”v -"..~v '! 1 v 'I‘A ’ . .J‘L -. {hwmnul¢‘-0 HI" u I ’ “V. ‘0’! to . .- u 11". . !I| '11.. u . n‘llq ‘Imnmoltmlfi. ,.m.l......uw...«. Q nu"! .| «km-Wig. .- 21. at um)..m..&m.uxfirfi.s .! 8 tag . . wig." .. sh! ‘I . a It, A ,Yl.‘ .Al. ’ ‘ 1’ ‘ tl . .- I.IQ . I».A.... VH1 -u Jul ‘ i 41|1¢i|f ‘1‘11 1 U v) [3‘ .‘l ’01! 00‘ . .( .3kvfl yin“ Y; : utc.‘ 5“ .v to; ov‘nwv‘ may . QIIJ 3.1.1. .3 ‘ ‘1‘“... VII. Dill” . .’ I'lvhlflluuft. JAI L431}- :HH‘E. ’I‘ICI' I I‘ll... I) ’03-‘71 I, o‘clbuwulsil . ‘\ i quG‘k 4.5; n u If?! .IELUILQ.‘ 1).- ‘tin 1.: 2‘ .-. . ‘» LLLK . . a........3 . . .V....£.....c..1o.n. ~ . 591. '2‘ -'K..’ '\ LIBRARY Michigan State University This is to certify that the dissertation entitled A Reconceptuaiization of the Innovativeness Construct presented by Wilhelmus Paulus Burgers has been accepted towards fulfillment of the requirements for Doctor of Philosophy degreein Marketing R. Dale Wilson Major professor Date January 19, 1988 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 MSU * LIBRARIES .—:—. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below.‘ ./ .L‘J’ (:3 U A RECONCEPTUALIZATION OF THE INNOVATIVENESS CONSTRUCT BY Wilhelmus Paulus Burgers A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Marketing and Transportation Administration 1988 ABSTRACT A 'RECONCEPTUALIZATION OF THE INNOVATIVENESS CONSTRUCT BY Wilhelmus Paulus Burgers This dissertation offers a reconceptualization of the innovativeness construct. The reconceptualization is derived from and presented in response to a broad range of criticisms and concerns regarding diffusion research found in the behavioral diffusion and consumer behavior literatures. The diffusion modeling literature conceptualization of innovativeness is integrated with contributions from the behavioral diffusion and consumer behavior literatures to form the reconceptualized innovativeness construct. Validation of the new construct takes place through con- struction and testing of a causal model which relates the construct to additional consumer behavior theoretical constructs. Specifically, the model relies on situa- tional variables (namely, enduring involvement and situational involvement) to explain innovativeness. The model is tested in three product-market set- tings: personal computers, restaurants, and movie theater attendance. The overall findings support the dissertation's central contention that innovativeness is a function of situational variables (i.e., situational and enduring involvement) and is not a personality trait. Overlap between innovativeness and opinion leadership is also examined. Hypotheses are presented and tested on the nature and direction of such overlap. ' The findings are shown to have potential implica- tions for new product (concept) testing and new product introduction strategies. Suggestions are also provided for future research on diffusion of innovations. To my father, Johannes Arend Burgers, and my mother, Wilhelmina maria Burgers van Eijk. iv ACKNOWLEDGMENT I would like to express my sincere gratitude and appreciation to Dr. R. Dale Wilson for his support, encouragement, and insistence on perfection during the course of this research and preparation of this manuscript. I would also like to thank committee members Drs. David K. Smith and Jack Allen for their time, comments, and helpful suggestions. In addition, Drs. Dale Duhan and Thomas Page generously helped me out of a few tight spots. My thanks and apologies also go to my perfectionist secretary Mrs. Joyce Stall who typed so many "final" versions of this product. Next, I wish to express my gratitude to the many people who took the time and trouble to fill out the questionnaire. Finally, a note to my son Craig who believes I build airplanes in my office: Craig, daddy's plane is done. TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . . Chapter I Introduction and Background . . . . . . . . . INTRODUCTION . PURPOSE . . . . CONTRIBUTIONS . ORGANIZATION‘-. II Literature Review . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . Section I. The Classical Paradigm . . . . ORIGINS . . . . . . . . . . . . THEORETICAL UNDERPINNING . . . . THE CLASSICAL PARADIGM IN THE MARKETING LITERATURE . . . . . . .g. . . . . EMPIRICAL RESULTS: SUMMATION . . . . . EMPIRICAL RESULTS: EVALUATION . . . . . Section II. Changing Perspectives. . . . . CONCEPTUALIZATION IN THE DIFFUSION MODEL- ING LITERATURE . . . . . . . . . . . . RECONCEPTUALIZATION IN THE CONSUMER DIF-. 'FUSION LITERATURE . . . . . . . . . . . . SUMMARY . . . . . . . . . . . . . . . . . vi Page 00000. H m 25 28 33 Chapter III Theoretical Framework and Hypotheses . . . . . Section I. Theoretical Framework . . . . . . DEFINITION OF INNOVATIVENESS . . . . . . . METATHEORETICAL PERSPECTIVES . . . . . . . Section II. Hypotheses . . . . . . . . . . . ENDURING INVOLVEMENT AND INNOVATIVENESS . SITUATIONAL INVOLVEMENT AND INNOVATIVENESS HEAVY USAGE, ENDURING INVOLVEMENT, AND. . EARLY ADOPTION . . . . . . . OPINION LEADERSHIP AND INNOVATIVENESS S MARY I O O O O O O O O O O O C O C IV 0 Me thOdo logy O O O O O O O O O O O O O O O O O 0 Section I. Data . . . . . . . . . . . . . . SWLE O O O O O O O O O O O O O SELECTION OF PRODUCT-MARKET SETTINGS . . . RESEARCH INSTRUMENT . . . . . . . . . . . . Section II. Variable Operationalization . . INNOVATIVENESS . . . . ENDURING INVOLVEMENT . SITUATIONAL INVOLVEMENT OPINION LEADERSHIP . . EARLY ADOPTION . . . . Section III. Analysis . . . . . . . . . . . HYPOTHESES H1 THROUGH H5 . . . . . . . . . HYPOTHESES H6 THROUGH H7 . . . . . . . . . V. Results . . . . . . . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . . Section I. Summary Results . . . . . . . . . SAMPLE . . . . . . . . . . . . . . . . . . RELIABILITIES . . . . . . . . . . . . . . . VALIDIN O O O O O O O O O O O O O O O O 0 vii Page 000000 uh ‘0 O O O O O m C . 75 Chapter VI. Section II. Hypotheses Hl through H5 . . . . RESULTS 0 O O O O O O O O O 0 0 DISCUSSION . . . . . . . . . . Section III. Hypotheses H6 and H7 RESULTS 0 O O O O O O O 0 O O 0 Conclusion . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . THEORETICAL CONTRIBUTIONS . . . ADDITIONAL THEORETICAL IMPLICATIONS FUTURE RESEARCH DIRECTIONS . . . . MANAGERIAL IMPLICATIONS . . . . . . CONCLUSION . . . . . . . . . . . . APPENDI CES O O O O O O O O O O O O O O 0 Appendix A. Pretest Questionnaire: Appendix B. Pretest Questionnaire: Appendix C. Pretest Questionnaire: Appendix D. Questionnaire . . . . Computers . Restaurants Movies . . Appendix E. Carlson-Grosbart Innovativeness Scale . . . . . . . . Appendix F. Zaichowski's Involvement Scale . . Appendix G. Major Movies of 1986 . BIBLIOGRAPHY O O O O O O O O O O O O O 0 viii Page . 93 99 .103 .104 .106 .108 .110 .115 .120 .126 O 134 .136 .137 .138 Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table II-1. IV-1. IV-2. IV-3. IV-4. IV-5. V-l. V'Z. V-3. v-‘4. V-S. V-6. V-7. V-8. V-9. V-lO. v-11 0 LIST OF TABLES Early Adopter Characteristics 18—19 Innovativeness Scale 58 Enduring Involvement Scale 60 Pretest Situational Involvement Scale 61 Pretest Opinion Leadership Scale 64 Model Equations 69 Reliabilities 76 Analysis of Variance - Innovative- ness by Product Category (3 groups) 78 Innovativeness - Confirmatory Factor Analysis 79 Analysis of Variance - Situational Involvement by Product Category (3 groups) 80 Parameter Estimates for Personal Computers 83 Cross-Sectional Approach: Parameter Estimates for Movie Theater Attendance 86 Direct Approach: Parameter Esti- mates for Movie Theater Attendance 89 Parameter Estimates for Restaurants 91 Parameter Estimates for Hypotheses H1 and H2 Tested Across the Three Product Categories Combined Results: Hypotheses Hl-HS Means and (Standard Deviations) ix 92 93 94 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure I-l. II-l. III-1. IV-1. IV-2. v-1. V-Z. VI-l. LIST OF FIGURES A Causal Model of Innovativeness Adopter Categories A Causal Model of Innovativeness LISREL Model of Innovativeness Cross-Tabulation of Innovators and Opinion Leaders Structural Model Results for Personal Computers Cross-Sectional Approach: Structural Model Results for Movie Theater Atten- dance Direct Approach: Structural Model Results for'Movie Theater Attendance Structural Model Results for Res- taurants Structural Model Results: Hypotheses H1 and H2 Tested Across the Three Product Categories Cross-Tabulation of Innovators and Opinion Leaders Enduring and Situational Involvement 14 44 68 71 82 85 88 90 92 97 107 Chapter I Introduction and Background INTRODUCTION Innovativeness is the central construct in several areas of marketing theory. First, according to Hirschman (1980), few constructs are as important to consumer behavior as innovativeness. The dynamic nature of the market place is a direct consequence of consumers' innovativeness. Second, assumptions on innovativeness underlie implicitly or explicitly all models in the diffusion modeling literature (e.g., Muller and Mahajan 1979; Fourt and Woodlock 1960; Mansfield 1961; Bass 1969). Third, to an extent new product management theory is, or should be, based directly on a general theory of innovative behavior (Midgley 1977, p. 161; see also Kleyngeld 1974). The current classical behavioral conceptualization of the innovativeness construct however has, as this dissertation will argue, hindered progress in diffusion research in marketing. From a managerial perspective, it must be pointed out that the great majority of prOducts, at any point in time, has a strictly limited life Span (Midgley 1977, p. 278). The continuous successful introduction of new products is thus essential for the very survival of many firms. Yet, failure remains common in the introduction of new products. Estimates of failure rates for new products, given an identified opportunity, range from 70% for industrial products to 80% for consumer products (Urban and Hauser 1980, p. 54). Hence, ample room exists for improvement in the practice of new product introduction. Currently, diffusion theory receives only a rather perfunctory treatment in the new product management literature (e.g., Urban and Hauser 1980) barring a few exceptions (e.g., Midgley 1977; see also Robertson 1971). An increase in the understanding of innovators and innovativeness, that would lead to greater applicability of diffusion theory to the theory and practice of new product management and introduction, will likely prove to be particularly helpful to managers in todayis fast-paced, increasingly competitive business environment. PURPOSE This dissertation, in an attempt to achieve such increased understanding, offers a reconceptualization of the innovativeness construct. This reconceptualization is designed to enhance both the theoretical meaningfulness and the managerial relevance