass media sources of information are more important than interpersonal sources in the earlier stages of the innovation decision process and interpersonal 40 sources are more important than mass media sources in the latter stages. By "more important” is meant a greater fre- quency of reported use. Studies by COpp and others (1958), Rogers and Beal (1958) and Rogers and Pitzer (1960) support this generalization. The second generalization is that cosmopolite sources (mass media and personal sources external to the receivers social system) are more important than localite sources in the earlier stages of the innovation decision process, and localite sources are more important than cosmopolite sources in the latter stages. Rogers (1962, p. 103) demonstrated this to be the case. Campbell (1960) and Leuthold (1960) also found support for the generalization. These findings serve to illustrate individuals' reliance upon multiple sources, or a configuration of sources. Additional evidence of individual source usage has been noted. First knowers of an innovation usually attribute their early knowing to the mass media or other cosmopolite sources. Later knowers tend to become aware through interpersonal localite contacts. Ryan and Gross (1943) found such to be the case among Iowa farmers. Deutschmann and Fals Borda (1962) noted a similar pattern in Colombia. It has also been found that earlier adopters tend to use more information sources (Copp, 1956; Emery and Oeser, 1958; and Fliegel, 1956). 41 While portraying important variations in channel orientation, thefbregoing findings are largely derived from research in more developed countries, notably the United States. The cross-national validity of these findings is largely untested. Initial research in developing countries suggests certain modifications. Deutschmann and Fals Borda (1962), Myren (1962) and Rogers and Meynen (1965) reported that the mass media are seldom mentioned as sources of in- formation in the innovation decision process of peasants. It should be noted that such a finding does not necessarily suggest the absence of the mass media. It simply may be that media content is not innovation-oriented. Or it may be that low levels of literacy and verbal facility inhibit mass media consumption. That the mass media do penetrate peasant villages and reach a sizeable audience has also been demon- strated by Rogers (1966), Frey (1964) and Deutschmann (1963). Absence of the media, lack of innovation-relevant in- :brmation in the media or an inability to consume media con- tent, forces information seekers to utilize the various in- terpersonal sources at their diaposal. Rogers and Meynen (1965) found that 44 percent of their reSpondents mentioned interpersonal-cosmopolite sources at the awareness stage and 15 percent mentioned such sources at the evaluation (persuasion) stage. These authors also found that 56 percent of the respondents 42 mentioned interpersonal-localite sources at the awareness stage and 85 percent mentioned these same source types at the evaluation stage. Thus, there is some validity for the notion that cosmopolite sources are more important during the earlier stages of the innovation decision process and localite sources are more important in the later stages of the process than they are in the earlier stages. However, overall, localite sources were mentioned more often at all stages than cosmopolite sources. We also note individual variations within stages. Rogers and Meynen (1965) reported the following per- centages of mention as first sources of information: (1) extension agents, 7.5 (percent); (2) farm store personnel, 36.6; (3) neighbors and friends, 40.8; (4) haciendadosl6, 10,03 (5) family, 3.3; (6) self, 1.8. The findings of Deutschmann and Fals Borda (1962) parallel those of Rogers and Meynen (1965) in Colombia. For the six innovations studied, only 2 percent of the respondents reported mass media sources as the source of first information about an innovation. The relative unimportance of the mass media in peasantcnmmunities is reflrted also in the most im- portant influence among adopters. Percentages of respondents l6Haciendados are owners of large land holdings or plantations. 43 nominating the mass media as the most important influence range from 10 percent to 1 percent from innovation to inno- vation. By far the most important influence, is the category, "What I saw."17 The foregoing evidence indicates dependence on a variety of kinds of sources. With a view of parsimony, the present analysis will attempt to empirically define clusters of kinds of sources. Data on exposure to a variety of sources has been obtained. Included in this list are general newspaper, radio and film exposure, attendance at agricultural demon- strations and lectures, talking with agricultural officers, viewing agricultural films, listening to special agricultural radio broadcasts and exposure to extension publications. Data on exposure to these various sources will be factor- analyzed to ascertain the nature of clusters (if any) of sources. From this point, individuals will be scored on each emerging factor to provide an analytic basis for determining the relationships between source orientations on the one hand, and knowledge levels and innovative behavior on the other. Since the various sources to be subjected to factor analysis 1 7See Deutschmann and Fals Borda (1962, p. 57). 44 are all of the cosmopolite or mass media type, it is hy- pothesized that degree of exposure to source types will be positively rented to information levels and innovativeness. Source utilization was indexed, also, by time. Data on sources attended to, at the awareness and conviction stages were obtained. Sources mentioned, varied on two dimensions: (1) mass media - interpersonal; and (2) cosmopolite - localite types. The mass media-interpersonal source distinction rests on the concept of interpersonal contact. Symbol manipulation and literacy skills may be important determinants of the ability of the individual to successfully utilize the mass media. Thus, one would anticipate that the more modern the individual, the greater will be‘hES-capacity to consume the available mass media and subsequently exhibit higher levels of information and adoption. The cosmopolite-localite distinction is based upon the contiguity of source and receiver. Contiguity may be viewed as geographical, occupational, social, cultural, etc. In fact, measures of the cosmopolite-localite dimension usually include several or all of these distinctions. In the present analysis the mass media and professional technical information sources (the agricultural officer) are considered cosmopolite. They constitute messages and persons who do not originate from the immediate social system (village) of the decision 1.5 maker. It is hypothesized that information levels and innovativeness vary directly with the degree of exposure to cosmopolite sources. Localite sources include friends, family, neighbors and kin. In summary, past research indicates decision-maker dependence upon multiple channels. The present analysis seeks to determine the configurations of sources to which decision makers attend. Degree of exposure to the factor analytically derived dimensions is hypothesized to be positively correlated with information levels and inno- vativeness. It is hypothesized that H6: Individual degree of innovativeness varies directly with the village level of leaders' relay capacity. H7: Individual information level varies directly with degree of exposure to source orientation dimensions. H8: Individual innovativeness varies directly with degree of exposure to source orientation dimensions. In addition, data on source exposure at awareness and conviction are analyzed. It is hypothesized that H9: Individual information level varies directly with degree of mass media exposure at awareness. H10: Individual innovativeness varies directly with degree of mass media exposure at awareness. 1+6 H11: Individual information level varies directly with the degree of cosmopolite source ex- posure at awareness. H12: Individual information level varies directly with the degree of cosmopolite source exposure at conviction. H13; Individual innovativeness varies directly with the degree of cosmopolite source exposure at awareness. H14: Individual innovativeness varies directly with the degree of cosmopolite source exposure at conviction. Attitudes Toward Innovations It has been suggested that attitudes toward innovations constitute constraints upon both information seeking and innovative behaviors. While the evidence is equivocal, there is some indication that voluntary exposure to information is such that content and the individual's attitudes are con- sistent (Sears and Freedman, 1967). In diffusion research little evidence on the relationship between information levels and attitudes toward the information object is available. Favorability toward the information object (the innovation) is generally positively (though not always strongly) related to adoption. The findings relating attitudes and information level in the area of public opinion are varied and ambiguous. Dervin (1967), in a review of research on attitudes and information levels, reports that about one-half of the 47 relationships were positive while the remainder showed no relationship. One of the limitations upon the comparison and explanation of many findings is the lack of measured situational factors. Dervin reported most of the positive correlations were found where: (1) neutral information was available; (2) the information was consonant with attitudes; and (3) low involvement issues were used. It is doubtful that the acceptance of new technology constitutes a low involvement issue for the decision maker. For those whom the acquisition of modern techniques and amenities is important, high involvement and favorable attitudes toward innovations should generally prevail. For those who rely on traditional behavior patterns and who lack the facility to c0pe with the modernizing information environ- ment, attitudes toward innovations are more likely to be neutral or negative. Thus, it is hypothesized that favorability toward innovations will be positively related to level of information about innovations. In a similar fashion, attitudes are viewed as a form of constraint upon behavior patterns. Knowledge of courses of action is not usually sufficient to predict behavior. An individual may possess similar amounts of information about a set of behavioral opportunities but differ on the evaluative 48 dimension. This statement requires qualification so as not to conflict with the foregoing attitude-information hypothesis. In more modern societies where information inputs are relatively high and the society is often referred to as being "media saturated", variation in information levels between indivi- duals may be relatively low compared with those in traditional cultures. Where such conditions exist, information levels may fail to discriminate behavioral categories whereas attitudinal states may have greater predictive power. In one-modern-alternative situations, such as the innovation- decision situation in Eastern Nigeria, where change and innovation are generally accorded a positive value, one might expect to find attitudes toward innovations directly related to both information levels and innovative behavior. In summary, attitudes toward innovations are viewed as constraints upon both information levels and innovative be- havior. It is hypothesized that attitudes toward innovations vary directly with both the levels of information about innovations and adOption of innovations. H15: Individual information level varies directly with degree of favorability toward innovations. H16: Individual innovativeness varies directly with degree of favorability toward innovations. 49 Innovativeness Innovativeness is defined as "the degree to which an individual is relatively earlier in adopting new ideas than other members of his social system" (Rogers, 1962, p. 20). Throughout the present analysis, innovativeness and infor- mation level are utilized as the two dependent variables. Information level is discussed and analyzed from the viewpoint of three classes of constraints. Innovativeness is sub- sequently viewed as a function of individual levels of in- formation as well as the various forms of constraints. Three innovations are studied: (1) fertilizer; (2) an insecticide (aldrin); (3) a new variety of maize (NS-l). Time of adoption data were obtained and standardized scores will be computed for each innovation as well as an average standard score of innovativeness across the three practices. It was argued that in one-modern-alternative situations, there is likely to be a direct relationship between individual levels of information and individual innovativeness. H17: Individual innovativeness varies directly with the individual's level of information about innovations. In addition to testing the innovativeness-information level hypothesis, the present analysis will seek to determine how variation in information levels is associated conjunctively with variation in innovativeness, levels of information inputs, 5O attitudes toward innovations and source orientations. Thus, the analysis becomes a multivariate problem. Information Constraints, Information Levels and Behavior: A Multivariate Analysis Throughout the discussion there has been both an express and implied sense of interrelatedness among the various classes of variables in the present analysis. These variables include: (1) information inputs; (2) source orien— tations; (3) attitudes toward innovations; (4) information levels; and (5) innovativeness. The diffusion and adOption processes are not sets of bivariate relationships; they are multivariate systems. Lerner (1958) characterized modernity as an "interactive behavioral system", a system of variables which variously affected each other to produce increasing states of modernity. The variables in the present analysis represent but a few of the variables which have been defined under the rubric of modernization. The foregoing discussion has attempted to demonstrate that they can be thought of as a related set of factors which affect levels of modernity. To admit only to a complex of variables is to stop short of the problem. The logical next step is to attempt to discern how the variables are interconnected and to what extent 51 they explain variation in the selected criterion variables. More specifically, there is a need to discern how a given variable operates in the presence of other variables; i.e., what values of one variable operate with what values of other variables to maximize explained variance. For instance, in the present analysis it would be interesting to know how the various indices of information inputs operate to affect information levels and what values of the input codes, when grouped, with the other variables account for maximum variance in information levels. At the very outset of the present discussion mention was made of the need to classify social systems on those variables which may affect variation in individual attributes. The present analysis will explore this course of action. In addition, the conjunctive effects of social system attributes and individual characteristics constitute a multi-level analysis of the interacting system of variables. Of the various multivariate techniques, sequential interaction analysis (Sonquist and Morgan, 1964) provides an empirical, descriptive, configurational analysis of the inter- relationships of a set of predictor variables with a dependent variable. The technique compartmentalizes those ranges of the different predictors to produce profiles or configurations which maximally explain variation in the dependent variable. 