of the innovativeness construct. Currently, innovativeness is defined in the classical behavioral diffusion literature as "The degree to which an individula or other unit of adoption is relatively earlier in adopting new ideas than the other members of a social system" (Rogers 1983, p. 22). This classical definition suffers from what Rogers (1976, 1983) has termed the "pro-innovation bias" that characterizes much of the diffusion literature. The classical definition implicitly assumes that "new ideas" are desirable and should be adopted. If an individual does not adopt early, he or she is not innovative according to this definition. However, individuals may not adopt early, or may reject an innovation altogether, for reasons other than a lack of innate innovativeness. For example, an innovation may be considered by an innovator and found to be inadequate and not an improvement over current offerings. Beyond possible inadequacy of the innovation, any number of situational variables (e.g., financial considerations) may also delay or prevent adoption of a given innovation by an otherwise innovative individual (Midgley 1977; Midgley and Dowling 1978). The current operational definition of innovativeness thus confuses the trait of innovativeness with adoption behavior (Midgley and Dowling 1970; see also Peterson 1973). In this dissertation, the innovativeness construct is defined as a continuous variable as follows: The degree to which an individual makes innovation— decisions independently of the decisions of other individuaISvin his or her social system. The term "innovation-decision", introduced by Rogers and Shoemaker (1971), implies that both adoption and rejection by innovator and non-innovator alike may occur upon consideration of the innovation. An individual's social system comprises co-workers , neighbors, relatives, and friends (Childers 1986) with whom the individual interacts, also independently of consideration of the innovation. The criterion then that is used to distinguish innovators from non-innovators according to the conceptualization in this dissertation is the independence of the innovator's decision making. Diffusion models (e.g., Bass 1969; Muller and Mahajan 1979) currently are based on this criterion. In addition, in the consumer behavior literature, Midgley (1977), Midgley and Dowling (1978), Hirschman (1980), and Gatignon and Robertson (1985) endorsed this criterion of independence in decision making. The new definition, which integrates contributions from the diffusion modeling and consumer behavior literatures, is presented as a response to and is derived from a broad range of concerns and criticisms found in the literature. These concerns and criticisms may be categorized as follows: 1. Methodological issues raised by Rogers (1976). These include the absence of consideration of causality, the pro-innovation bias found in most studies, and the general lack of a process orientation in diffusion research. 2. The inconsistency and or weakness of empirical findings in the literature on innovativeness and innovators (Robertson 1971; Kohn and Jacoby 1973; Downs and Mohr 1976; Taylor 1977; Midgley and Dowling 1978; Gatignon and Robertson 1985). 3. The lack of integration of the behavioral diffusion literature with the theory and practice of new product (concept) testing (Kleyngeld 1974) and new product management (Midgley 1977). 4. The lack of integration of the behavioral and modeling diffusion literatures (Gatignon and Robertson 1985). S. The need for advances in consumer diffusion theory beyond merely applying concepts from the general diffusion literature (Gatignon and Robertson 1985). The latter two concerns, or suggestions, provided the direction which this dissertation has taken in its search for an answer to the concerns and criticisms aimed at methodologies, empirical results, and managerial relevance of diffusion research. Hence, through integration of the behavioral and diffusion modeling literatures and by drawing upon consumer behavior theory, it is proposed that methodological and empirical weaknesses may be resolved. The successful resolution of these weaknesses should result in an increased potential for managerial relevance of consumer diffusion theory. For example, the key concern in new product management and new product (concept) testing is whether innovators will adopt a new product. The current classical conceptualization of innovators as "early adopters" by definition precludes consideration of that question. The focus, therefore, of the dissertation's efforts at integration and reconciliation of prior theoretical and empirical contributions and criticisms thereof lies in its reconceptualization of innovativeness. Thus, the contention of this dissertation is that the key issue underlying the above concerns and criticisms involves the classical operational definition of innovativeness in the behavioral diffusion literature. Validation of the new innovativeness construct will take place through construction and testing of the causal model depicted in Figure I-l. This model shows the variables, enduring and Situational involvement, that are hypothesized to influence the individual's innovativeness. Enduring involvement has been defined as "the ongoing concern with a product the individual brings into a purchase situation" (Bloch and Richins 1983, p. 71). Situational involvement has been defined as "the degree of involvement evoked by a particular situation, such as a purchase occasion" (Bloch and Richins 1983). These variables, enduring and situational involvement, do not represent personality traits of the consumer. In Belk's (1975) classification of situational variables, these would be categorized as "task definition" variables. This focus on Situational variables represents a significant departure from tradition in the diffusion literature. The classical approach conceives of innova- tiveness as a personality trait, defined as a "persisting characteristic by which individuals can be distinguished from one another" (English and English 1958; Wolman 1973; see Midgley and Dowling 1978). Thus, according to the classical approach, some individuals will be innovative all the time, and research efforts have been directed at the identification of various characteristics of such innovative individuals. The model proposed in this Enduring Situational ”$099 + ‘ Involvement Involvement Early Adoption + .nnovativenes Figure I-1. A Causal Model of Innovativeness dissertation on the other hand suggests that all consumers can sometimes be innovative and therefore endeavors to investigate situational variables that do or do not lead to innovative behavior. Early adoption and usage are also included in the model, linking the alternative approach of this dissertation to traditional diffusion research. Usage, it may be noted, has consistently been positively related to early adoption and is hypothesized to also be related positively to enduring involvement. Specifically, the following hypotheses will be tested in the context of the causal model depicted above: H1: Enduring Involvement will have a positive impact on innovativeness H2: Situational involvement will have a negative impact on innovativeness H3: Usage will have a positive impact on enduring involvement H4: Usage will have a positive impact on early adoption H5: Innovativeness will have a positive impact on early adoption Further (nomological) validation of the reconceptualized innovativeness construct is attempted through the application of the new construct in a related area. The relationship between early adoption and opinion leadership has been extensively investigated in past research. According to the literature, opinion leadership tends to be a characteristic of early adopters and early adoption tends to be positively related to opinion leadership (King and Summers 1970; Summers 1971; Baumgarten 1975; Engel, Kegereis, and Blackwell 1969). According to Baumgarten (1975), "this very substantial similarity between opinion leaders and early adopters leads to the question of the extent to which opinion leaders are. early adopters, and vice versa". This dissertation addresses that question, be it in the context of the reconceptualized innovativeness construct rather than early adoption, through the investigation of the following two hypotheses: H6: All innovators are likely to be opinion leaders H7: Not all opinion leaders are necessarily innovators In other words, it is expected that innovators are a subset of opinion leaders. By addressing separately the degree of opinion leadership among innovators and the degree of innovativeness among opinion leaders, (stronger) results can be obtained than would be indicated by simple correlation of the two constructs. Specifically, a weak relationship would obtain if a large proportion of opinion leaders were found to be non-innovative, even if meanwhile innovators consistently were high in opinion leadership. CONTRIBUTIONS Hence, the major contributions that this dissertation intends to make to consumer diffusion theory may be stated 10 formally as follows: 1. A reconceptualization of the innovativeness construct in response to criticisms and concerns in the literature. 2. The building and testing of a causal model that explains innovativeness on the basis of situational rather than personality variables. 3. Refinement of the opinion leadership construct. In addition, the following contributions derived directly and or indirectly from the first two major contributions may be listed: * The introduction of causality in the specification of relationships between relevant constructs in the diffusion of innovations. * The reconciliation of conflicting and or weak empirical results through the reconceptualization of innovativeness on the basis of explanations in the diffusion literature for these conflicting results. * The integration of the behavioral assumptions underlying diffusion modeling theory with the behavioral diffusion literature through the reconceptualization of the innovativeness construct on the basis of these behavioral assumptions. * The advancement of consumer diffusion theory beyond mere application of diffusion theory in consumer behavior through the use of consumer behavior theore- tical constructs in the explanation of innovativeness. A lack of integration of the behavioral diffusion literature and the theory and practice of new product (concept) testing and new product management presents the final challenge in this dissertation. Ultimately, good theory should lead to good practice. If this dissertation is to be considered successful, it will have to Show how 11 and why its theoretical contributions will lead to good or better practice. Hence, implications for the theory and practice of new product (concept) testing and new product management will be explored in the final chapter on the basis of the results of this dissertation. ORGANIZATION The balance of the dissertation is presented in Chapters II through VI. Chapter II contains the literature review. Chapter III develops the theoretical framework underlying the propositions to be tested. Chapter IV deals with the research design. The methodology and a framework for data collection are developed in this chapter. Topics include scale development, tests for reliability and validity, data analysis, and survey procedures. The main findings and any supplementary findings are dealt with in Chapter V. In Chapter VI the conclusions are summarized and theoretical and managerial implications are explored. Suggestions for further research are also included. Chapter II Literature Review INTRODUCTION The literature review is contained in two separate sections. The first section surveys theoretical and empirical contributions that have been made in the context of what may be called the classical paradigm of diffusion research. The origins and theoretical underpinning of the classical conceptualization of innovativeness and attendant research methodologies are examined first, followed by a summation and evaluation of the collateral body of empirical evidence, generated in the course of over four decades of diffusion research. The second section of the literature review investigates several alternative conceptualizations of innovativeness. These include the conceptualization of innovativeness that underlies the diffusion modeling literature and recent alternative conceptualizations that have been developed in the consumer behavior literature. Section I. The Classical Paradigm ORIGINS The classical conceptualization of innovativeness traces its theoretical roots to the rural sociology literature. Specifically, the classical conceptualization and attendant research methodologies can be traced to the seminal article 12 13 by Ryan and Gross (1943) on the diffusion of hybrid-seed corn among Iowa farmers. ‘Their study provided the central paradigm for subsequent diffusion research in a variety of disciplines including marketing (Rogers 1983, pp. 51-55; Robertson 1971, pp. 22-23). The following, brief description of the Ryan and Gross study serves to illustrate this central paradigm. Respondents in the study, 259 farmers in two small Iowa communities, were interviewed about when they decided to adopt hybrid-seed corn. This time of adoption from the year of introduction of hybrid-seed corn became the main dependent variable in the study. In addition, respondents were asked about their education, age, income, Size of operations, sources of information on the innovation, and other variables. Following the collection of these data, the relationship between time of adoption and all other variables was examined. The publication of this study led, in a variety of disciplines, to a virtual explosion of research effort on the diffusion of innovations. Further theoretical and methodological development, however, remained extremely limited (Rogers 1976). THEORETICAL UNDERPINNING As regards subsequent theoretical development, one may point to Rogers (1962; Rogers and Shoemaker 1971; Rogers 1983) as having almost singlehandedly defined the field of diffusion research. Among the many contributions by Rogers to the field of diffusion research is his classification of 14 members of a population of potential adopters into separate categories. Five adopter categories were defined by Rogers, including: innovators, the first 2.5% of a population to adopt; early adopters, the next 13.5%; early majority, the next 34%; late majority, the next 34%; and laggards, the remaining 16%. This classification, in conjunction with the normal curve depicting diffusion over time (see Figure 11-1), became widely accepted. Innovators Early Earlg Late Adapters Majority Majoritq Laggards 1 3 5% 34% 34% 1 6% Figure II-1. Adopter Categories This presentation of the normal diffusion curve signified an important theoretical development in the classical conceptualization of the innovativeness construct. The assumption of normality for the diffusion curve was based on an explicit new assumption on the nature of innovativeness, heretofore synonymous with early adoption behavior. The degree to which an individual would be relatively earlier in adopting an innovation was now 15 assumed to be a function of the speed with which that individual would be able to complete some sequence of information processing activities that presumably precede adoption. The classificational scheme was not meant to imply that there be pronounced breaks in the innovativeness continuum between these categories. For example, early adopters were not considered to be "non-innovative" but rather "less innovative than innovators". Innovativeness thus became conceptualized as a human trait, akin to traits such as human intelligence or the learning of information (Rogers 1983, p. 244). This conceptualization of innovativeness as a human trait rather than behavior became part of the classical paradigm widely adopted by diffusion researchers (Midgley and Dowling 1978). Yet, the accepted operational definition of innovativeness, though conceptualized as a human trait, is behavioral (viz., "the degree to which an individual or other unit of adoption is relatively earlier in adopting new ideas than the other members of a social system" Rogers 1983, p.22). The time of adoption of an innovation (from the time that the innovation had been introduced) therefore continued to be the central element of the classical paradigm of diffusion research. No less than 60% of all diffusion research studies focused on early adoption as the main dependent variable (Rogers 1976). Research designs mainly followed the Ryan and Gross (1943) approach, described earlier, consisting of correlational analyses of, l6 cross-sectional data, generally gathered in a single survey (Rogers 1976). THE CLASSICAL PARADIGM IN THE MARKETING LITERATURE The marketing tradition of diffusion research, modeled on the rural sociology tradition described above, emerged during the early 1960's (e.g., Bell 1963; King 1963, 1965; Cunningham 1966; Arndt 1967; Robertson 1967). The subsequent, near explosive, growth in research effort (e.g., see Robertson 1971, p. 22) led marketing diffusion research to be ranked among the more prominent of major diffusion research traditions (Rogers 1983, pp. 52-53). Yet, conceptual and methodological contributions remained rather limited. Research, for the most part, focused on applications of the classical paradigm (i.e., its concep- tualizations and methodologies) to marketing problems (Gatignon and Robertson, 1985). The following two sub-sections summarize and evaluate the empirical evidence concerning early adoption behavior and assorted related variables accumulated across a variety of disciplines, including marketing. Following that, the second major part of the literature review examines theoretical and conceptual developments in the diffusion modeling and consumer diffusion literatures. Developments in the latter literature were motivated by deficiencies in the nature of the empirical evidence to be surveyed and represent additions to -- rather than applications of -— diffusion theory in marketing. 17 EMPIRICAL RESULTS: SUMMATION The amount of research on variables related to innovativeness, that has accumulated since the Ryan and Gross (1943) study was first published, is truly impressive. By 1983, the total number of empirical publications on characteristics of innovators could be estimated at approximately 1800 studies (see Rogers, 1983, p. 261). A summary of generalizations on characteristics of early adopters, derived from the accumulated empirical evidence across a variety of disciplines, including marketing, is presented below in Table 11-1, taken from Rogers (1983, pp. 260-261). EMPIRICAL RESULTS: EVALUATION A variety of criticisms regarding these generalizations, however, can be found in the literature. First, it may be noted here that the majority of generalizations supported by more than 75% of relevant studies tend to be those that have received less attention. Different studies, moreover, often report opposite findings on several of the characteristics presumed to be related to innovativeness. Robertson and Meyers (1969) concluded that personality variables had little if any relationship to innovative behavior. Taylor (1977) comments that "the overall result of these empirical studies is characterized by finding correlations that are so weak as to be questionable or meaningless." Midgley and Dowling (1978) point to the 18 Table II—l. Early Adopter Characteristics Studies: Not Percentage Direc- Genera- Suppor- Suppor- of Studies tion. lization. ting. ting. Supporting I. Socioeconomic Characteristics 2-1 Age (not related) 108 ‘120 48 2-2 + Education 203 72 74 2-3 + Literacy 24 14 63 2-4 + Higher social status 275 127 68 2-5 + Upward social mobility 5 0 100 2-6 + Larger-sized units 152 75 67 2-7 + A commercial, rather than a subsistence, economic orientation 20 8 71 2-8 + A more favorable atti- tude toward credit 19 6 76 2-9 + More specialized operations 9 6 60 II. Personality Variables 2-10 + Empathy 9 5 64 2-11 - Dogmatism l7 19 47 2-12 + Ability to deal with abstractions S 3 63 2-13 + Rationality ll 3 79 2-14 + Intelligence 5 0 100 2-15 + A more favorable attitude toward change 43 14 75 2-16 + Ability to cope with uncertainty 27 10 73 2-17 + A more favorable atti- tude toward education 25 6 81 2-18 + A more favorable atti- tude toward science 20 7 74 2-19 - Fatalism l4 3 82 2-20 + Achievement motivation l4 9 61 2-21 + Higher aspirations for education, occupations 29 10 74 Source: Adapted from DIFFUSION OF INNOVATIONS, Third Edition, by Everett M. Rogers, pp. 260-261, 1983. Copyright c 1962, 1971, 1983 by The Free Press. A Division of Macmillan, Inc. Reprinted with permission from The Free Press, A Division of Macmillan, Inc. 19 Table 11-1 (Cont'd) III. Communication Behavior 2-22 + Social participation 109 40 73 2-23 + Interconnectedness . with the social system 6 O 100 2-24 + Cosmopoliteness 132 42 76 2-25 + Change agent contact 135 21 87 2-26 + Mass media exposure 80 36 69 2-27 + Exposure to inter- . personal communica- tion channels 46 14 77 2-28 + More active informa- tion seeking 12 2 86 2-29 + Knowledge of inno- . vations 61 19 76 2-30 + Opinion Leadership 42 13 76 2-31 + Belonging to highly interconnected systems 8 7 ' 53 Source: Adapted from DIFFUSION OF INNOVATIONS, Third Edition, by Everett M. Rogers, pp. 260-261, 1983. Copyright c 1962, 1971, 1983 by The Free Press. A