52 In this way it may be possible to ascertain the relative effects of the different predictors and thus provide an analytic base for describing typologies of social system and individual attributes. Two dependent variables will be utilized: (1) individual information level; and (2) individual innovativeness. The first analysis will be concerned with an explanation of the variation in individual information levels. Predictor variables include: (1) non-specific input, facilities; (2) specific input, facilities; (3) leaders' relay capacity; (4) source orientations; and (5) attitudes toward innovations.18 The second analysis will utilize innovativeness as a de- pendent variable. Predictors include all of the foregoing variables, plus information levels. Of particular interest here is the manner in which information levels become partitioned to explain innovativeness. In summary, the process of modernization is viewed as an interacting system of variables. Figure 2 is a schematic representation of the process. The analysis plan will attempt to determine the conjunctive manner in which the various predictor variables, taken from both the level of the village social system and the individual respondent, 18Innovativeness is not used as a predictor variable since it is considered a consequent rather than an antecedent of information levels. 53 explain variation in both information level and innovativeness. This latter segment of the overall analysis plan is largely exploratory. Few attempts have been made within the context of diffusion-adoption research to simultaneously discern the effects of various levels of constraints upon information levels and innovative behavior. Perhaps the present analysis will suggest additional ways in which to describe and analyze the processes of diffusion and adoption. soapmmflcamooz one soapmsaomcH no smflomamm .H maswflm _ . . mQOHpm>occH sofipofl>ooo . . pamZOp .nmooa50m _ . mooepaee< .m Hesse _ oeaaoa _ _ coapmsaom _ nosmoo .2. _ IQfl HMPOB .N. . _ _ QOHPOH> _ _ soap _ usoo _ . nQOpo .ummoaSOm _ _ mo mmmc _ tacos _ _ IHHHGm a mmwz .m. _ u. o>HpmH omfiws _ _ _ hpfiomgmo er nmm .m Humz on _ mmms . . beams combos undo . leases _ _ uaoecfi .mampmoq .: pmoo< Canvas ADV .umooMSOm . . eoz newness _ meaaod _ _ menace coarse -edoo< .H -nma Amy _ -oSwoo .m. _ -noeca oaaeooom .m mNflmE Hlmz o COHPM> _ _ _ pmdp Canada 9 noncfi _ mmms . _ mpSQQH soap amNHHHpamm m oamaoomm _ sonnet _ .nmsaomcfi oflwfioomm doflpm>ossfl do .ammoa50m _ .ucos pnoppflsampsH .m onHoon mao>oa _ magma ..pomomxo we _ pamzoo coapws _ mmwz .H. oco £0H£3 _ magnum coapmsaog woodpfipp< .H uaoncH .H. .pOp mova50m _ Ina oamaoomm _ . mo apofiam>..a. icon pcwpmcoo .H aoa>wnmm woodpfipp< mam>mq sopmmm Hmfloom mcHNHanooz dowmeaomcH CH weapo< maw5©H>deH Smpmam Hmfloom mopsranpp< assessaoce-onecH COprOflMHmmeo manmahm> CHAPTER III METHODOLOGY Phase I Village and Respondent Selection Procedures Sampling for Phase I was conducted in two stages. A sample of 71 Igbo and Ibibio villages was selected from the Eastern Region of Nigeria. Following village selection, 13 to 16 leaders were selected for personal interviewing from each village. Village Selection Due to the inadequacy of census figures and the fact that no complete lists of Eastern Nigerian villages existed, it was impossible to adhere strictly to a random sample of villages. Counties were selected as a first sampling step since they are geographically defined. Thirty counties were randomly selected. From each county two (or morelg) villages were selected. Several criteria were used for selecting villages. The change agent must have worked in the village for at least nine months. One "success" viflage and one 19In some instances a county and certain villages had to be replaced. In the final analysis 71 villages were included. 55 56 "failure"20 village were designated by the extension agent and his supervisor. Where there was disagreement between the two, the agent's nomination was accepted. ’Villages were stratified on the basis of cultures and sub-cultures (Igbo and Ibibio) and population estimates from 1953 were used to determine proportions. Of the 71 villages, 52 were Igbo- Speaking and 19 were Ibibio—speaking. Of the 71 villages, 34 were success villages and 37 were failure villages. It should be noted that agents and supervisors were generally reluctant to designate failure villages. Thus, the villages listed as failure villages may not represent the population of failure villages but rather those only somewhat less successful than the success villages. Another problem inherent in the selection process is that counties were not ranked on relative success. Therefore, a success village in one county may be a failure village in another county. ReSpondent Selection Village leaders and change agents were interviewed in Phase I. Leaders were selected on the basis of sociometric nominations by informants (some of whom were leaders them- selves). Six formal roles were defined: (1) civic; 20"Success" and "failure" refer to the degree to which the villages had adopted the innovations promoted by the Eastern Nigerian Ministry of Agriculture. 57 (2) religious-traditional; (3) religious—modern; (4) adminis- uative-traditional; (5) administrative-modern; (6) educational. Two informal leadership roles were defined: (1) village affairs opinion leadership; (2) village farming opinion leadership. In addition to leaders, innovators (usually three per village) were selected to be interviewed. Thus, there were 13 to 16 leaders and innovators interviewed in each village. The interviewer first spoke with the village school headmaster who indicated potential leaders. Chiefs were then asked to do likewise. Those nominated were then asked to name persons for each of the roles. This process was repeated until ten or so informanats had been contacted. For each role the leader selected to be interviewed was the one scoring the highest number of sociometric mentions. Phase I Interview Schedules Phase I data were obtained from: (1) leaders and innovators; (2) change agents; (3) interviewer's observation of the villages. The leader/innovator schedule was translated into Igbo and Ibibio. The interviewing staff and study dir- ectors reviewed and revised the wording of items. Two persons, both unfamiliar with the study, translated the two schedules 58 back into English. Ambiguities were resolved. Where inter- viewing staff could not agree, multiple wording was used in order to best approximate the meaning from village to village. A two—week pretest was conducted. Based on difficulties that were encountered, certain questions were reworded or deleted. Where the data-analysis revealed no variation, items were reworded or discarded. Phase I Interviewer Selection and Training Interviewers were selected on the basis of their education, age, cultural origin and rural experience. All interviewers were school teachers (or similar professions) with experience in rural Eastern Nigeria who possessed at least the Grade II Teacher Training Certificate. They were 25 years or older and Spoke the dialect of the villages in which they worked. Training consisted of a one—week orientation at head- quarters (Enugu), a two-week pretest in the field and one- week at headquarters revising the interview schedule and discussing field problems. Phase I Data Collection The lea‘chr/innovator data were collected by personally interviewing the designated village leaders and innovators. 59 Interviewers averaged three completed interviews per day; each interview took from one hour to one and one-half hours to complete. Change agents were also interviewed (in English) by the village interviewers. During the interviewer's stay in a village he also completed the village description schedule. Phase II Village and Respondent Selection Procedures Sampling for Phase II was conducted in two stages: (1) village selection; and (2) respondent selection. Village Selection The population of villages from which the Phase II sample was drawn, consisted of the 34 "success" villages from Phase I. Phase I "failure" villages were excluded since one of the primary objectives of Phase II was the determination of correlates of high innovativeness and other modern attri- butes. Each of the 34 villages was then scored on two criteria: (1) "access to the outside world," and (2) "institutional development.” Degree of external access was defined as the presence or absence of modern amenities and proximity to tranSport routes and governmental officer (see Appendix E for a list of indicants of external access). Village amenities (items 1-9) were scored zero or one (absent or present). Distances were scores as: 2(lower quartile-closest); 6O 1 (intermediate quartiles; 0 (highest quartile - farthest). Saares were Simmed across all items to obtain a village score. Scores ranged from 2 to 13. The scores were ordered and villages categorized as high, intermediate or low on external access (upper quartile, intermediate quartiles and lower quartile, reSpectively). Institutional development was indexed in much the same manner. Twenty-four modern amenities were taken as indicants of institutional development. Villages were scored zero or one on each item based upon the presence or absence of the amenity (see Appendix E for a list of amenities). Village scores were computed and the total scores ordered. The dis— tribution was partitioned in the same manner as was the dis- tribution of external access, i.e., upper quartile - high development, intermediate quartiles, lower quartile - low development. Subsequent to the foregoing analyses, village ranks on both criteria were summed to produce an external access/in- stitutional develOpment continuum. The actual selection of pretest and main study villages was based upon variation in access and institutional develop- ment. However, the formal selection criteria were necessarily compromised by geographical-cultural capacities of the inter- viewers, physical accessibility of the villages during in- clement weather and the extent to which interviewers could 61 be supervised. Nine villages were selected for the pretest and eighteen for the main Phase II study. Respondent Selection Three samples of farmers were selected. In each case the respondent was a male, head of household, 20 years of age or older, farming any amount of land, under any kind of tenure system. The three samples were: (1) general sample in which respondents were randomly selected from census data; (2) innovators, of which nine were selected at random; and (3) group leaders of development groups, from which six, designated by sociometric choice, were selected (see Appendix F for village numbers, names and number of respondents). Inno- vators were classified in three types. "Agricultural program innovators" were those individuals who actively participated in government development schemes. "Agricultural practice innovators" were those employing new practices or acquiring new equipment, supplies, etc. "Home innovators" were those who registered childbirths, whose wives used maternity centers, took malaria pills, etc. In all three cases designation of such individuals was made by the local extension agent. Group leaders were classified as "introductory leaders," "sanctioning leaders" and "implementing leaders." Sociometric 0\ l0 nominations provided the basis of selection of group leaders. Two of each type of leader were selected. A total of 1,347 reapondents were interviewed. Phase 11 Interview Schedule Based upon two pilot studies, one in Western Nigeria and the other in the Midwest region, selected modernization concepts were considered for inclusicn in the Phase II study. Expertise from governmental sources and other social scientists were utilized to construct the initial interview schedule. A fivevweek pretest of the interview schedule was con— ducted in the hire Phase II pretest villages. Techniques for mapping villages and enumerating the pOpulations were developed. Items were selected for the final Phase II schedule on the base of their discrimination power, their case of understanding and the ease with which they could be administered by interviewers. The interview schedule consisted of several types of questions: (1) structured response items; (2) unstructured or Open-ended response items; and (3) interviewer ratings. 63 Phase II Interviewer Selection and Training Nine senior interviewers were selected from the group of Phase I interviewers. These nine had at least the Grade II teacher training certificate, were twenty-five years old or more, had rural experience and were culturally and linguistically suitable for the Phase II study areas. In addition to the nine senior interviewers, nine junior interviewers were selected. They had somewhat less education and were younger. Thus each interviewing team consisted of one senior and one junior interviewer. Techniques of rapport-building, field behavior and familiarity with Ministry of Agriculture programs were one facet of the interviewers' training. Two weeks were Spent on training in the methods of respondent selection and rapport- ‘building. One week was Spent on the problems of translation and administration of the schedule. Three weeks were Spent in the field where interviewers received supervision. One week was again Spent on additional translation and interview schedule administration problems. Phase II Data-Collection Personal interviews were conducted with the selected respondents. Interviewer observation of the village and respondents' property was the other form of data-collection. 64 Administration of the interview schedule averaged one and one-half hours. On the average, two or three completed inter- views per day per interviewer were obtained. Interviewers spent their first two weeks in the village creating village maps, locating potential rSSpondents (general sample, inno- vators and group leaders) and building rapport. Four weeks were Spent in each village obtaining the individual respondent data. Operationalization of variables The discussion now turns to consideration of the manner in which the dependent and independent variables were empirically indexed. The various techniques of scoring, recoding, weighting and scale analysis are documented. Appendix B contains the items used in the present analysis. In outlining the operations to be performed, the dependent variables will be discussed :flrst, followed by a restatement of the hypotheses and a descrip- tion of the operational procedures for the independent variables. Dependent variables Individual Information Level Individual information level is defined as the amount of knowledge a respondent has about three innovations. 65 Individual information levels: fertilizer; aldrin dust; NS—l maize: Knowledge about three innovations was elicited in some detail. Fertilizer, aldrin dust and NS-l maize, three generally applicable and important innovations were the focus, in an attempt to determine the reSpondents' degree of correct information about the innovations. Probing techniques21 were used by the interviewers to determine what was known about each of the three innovations. Each item of correct infor— mation given by the respondent was coded as "1". In addition, a specific question was asked of each respondent about each of the three innovations. The three questions were as follows: 1. IS fertilizer a plant food or a medicine? 2. How is aldrin dist applied to crops? 3. What do the agricultural people22 do for farmers who plant agricultural maize? Responses were originally coded incorrect ("0"), partially 21By probing is meant that the interviewer asked the respondent after each response, "IS there anything more you can tell me about the innovation?", or "What else do you know about the innovation?". 22"Agricultural people" refers to personnel from the Eastern Nigeria Ministry of Agriculture. 66 correct ("1") and correct ("2"). For purposes of the present analysis only the category "correct" was considered to repre- sent additional information about the innovation. Thus, these three items were recoded. The zeros and ones were scored as "0" and the twos as "1". An information level score was computed for each respondent on each innovation. Those unaware of an innovation were Scored as zero. Those aware but possessing no correct information were scored as one. Those possessing correct in- formation were scored as the number of correct items known (plus the awareness score). To this score was added the :espondent's score for each of the three Specific questions. The possible ranges of scores are as follows; fertilizer, 0—9; aldrin dust, 0-8; NS-l maize, 0-10. A composite information score across the three innovations (fertilizer, aldrin dust, NS-l maize) wasiflxnlcomputed for each respondent by summing the information level scores on each of the three innovations. Scores may range from O to 27. Innovativeness and Adoption Innovativeness is "the degree to which an individual is relatively earlier in adopting new ideas than the other members of his social system" (Rogers, 1962, p. 20). 