Division of Macmillan, Inc. Reprinted with permission from The Free Press, A Division of Macmillan, Inc. 20 "confused and contradictory nature" of these findings. Similar comments are made by Robertson (1971), Kohn and Jacoby (1973), Ostlund (1974), and Downs and Mohr (1976). It may be noted, in defense of Rogers' continuing presentation of generalizations that are based on arguably less than perfect empirical evidence (Rogers 1962; Rogers and Shoemaker 1971; Rogers 1983), that the burden of evidence in the social sciences is not quite the same as in "harder" sciences. Several instances of falsification need not lead to outright rejection of stated hypotheses in the scientific realist (Hunt 1982) or relativist (e.g., see Anderson 1983) metatheoretical perspectives that typify the social sciences, including marketing. Instead, occurrences of falsification provide impetus for further refinement and explanation regarding extant bodies of theory and evidence. A variety of explanations for the relative inconsis- tency of empirical evidence is accordingly found in the diffusion literature. First, it has been suggested that the classical paradigm was developed for major (discontinuous) innovations rather than the minor (continuous) innovations to which the classical paradigm was subsequently applied in many studies, particularly in marketing and consumer behavior (Midgley and Dowling 1978). However, it is not made clear how or why the classical paradigm is relevant to major innovations only. In addition, the logically concomitant argument is not made 21 that previous findings would gain in consistency if publications on minor innovations were to be eliminated from the body of evidence. According to Midgley and Dowling (1978), conflicting results are more likely due to differences between operationalizations of the innovativeness construct in different studies, and to the operational definition of the innovativeness construct per se. For example, in marketing innovativeness has been measured on the basis of: (1) early adoption of one particular innovation (e.g., Donnelly and Ivancevich 1974; Peat, Gentry and Brown 1975; Robertson 1968; Engel, Blackwell and Kegereis 1969; Robertson and Kennedy 1969; Warren 1985); (2) ownership of a range of new products in a particular product category (Robertson and Meyers 1969; Summers 1972; Ostlund 1972; Darden and Reynolds 1974; Kohn and Jacoby 1973; King and Sproles 1973); and (3) ownership of new products or stated preference for new products across a range of product categories (Summers 1971). Additional confusion in measurement is generated through the divergent manner in which researchers deviate from Rogers' specification of innovators as "the first 2.5% to adopt". For example, in the marketing literature innovators are defined by Robertson and Kennedy (1968) as the first 10% to adopt while Ostlund (1972) defines innovators as the 33% of his subjects who had tried the greatest number of new products among a prescribed 22 category. Uhl, Andrus and Paulson (1970) define the first 16% to be innovators, the next 24% as laggards, and the remaining 60% as "other adopters". It is therefore possible that conflicting findings regarding relationships between innovativeness and other variables are due to differences in operationalization (Kohn and Jacoby 1973). An additional explanation of conflicting findings centers on the definition of the innovativeness trait as early adoption behavior. That definition necessarily neglects the effect of situational factors that may intervene between the trait of innovativeness and resulting early adoption behavior. A prominent example of such situational factors is represented by a consumer's interest, or lack of interest, in the product category in which the innovation takes place. Robertson (1971) concluded that innovativeness is product category specific (see also Midgley and Dowling 1978; Summers 1971). Product category interest may be a necessary though not necessarily sufficient factor in the early adoption of innovations. As a result, samples of later adopters will include individuals that would adopt early, but for their lack of interest in the product category as a whole. Also, members of a population of potential adopters may not all receive information on an innovation at the same time. The latter constitutes an additional example of a situational factor intervening between trait and behavior (Midgley and Dowling 1978, see also Robertson 1971). 23 Consequently, samples of early adopters may include individuals who received information on the innovation earlier than other members of the population and adopted earlier for that reason, rather than from an inner inclination to embrace innovations sooner than other members of the population. The relative lack of consistency in the accumulated empirical evidence is however in the view of this dissertation less problematic than weaknesses in the theoretical underpinnings of the hypothesized relationships per se. Specifically, while all variables are generally treated separately in their relationship to innovativeness, many of these variables are closely, possibly causally, related to one another while others seem tautological (e.g., education, literacy, social status, and upward social mobility; or upward social mobility, achievement motivation, and higher aspirations; or interconnectedness, belonging to interconnected systems, social participation, and exposure to interpersonal communication channels). Yet, issues of causality remain largely unaddressed (Rogers 1976), though causality is often implied (e.g., Eveland 1979; see Rogers 1983, p.263). The causality issue will receive additional attention in the third chapter of this dissertation. Summarizing, it may be concluded first that widespread agreement exists regarding the disappointing nature of empirical results (Robertson and Meyers 1969; Taylor 1977; 24 Robertson 1971; Downs and Mohr 1976; Ostlund 1969, 1974; Midgley and Dowling 1978; Kohn and Jacoby 1973). Second, these disappointing empirical results can be attributed not only to differences in the operationalization of the innovativeness construct, but also to the definition and conceptualization of innovativeness itself. This is so because the tautological definition of innovativeness as early adoption behavior does not account for the influence of situational variables. Consistency in the operationalization of an improperly conceived construct would therefore not suffice to address deficiencies in the nature and quality of empirical evidence. Third, diffusion research has largely ignored the issue of causality. For example, it is not clear whether innovative behavior is caused by a consumer's personality or whether it is caused by situational variables. The following, second section of the literature review examines the conceptualization of innovativeness developed in the diffusion modeling literature and surveys recent efforts toward a reconceptualization of innovativeness in the consumer behavior literature. It is these contributions, both from the diffusion modeling and the consumer diffusion literatures, that provide the basis for this dissertation‘s reconceptualization of the innovativeness CODStI‘UCt . 25 Section II. Changing Perspectives CONCEPTUALIZATION IN THE DIFFUSION MODELING LITERATURE The objective of diffusion models is to represent the diffusion of an innovation among a population of potential adopters as a mathematical function of time (Mahajan and Muller 1979). This allows the modeler to forecast future sales. The diffusion modeling literature (Bass 1969; Fourt and Woodlock 1960; Mansfield 1961; Mahajan and Muller 1979) therefore focuses mainly on prediction rather than explanation or understanding. Yet, the behavioral assumptions underlying diffusion modeling are of great interest to this dissertation. The intriguing behavioral assumption underlying diffusion modeling is that much of new product acceptance is an imitation process (Bass 1969; Rogers and Shoemaker 1971; Muller and Mahajan 1979). Specifically, Bass (1969) defines innovators as "individuals who decide to adopt an innovation independently of the decisions of other individuals in a social system" (p.216). Non-innovators on the other hand are defined as 'imitators'. It is the variation in the proportions of innovators and imitators among adopters in subsequent time periods that underlies the mathematics of diffusion models. Simply put, the shape of the diffusion curve is calculated on the basis of the proportions of innovators and imitators that are expected to adopt an innovation over subsequent time periods. The total number of innovators that adopt in a 26 given time period decreases over time as the pool of nonadopters, of which a constant proportion is assumed to innovate during any given time period, becomes smaller; the total number of imitators that adopt in a given time period first increases with the number of earlier adopters that becomes available for imitation and then decreases as the pool of nonadopters becomes smaller. The diffusion modeling literature definition of innovators therefore represents a radical departure from the widely accepted definition of innovators as earlier adopters. The conceptualization by Bass distinguishes innovators from other adopters on the basis of differences in communication behavior. That is, innovators do not need supportive (socially or otherwise) communication from the social system. Innovators, according to the diffusion modeling literature conceptualization, can be found both among earlier and among later adopters. The average time of adoption from the time of introduction of an innovation will, however, be less for innovators as a consequence of a lack of a need for communication from other consumers. Relatively later adoption by imitators, on the other hand, is caused by their need for supportive communication from those who have adopted earlier. The conceptualization of innovativeness in the diffusion modeling literature therefore introduces causality into the explanation of earlier and later adoption by innovators and imitators respectively. 