67 An average innovativeness score is computed for each respondent. Percentile scores are assigned to individuals for each practice, based on the distributions of adoption dates (see Table 2). Non-adopters are assigned the percentile score of the midpoint of the remainder of the adoption curve in cases of less than 100 percent adoption. Since percentile scores are not equal-interval scores, these data are trans- formed to z-scores. In order to avoid negative z—scores, the smallest z—score is added to all other z—scores. Respondents' average 2 score are then computed. Table 2 shows the dis- tributions of percentile scores and standard scores for the three innovations. Hypotheses and Independent variable Operationalizations Three classes of independent variables are utilized in the present analysis: (1) information inputs; (2) source orientations; and (3) attitudes toward innovations. Information Inputs Three categories of information inputs are considered to have direct relationships with levels of information and innovativeness. The three categories are: (l) non—Specific information inputs; (2) Specific information inputs; (3) leaders' relay capacity. The various hypotheses will now be restated, with a discussion of the operations used to define each of the independent variables. .mafipsmoamm Spam esp CH cmaoom mam dzozm amok pmaflm Op QOHMQ mQHpQOGS mQOmhmmH oo.o H:.o- :m oo.o mm.o- mm oo.o Hm.o- Hm mntnooem :oz ww.o as.o we mo.a se.o ea mm.o nm.o no mean wa.a ss.o we ma.a mn.o Hm an.o om.o mm meme sm.a ma.a an em.a oo.o an mm.o sm.o me some mm.a m:.e mm a:.a mH.H an ma.a Hw.o ms meme m:.m mo.m mm mm.H H:.H mm mm.a mo.H mm mama ss.m mm.m mm mm.a mo.a mm ms.a ma.a mm Heme . ©H.m ww.a so mm.H mm.H om coma @ noun men we no; em; a an? Hm.m mm.m Ham ms.a ms.a mm mmma am.a mm.a so smma mm.a mm.a mm mmma mo.m ms.a mm mmma mH.m mm.a em smma mm.m mo.m mm mmma zm.m mm.m Ham mmma monoom n mmaoom mmaoom mmaoomnn mmaoom moaoom mmaoomnn mmaoom moaoom,. coapmoo¢ um um mHHpcmo copmSn um maapdmo ompmSm In oaapdoo go cmpmsh unmm nc< uamm no< namm awmw -e< one»: H-mz page canoae nmnaaasnme .mnoapw>oscH mange pom menoom mmocm>fipw>ocsH mo GOHpm>HamQ .m mapwe 69 Theoretical Hypothesis 1: Source Orientations Individual information levels vary directly with the village level of non— specific information inputs. Prior to the statement of the relevant empirical hy- potheses, it is necessary to state the elements which comprise the concept of non-Specific information inputs, and Secondly, determine what scalar properties these items may have. The category of non-Specific information inputs includes the presence (or absence) of the following institutions, facilities, and organizations in the village: 1. Daily market . Weekly market Weekly market in area Post office . News agent 2 3 4 5. Post office in area 6 7 News agent in area 8 . News vendor 9. Lorry (truck) park 10. 11. 12. 13. 14. 15. l6. l7. Lorry park in area Permanent strangers Lorry stop once a week Taxis stop once a week Advertising van stops once a week Sons abroad organizations Number of families receiving newspaper Village has a student at a university . Meetings of leaders with other villages Each of these facilities is considered to be a potential in- formation carrier or relay mechanism from the external society to the village. 70 The data for 71 VillaS8823 were subjected to factor analysis to ascertain the structure of relationships between the 18 items. The loadings on the first factor of the principle axis solution were not sufficiently compelling to warrant treatment of all items as a unidimensional variable, hence, a varimax rotation was performed to more clearly dis- cern the factor structure. The two factor solution yielded fairly readily interpretable factors. Table 3 contains the factor loadings of each item on the two factors. The two factors each explain 15 percent of the total variance. From Table 3, items no. 1, 4, 6, 9, ll, 16 have loadings on the first factor ranging from -44 to .71. Loadings on the second factor range from .01 to .40. These six items appear to reflect the degree of presence in the village of the various institutions, facilities, etc. In some sense, these items reflect a dimension of permanence or continuous existence in the village of such input facilities. By this is meant that the presence of these facilities is constant rather than present sometimes and not others. This factor and the selected items are therefore referred to here- after as "constant information inputs." 23Though only 18 villages of the 71 were utilized in Phase II, the factor analyses were conducted on all 71 villages to provide a more stable factor structure. Even when N = 71, the use of factor analysis is somewhat questionable. For com- parative purposes, the data for only the 18 Phase II villages were factor analyzed and the results are presented in Appendix C. Though differences between the factor structures are evident, Similarities are also present. 71 The second factor would appear to have but three items loading sufficiently highly on the factor. They are items no. l2, l3, and 14: lorry stops at least once a week, taxis stop at least once a week and advertising vans stop at least once a week. The respective loadings of these items on Factor II are; .76; .77; .71. Their respective loadings on Factor I are: -.O5; .05; .15. These items might be said to reflect a discontinuous or intermittent input pattern. Thus this factor is hereafter referred to as, "intermittent information inputs." Empirical Hypothesis 1-1: Individual information level about three innovations varies directly with the village level of constant information inputs. Each of the six items of the input variable is scored as zero or one; zero if absent and one if present. Of the 71 villages, 32 were reported as having no families receiving newspapers. The range of families in the remaining 33 villages getting newspapers is from 1 to 99. Thus, villages with no families receiving a newspaper were coded as zero and all other villages scored as one. A total score for constant information input was computed by summing the codes across the six items. Empirical Hypothesis 1-2: Individual information level about three innovations varies directly with the village level of intermittent in- formation inputs, under con- ditions of low constant in- formation inputs. 72 Table 3. Rotated Factor Loadings for 18 Non-specific In- formation Input' Items on 71 Villages. Item Factor I Factor II 1 Daily market .g;? .10 .27 2 Weekly market .03 .27 .08 3 Weekly market in area -.23 .36 .18 4 Post office .71 .08 .51 5 Post office in area -.40 .22 .21 6 News agent .64 .26 .47 7 News agent in area .18 .05 .03 8 News vendor .41 .40 .33 9 Lorry park .51 .40 .42 10 Lorry park in area .23 .28 .13 11 Permanent strangers in village 49 .01 .24 12 Lorry stopsonce a week -.05 .76. .58 13 Taxis stop once a week .06 .77 .60 14 Advertising vans stop once a week 15 .71 .53 15 Sons abroad organization in village .17 .25 .09 16 Families getting newspaper .44 .13 .21 17 Village has university students .47 .39 .37 18 Village leaders meet with other villages -.36 .24 .19 Proportion of variance Accounted for .15 .15 Ii Underlined factor loadings represent those items used to represent that factor in the resultant indices. 73 It is hypothesized that where the level of constant information inputs is high the relative effects of inter- mittent inputs will be considerably less. Thus, under con- ditions of high constant inputs there Should be little difference in information levels between villages with low and high intermittent inputs. However, where constant inputs are low the effects of intermittent inputs on information levels should be more pronounced. Figure 2 illustrates the hypothesis. Intermittent inputs consist of lorry, taxi and advertising vans stopping in the village at least once a week. The three items are scored as zero (absence) and one (presence), hence, village scores could range from zero to three. The sample of 18 villages are partitioned on both constant and intermittent inputs. The distribution of villages on intermittent inputs precludasa median Split. Of the 18 villages, 12 have an intermittent input score of three, two have a score of two anddfour have a Score bf one(See Table 4).' Thus the Sample will befiSplit 1236» A skewed distribution is also noted for donatantainput?facilities (See Tablec5)3 Thus,»avscore of two animate is indebted-as ‘~3'high£0an‘diMillage/s: w-ithoscores efr'cne and zerOKadeflpw>oscH Hum .m.m s:m.a s as ommomxm moopSOm pcopomwflo mo sonadz + Hm>mH cowmeaomcH Hub .m.m quna a zpfiommmo zmaos smowmq + mmmcm>pr>occH Hum .m.m nqmna a hmam$pmwmmmw + Hm>mH coHmeaoch Hum .m.m ssm.a s mmfipaafiomm pSQQH oflmfiomam + mmoco>flpm>occH an: .m.m s:m.a s mmflpfiafiomc uzmcfi oflgfiommm + Hm>ma coflmesomcH Hum .m.m sqm.fi pm®p-p mmapflflsome psmcfl unoppflegoch + mmocm>fipm>oscH mum .m.m s:m.a s moHpHHaome psmcfi pcwpmsoo + mmmcm>fipw>oscH Hum .m.m s:m.a pm®p-p mmflpafiflomo pug ucfi pamppfissmpsH + Hm>ma coapwanmcH NIH .m.m szna a wanna pswpmcoo + Hm>oa cowp58aoecH HIH .m.m z oapmfipmpm manmflsw> QHQmGOprHmm manwflpw> mommnpomhm pcmocmmmocH omNHmmSpoqmm pcmocmmmm mommams< Hmofipmapmpm cam mmfismcoapmamm oomflmmsuoghm .5 manna 91 bzmaa p Hm>mH COHmesomcH + wmoco>flpw>occH HINH .m.m Ndmna A mcoapm>occfl osmBOp hpflaflnwpo>mm + mmoco>flpw>occH Huma .m.m bzmna p mcoflpm>occfi osmBOp zpfiafinmao>mm + Hm>oa coflpwEsomcH HumH .m.m Ham 9 coapofi>coo ”chamomxo mosSOm opfiaomosmoo + mmoco>flpm>occH anza .m.m mmmwa s mmmsmsmzw "madmOQXo mosSOm mpwaogoEmoo + mmmco>flpm>occH HIMH .m.m Haw p coapofl>coo "madmog uxm opfiaomosmoo + Ho>oa coamesomcH Huma .m.m wmmqa m mmmcosmzm ”madmogxm mosSOm opfiaomoEmoo + Hm>oa cospmasomcH Huaa .m.m wmmqa a mmocmsmzm "mszmom uxm aflome mmmz + mmocm>flpm>oscH HIOH .m.m mmmaa s mmoCmamsw “whomog uxm wanes mmmz + Ho>oa :oflpwapomCH Hum .m.m z ospmfipwpm mfinmfism> gflgmcoapmamm manmflsm> msmmgpogsm pcmocmmmocH bomfimmgpomhm pcmbcogmm omzcfipcoouuw canoe 92 variables on the dependent variable(s). One may wish to know how the variables in question can be both internally partitioned and grouped such that the variance explained in the dependent variables is maximized. In other words, one wishes to know how the sample can be reconstructed as a series of subconfigurations, each of which is internally homogeneous and maximally different from the other subsets with respect to the dependent variable. The present analysis seeks to determine such relationships between three levels or classes of variables. Attributes of the social system, of the individual's utilization of his information environment and certain of his attitudinal states constitute the three classes. The objective of the present section of the analysis is to discern how the variables representing these various classes are differentially effective in explaining variation in levels of information about innovations and levels of innovativeness. One procedure for analyzing data with such objectives in mind is sequential interaction analysis (Sonquist and Morgan, 1964). Sequential interaction analysis is a technique for deriving structure. lIn the present analysis a structure of relationships is sought for the effects of information inputs, source orientations and attitudes toward innovations upon information levels and subsequently for the effects of all these factors upon levels of innovativeness. The 93 technique does not assume linearity, additivity or prOperly scaled variables. The term sequential refers to the sequential selection of code values of a single variable to determine that partitioning point in the range of values where explained variance is maximum. It also refers to the iterative process in which each iteration includes all variables being paired with the dependent variable, regardless of whether a given variable has been selected previously as the optimum explana— tory variable. The analysis proceeds as follows: 1. Examination of the mean of the dependent variable against each possible partition point in the range of a variable code. The partition point selected is that point which best divides the sample of respondents in terms of variance explained. 2. This same process is repeated for each variable. 3. Having done this,the technique selects that variable which best explains the dependent variable. 4. At this point the sample is dichotomized on that variable at the derived partition point. 5. The next step is to take that subgroup with the largest unexplained variance, and once again repeat the sequential within- variable partition process across all variables, with the residual variance. 94 6. Again, that variable which best explains the residual variance is utilized to further classify (by dichotomization) that segment of the sample being scrutinized. 7. The process is allowed to continue until no way can be found to reduce the residuals by a sufficient amount or the subgroup size becomes too small.31 Sequential interaction analysis configurations are obviously a function of the variables selected for analysis and the manner in which they have been measured. Any derived configuration is subject to change with the inclusion of new variables, if those variables are sufficiently related to the criterion variable and at the same time are not highly correlated with other predictor variables. Measurement, too, is important. It may be the case that finer gradations in the instrument and thus, increased differentiation may pro- duce different profiles. The between-variable sequential process also copes with the question of intercorrelations among explanatory variables. Where two variables are highly intercorrelated the process will select the better of the two in terms of variance explained and likely discard the other leorgan and others (1966) suggest a level of 0.5 percent of the original total sum of squares for the variance reduction criterion and a group "n" no less than .25. In addition, no groups will be considered that contains less than 1.5 percent of the total original sum of squares. 95 since it can add little by virtue of its small Specific variance, to variance explained. The present analysis will utilize, first, information levels about three innovations as the dependent variable. Independent variables include: (1) constant information inputs; (2) intermittent information inputs; (3) specific information inputs; (4) mass media exposure at awareness; (5) cosmopolite source exposure at awareness; (6) cos- mopolite source exposure at conviction; (7) source variety exposure; and (8) attitudes toward innovations. The second problem utilizes all those explanatory variables used in the initial problem, plus level of in- formation, in an effort to discern the structure of re- lationships that best explain innovativeness. The foregoing multivariate analyses are utilized primarily as an exploratory approach. No hypotheses are advanced as to the nature of the structure. Such analyses will hopefully suggest: (1) how village social systems may be characterized with respect to information input types; (2) how such inputs affect levels of information, innovativeness and source orientations; (3) hypotheses to be tested. CHAPTER IV FINDINGS Information Levels Descriptive Findings Most respondents in the present analysis are aware of at least some innovations. Only 2 percent are not aware of any of the 14 innovations (Table 9). Twenty- five percent are aware of four or fewer innovations and 50 percent are aware of six or fewer innovations. Marked differences in awareness emerge when the innovations are considered individually. Table 10 reveals two important points on the distribution of the percent who are aware. The range of awareness on the four most widely known innovations is from 77 percent to 87 percent. From this point, however, levels of awareness drop to 45 per- cent. The next seven innovations by order of degree of awareness, range from 45 percent to 33 percent. The re— maining three innovations are relatively unknown among respondents; levels of awareness range from 19 percent to 15 percent. The distribution of number of innovations correctly known follows the distribution of awareness closely as 96 97 can be seen from Table 11. Only three percent of the respondents failed to know any correct item of information about any of the 14 innovations. One quarter (26 percent) correctly knew three or fewer innovations and 50 percent knew five or fewer innovations correctly. Possessing information about any one of the three innovations (fertilizer, aldrin dust, NS—l maize) is related to having information about the others. Table 12 shows that awareness among the innovations is significantly, but only slightly, interrelated. The median correlation is .23. Likewise, Table 13 shows having correct infor— mation about innovations is significantly, but slightly, interrelated. The median correlation is .22. The correlations between information level on each innovation are also significantly correlated with each other (Table 14). In addition, we note (Table 15) that adoption among the three innovations is significantly correlated. These findings lend credence to the decision to utilize total '(composite) scores for information level and inno- vativeness, across the three innovations. 98 Table 8. Distribution of Awareness of 14 Agricultural Innovations. Number of Innovations of Which Respondents Cumulative Are Aware Percent Percent o 2% 2% l 4 6 2 ll 3 l7 4 9 26 5 13 39 6 ll 50 7 10 6o 8 9 69 9 8 77 10 7 84 ll 6 9O 12 4 94 13 4 98 14 2 100 N = 1,347 99 Table 9. Percentages of Awareness of 14 Agricultural Innovations (N = 1,347) Innovations Percent ___ Aware 1. Poultry 87 2. Oil palm scheme 87 3. Fertilizer 83 4. NS-l Maize 77 5. Cocoa scheme 45 6. Citrus fruits 44 7. Rubber scheme 44 8. Cassava (improved) 41 9. Aldrin dust 39 10. Community plantations 36 11. Rice 33 12. Vegetable seeds l9 l3. F.A.I.D. loans/credit 19 14. Cashew 15 100 Table 10. Distribution of Number of Innovations Correctly Known. Number of Innovations Cumulative Correctly Known Percent Percent O 3% 3% 1 IO 2 6 l6 3 IO 26 )4 12 38 5 12 50 6 ll 61 7 IO 71 8 7 78 9 6 84 10 7 91 11 4 95 12 2 97 13 2 99 14 l 100 N = 1,347 lOl Table 11. Intercorrelations Among Awareness Scores for Agricultural Innovations (N = 1,347) variable Variable 2 3 l. Aware: fertilizer .19* .38* 2. Aware: aldrin dust - .23* 3. Aware: NS-l maize — *Significant at the one percent level. Table 12. Intercorrelations Among Correct Information Scores for Agricultural Innovations (N = 1,347) Variable Variable 1. Correct: fertilizer .18* .38* 2. Correct: aldrin dust - .22* 3. Correct: NS—l maize - *Significant at the one percent level. 102 Table 13. Intercorrelations Among Individual Information Level Scores by Innovations (N: 1,347) variable Variable 2 3 1. Information level: fertilizer .23* .48* 2. Information level: aldrin dust - .27* 3. Information level: NS-l maize - *Significant at the one percent level. Table 14. Intercorrelations Among Adoption Scores for Agricultural Innovations. (N = 1,347) Variable Variable 2. 3 1. Adopt: fertilizer .24* .37* 2. Adopt: aldrin dust - .l8* 3. Adopt: NS-l maize *Significant at the one percent level. 103 Of the three innovations dealt with in greater detail, fertilizer and NS-l maize are most widely known. Eighty-three percent of the reSpondents are aware of fertilizer and 77 percent are aware of the new variety of corn. Only 39 percent are aware of aldrin dust. With reSpect to knowing correctly these three innova— tions, 81 percent and 74 percent have some correct in- formation about fertilizer and NS-l maize, reSpectively. Only 38 percent have some correct information about aldrin dust. Table 16 summarizes these findings. Table 15. Percentages of Awareness and Correct Infor— mation About Three Agricultural Innovations (N = 1,347) Percent Innovations Aware l Aware of fertilizer 83 2 Aware of Aldrin dust 39 3 Aware of NS-l maize 77 4. Correctly knows fertilizer 81' 5 Correctly knows aldrin dust 38 6 Correctly knows NS—l maize 74 104 Table 17 depicts the ranges of information levels about three innovations and respondents! total information score. The range of information levels for fertilizer and aldrin dust is from zero to seven, while for NS—l maize it is from zero to nine. The range for total information level (three innovations) is from 0 to 21. Of the total sample, 120 reSpondents are not aware of any of the three innovations. One half (51 percent) of the sample know only six items of information about the three innovations. Table 16. Distributions of Individual Information Levels by Innovation. (N = 1,347) Fertilizer Aldrin Dust NS-l Maigg Total Information Cum Cum Cum Cum Level % 1% .2_ % z. % % .% 0 l7 17 61 61 23 23 9 9 l 2 l9 1 62 3 26 l 10 2 l9 38 9 71 23 49 7 l7 3 29 67 16 87 26 75 6 23 4 27 94 10 97 18 9g 8 3l 5 4 98 2 99 5 9 9 4O 6 l 99 * 99 l 99 ll 51 7 * 100 * 100 * 99 ll 62 8 * 99 12 74 9 * 100 9 83 10 6 89 ll 4 93 12 3 96 l3 2 98 14 l 99 15 * 99 16 * 99 17 * 99 18 * 99 19 * 99 20 * lOO *Less than one percent. 105 Tests of Hypotheses Dealing with Information Level Prior to testing the hypotheses the intercorrelations among the independent variables were inSpected. In- 'spection of Table 17 indicates that some of the independent variable are relatively highly correlated among themselves. Thus, in order to determine the Specific effects of each independent variable with the dependent variables, partial correlations were computed for each hypothesized relation- ship. Thus, each relationship is tested, removing the effects of all other independent variables. The findings for each hypothesis are now presented. Those dealing with information level are presented first, followed by those concerning innovativeness. Theoretical Hypothesis 1: Individual information level varies directly with the village level of non-specific in- formation inputs. Empirical Hypothesis l-l: Individual information level about three innovations varies directly with the village level of constant information inputs. The partial correlation between individual in— formation levels and the village level of constant in- formation inputs is -.04 (Table 18), which is not significant at the five percent level. Empirical hy- pothesis 1—1 is not confirmed. 106 Table 17. Intercorrelations Among Dependent and Independent variables Variables 2 3 4 5 6 7 8 9 10 ll 1. Constant inputs .20 .28 .32 .07-.09 -.11 .04.12 .07 .20 2. Intermittent inputs -.10 .24 .11 .06 .06 .10.06 .02 .10 3. Specific inputs .55 .05-.02 -.l2 .o8.33 .30 .25 4. Leaders' relaytf capacity .10 .07 -.O9 .O5.29 .26 .21 5. Mass media awareness .06 - O5—.03.02 .07 .03 6. Cosmopolite Sources- awareness .71 .37.37 .22 .35 7. Cosmopolite Sources- conviction .42.30 .06 .28 8. variety of sources .42 .53 .55 9. Attitude toward innovations .80 .50 10. Information level .58 11. Innovativeness 107 Empirical Hypothesis l-2: Individual information level about three innovations varies directly with the village level of inter- mittent information inputs, under conditions of low constant in- formation inputs. The mean information level for those with high intermittent information inputs is 6.48. The mean information level for those with low intermittent information inputs is 6.08. The t-value for the difference between the two means is 1.38, which is not Significant at the five percent level. Empirical Hypothesis l-2 is not confirmed. Thus, Theoretical Hypothesis 1 is not confirmed. Theoretical Hypothesis 3: Individual information level varies directly with the village level of Specific information inputs. Empirical Hypothesis 3-l: Individual information ‘7 level about three inno- vations varies directly with the level of ex- tension agent's use of communication strategies. The partial correlation between individual informa- tion levels and the village level of extension agent's use of communication strategies is —.04 (Table 18), which is not significant at the five percent level. Empirical hypothesis 3—1 is not confirmed. 108 Table 18. Correlations Between Independent Variables and Individual Information Level and Innovativeness (N = 1,347) Individual Information Level Innovativeness Independent Zero-order Partial Zero—order Partial Variables jCoppelation Correlation Correlation Correlation l. Constant inputs .07* —.04 .2o** .l6** 2. Intermittent inputs —.O2 -.l9** .lO** -.O2 3. Specific inputs .30** -.04 .25** .02 4. Leader relay capacity .26** .10** .21** —.01 5. Variety of sources .53** .23** .55** .30** 6. Mass media- awareness .O7* —.03 .03 -.O5 7. Cosmopolite source- awareness .22** .13** ,35** ,13** 8. Cosmopolite source- conviction .O6* -.O9** .28** .O4* 9. Attitudes toward innovations .80** .62** .50** .21** 10. Information level - .58** .16** *Significant at the 5 percent level. **Significant at the 1 percent level. 109 Thus, Theoretical Hypothesis 3 is not confirmed. Theoretical Hypothesis 5: Individual information level varies directly with village leaders' relay capacity. Empirical Hypothesis 5-l: Individual information level about three innovations varies directly with the leaders' external contact and opinion leadership. The partial correlation between individual infor- mation level and village leaders' relay capacity is .10 (Table 18), which is significant at the one percent level. Empirical hypothesis 5-1 is confirmed. Thus, Theoretical Hypothesis 5 is confirmed. Theoretical Hypothesis 7: Individual information level varies directly with the variety of sources to which one is exposed. Empirical Hypothesis 7-1: Individual information level about three innovations varies directly with the number of different sources to which one is exposed. The partial correlation between individual in- formation level and the number of different sources to which one is exposed is .23 (Table 18), which is signifi- cant at the one percent level. Empirical hypothesis 7—1 is confirmed. 110 Thus, Theoretical Hypothesis 7 is confirmed. Theoretical Hypothesis 9: Individual information level varies directly with degree of mass media exposure at awareness. Empirical Hypothesis 9—1: Individual information level about three innovations varies directly with degree of mass media exposure at awareness. The partial correlation between individual in- formation level and the degree of mass media exposure at awareness is -.03 (Table 18), which is not Significant at the five percent level. Empirical Hypothesis 9-1 is not confirmed. Thus, Theoretical Hypothesis 9 is not confirmed. Theoretical Hypothesis 11: Individual information level varies directly with degree of cos- mopolite source ex- posure at awareness. Empirical Hypothesis ll—l: Individual information level about three innovations varies directly with degree of exposure to cos- mOpolite sources at awareness. The partial correlation between individual information level and degree of cosmOpolite source exposure at aware- ness is .13 (Table 18), which is significant at the one jpercent level. Empirical Hypothesis 11-1 is confinmed. 111 Thus, Theoretical Hypothesis 11 is confirmed. Theoretical Hypothesis 12: Individual information level varies directly with degree of cos- mopolite source ex- posure at conviction. Empirical Hypothesis 12-1: Individual information level varies directly with degree of exposure to cosmopolite sources at conviction. The partial correlation between individual informa- tion level and degree of cosmopolite source exposure at conviction is -.09 (Table 18), which is significant at the one percent level but in the unpredicted direction. Empirical Hypothesis 12-1 is not confirmed. Thus, Theoretical Hypothesis 12 is not confirmed. Theoretical Hypothesis 15: Individual information level varies directly with degree of favorability toward innovations. Empirical Hypothesis 15-1: Individual information level about three innovations varies directly with degree of perceived advantages of innovations. The partial correlation between individual in- formation level and degree of perceived advantages of innovations is .62 (Table 18), which is significant at the one percent level. Empirical Hypothesis 15-1 is confirmed. Thus, Theoretical Hypothesis 15 is confirmed. 112 Adoption and Innovativeness Descriptive Findingsw Adoption of agricultural innovations is generally low. From Table 19, we note that 37 percent of the sample have not adopted gpy_of the 14 innovations. Eighty per- cent have adopted three or fewer innovations. The most frequently adopted innovations are: (l) fertilizer (39 percent); (2) NS—l maize (32 percent); (3) aldrin dust (23 percent). Table 20 shows the percent of people adopting each of the 14 innovations. Table 21 demonstrates the manner in which adoption over time is distributed among the reSpondents for three innovations: (l) fertilizer; (2) aldrin dust; (3) NS-l maize. While the rate of adoption is increasing for all three innovations, it is increasing at a greater rate for NS-l maize, followed by fertilizer and aldrin dust in that order. 113 Table 19. Percentages Adopting Agricultural Innovations. Number of Innovations Percentages of Adopted Respondents 0 37% l 19 2 l4 3 10 4 7 5 5 6 3 7 2 8 2 9 or more 1 Total 100 N = 1:347 114 Table 20. Percent of Respondents Adopting Each of 14 Agricultural Innovations. Percent of ReSpon— Innovation dents Adopting 1. Fertilizer 39% 2 NS-l maize 32 3 Aldrin dust 23 4 Poultry l9 5. Oil palm rehabilitation l7 6 Community plantations l3 7 Citrus lO 8 Rice 8 9 Improved cassava variety 7 10. Vegetable seeds 7 ll. Cocoa planting scheme 5 12. Rubber planting scheme 4 13. Cashew 3 l4. F.A.I.D. credit / loans 2 100 N = 1,347 115 Table 21. Adoption Distributions for Three Innovations in Eastern Nigeria: Fertilizer, Aldrin Dust and NS-l Maize (N = 1,347)_ Innovation ;‘« Fertilizer Aldrin Dust NS-l Maize Year of Cum Cum Cum Adoption % % % % % % 43 0.1 45 0.2 .1 46 0.1 48 0.2 .l 49 0.2 50 0.5 l 51 0.2 1 1 .1 52 0.5 1 .l 53 0.5 2 54 1.0 3 0.1 55 0.5 4 0 1 56 1.0 5 0.2 1.0 57 1.0 6 - 1.0 58 0.5 7 0.3 1.0 59 2.0 9 1 0 2.0 60 1.0 10 l 0 3.0 0.3 61 2.0 12 2 0 5.0 0.3 .0 62 2.0 14 3 0 8.0 2.0 .0 63 7.0 21 5.0 13.0 5.0 7.0 64 4.0 25 3 0 16.0 6.0 13.0 65 6.0 31 3.0 19.0 9.0 22.0 66 8.0 39 4 0 23.0 10.0 32.0 Aware/Non- Adopters 45.0 84 18.0 41.0 44.0 76.0 Not Aware 16.0 100 66.0 100.0 24.0 100.0 116 Usage in 1966 of fertilizer, aldrin dust and NS—l maize are intercorrelated. Table 22 presents the inter- correlations among... the three innovations. All are significantly correlated with each other. Table 22. Intercorrelations on Usage of Three Inno- vations in 1966 (N = 1,347). Innovations Aldrin dust NS-l Maize l. Fertilizer .27* .37* 2. Aldrin dust - .18* 3. NS-l maize - — *Significant at the one percent level. Tests of Hypotheses Dealing with Innovativeness. Theoretical Hypothesis 2: Individual innovative— ness varies directly with village level of non-Specific informa— tion inputs. Empirical Hypothesis 2—1: Relative earliness of adoption of three inno- vations varies directly with the village level of constant input facilities. The partial correlation between individual innovativeness and the village level of constant infor- mation inputs is .16 (Table 18), which is significant at 117 the one percent level. Empirical Hypothesis 2-1 is confirmed. Empirical Hypothesis 2-2: Relative earliness of adoption of three innovations varies directly with the village level of inter- mittent information inputs, under conditions of low levels of con- stant information inputs. The mean on innovativeness for high intermittent infor- mation inputs is .31. The means on innovativeness for low intermittent information inputs is .38. The t—value for the difference between the two means is 2.12, which is significant at the 5 percent level. Empirical Hypothesis 2-2 is confirmed. Thus, Theoretical Hypothesis 2 is confirmed. Theoretical Hypothesis 4: Individual innovative- ness varies directly with the village level of specific information inputs. Empirical Hypothesis 4—1: Relative earliness of ’ adoption varies with the level of extension agent's use of communi- cation strategies. The partial correlation between individual inno- vativeness and the village level of extension agent's use of communication strategies is .02 (Table 18), which is not significant at the five percent level. Empirical Hypothesis 4-1 is not confirmed. Thus, Theoretical Hypothesis 4 is not confirmed. 