27 This conceptualization also provides a theoretical basis for its distinction between innovators and non- innovators. This is so because diffusion theory is essentially a theory of communication (Gatignon and Robertson 1985). A classification of adopters that intends to explain diffusion should consequently be related in some manner to communication behavior by adopters. Note in this respect that the ultimate criterion as regards proper classification schemata concerns the usefulness of a classification as regards explanation (Hunt 1983). The arbitrary (Robertson 1971) classical categorization of adopters is essentially tautological (Midgley and Dowling 1978) and is not, and for that very reason could not be, explanatory. The diffusion modeling literature allows for the possibility that there may be no innovators among a population of potential adopters (Mansfield 1961) or that all potential adopters be innovators (Fourt and Woodlock 1960), or any variation in between. The diffusion modeling literature conceptualization of innovativeness is therefore entirely innovation specific and does not suggest the existence of certain characteristics by which innovators might be consistently identified since individuals will not consistently innovate across different innovation-decision situations. Any consumer, depending on the situation, can sometimes be innovative according to the diffusion modeling definition of innovativeness, a notion also found with 28 Hirschman (1980). The conceptualization of innovativeness as a personality trait on the other hand, if it is to lay claim to validity, should necessarily be connected somewhat consistently to additional personality variables. The accumulated body of empirical evidence reviewed earlier hardly indicates that such has been or can be succesfully accomplished. It may also be noted here that the conceptualization of innovativeness in the diffusion modeling literature provides a good explanation as regards the lack of integration between the diffusion modeling literature and the behavioral diffusion literature, remarked upon by Gatignon and Robertson (1985). The very existence of diffusion modeling in its attempt to predict the future shape of diffusion curves denies the central tenet of the behavioral diffusion literature which assumes the existence of a normally distributed innovativeness personality trait, the latter having been connected explicitly (Rogers 1983, p.20) with the existence of normally distributed diffusion curves . RECONCEPTUALIZATION IN THE CONSUMER DIFFUSION LITERATURE The diffusion modeling literature developed independently from the general behavioral diffusion literature. Hence, it did not itself Specify or examine the theoretical implications for the behavioral diffusion literature as regards its conceptualization of the innovativeness 29 construct. Several authors in the consumer diffusion literature have however attempted to specifically explain and address the failure of empirical evidence to confirm the existence of some set of defined characteristics Shared by innovators. First the seminal contribution by Midgley (1977) and Midgley and Dowling (1978) is examined. The conflicting nature of empirical evidence in the diffusion literature may, as noted above, be attributed not only to differences in operationalization of the innovativeness construct, but, more importantly, to the very conceptualization of innovativeness itself. Midgley and Dowling (1978) provide the following cogent summary on that issue. "Researchers in this [diffusion] area make two implicit assumptions. First, that innovativeness is a personality trait possessed, to a greater or lesser degree, by all members of a society, and second that what is being measured [early adoption) is, in fact, this trait" ---"By anchoring the construct [innova- tiveness] directly to its measurement [early adoption] researchers --- have rendered their version of innovativeness innovation-specific, leading to severe problems of inter-study comparison." Their solution to this problem centers on a reconcep- tualization of the innovativeness construct, rendering it distinct from its measurement. They argue, referencing Rozeboom (1966, p.206), that when a construct is central to a theory, as is innovativeness to diffusion theory, that construct and measurement should be logically distinct. They extract that necessary distinction by first 30 pointing to researchers who measured innovativeness across a range of products within a product category (e.g., Robertson and Meyers 1969; Darden and Reynolds 1974; Kohn and Jacoby 1973; King and Sproles 1973; see Midgley and Dowling 1978) or even across a range of product categories (e.g., Summers 1971). The innovativeness construct when measured through such cross-sectional methodologies is presumably more likely to reflect some persisting characteristic of the individual, since its presence is measured across a series of innovation-decisions confronting the individual, thereby eliminating the influence of situational factors. The cross-sectional methodology is therefore appropriate to measure the innovativeness personality trait according to Midgley and Dowling (1978). That measurement methodology may however not serve to define the innovativeness construct, since the definition of that central construct should be logically distinct from its measurement. Such a definition moreover should be logically consistent with the theory in which it plays a central role. Midgley and Dowling (1978) approach the problem of defining innovativeness by pointing to the key role in diffusion theory played by communication behavior (see also Rogers 1976; Gatignon and Robertson 1985). Indeed they argue that it would be difficult to account for observed non-linear cumulative adoption curves without the existence of communication processes. An assertion, it may 31 be noted, that is contrary to theoretical and empirical assumptions and empirical evidence found in the diffusion modeling literature. Specifically, Fourt and Woodlock (1960) present a diffusion model where all adopters are assumed to be innovators (i.e., assumed to adopt independently of communication with others). The general tendency of the Midgley and Dowling (1978) argument however closely follows assumptions implicit in diffusion modeling. Thus, Midgley and Dowling offer the following definition of innovativeness as "the degree to which an individual makes innovation-decisions indepen- dently of the communicated experience of others." They explicitly present this definition however while conceptualizing innovativeness as a trait, a dimension of the human personality, and a function of additional dimensions of the human personality. Additionally, it is suggested by Midgley and Dowling (1978) that a receptivity to new ideas is part of the essential notion of innovativeness. Individuals who adopt very late because they lack receptivity to new ideas may therefore not be real innovators even if they adopt independently of interpersonal communication. This, as was seen earlier, directly contradicts the notion of innovativeness that underlies diffusion modeling, in spite of the close resemblance between the definitions offered by Midgley and Dowling (1978) and Bass (1969) respectively. Midgley and Dowling (1978), quoting Rogers and Shoemaker 32 (1971), express the expectation that receptivity to new ideas will be closely linked to independent decision making, but also suggest that such should be empirically examined. Hirschman's (1980) reconceptualization of innovativeness closely follows Midgley and Dowling's metatheoretical stance. That is, she rejects the operationalist definition of innovativeness and seeks to replace it with an axiomatic definition that is valid in the context of the theoretical framework to which the construct is central. However, she argues that a consumer‘s "receptivity to new ideas" should serve to define that consumer's degree of innovativeness. She points out that Midgley and Dowling did not identify factors that cause independent decision making, be it within or across product categories. She combines theories on consumer creativity, role accumulation, inherent novelty seeking (the inner inclination for stimulation by the new and different), and actualized novelty seeking (actions to experience, possibly vicariously, the new and different) to arrive at a causal explanation of early independent adoption and/or rejection behavior. Her explanation of early adoption behavior, simply put, proposes that early adoption is caused by a consumer's ability to recognize superior solutions to (sometimes novel) consumption problems. That superior ability is ultimately, primarily determined by novelty 33 seeking self-fulfillment. _The latter concept is connected to some internal drive or motivating force to seek out new and potentially discrepant information, and to an inherent need for variety in experiences. In addition, Hirschman (1980) advances a teleological notion underlying novelty seeking. Specifically, she proposes that such behavior might be explained by a consumer's need for knowledge in a complex environment. The latter notion is metatheoretically quite different from established practice in the diffusion literature, as will be seen in the next and third chapter of this dissertation. SUMMARY In view of the above, the following conclusions can be listed summarizing the current status of diffusion research in marketing: 1. The nature of empirical evidence on characteristics of early adopters is considered quite unconvincing. It is particularly doubtful that any conclusions can be drawn regarding relationships between personality variables and early adoption. 2. A consensus has emerged in the marketing literature that the operational definition of innovativeness as early adoption behavior is tautological. That is, to equate innovativeness with innovative behavior makes it impossible for the former to explain the latter. 34 3. Research efforts attempting to address that deficiency have focused on establishing an independently defined innovativeness trait that, though mediated by Situational factors, explains early adoption behavior. 