118 Theoretical Hypothesis 6: Individual innovative- ‘ ness varies directly with the village level of leaders' relay capacity. Empirical Hypothesis 6-1: Relative earliness of adoption of three innovations varies direttly with the level of village leaders' external contact and Opinion leadership. The partial correlation between individual inno- vativeness and the village level of leaders' relay capacity is -.01 (Table 18), which is not significant at the five percent level. Empirical Hypothesis 6-1 is not confirmed. Thus, Theoretical Hypothesis 6 is not confirmed. Theoretical Hypothesis 8: Individual innovative- ness varies directly with the variety of sources to which one is exposed. Empirical Hypothesis 8—1: Relative earliness of adoption varies directly with the number of different sources‘to which one is exposed. The partial correlation between individual inno- vativeness and the number of different sources to which one is exposed is .30 (Table 18), which is significant at the one percent level. Empirical Hypothesis 8-1 is confirmed. Thus, Theoretical Hypothesis 8 is confirmed. 119 Theoretical Hypothesis 10: Individual innovative- ness varies directly with degree of mass media exposure at awareness. Empirical Hypothesis lO-l: Relative earliness of 7' ad0ption of three innovations Varies directly with level of mass media exposure at awareness. The partial correlation between individual inno- vativeness and degree of mass media exposure at awareness is -.05 (Table 18), which is not significant at the five percent level. Empirical Hypothesis 10-1 is not confirmed. Thus, Theoretical Hypothesis 10 is not confirmed. Theoretical Hypothesis 13: Individual innovative- ness varies directly with degree of cos— mopolite source ex- posure at awareness. Empirical Hypothesis 13-1: Relative earliness of adoption of three innovations varies directly with degree of exposure to cosmo- polite sources at awareness. The partial correlation between individual innovativeness and degree of cosmopolite source exposure at awareness is .13 (Table 18), which is significant at the one percent level. Empirical Hypothesis 13-1 is confirmed. Thus, Theoretical Hypothesis 13 is confirmed. 120 Theoretical Hypothesis l4: Empirical Hypothesis 14-1: Individual innovative- ness varies directly with degree of cos- m0polite source ex- posure at conviction. Relative earliness of adoption of three innovations varies directly with degree of exposure to cos- mopolite sources at conviction. The partial correlation between individual inno- vativeness and degree of cosmopolite source exposure at conviction is .04 (Table 18), which is not significant at the five percent level. Empirical Hypothesis 14-1 is not confirmed. Thus, Theoretical Hypothesis 14 is not confirmed. Theoretical Hypothesis l6: Empirical Hypothesis l6-l: Individual innovative- ness varies directly with degree of favorability toward innovations. Relative earliness of adoption of three innovations varies directly with degree of perceived advantages of innovations. The partial correlation between individual inno- vativeness and degree of perceived advantages of innovations is .21 (Table 18), which is Significant at the one percent level. Empirical Hypothesis 16-1 is confirmed. 121 Thus, Theoretical Hypothesis 16 is confirmed. Theoretical Hypothesis 17: Individual innovative- ness varies directly with levels of infor— mation. Empirical Hypothesis l7-l: Relative earliness of adoption varies directly with individual in- formation level about three innovations. The partial correlation between individual innor vativeness and individual level of information is .16 (Table 18), which is significant at the one percent level. Empirical Hypothesis 17-1 is confirmed. Thus, Theoretical Hypothesis 17 is confirmed. Multivariate Analyses In an attempt to determine the conjunctive effects of the independent or constrant variables upon the two dependent variables, the data were analyzed using the technique of sequential interaction analysis (Sonquist and Morgan, 1964). Sequential interaction analysis pro— vides the analyst with a configuration of variables organized in such a way as to demonstrate how variables combine to maximally explain variation in a dependent variable. In the present analysis two dependent variables are investigated. In the first instance the analysis seeks to determine how the selected constraint variables explain variation in individual information level. In the 122 second case, the analysis seeks to ascertain how the selected constraint variables and individual information level together explain variation in individual innovative- ness. The findings for each dependent variable are pre- sented separately. Dependent variable: Individual Information Level. From Figure 3, it can be seen that the sample initially Splits on the variable attitudes toward inno- vations, in attempting to explain variation in informa- tion levels. The more favorable one is toward the three innovations, the greater is one's level of information about those innovations. Generally, speaking, the degree of exposure to the mass media and degree of cosmopolite source exposure tend to discriminate individuals with lower levels of infor- mation. For those individuals with higher levels of information, attitudes toward innovations and the variety of sources to which one is exposed become im- portant discriminants. Those individuals with the highest information levels had exposure to five or more sources of information and were highly favorable toward all three innovations. The only social system variable to appear in the final configuration is intermittent information inputs. However, it should be noted that the relationship be- tween intermittent information inputs and information level is negative for the particular subgroup in question. 123 Intermit- tent Inputs X = 10.41 N = 66 Codes = 1,2 15 Intermit- tent Inputs x = 8. 60 N = 199 Codes = 3 ources** 11.15 —l225 odes = Sources =9 05 5 = 265 Attitude Codes = O- 1 Toward ///;fi Innova- ’ ttitude tions Mass Media oward X = 9,71 Exposure: nnova- 1N = 387 Conviction ions 8 Codes = 6 = 7-14 = 5.9 - N = 289 = 90 Codes = 1.C ‘ Odes = 5)6 Attitude / 2.0 Toward 4 l/fa Innova- 0 tions ass Media 1 = 6.64 xposure: Infor- = 504 Conviction mation Codes -.= 5 = 5,98 Level - N = 215 x = 6.25* Codes = 0 N = 1,347 ‘ I3 Cosmopolite Sources: Awareness = 4.88 N = 120 Codes = 1.3 2. Figure 3. Configurational Analysis of Prediction of Individual Information Levels. 124 7 Mass Media ‘ Exposure Awareness 2 X = 3.90 12 Attitude N = 337 Cosm0polite Toward Codes = 1.0 W Source: Innovations 2.0 Awareness X = 2.88 X = 3.36 N = 456 N = 217 Codes = 1-4 . 6 Codes = 1.0 Mass Media Exposure: Awareness X = 0.00 N = 119 Codes = 0 Figure 3--continued *Means reported in Figure 3 represent the subgroups mean value on the dependent variable of infor- mation level. **Sources refers to the "variety of sources to which one is exposed." 1 7.111 . 125 In other words, higher levels of intermittent inputs are associated with lower levels of information. Terminal subgroup means on information level range from 0.00 (those unaware of the innovations) to 11.15. In all cases except that of intermittent information inputs the relationships between the independent or con- straint variables and the dependent variable is in the predicted direction. Information Level Typologies Four possible typologies emerge with reSpect to variation in the dependent variable, information level. The first type are those who can cite advantages to all three innovations and utilize a variety of sources to acquire their information. Persons of this type have the highest levels of information. This type consists of 387 reSpondents. The second type are those with moderatey high in- formation levels. Such persons are characterized by relatively highly favqrable attitudes toward the inno- vations and indicate they have been convinced of the usefulness of the innovations by either the mass media or interpersonal sources. Those with favorable attitudes and who have not yet reached the point of being con- ‘Vinced (Subgroup No. 10) are only slightly less knowledge- able. This type consists of 504 reSpondents. 126 A third type are those who are relatively un- favorable toward the innovations, i.e., could cite few advantages of the innovations. Among this group, those attributing awareness to cosmopolite rather than localite sources, were Slightly more knowledgeable. This type consists of 337 respondents. The fourth type are those who are unaware of any of the three innovations. (Subgroup No. 6). Of the total sample, 179 are unaware of any of the three inno- vations. Figure 4 summarizes the characteristics of these information level typologies. Dependent Variable: Individual Innovativeness Figure 5 is an explanatory configuration of in- dividual innovativeness.. The sample is first split on the degree of mass media eXposure at the conviction stage. 0f the total sample, 660 persons indicated nothing or no one had convinced them to use any of the three innovations. By virtue of the wording of the question it is possible for someone to indicate "not being con— vinced by someone or something" and still have adopted. The data bear this out, since the mean value of subgroup #2 is greater than zero. It may be that the question failed to discriminate those who adopted on the basis of convincing themselves rather than being convinced. 127 moflmoaomza Ho>oq soapmahomcH .: oasmflm zoq - I - mad oo.o case - soq swam - - soq saw mm.m case - swam swam I I soq omH mm.s - I - son I swam mam mm.m ooze I I I Qwflm I swam mwm #H.N swam - I » soq swam mmw. om.w sou - - - sou swam mo H:.oa H some I I I I swam swam NNH mH.HH 6mm: sown Imamz< IoH>soo mCOHp QSOHw Hm>oq soap Hmsmq_c0Hp mpSQGH moadom mums madmomxm mochsom Iw>occH Imam ImEHOMGH ImagomcH unoppwa mafiaom Imamz¢ oaomz mo oaozoe mo no :60: so Iaoch Ioamoo mmoz hpmfiam> oodpapp< ouwm msoanSm zwoaomze manwfiam> pcoocmmoc:H mmoco>flpm>oqu stofl>fiocH mo mQOpofiomam mo mfimhamc< Honoapwssmflmnoo .m onsmflm 128 o.H “moooo as u z Hm.o n M COHpOH>coo mIo "wmooo madmoaxm . IIIIIIIIIII 4 mm H z mPHH _ m.H.m.H Hm.o n M IoaoEmo ”m . “moooo. Ho>mq UNA . L mmmmomom - N. H z. COHFME OoN I'D "mmfivoo omoo H I“ . . I . lpHO G 55H N Z I . mmowpmam. m HmH Im.H"ooooo 35.0 n x moose . . . . fidHHdIJddddU4 mm H z mooasoml Im>occH Icoo . I , .b- @9030 . madmoaxm. 1 mm u z :m o I_M B . m a on. mm.o u x COHHOH>COQ hum "moooo mUSppr< oi. so? 33 34 u m . IIIIIIIIIIII IMEhOmchH IOQOEmOUmH. wOOHSOmb 94w "moooo me u z :Io "moooo mo.a n m OM H z WQUQDOWJH NN.H n M . mpsmcH scopwcoo mm s "moooo u-a.m-mmmaaou mm H z . s u 2. www M I m . om.a u m. a ommH . opeacH. . esopmcoomm. 129 o.~ “moooo mm H z Hm.o u m coHp0fi> Icoo madmoaxm opaaoa IOEmoumm .mmmco>fipw>occa stofi>aona .manwahw> pcoocmmov 059 so 05Hw> some m_msoam95m 05p pcmmoamma m ohsmfim CH bophomoh mcwsz emanapcoo.ru m ouswam .muouooooo mud n z mma.o u m U . NIQDOWQH o.a .o "moooo ma: u z. :o.o u MA mmmcmamz< chamomxm mafia OU film "WMU mm.ert.m IoaoEmoom a V...” mmowsomsa mHIo “moooo mma u z oo.o u M Ho>wq soap ImELomcHom .wO#w¥uxz .mm.o mam mowQSOmoH «no "moooo. mHImHHmouou m ",2 mm.H n m Ho>mq soap IMELOMCHHNH ----J ———-— blmquUOU. 09.9 n M moohdomaa o.m m.Hummooo Hzm u z mH.o n M mmocohmz< madmomxm oped IoaoEmoom o “oo6004 coo u z oo.o u m cowp I0fi>coo mgjm dogxm taco: wmmzm mma.u,z; I u . . 1 vi. I‘ldtu- 31!. ‘v _ . , L . . . pl. . IL A , fail-um. 4 1.1.1.. .I‘Il ,. (I .....- , (Ci. mIanmoooo mam u z o:.o u m mCOflp Im>occH UQMBOH oospaupflpm> IoccHH 130 The prevalence in the configuration of the variable called "variety of sources to which one is exposed," is worthy of note. Three terminal subgroups split on this variable. Seven terminal subgroups, at one stage in the analysis, derived from this variable. Only one terminal subgroup failed to end in or be derived from the "variety of sources" variable. The group in question is the least innovative group, those 419 people who are either unaware (N = 119) or who have become aware only through localite sources. It should be recalled that all sources utilized in the measure of "variety of sources” were in fact cosm0polite sources. It should be noted that certain of the terminal subgroups are made up of too few respondents to be descriptively useful. One of the terminal groups has two respondents and two others have only seven. With cell sizes this small, the generalizability and stability of such subgroups is questionable. The remaining subgroups vary in Size from 28 to 419. Innovativeness Typologies From Figures 5 and 6 four possible typologies emerge. Those who are generally most innovative are typified by having been convinced of the innovations' usefulness, having very favorable attitudes toward the innovations and being exposed to several sources. This 131 description generally fits the four subgroups with the highest means on innovativeness. The seven most inno- vative respondents, in addition to being high on the three aforementioned variables, were in communities with high constant information inputs. In one subgroup, the reSpondents were exposed to a few rather than several sources, however, they had relatively high levels of in- fbrmation and attributed their being convinced to cos- m0polite sources. A second type are those who could be said to be moderately innovative. Again, these respondents indicate having reached the conviction stage. Approximately 40 percent (N = 114) of this type were highly favorable toward the innovations, but were exposed to fewer sources, while approximately 60 percent were somewhat less favorable but were exposed to a greater number of sources. For two of these subgroups, respondents had relatively low levels of information. The third type were only slightly innovative. While 240 reSpondents of this type reported having reached the conviction stage they were relatively less favorable toward the innovations and were exposed to relatively fewer sources of information. The remainder of this third type were generally less favorable toward the innovations, however, they tended to be exposed to a greater number of sources. 132 .mmososflpm>occH mo moflwoaogme .@ magmas 30a I I I I I 30A ma: :0. make swam I I I sou I sou msu mu. emum I I I emum I sou mm mm. m case I I I I sou sou swam oam cm. I sou I I swam sou swam mmu om. I ewum I I emum sou emum m mm.u case I I sou I sou swam emum as up. I sou tour I sou ewum swam mm um. I ewum swam I sou smum sous mo so. I I I I emum cwum ewum ms mo.u case I I I sou swam sous swam om mm.u I I I emum swam ewum smum s om.u soap mcofip msoum mama Ho>ou Iofl> .mpsg .woo. ..st soap Indm,.Io>Hpm>oc mmomo>ap mmoc SOprE Icoo IcH Ihsom IoccH IoH>coo to IQH co Im>OQCH Ioswz< IsomcH moo pcopm Mo ohmzoa .mnSmom omflm com: mo mooasom Issom Icoo apofism> 065p Ixm msonw maze ouuuoa ouuu Iusou ouooz Insm Ioamoo Iogoa mmmz ImOO 133 The fourth type are the least innovative group. This group includes the 119 reSpondents who are not aware of any of the three innovationsa This type is characteri- zed by not having reached the conviction stage and being exposed to few sources. Most of those who are aware attribute their awareness to localite rather than cos- m0polite sources. Those who do attribute awareness to coosmppolitewsources((Nw=1172)mhavenonly a slightly higher level of innovativeness. CHAPTER V SUMMARY AND DISCUSSION Summary The present analysis explored the relationships between information one possesses about a course of action and one's decisions with reSpect to that course of action. In particular the present study is concerned with in- dividual levels of information about innovations and decisions to adopt or not adopt those innovations. Information level is defined as the amount of knowledge an individual has about agricultural inno— vations. Innovativeness is, "the degree to which an individual is relatively earlier in adopting new ideas than other members of his social system" (Rogers, 1962). Information is viewed as a necessary condition for adoption behavior. Constraints operate to affect both information level and adoption behavior. Three levels of constraints are considered. The first of these are characteristics of the social system. Information inputs to a social system limit that which can be known by members of that system. Three forms of information inputs are defined: (1) constant information inputs; (2) intermittent in- formation inputs; and (3) Specific information inputs. 