4. Two distinct directions have been taken in the conceptualization of such a separate innate innovativeness - trait. First, innovativeness as a personality trait has been equated with an inner inclination to independent judgment making. Secondly, innovativeness as a personality trait has been equated with an inner inclination to embrace that what is new and different. The following chapter presents this dissertation's theoretical framework and hypotheses. Chapter III Theoretical Framework and Hypotheses Section I. Theoretical Framework DEFINITION OF INNOVATIVENESS This dissertation defines innovativeness as the degree to which an individual makes innovation- decisions independently of the decisions of other individuals in his or her social system. The term "innovation-decision", introduced by Rogers and Shoemaker (1971), implies that both adoption and rejection by innovator and non-innovator alike may occur upon consideration of the innovation. An innovation is defined as any product, service, or idea that is perceived as new by the individual (Rogers 1983, p. 11). An individual's social system comprises co-workers, neighbors, relatives, and friends (Childers 1986) with whom the individual interacts, also independently of consideration of the innovation. The definition above traces its roots to the conceptualization of innovativeness in the diffusion modeling literature. That is, it implies that non- innovators are imitators who (in a given product category) do not make innovation-decisions independently, and it regards early adoption as neither sufficient nor necessary to infer innovativeness. The use of the term innovation- decision emphasizes that adoption is not a necessary 35 36 conclusion of the adoption process or innovation-decision process (Rogers and Shoemaker 1971). The definition thereby incorporates Hirschman's (1980) conceptual distinction between vicarious innovativeness and actualized innovativeness. That is, the making of an innovation- decision is conceptually distinguished from acting on that decision. The latter can include adoption or rejection and/or impartment of (positive or negative) recommendations regarding an innovation. Note that, strictly speaking, mere rejection does of course not constitute a physical act, but rather the absence thereof. That absence is included however in this dissertation among the possible actions on a decision since, given the high failure rate of new products, it is not unlikely to be the most prominent of "actions" taken in innovation-decision situations and of consequent great interest to managers. The key difference between the proposed definition and the definition by Midgley and Dowling (1978), to which it bears a close resemblance, lies in its conceptualization of innovativeness as behavior. This behavior is admittedly not readily observable, nevertheless it is potentially measurable. Midgley and Dowling (1978) criticized the operational definition of the innovativeness personality trait as early adoption behavior. Accordingly, they proposed the following definition of innovativeness as a personality trait: "the degree to which an individual makes 37 innovation-decisions independently of the communicated experience of others." This definition of the construct is separate from its behavioral operationalization. Operationally, the presence of this trait should be inferred, according to Midgley and Dowling (1978), from observations of early adoption behavior across a series of innovation-decision opportunities in order to eliminate the influence of Situational factors. Closer inspection of the Midgley and Dowling (1978) definition however reveals that it too is implicitly behavioral and tautological. This is so because their definition, though said to be so, is not in fact that of a personality trait. To see this, consider the following. The degree to which a consumer selects new service stations for car repairs "independently of communicated experience of others" is a function of that consumer's trust in his or- her ability to judge the character and expertise of management and mechanics in the car repair business. It is hardly inconceivable that the same consumer who confidently decides to give the new service station in the neighborhood a try may call on friends and relatives when contemplating the use of a new recipe. Therefore, the degree to which an individual makes innovation-decisions independently of the communicated experience of others cannot be a personality trait, which is defined as a "persisting characteristic or disposition by which one individual can be distinguished from another" (English and English 1958, Wolman 1973, see 38 Midgley and Dowling 1978). Instead, it constitutes behavior. Such behavior may or may not be explained additionally through some inner inclination to independence of decision making, or related to additional psychological traits such as empathy, dogmatism, achievement motivation, etc. (see Midgley and Dowling 1978). Thus Midgley and Dowling (1978) offer a behavioral definition of innovativeness ("the degree to which innovation-decisions are made independently") and explain it tautologically by assuming an underlying personality trait (an inclination to make innovation-decisions independently). In view of the example provided above, one may wonder about the explanatory power of the assumed inner inclination to independent decision making. Specifically, in explaining the occurrence of independent decision making in a product category, situational factors such as consumers' familiarity with the product category (e.g., cars or cooking) seem rather more important than consumers' inner inclinations. AS stated, this dissertation's definition of innovativeness is essentially behavioral, similar to the classical paradigm, and similar in fact to Midgley and Dowling (1978). But it does not, implicitly or explicitly, equate that behavior with a personality trait. It also does not regard that behavior directly connected to, caused by, or mirrored by a corresponding personality trait. Such may appear to be somewhat radical, yet this perspective is 39 sanctioned by the lack of results of prior empirical evidence (Taylor, 1977; Robertson and Meyers 1969; Downs and Mohr 1976; Ostlund 1969, 1974; Midgley and Dowling 1978; Robertson 1971). In its attempt to explain innovative behavior, as defined, this dissertation views innovative behavior as situationally determined. Belk (1975) has pointed to the need for examination of Situational variables to explain variance in consumer behavior. Lavidge (1966) similarly noted that differences in behavior between consumers may be based on differences in situational variables. METATHEORETICAL PERSPECTIVES This dissertation, in the explanation of innovative behavior, views human action as consciously motivated to achieve certain goals that are relevant to the individual. Thus a purposive, teleological perspective is taken as regards the causality that presumably underlies human action. The dissertation thereby breaks new ground, leaving the metatheoretical confines of the classical paradigm of diffusion research. Confines that also characterize the conceptualizations by Midgley (1977), Midgley and Dowling (1978), and (to an extent) Hirschman (1980). To see this, a necessarily concise survey of current thinking on metatheoretical issues in social science, especially marketing, may be appropriate and is provided below. Two basic philosophical paradigms dominate theory 40 construction in social science. In psychology these are characterized as behavioristic and phenomenological (Runyon 1980). Deshpande (1983) characterizes the approximately similar metatheoretical dichotomy in marketing as positivist versus idealist. Logical empiricism and scientific realism (see Hunt 1982) have been predominant in marketing (Peter 1982; Arndt 1985), and more nearly reflect the positivist perspective of science. Positivism originated as the paradigm of natural science. In the extreme it rejects any consideration of causality; in its more moderate form it considers causality factual (Deshpande 1983), operating from deterministic necessity (Rapoport 1969). That perspective is exemplified in previous approaches where the explanation of adoption is akin to explanations of chemical reactions. That is, if the elements are present, an innovator and an innovation or an inclination to make decisions independently and the opportunity to make a decision independently, and the environment is appropriate, adoption or an independent decision occur. Accordingly, positivism is said to hold to the existence of a reality independent of theory, or what might be called a mechanistic reality, which theory attempts to verifiably explain (Hunt 1982). Its methodologies are consequently quantitative in nature, uniquely qualified to verify, but not to discover, hypotheses (Deshpande 1983, see also Reichardt and Cook 1979). 41 The polar opposite of positivism has been termed relativism (Long 1985) or idealism (Deshpande 1983). Its approach to the explanation of human behavior can be considered teleological. Specifically, it attempts to understand human action from the actor's frame of reference (Deshpande 1983). A prominent example of the non- positivist school of thinking is found in the Austrian economists' approach to explanation. According to their perspective (see Kirkpatrick 1982) the actor's free will is considered axiomatic, and the causes of the actor's behavior must be sought in the actor's purposes. Hence, the idealist school may adopt a purposive or teleological understanding as regards the meaning of causality. A prominent application in marketing of this paradigm is found in Alderson's functionalism. The two examples below may clarify the meaning of causality according to deterministic and teleological prehensions respectively. I. Q. Why does the earth move around the sun? A1. (Deterministic necessity) because of a combination of gravitational and centrifugal forces. A2. (Teleological) because it can keep an even tempera- ture in that manner. II. Q. Why did the general execute the tallest men among his captives? A1. (Deterministic necessity) Because his father, who was tall, beat him as a child. A2. (Teleogical) Because he feared that the tallest among prisoners would more likely prove troublesome. 42 The debate between supporters of the positivist and idealist schools of thought has often been sharp, yet they are also somewhat complementary in nature. Deshpande (1983) strongly advocates that both approaches be applied in marketing in a balanced way (see also Long 1985). This dissertation will follow that recommendation in the generation and testing of its theory and associated hypotheses. The conceptualization of innovativeness, and the generation of the accompanying causal model and associated hypotheses pay tribute to the idealist view of human behavior (i.e., behavior is explained from the actor's frame of reference). Subsequent verification of the theoretical framework will take place according to quantitative methodologies that more nearly reflect the positivist paradigm. Section II. Hypotheses ENDURING INVOLVEMENT AND INNOVATIVENESS In view of the above, this dissertation therefore does not ask who the innovative consumer is. This dissertation asks instead why the consumer is innovative. Moreover, when it asks the latter question, the answer (contrary to previous efforts) is not sought in deterministic necessity. That is, it is not assumed that the consumer is innovative because the consumer is innately inclined to embrace the new and different, or innately inclined to independence in decision making, or innately curious. The answer instead 43 is sought in the motivation of the innovative consumer. That is, this dissertation asks why the consumer would want to be innovative (i.e., want to make innovation-decisions and forgo the communicated experience of others). The answer to that question, according to the causal model presented in Chapter 1 and reproduced below (Figure III-1), is sought initially in the consumer's enduring involvement with the product category in which the innovation takes place. Enduring involvement is defined as "the ongoing concern with a product the individual brings into a purchase situation" (Bloch and Richins 1983, p. 71). Enduring involvement is a situational variable akin to Midgley and Dowling's (1978) "interest in product category" variable. In Belk's (1975) taxonomy of situational variables, enduring involvement would be classified as a "task definition" variable. Task definition variables include such variables as "an intent or requirement to select, Shop for, or obtain information about a general or specific purchase" (Belk, p. 159). The proposed model and the proposed conceptualization of innovativeness argue that enduring involvement does not serve as an intervening variable, allowing or disallowing (in its absence) the expression of innate innovativeness. Instead, it is argued that innovativeness (independence of innovation-decision making) is directly caused by enduring involvement. 