134 135 Constant information inputs are defined as those facilities, continously present in the village which enhance the flow of information into the village. Intermittent informa- tion inputs are those facilities, discontinuously pre- sent in the village which enhance the flow of information intotthervillage. Specific information inputs are those facilities and activities which represent the Spread of Specifically, innovation information. The capacity of leaders of the social system to acquire and relay in- formation to other members of the social system is yet another attribute of social systems considered in the present analysis. It was hypothesized that levels of information inputs and leaders' information relay capacity varied directly with individual information level and innovativeness. A second level of constraints included variables representing the individual's participation in his infor— mation environment. Degrees of mass media exposure and cosmopolite source exposure and the variety of sources attended to, were taken as indicants of the individual's participation in the information environment. Mass media exposure, cosmopolite source exposure and variety of sources used, were hypothesized to vary directly with individual information level and innovativeness. The third type of constraint upon information level and innovativeness is attitudes toward innovations. 136 It was hypothesized that attitude toward innovations vary directly with individual information level and innovativeness. In addition to the aforementioned hypotheses, the study sought to determine multivariate configurations of predictor variables for two dependent variables: in— formation levels and innovativeness. The data were collected from male heads of house- holds and village leaders in rural Eastern Nigeria. A total of 1,347 respondents in 18 villages were personally interviewed to provide the data on information levels, attitudes, and source exposure and innovativeness. Data on the social system variables were derived from reSponses by selected village leaders. Only two percent of the sample were unaware of any of the 14 selected innovations. Fifty percent of the sample were aware of six or fewer innovations. Fifty percent of the sample possessed some correct information about five or fewer innovations. For the three inno— vations, fertilizer, aldrin dust and NS-l maize, aware— ness, information levels and adoption were each signifi— cantly intercorrelated. Individual information level about three innovations was positively and significantly related to: (1) leaders' relay capacity; (2) variety of sources exposed to; (3) cosmopolite sources at awareness; (4) attitudes 137 toward innovations. Information level and intermittent inputs were significantly and negatively related. Of the total sample (Ne1,347), 37 percent had not adopted any of the 14 innovations. Fifty percent of the sample had adopted from five to one innovation. For the three innovations upon which the present analysis focused, 39 percent adopted fertilizer, 32 percent adopted NS-l maize and 23 percent adopted aldrin dust. Individual innovativeness across three innovations was positively and significantly rleated to (l) constant information inputs; (2) variety of sources exposed to; (3) cosmopolite sources at awareness, (4) cosmopolite sources at conviction; (5) attitudes toward innovations; and (6) individual information level. Two multivariate analyses were performed. The method of sequential interaction analysis was used to explain variation in individual information level and individual innovativeness. Attitude toward innovations, and variety of sources used best distinguishes those individuals with high levels of information. Being aware (and not aware) of innovations and utilization of localite sources are the variables best characterizing those persons with low levels of information. Generally Speaking, characteristics of the social system did not significantly differentiate the sample on the basis of knowledge levels. The exception was intermittent in- formtion inputs, which for a subset of the sample, exhibited a negative relationship with information level. 138 From Figure 5 we note that the variety of sources to which one is exposed appeared in all but one branch of the configuration of predictors of innovativeness. The social system variable of constant information inputs, in addition to variety of sources, attitude toward innovations and being convinced of the value of innovations, characterized the most innovative subgroups. Information level served to discriminate some of the intermediate or moderately innovative subgroups. Discussion Levels of Information and Innovativeness The data on "awareness" indicate that most of the respondents are aware of some, though not many, innovations. Fifty percent could identify only five innovations or fewer. A word of caution must be injected at this point. The data are based upon aided recall rather than unaided recall.32 One does not know to what extent, if at all, presenting the reSpondent with the names of the inno- vations prqduced a bias in reSponses which had the effect of inflating levels of awareness. From Table 10, it can be observed that four innovations (domestic poultry, oil palm scheme, fertilizer, and NS—l maize) are much more 32By aided recall is meant that the reSpondents were cued in some way, in this instance by asking the reSpondent if he knew of a particular innovation. By the method of unaided recall reSpondents named innovations they knew without their being mentioned by the interviewer. 139 widely known than others. This is due, to a considerable degree, to the emphasis these innovations have received from the Ministry of Agriculture. Aldrin dust, though actively promoted by the Ministry, is a more recent innovation and is applicable only for yarn producers. These two conditions may account for lower levels of awareness. The pattern of responses for "correctly" knowing fourteen innovations follows closely the awareness dis- tribution. Fifty percent correctly identified five or fewer innovations. Only three percent failed to give correct information about any of the fourteen innovations. It must be remembered at this point that being able to state but one attribute of the innovation constitutes hav mg correct information. By such a minimum definition, an individual may be said to have correct information about a course of action, but this is not to say he has a sufficient amount or adequate kinds of information for innovation decision-making purposes. As previously noted, the innovations of fertilizer, aldrin dust and NS-l maize were selected for greater depth of study. Reasons for choosing these three were: (1) gen— eralized applicability to the study area; and (2) major emphasis by the Eastern Nigerian Ministry of Agriculture. Awareness among the three innovations is intercorrelated, as is the case with having correct information. Likewide, information levels and adoption among the three innovations are each intercorrelated. 140 Though these three innovations constitute a major focus from the VieWpoint of the Ministry of Agriculture, information levels about the three innovations are relatively low. Sixty—seven percent knew three items of information or less about fertilizer. Eighty—seven per- cent knew three items or less about aldrin dust and seventy-five percent knew three items or less about NS-l maize. When information levels for each innovation were summed to get a total index, fifty—one percent knew six items or less about the three innovations combined. It must be remembered, however, that individual information levels were obtained by having interviewers probe for statements of information about their innovations. Had respondents been given some form of information test the results might have been different. That is, respondents might have exhibited higher levels of information. Whether or not the rank-order of individuals on infor— mation levels would have changed is also uncertain. In the absence of such evidence, one might assume that it would not; i.e., that either technique would result in the same distribution and rank-ordering of respondents. Low levels of information might also be attributed to limited mass media facilities; extension agents must rely on interpersonal communication methods. While such methods may be generally effective in inducing change, 141 the fact that there are insufficient personnel to contact most of the rural population, places limits upon that which peOple can be expected to know and adopt. The criteria of generalized applicability and emphasis in promotion of the three selected innovations would appear to be born out again by the data on adOption. The three most widely-adopted innovations are fertilizer, NS—l maize and aldrin dust, in that order (Table 20). The poultry program, though widely known, has proven less than satisfactory from a production viewpoint and the oil palm scheme requires that the adOpter possess five acres of land or more, which most farmers do not have. As was the case with information levels, the level of adoption of innovations is low. Obly 39 percent of the respondents had adopted commercial fertilizer, the most frequently adOpted innovation. It would seem that Eastern Nigerian farmers can be typified as being generally aware of a few innovations, possessing little information about them, and having adopted very few of them. Bivariate Relationships: Information Level As a first step in the investigation of the relation- ships in the paradigm of information and modernization (Figure 2), the analysis sought to determine how infor- lu2 mation lavel and innovative behavior were related and how each of the various constraint variables were related to information level and innovativeness. The zero—order correlation between information level and innovativeness is .58; the partial correlation is .16. While both are significant at the one percent level, the squared partial correlation indicates that only 3 percent of the variance is accounted for. It should be pointed out, however, that one reason for such a low partial correlation may be the very high correlation between attitude toward innovations and information level. The same argument may account for the low-partial correlation between attitude toward innovations and innovativeness (.21). It should be noted that the squared partial correlations represent the specific variance accounted for by each independent variable. The partial corre- lation between information level and attitude toward innovation is .62. In other words "attitudes" account for approximately 36 percent of the variation in infor- mation level. It has been previously pointed out that the measure of attitude toward innovations is con:“, ----- -” taminated probably to a large degree by a cognitive component. The social system variables are not strongly re- lated to individual information level. Intermittent inputs is significantly and negatively correlated with in- 143 formation level (r = .l9). The concept itslef, as well as its measurement, presents problems. The index of intermittent inputs consists of only three items and the distribution was not normal. Of the eighteen villages, 12 possessed three inputs, two had two and four had one. No villages were without any intermittent inputs. Thus, the limited range of values of the variable plus the skewed distribution contributed, in part, to the low correlation and the unanticipated negative relationship. Leader relay capacity, while significantly related to information level, only accounts for one percent of the variance and thus, has little predictive or eXplanatory utility. Constant inputs and Specific inputs were not significantly related to information level. Thus, the thesis that information inputs to a social system affect individual information levels is somewhat suSpect. How- ever, it should be noted that the attitudinal variable has been partialled out and this is likely to affect the strength of the partial correlations. Of the variables indexing the reSpondents‘ participation in the information environment, variety of sources exposed to, and cosmopolite source exposure at awareness were significantly related to information level as hypothesized. The variance eXplained is low in both instances; five percent and two percent reSpectively. Degree of mass media expOsure at awareness was not saignificantly re1ated to information level. Cosmopolite 144 source exposure at conviction was significantly related to information level in the unpredicted direction. Only four percent of the sample attributed awareness of any of the three innovations to the mass media. Eighty-nine percent attributed their awareness to interpersonal sources. Fifty one percent of the sample attributed their aware- ness to one or another cosmopolite sources. Forty per- cent became aware via localite sources. Nine percent were unaware of the innovations. The partial correlation between attitude toward innovations and information level is .62; i.e., 36 percent of the variance is accounted for. The measure of attitudes is undoubtedly weak, in that it would appear to, too directly, measure information about the innovations. The extent to which attitude toward innovations influences the findings can be demonstrated otherwise. The multiple correlation of all predictors with information level is approximately .85 which is to say approximately 72 percent of the variance is explained. The attitude-information level relationship alone accounts for one half of that (36 percent; partial r = .62). Bivariate Relationships: Innovativeness Only one social system characteristic is slightly related to innovativeness, namely, constant information inputs. The partial correlation (.16) while significant 145 accounts for only three percent of the variation in innovativeness. As a general predictoriflfisutility would seem dubious. The partial correlation between variety of sources exposed to and innovativeness is .30, thus nine percent of variance is accounted for. Lack of exposure to mass media accounts for the little variation that the mass media exposure variable explains. Thus, in general the best predictor of innovativeness, among the environmental participation variables, is the number of different sources to which one has been exposed. This finding is consistent with Copp (1956), Emery and Oeser (1958) and Fliegel (1956), and thus, adds some generalizability to the relationship. The partial correlations between both attitude toward innovations and information level with innovative— ness are .21 and .16, respectively. While both are significantly they only account for approximately four and three percent of the variance, reSpectively. Once again, however, the high correlation between these two independent variables may account for their low partial correlation with innovativeness. The multiple correlation of all the independent variables with innovativeness is .75. Thus thses nine variables account for approximately 56 percent of the variation in innovativeness. Squaring and summing the partial correlations for innovativeness results in the explanation of 25 percent of the variance 1H6 (specific). The comparison of this figure with the squared multiple correlation as well as inspection of the differences between the zero order and partial correlations indicates these nine variables have considerable "in common". Multivariate Relationships: Information Level Figure 3 is a configurational representation of predictors of individual information level. Of the ten variables included in the analysis, six are retained by the analysis in the derived configurations. 