44 Enduring Situational Usage + Involvement Involvement Early Adoption + nnovativenes Figure III-l. A Causal Model of Innovativeness 45 To see this, consider an individual who is less inclined to rely on interpersonal communication. Such an individual likely believes either that interpersonal communication will not yield additional useful information or that the cost of waiting for such additional information exceeds its potential value. AS regards the former, the individual may learn through experience that his or her peers or near-peers generally do not know as much or do not know much more than he or she does about innovations in the product category of interest. Innovators, therefore, should be expected to know more and know earlier about innovations. Greater knowledge of innovations is indeed positively related to earlier adoption according to 61 out of 76 studies examined by Rogers (1983, see Table II-l, generalization 2-29, p. 19 of this dissertation, see also Engel, Kegereis, and Blackwell 1969; Hirschman 1980; Dickerson and Gentry 1983; Gatignon and Robertson 1985). Greater and earlier knowledge of innovations in a product category is precisely the condition that is brought on by enduring involvement, or involvement with products (Howard and Sheth 1969). This is so, since enduring involvement is acted upon or expressed by a consumer's continuing search for information on products in the product category of interest, independent of the need to purchase (Bloch, Sherrell and Ridgway 1986). Enduring involvement, reflec— ted in a continuous monitoring of the marketplace for new 46 and interesting information, is therefore hypothesized to cause independent decision making. Knowledgeability per se however, though a charac- teristic or condition likely brought on by enduring involvement, is not thought to necessarily lead in and of itself to continuous independent decision making. Since knowledgeability may also be a temporary condition brought on by a recent or current experienced need to purchase and may have been acquired on the basis of the experience of others. Hence, it is hypothesized that H1: Enduring involvement will have a positive impact on innovativeness. This first hypothesis embodies the central argument con- tained in this dissertation and reflects this disser- tation's teleological understanding of causality in the explanation of the adoption of innovations by consumers. It may therefore be appropriate to interrupt the presen- tation of the hypotheses in order to further consider this causality issue in connection with the presentation of the hypothesis above. Enduring involvement is not considered a dimension of the human personality, or as being caused by a dimension of the human personality, with the exception of the axiomatic assumption that human action is aimed at the achievement of certain goals. This means that it is assumed that the adoption of innovations is a consequence of a desire for better solutions to consumption problems. New products 47 will be adopted if, and more importantly because, they present superior solutions. They are not, in the view of this dissertation, adopted because an innovative consumer is inherently predisposed to embrace that which is new and different. The latter notion, which reflects a deterministic necessity understanding of the meaning of causality, has directly or indirectly served to explain innovativeness in previous conceptualizations (except for certain considerations advanced by Hirschman (1980) as reviewed). The high failure rate of new products in fact suggests that innovators are, if anything, inherently more likely to reject rather than adapt new products. If innovators were to be defined on the basis of adoption behavior, they would be termed more accurately early rejectors rather than early adopters. Implications of that consideration as regards the theory and practice of new product (concept) testing and new product management, and the lack of integration of these with diffusion research, will be examined in the final chapter of this dissertation. Here it is necessary to return to a consideration of additional hypotheses to be tested in this dissertation. SITUATIONAL INVOLVEMENT AND INNOVATIVENESS As stated, a consumer may have concluded on the basis of experience that additional useful information is not likely to be gained from his or her social system, or alternatively a consumer may feel that the value of such 48 information does not exceed the cost of waiting. One example of the latter is innovativeness in the face of desperation, for example, a terminal disease or (more mundane) a car failure in a strange town. Another example is when the cost of an incorrect innovation-decision is quite low. According to Houston and Rothschild (1978) product-related stimuli and social psychological stimuli combine to reflect the perceived severity of the consequences of an inappropriate purchase decision. That perceived severity induces situational involvement "evoked by a particular situation, such as a purchase occasion" (Bloch and Richins 1983). Involvement with a purchase causes individuals to spend more time gathering information and to gather greater amounts of information (Clarke and Belk 1979). Hence, it is suggested that situational involvement may cause an individual to be more willing to delay an innovation-decision until additional information in the form of communicated experiences of others becomes available in his or her social system. That perspective is supported by Arndt (1967) and Ostlund (1974). They found that perceived risk is negatively related to early adoption. The situational involvement construct captures the "consequence" component of perceived risk according to Houston and Rothschild (1978). Hence, H2: Situational involvement will have a negative impact on innovativeness 49 HEAVY USAGE, ENDURING INVOLVEMENT, AND EARLY ADOPTION The alternative approach to innovativeness is linked to the traditional approach through inclusion of early adoption in the model. Heavy usage, for two reasons, is also included in the model. First, consumer diffusion researchers have consistently verified, across a range of product categories, that early adopters are found among heavy users within a product category (Frank, Massy, and Morrison 1964; Robertson 1971; Taylor 1977; Danko and MacLachlan 1983; Dickerson and Gentry 1983). Second, heavy users as a group are of great managerial importance and are generally readily identifiable. The strong relationship between early adoption behavior and heavy usage is hypothesized to operate in two different ways, as shown in the model. Heavy usage may directly influence early adoption, since purchase occasions on average will arrive sooner for the heavy user, or heavy usage may influence early adoption through its enhancement of the individual's level of enduring involvement (Houston and Rothschild 1978), and consequently the individual's inclination to make innovation-decisions independently. Hence, H3: Heavy usage will have a positive impact on enduring involvement H4: Heavy usage will have a positive impact on early adoption Finally, independent innovation-decision making on average 50 leads to earlier adoption, since independent innovation- decision makers need not wait for adoption by other individuals in their social systems as discussed in the examination of the diffusion modeling literature. Hence, H5: Innovativeness will have a positive impact on early adoption OPINION LEADERSHIP AND INNOVATIVENESS Opinion leaders are defined in the marketing literature as "individuals who influence the general and purchase behavior of other people" (Engel, Kollat and Blackwell 1982, p. 354). The construct originated with the Lazarsfeld, Berelson, and Gaudet (1948) study of the 1940 presidential elections. That study proposed a two-step flow of communication hypothesis, meaning that information flows from the mass media to opinion leaders and from them to the rest of a population of potential adopters. The latter conceptualization distinguishes opinion leaders from non-opinion leaders on the basis of the nature of information (impersonal versus personal) used to determine their respective opinions. That conceptuali- zation is directly related to the conceptualization of innovators versus imitators found'in the diffusion modeling literature and adopted in this dissertation. That is, it regards opinion leaders not only as disseminators of opinions, but also as originators of opinions within their social system. Thus, it may be said to conceptualize 51 opinion leaders as individuals who make decisions independently of decisions by others. The accepted conceptualization of the opinion leader in marketing (and in the diffusion literature, see Rogers 1983, p. 27) focuses on influence only, regardless of whether the opinion leader was in turn influenced by others. That definition therefore includes a much larger proportion of individuals in a target market. The distinction between the two interpretations of the opinion leadership construct serves to develop the final two hypotheses to be tested in this dissertation. Specifically, researchers in marketing repeatedly report findings of a positive relationship between opinion leadership and earlier adoption (King and Summers 1970; Summers 1971; Baumgarten 1975; Engel, Kegereis, and Blackwell 1969; Mancuso 1969) although the relationship is not always found to be very strong (Meyers and Robertson 1972; Summers 1971; Mancuso 1969). Opinion leadership is also reported to be