'These six variables account for approximately 72 percent of the variation in individual information levels. Based on the partial correlations concerning in- formation level (see Table 18) we would expect those re- lationships with the highest partial correlation (most variance accounted for), to be nretained in the analyses. Attitude toward innovations, variety of sources, intermittent inputs (negative) and cosmopolite sources at awareness, were most strongly related to information level, in that order. Figure 3 reveals that these variables, in fact, do feature in the configuration. In addition, some of less strongly related variables appear in the configuration. This can be taken as evidence that certain interactions may exist; i.e. such variables may be predictive for some types of respondents but not others. Interactions are also noted among the variables which were more strongly related to information level for instance, variety of sources 147 discriminates among the more knowledgeable reSpondents but not among the less knowledgeable ones. Degree of mass media exposure at awareness discrim- inates among the less knowledgeable. The reader is advised to recall the coding scheme employed for this variable. Those scored zero are unaware of the three innovations. Those scored one, attribute awareness to interpersonal sources. Those scored two, attribute awareness to the mass media. Thus subgroup # 6 consists of the 119 reSpondents who are unaware of the innovations. Subgroup # 7 consists of 337 of those who are aware regardless of the source of information (i.e., codes 1.0 to 2.0). Four information typologies were suggested in Chapter IV. Type # l, the reSpondents with the highest levels of information were characterized by very favorable attitudes toward the innovations and exposure to several sources of information. It was noted, previously, that variety of sources differentiated the more knowledgeable but not the less knowledgeable. The less knowledgeable include those who are unaware of the innovations and those who are aware, but have little information about them. These respondents would seem to lack the environ— ment, and sources in the environment, to acquire information. Such being the case we would anticipate that the diversity of sources to which one is exposed would not discriminate 148 among the less knowledgeable. For the more knowledgeable, an inverse relationship was found between intermittent inputs and information level. The adequacy of measure- ment and the conceptual utility of this variable has been questioned previously. The second type, those with moderate information levels, are also characterized by quite favorable attitudes toward the innovations. Some indicate having been con- vinced of the innovations' utility (N = 289) while others indicate they have not yet reached that point (N = 215). Those who say they have reached the conviction stage are slightly more knowledgeable than those who have not. Once again, the attitudinal variable may well be detrimentally affecting the results and interpretations. The fact that only one scale point difference in attitude corresponds ‘ to differences in information level is additional evidence in this regard. The third type are only slightly informed, in fact, they are not much more than aware of the innovations (subgroup means one 4.88 and 3.36). Less favorable attitudes, and to some extent awareness attributed to cos- mopolite sources, characterize this type. They may well be situated in less accessible places making contact by diffusion agencies more difficult and occurring later in time, than in less remote areas. They may also be more traditional individuals or later knowers and adopters. 149 The fourth type are those who are unaware of the three innovations. These are typically the "laggards", those who know last and adopt last, if at all. Consideration of the information level typologies must be tempered by the suspected effects of the attitudinal variable. These types are at best tentative and suggest that further studies and similar analyses be performed. InSpection of the configuration suggests that the individual's participation in his information environment is of considerable importance in explaining his level of information. The correlations between the social system variables and information level were low. This lack of strength plus the absence of most of the social system variables from the configuration suggests that the manner in which the individual structures his information environ- ment and the ways he participates in that environment are of importance in explaining and predicting information level. Participation in one's information environment may be characterized on various dimensions. One such dimension is the distance between place of origin and the source and the receiver. Another dimension may be the social, psychological and cultural differences that exist between sources and receivers. Yet another dimension is the intensity with which an individual either seeks information or orients himself to potential sources of information. The present data do not directly 150 consider all of these dimensions, however, they may prove useful for further research concerned with information environments. Multivariate Analyses: Innovativeness InSpection of the partial correlations for inno- vativeness suggests that constant inputs, variety of sources, attitudes toward innovations and information level are likely to appear in the explanatory configuration for innovativeness. From Figure 5 we note this to be the case. Variety of sources exposed to, which itself accounted for more variation in innovativeness than any other variable (9 percent) appears in almost every branch of the configuration. Constant inputs, on the other hand, discriminate only among the more innovative. Likewise, attitude toward innovations and information level tend to discriminate among the more innovative rather than the less innovative. Not being convinced by someone or some— thing, of the innovations' utility serves to initially describe the less innovative respondents. Such persons are likely to not participate in their information environ- ment or be information seekers. If such is the case, variables which reflect such participation are not likely to differentiate among the less innovative but should do so between the more and the less innovative reSpondents. The data generally fit such a pattern. 151 Four possible typologies of innovativeness are discernible. The first type, those who are most innovative, are characterized by (1) having been convinced of the innovations' utility; (2) having a very favorable attitude toward the innovations; and (3) being exposed to several sources of information. Among this innovative group, the most highly innovative individuals (N = 7) also, were from villages with high constant information inputs. How- ever, with a cell-size as low as those, such results are little more than interestingly Speculative and suggestive. Some evidence of use of cosmOpolite sources at the con- viction stage is found for some of the respondents. Thus, Type # 1 would appear to fit modern typology. Favorable attitudes toward change, high information seeking activity, the use of cosmOpolite sources and higher levels of in- formation in combination with a greater degree of innovativeness suggests a modern type of individual. Type # 2, or the moderately innovative individuals, exhibit greater variation in make-up than do the more innovative reSpondents. This second type, like the first type have been convinced of the innovations' utility. However, attitude toward innovations and variety of sources seem to be substitutes for each other. Those with more favorable attitudes are exposed to fewer sources while those with less favorable attitudes are exposed to a greater number of sources. These two apparently negatively relation— 152 ships would seem to be cases of substitutibility. In other words, one compensates for the other. Information levels appear to range from low to moderate. Thus, the moderately, innovative have, of course, passed the con- viction stage and have either very favorable attitudes toward the innovations or have been exposed to several sources of information. Information levels are low to moderate. The image presented is perhaps that of a transitional person, one who does not fit the modern stereotype but is significantly different from the traditional typology. The third type are only slightly innovative. Some have reached the conviction stage but are less favorable toward the innovations and have been exposed to fewer sources. Others have yet to be convinced though the tendency exists for these persons to seek cosmopolite sources. The third type appear to be more traditional. This may be a function of their environment, i.e., they may live in more remote villages where access to infor- mation and ideas is difficult. On the other hand they may represent the late adopters, those who are customarily wary of change. Type # 4 are the laggards and non-adopters. They have not been convinced of the advantages of innovations. They are generally exposed to few sources of information. ReSpondents in this typology represent the laggards or 153 traditional individuals who either adopt at some very late point in time relative to other members of the social system, or who seldom, if ever, alter their occupational behavior patterns. One might ask, "Why does information level appear to be a differentiating factor in the middle range of innovativeness as represented in the present analysis?" Failure to discriminate among the less innovative re- spondents is likely due to such a large proportion of these peOple being either unaware (N = 119) or having limited awareness and information. Approximately one- quarter (23 percent) of the sample possessed only three items of information or less about the three innovations. Now, one might ask why information level discriminates among the moderately innovative and not among the more highly innovative. Again, one can postulate variation in information levels as a possible explanation. From Figure 3 we note little difference in mean information level scores be- tween the two highest subgroups. In Figure 5 we note in— formation level as a discriminator does not appear in connection with the two highest sub-groups but rather, a social system variable constant inputs, appears in the configuration. Among the more highly innovative persons, levels of information, as measured, would seem to be very similar. Attitudes toward innovations are most favorable 154 and high levels environmental participation are in evidence. Since information levels between persons would appear to exhibit low variation, discrimination on this variable is less likely. In the moderately innovative range there would appear to be greater variation in information. Furthermore information level and innovativeness were relatively strongly related. Thus, where variation in information level occurs we would expect it to feature as a discriminating predictor. A question to be asked now is, ”Why does a social system variable appear in the configuration as a dis— criminator among the more highly innovative reSpondents?” It should be pointed out that the results in this regard are not entirely compelling. The size of the reSpective subgroups (Nos. 22 and 23) is small. At most, the findings are suggestive. In the present analysis the decisions to adopt or not adopt are individually based. They do not represent group decisions. This is not to say group pressures have been non-existant. Such pressures may well have been operating, whatever the nature of the reference groups. The relationship then, between social system characteristics and individual behavior may reflect individual differences in orientation to institutionalized information inputs. As social systems become increasingly more modern, the impact of the modern institutions may progress at a similar rate. That is to say, that it is unreasonable to 155 expect all segments of a social system to be similarly disposed to novel facilities in that social system. The development of Specialized roles and information input systems and their use, is dependent in part, upon the capacity of some members of a social system to utilize such facilities. For example, without the skills of literacy being present, media institutions may exert little influence upon members of a social system. The impact of such institutions may be delayed until such time as some of the members of the social system are able to participate in that segment of their information environment. Thus, effective utilization of institutionalized information input facilities, would seem to require that some portion of the social system, at least, be literate, educated perhaps, and physically and socially mobile. Prior to these conditions, modernizing institutions can be ex— pected to have little discriminatory impact upon members of a social system. However, once certain segments of the society can meaningfully participate in the system of information inputs, such inputs may exert both direct and indirect effects on members of the social system. At such a time information input institutions may operate more widely to facilitate or constrain innovative behavior. It may be the case that the present study does not sufficiently represent the continuum of traditional to modern (or modernizing) communities. Village selection 156 criteria dictated the selection of rural villages and villagers. Were one to select a sample of communities both rural and urban and utilize some apprOpriate body of information and behaviors (e.g., health, education, politics), it might be possible to more definitively assess the range and nature of the impact of social system characteristics upon both individual and collective decisions and behaviors. Implications for Research One of the directions which analyses of the type used in the present study leads to is that of multiple correlational analysis of predictors with the dependent variable. Using the derived configurations one could then sub-divide the sample into meaningful subsections of the configuration and multiply correlate those predictors in the various segments of the configuration with dependent variable. Such procedure would seem to assume a relatively stable and valid configuration. The configurations de- rived from the present analyses must be considered as tentative. The measure of attitude toward innovations is not entirely adequate, and thus renders the configurations somewhat suSpect. Relatively small subgroup Sizes in certain instances and unanticipated relationships suggest that further research be done to determine whether similar variables on similar respondents produce Similar configura— tions. Once one is relatively certain of the stability of configurations, multiple correlational analyses should be attempted. 157 The present analysis utilized only intra-individual variable, namely, attitude toward innovations. The inadequacy of the measurement of this variable has been discussed. Future research might consider additional variables. Such attributes as achievement motivation and tolerance for uncertainty or risk-taking might be fruit— :mlly included in subsequent investigations. Each of these variables would seem to bear some relationship to the manner and extent of one's information seeking or acquisition and the degree to which one is innovative. The apparent importance of the environmental participation variables suggests further investigation. In addition to considering the physical distance of sources from receivers it may prove useful to define the social distance between the sources and receivers, in the context of the innovation decision process. One would expect sources to differ for various types of receivers. One might hypothesize that the relationship between social distance of the source-receiver dyad will vary curvilinearly with innovativeness. The least innovative people will tend to utilize sources more like themselves (homOphily) than will moderately innovative individuals. The most innovative persons will also tend to utilize sources at a minimum social distance (i.e., cosmOpolite sources since they will tend to view themselves and be viewed as cos- mopolite persons. Moderately innovative persons will 158 tend to utilize sources seen at a greater social distance from themselves than will either the least or most inno— vative individuals. In addition to social distance, future research might well attempt to determine more conclusively the nature of the effects intensity of participation (e.g., differential degree of exposure to each of several sources) and the conditions under which, if at all, intensity Significantly affects level of information and innovative- ness. One area of the paradigm of information and moderni- zation requiring additional conceptual clarity is that of the social system characteristics. As has been mentioned, the viability of the intermittent-constant distinction needs clarification. In addition, the suggestion was made for some consideration of the broader concept of institutional development. Not only could this concept be investigated to ascertain its component parts and their relationship to information levels and innovativeness, but research might attempt to determine those conditions (e.g., level of modernization) under which information input facilities differentiate among reSpondents. The information-attitude relationship might be investigated within the framework of the innovation decision process. The relationships between information about a course of action and one's attitude toward that 159 course of action are not well understood. One might take, for instance, a situation in a developing society, such as Eastern Nigeria, where highly positive values are accorded to change, betterment and innovation where an innovation has failed. Noting the reasons or failure it would be interesting to determine the nature of the effects of negative feedback or information upon attitudes and the effects of such attitudes on further information seeking patterns and innovative behavior. The attitude-information relationship brings into question once again the concept of information—acquisition and evaluation formation as concommitant subprocesses eventuating in an information-behavior threshold. The data in the present analysis by virtue, partly, of the nature of the distinction of information level preclude consideration of the question of a threshold. It may be necessary to utilize some form of aided recall such as an information test, especially in circumstances where levels of information are low and the ability to recall information about objects and practices which are culturally alien, a re limited . Thus, there are various directions future research might take. However, the suggestion is made that such research be conducted within the framework of the in- formation-modernization paradigm in Figure 1. Both the bivariate relationships and the multivariate analyses 160 offer some degree of credence to the conceptual frame- ‘work. Additional research Should be directed to de- termining the efficacy of both the predictive and ex- planatory aspects of the paradigm. Implications for Practice The foregoing findings underline the importance of the interpersonal communication systems in developing societies with low levels of education and literacy. Thus, practice must be adapted in such ways as to maximize the effectiveness of the limited trained sources of in- ibrmation. In other words, means must be found to utilize extension personnel and other with suitable qualifications such that as much information is diffused to as many persons as possible. Several strategies might be con- Sidered. Perhaps the obvious strategy, though most difficult and time-consuming to achieve, is to markedly increase the numbers of extension agents in the field. Such a strategy is difficult because of limited training facilities and personnel and time consuming because of the level of com- petency required to ensure quality service. Thus, other strategies may prove to be of more immediate utility. One possible strategy is the development of radio farm forums. Rogers and Bebermeyer (1966) summarize much of the relevant research and writing on the communication 161 strategy. Radio forums combine the effects of external contact and its attendant competence with intimate par- ticipation at the locallevel. Though the acquisition and maintenance of the necessary facilities and equipment poses no small problem, the extension of the information environ- ment via the radio to the community and its people allows for a level of individual participation on a regular basis which for many would be nonexistant otherwise. Radio forums allow producers to bring the most com— petent thought available to large numbers of people at any one time. In addition to a regular fare of instrumental information, it is possible to make use of such strategies for Specialized services, such as impending crop hazards, etc. For some communities the advant of a radio forum might well mark the beginning of and the necessity for additional and related institutions which would serve to modernize the community. Roads may have to be improved to provide adequate service and tranSport and mail facilities must be provided. The evidence compiled and reviewed to date warrants consideration of radio forums as an important means of enlarging local information environments. Where the numbers of extension agents are too few to adequately serve a given area or areas, training and utilization of local personnel may serve to alleviate the information and motivational barriers to change. The 162 relationship in the present analysis between leaders' relay capacity and both information level and innovativeness suggests that village leaders and in particular informal or Opinion leaders may be instrumental individuals in the diffusion process. The channelling of change agents' efforts towards such persons may increase the range of the agents' effectiveness. In a Similar manner, techniques which provide some form of training and sensitization to change for selected villagers may be useful developmental strategy. Hapgood (1964), in outlining a program of "animation" in Senegal suggests such would be the case. A community designates certain individuals who are to receive some form of training regarding the communities needs. It may be agricultural methods, health practices, etc. Care is to be exercised such that, that which is taught is both com- prehendible and capable of being implemented within the village setting. Training usually takes place at Specially located centers which approximate village con— citions. The intent is to keep both thought and action as relevant as possible to the conditions as they exist in the participants villages. Once individuals within a community are established as informants, liason persons or opinion leaders, it may be possible, where they are literate, to provide written reference materials for use in group discussions, etc. 163 Such a strategy was in action in Eastern Nigeria when the Diffusion Project, the larger investigation from which the present data were obtained, was in operation. The project, in cooperation with the Ministry of Agriculture prepared a handbook of agricultural information. COpies were distributed to village discussion groups. Clarifi— cation, interpretation and other forms of assistance *were solicited by the group from the extension agents, when necessary. Once again, the emphasis is upon local participation and extension of the information environ- ment. Permanent acquisition of the reference material, in essence a form of Specialized media, assures ready ‘ availability of information, the lack of which may often result in lack of interest and motivation in change. These few strategies are not intended to be exhaustive, but rather to point out how information might conceivably be extended to the village level with a minimum of additional strain placed upon the already heavily-burdened extension agent. 10. BIBLIOGRAPHY Ackoff, R. L. (1958), "Towards a Behavioral Theory of Communication," Management Science, 4, 218—234. Brim, O. G. Jr., and others (1962), Personality and Decision Processes. Stanford, California: Stanford University Press. Campbell, H. L. (1959), Factors Related to Differential Use of Information Sources, M.S. Thesis, Ames, Iowa State University. Copp, J. H. (1956), Personal and.Social Factors Associated with the Adoption of‘Recommended Farm Practices Among Cattlemen,lManhattan,Kansas Agricultural ExperimenffStation Technical Bulletin 83. Copp, J. H. and others (1959),'The Function of In- formation Sources in the Farm Practice Adoption Process;'Rural Sociology, 23, 2, 146-157. Dervin, Brenda (1967), A Critical Review of Research Relating Attitudes and Attitude Change to Prior Information, Information Gain and Retention, Unpublished paper, Department of Communication, Michigan State University, East Lansing. Deutschmann, Paul J. (1963), "The Mass Media in an Underdeveloped Village,‘ Journalism Quarterly, 40, 27-35- Deutschmann, Paul J. and Fals Borda, Orlando (1962), Communication and Adoption Patterns in an Andean Village, Programa Interamericano'de Informacion Popu ar, Mimeo Report, San Jose, Costa Rica. Emery, F. E. and Oeser, O. A. (1958), Information Decision and Action, New York, Cambridge—Uni- versity‘Press. . Festinger, Leon (1957), A Theor¥ of Cognitive Dissonance, Evanston, no S, Row, Peterson and Co. 11. Fliegel, F. C. (1956), "A Multiple Correlation Analysis of Factors Associated with Adogtion of Farm Practices," Rural Sociology, 21: 2 4-292. 164 165 12. Frey, F. W. (1964), "The Mass Media and the Peasant," A paper presented at the Association for Education in Journalism, Austin, Texas. 13. Hapgood, D. (1964), "Rural Animation in Senegal," International Development Review, 6:15—18. 14. Hovland, C. I. (1959), "Reconciling Conflicting Results Derived from Experimental and Survey Studies of Attitude Change." American Psychologist, 14, 8-17. l5. Jacobson, Eugene (1967), "An Approach to Measuring Communication Environments,I A paper presented to the American Psychological Association: Annual Convention, Washington, D.C. 16. Johnson, Glenn L. and others (1961),.A.Study of Managerial Processes of Midwestern Farmers, Ames, Iowa, The Iowa StateGUniversity Press. l7, Katz, Daniel (1960), "The FUnctional Approach to the Study of Attitudes," Public Opinion Quarterly, 24, 2, 163-204. 18. Lerner, Daniel (1958), The Passing of Traditional Societ , Glencoe, Illinois, The Free Press of GIencoe. , l9. Leuthold, F. O. (1960), Demonstrators and the Diffusion of Fertilizer Practices, Paper presented at the Rural Sociological Society, University Park, Pennsylvania. 20. Mason, Robert G. (1964), "The Use of Information Sources in the Process of Adoption," Rural Sociology, 29, 40-52. 21. Morgan, J. N. and others (1966), Productive Americans, Survey Research Center Monograph 43, Institute for Social Research, University of Michigan. 22. North Central Rural Sociology Subcommittee for the Study of Diffusion of Farm Practices, (1955), How Farm PeOple Accept New Ideas, Ames, Iowa Agri-_—_ culturalIExtension Service Special Report 15. 23. Neurath, Paul M. (1960), Radio Farm Forum in India, Delhi, Government of India—Press. 24. Pye, Lucien W. (1961), Communications and Political Develgpment, Princeton, N.J., Princeton University Press. 25. 26. 27. 28. 29. 30. 31. 32. 33- 34. 35. 36. 37. 166 Rogers, Everett M. (1967), The Diffusion of Innovations, New York, Free Press. Rogers, Everett M. (1966), "Mass Media Exposure and Modernization and Colombian Peasants," Public Opinion Quarterly, 29, 614-625. Rogers, Everett M. with Shoemaker, F. Floyd (1968), Communication of Innovations: A Cross-Cultural Approach, New York, FreeFPreSS of Glencoe. Rogers, Everett M. and Beal, George M. (1958), "The Importance of Personal Influence in the Adoption of Technological Changes," Social Forces, 36, 329- 335- Rogers, Everett M. and Bebermeyer, James P. (1966), Mass Media and Interpersonal Communication in National Development, Working Paper 16, AID Diffusion Project, Department of Communication, Michigan State University, East Lansing. Rogers, Everett M. and Me nen, W. L. (1965), "Communi- cation Sources for 2, -D Weed Spray among Colombian Peasants," Rural Sociology, 30, 2, 213-219. Rogers, Everett M. and Pitzer, R. L. (1960), The Adoption of Irrigation by Ohio Farmers, Wooster, Ohio Agricultural Experiment Station Research Bulletin 851. Ryan, B. and Gross, N. C. (1943), "The Diffusion of Hybrid Seed Corn in Two Iowa Communities," Rural Sociology, 8, 15—24. Schultz, Theodore W. (1964), Transforming Traditional Agriculture, New Haven, Yale University Press. Sears, David O. and Freedman, J. L. (1967), "Selective Exposure to Information: A Critical Review," Public Opinion Quarterly, 31, 2, 194-213. Shannon, Claude E. and Weaver, Warren (1949), The Mathematical Theory of Communication, Urbana, The University of Illinois Press. Smith, M. B. and others (1956), Opinions and Person- ality, New York, John Wiley and Sbns. Sonquist, J. A. and Morgan, J. N. (1964), The Detection of Interaction Effects, Survey—Research CenterFMonograph NF. 35, Ann Arbbr, Institute for Social Research, The University of Michigan. 167 38. Spector, Paul and others (1963), Cpmmunication and Motivation in Community Development: An Experiment. Institute fOr International BerVICes, The American Institute for Research, Washington, D.C. 39. Uchendu, Victor C. (1965), The Igbo of Southeast Nigeria, New York; Holt, Rinehart and Winston. 40. Walker, Helen M. and Lev, Joseph (1953), Statistical Inference, New York; Holt, Rinehart and WinSton, Inc. 41. Westley, Bruce and MacLean, Malcolm Jr. (1957), "A Conceptual Model for Communications Research," Journalism Quarterly, 34, Winter, pp. 31—38. APPENDIX A LIST OF INNOVATIONS STUDIED 169 List of Innovations Studied in Phases I and Phase II in Eastern Nigeria. Name of Innovation Percentppafsgepple figgggegl l. Fertilizer 22 39 2. Oil palm (rehibilitatun scheme) 2O 17 3. NS-l maize (hiphridvanety) 16 32 4. Aldrin dust (insecticule) 12 33 5. Poultry (improved breeds) 6 l9. 6. Rice 5 8 7. Rubber 5 4 8. Vegetables 5 7 9. Cocoa 4 5 10. Community Plantations 4 13 ll. Cerssava (improved variety) 2 7 12. Livestock l N.A. 13- Cashew l 3 l4. Pineapple * N.A. 15. Citrus * 10 16. F.A.I.D * 2 17. Farm settlements * N.A. l8. Stork oil press * N.A. Joined no programs 53 37 * Less than one percent. N.A. Not studied in Phase II. 1. Data are taken from G. D. Hursh pp, al., "Phase L Preliminary Report of Selected Descriptive Findings for Leaders and—Progressive Farmers in 71 Eastern Nigerian Villages“, Economic DeveIopment InStitute, University oleIgeria, Enugu, March, 1967; and G. D. Hursh, et. al., "Phase II Preliminary Report of Selected D3?Eriptive“Findingsfor 1,347 Rural ‘ Farmerpwand Innovators inil8'EaStern Nigerian vmnlages,FfiEconomic Development Institute, Uni- versity of Nigeria, Enugu, April, 1967. APPENDIX B Se lec ted Items From PHASE II: FARM FAMILY SCHEDULE 171 46. Did you read (did anyone read to you) any neWSpapers during the past four weeks (How many?) O--no papers read/read to him l--can't read, but had papers read to him (No. ) 2--can read and read one or more papers (No. ___) 47. Did you listen to the radio in the past two weeks (How many days?) /CIRCEE7 O--no/none/didn't listen 2--1istened one or more days (RECORD NUMBER ) 48. Did you see any cinema films during 1966? (How many) [2239;57 O--no/none/didn't see 2--Saw one or more films (RECORD NUMBER:___) 51. During 1966 did you: EXPOSURE SOMEONE FARMING ACTIVITY IN 1966 yes:gk;pp_ yes:gk;pp. yes:gk:pg a. read (have read to) agric news- letter/pamph. 2 l O 2 1 O 2 l O squamSAOJd -mt otgtoeds b. see agric demonstr. 2 l O 2 l O 2 l O c. hear agric radio prg. 2 1 O 2 1 O 2 1 O d. see agric cinema 2 1 O 2 l O 2 l O e. attend agric lecture 2 l O 2 1 O 2 l O f. talk with ag. officer 2 l O 2 l O 2 l O ethoseq 172 o m ”ma 0 m "0H 0 m mNHmE H-mz Seema mH o m "mH o m .mH o m ease emmmHa NH o N "ma 0 m .0H m HmNHHHpHmm HH 0 m "0H 0 m .ma 0 m memoa so pHeeee oHam .0H 0 m "0H 0 m "mH o m mQOHeeeeeHe szGSEEoo .m o m 4mm. o m "ma 0 m msoonQo omhw< .m o m "0H 0 m “mH o m zmnmeo \msspHo oHsm< .5 o m "ma 0 m ”ma 0 m moomm mHQmP Immo> oflsw< Mm o m ”ma 0 m .ma 0 m m>wmmwo oflsm< .m o m "ma 0 m .ma 0 m ooHs 0Hmm< .H o m "0H 0 m ”ma 0 m oEoSom thQSM 0st< .m o m "0H 0 m ”0H 0 m oEogom moooo low< .m o m "ma 0 m ”ma 0 m oEonom .QSSms , sHee HHo .H .opo arranges oewmms \eHep\eoHe .eem\eeHee. eeo\emmemeem em mflfl Emam comb nmnpmeoemu \mwmaHH> CH \soum>occH \ocmHsm zm mm: ma? Hm>m \memcflo mooo\c30p \smehmw\aoo COm\am£posn EmmHm 2302M m2 SH [Hmmo .oHawe \e>HemHee mm 50h UHQ .Hm eommmoo m.0SZH mHm mOHm3 mom m2