product category specific (Katz and Lazarsfeld 1955; King and Summers 1970; Montgomery and Silk 1971). As therefore might be expected, the search for characteristics that can consistently (across product categories) identify opinion leaders has not been very successful. yielded low or inconclusive correlations (Montgomery and Silk 1971). However, characteristics of innovators and opinion leaders within a Specific product category (fashion) exhibited remarkable similarity 52 (Baumgarten 1975). The considerable overlap between early adoption behavior and opinion leadership in general (and between characteristics of innovators and opinion leaders in fashion in particular) led Baumgarten to suggest that the question arises of the extent to which opinion leaders are innovators and vice versa. It is precisely that question this dissertation intends to answer. That answer will however be provided in the context of the definition of innovators as proposed in this dissertation. As mentioned, the two-step flow hypothesis of communication regards opinion leaders as those who collect and interpret information from mass media at the behest of others. That conceptualization of opinion leadership concurs with this dissertation's conceptualization of innovativeness. It suggests the following hypothesis: H6: All innovators are likely to be opinion leaders The accepted conceptualization in marketing of opinion leaders as influencers however includes also those individuals who initially were influenced by others. In other words, it includes also imitators. As noted in the earlier discussion of the diffusion modeling literature, non-innovators imitate those that adopted earlier. Both innovators and imitators will be found among these (influential) earlier adopters according to the diffusion modeling literature conceptualization of innovativeness, similar to the conceptualization of innovativeness in this 53 dissertation. This suggests the following hypothesis: H7: Not all opinion leaders are necessarily innovators Therefore, in response to the question posed by Baumgarten (1975) as regards "the extent to which innovators are opinion leaders, and vice versa", this dissertation hypothesizes that the answer may be that innovators are opinion leaders, but not necessarily vice versa. SUMMARY This chapter has conceived of innovativeness as conscious- ly motivated by a desire to recognize superior alternatives to consumption problems. The innovative consumer has been distinguished from the non-innovative consumer on the basis of the former's ability and willingness to independently evaluate and decide on the merits of proposed alternative solutions to consumption problems. This conceptualization formed the basis for hypotheses linking the innovativeness construct to well defined additional consumer behavior constructs. In addition, the proposed conceptualization of innovativeness served to establish a theoretical foundation on the basis of which a distinction could be made between opinion leaders that are and opinion leaders that are not innovators. Chapter IV Methodology Section I. Data SAMPLE A mail survey of 333 clerical and administrative staff members at a major university was used to collect the data to test the research hypotheses. The selection of this non-random population is appropriate since the research at hand entails the application of general scientific theory. That is, the effects observed in the research are employed to assess the status of theory. They are not intended to be generalizeable to other settings or other populations. Instead, the theoretical considerations of the study are intended to be generalizeable to alternative settings (for additional discussion see Calder, Phillips, and Tybout 1981). The selection of clerical and administrative staff as subjects was motivated by the expectation that they are likely to have a certain degree of experience with each of the three product categories investigated (personal computers, movie theater attendance, and restaurants). In addition, they are expected to be relatively homogeneous with respect to factors that are not included in the model (e.g., income or lifestyle). The latter consideration led to the decision not to include, for example, faculty members or students. This approach 54 55 provides for a stronger test of the theory, Since it reduces the likelihood of type II errors occurring due to sample heterogeneity (Cook and Campbell 1975; see also Calder, Phillips, and Tybout 1981). For example, as regards inclusion of faculty members, it is not inconceivable that the nature of the academic life style leads to greater enduring involvement with personal computers or with restaurants. Findings of a positive relationship between enduring involvement, earlier adoption, and higher usage might then be attributable to relatively higher incomes earned by faculty members. SELECTION OF PRODUCT-MARKET SETTINGS The variables of interest are measured across three dif- ferent product-market settings: personal computers, movie theater attendance, and restaurants. Several considera- tions led to the selection of these categories. First, the categories are presumed to vary with respect to the degree of involvement evoked by a purchase occasion (situational involvment). This is important since variance in Situational involvement is likely to be greater between product categories than between subjects (Houston and Rothschild 1978). Second, as regards movie theater attendance and restaurants, services have received comparatively little attention in the marketing diffusion literature (Warren 1985). Third, the timing of adoption or purchase in the selected categories is presumed to be quite recallable, an important consideration in diffusion 56 research (Rogers 1976). Finally, movie theater attendance and restaurants continuously present innovation-decision opportunities. RESEARCH INSTRUMENT Three separate questionnaires, each addressing a single product category (personal computers, movie thetaer attendance, and restaurants) were used to pretest the operationalization of the constructs. A direct comparison of the results across the three product categories is intended. Hence, it was decided that data on each of the three product categories should be collected from the same group of subjects. As a result, a concern existed that subject cooperation might be imperiled due to the lengthy and repetitive nature of their intended task. The pretest allowed an assessment of that issue. Thirty-eight subjects were requested (in person by the researcher's assistant) to fill out all three questionnaires (see Appendices A-C). Eighteen of thirty- eight respondents returned all three questionnaires within one week. Questionnaires from two additional subjects werre received several weeks later. Of these (sixty) questionnaires all but one were completed. This response rate was felt to be quite low given the personal nature of the request for cooperation. Comments by prospective respondents confirmed that the task of filling out all three questionnaires was felt to be quite burdensome. The results of the pretests however led to significant 57 reductions in the size of the research instruments. This, in combination with changes in lay-out and design, allowed the construction of a single questionnaire (appendix D). That questionnaire measures all variables across each of the three product categories without, it was expected, unduly burdening respondents. The response rate of 43.2% (within a three week cutoff period) to the final mail survey bore out that expectation. Section II. Variable Operationalization INNOVATIVENESS The items used to measure innovativeness are adapted from a scale developed by Carlson and Grosbart (1984). Their 13- item scale, developed in accordance with suggestions by Hirschman (1980) and Midgley and Dowling (1978), measures the degree to which an individual makes innovation- decisions independently (see Appendix E). Such indepen- dence need not imply an absence of interpersonal communication. Indeed, Kohn and Jacoby (1974) found early adopters to be more likely to obtain information from friends. Instead, the items are designed to measure the degree to which the innovator's decision is not dependent on'such communication. The Carlson-Grosbart scale, following Midgley and Dow- ling's (1978) recommendation, measures innovativeness as a personality trait, generalizable across product 58 categories. The present approach however considers innovativeness to be product category specific. Hence, the items of the Carlson-Grosbart scale were adapted to fit each of the three product categories investigated (see Appendices A-C and Table IV-l below). Pretests led to the elimination of all but five items. These five items are used for each of the three product categories investigated. The five-item innovativeness scale for movie theater attendance is presented below in Table IV-l. The essentially identical innovativeness scales that fit computers and restaurants can be found in Appendix D. Reliability scores obtained during pretesting for the five- item scale equal .8414 (computers), .9444 (movie theater attendance), and ,8809 (restaurants). Table IV-l. Innovativeness Scale Strongly Strongly 1. I talk with others who Agree Disagree have seen new movies before I decide whether to go see them. 1 2 3 4 5 6 7 2. I seek advice from other people who have seen a new movie before I go see it. 1 2 3 4 5 6 7 3. I find it hard to decide whether to go see a new movie before I learn the opinions of those who have already seen it. 1 2 3 4 5 6 7 4. I wait to see new movies until I know whether friends whO'have tried them think they are "0k". 1 2 3 4 5 6 7 5. I am one of those people who would decide to go see a new movie without consulting others who had previously tried it. 1 2 3 4 5 6 7 59 ENDURING INVOLVEMENT Enduring involvement has been defined as "the ongoing concern with a product the individual brings into a purchase occasion" (Bloch and Richins 1983, p. 71). In other words, enduring involvement exists independently of the purchase occasion. This construct is operationalized in this dissertation on the basis of an adaptation of Zaichowski's (1985) 20-item bipolar adjective scale (see Appendix F). Machleit (1986) found that the twenty items in the Zaichowski scale possibly represent three different dimensions: importance, utility, and interest. The latter involvement dimension was taken to represent the domain especially pertinent to enduring involvement. Hence, the enduring involvement construct was operationalized on the basis of items that constituted the interest dimension of the Zaichowski scale. A content analysis of the 20-item scale suggested items 10 (uninterested-interested), 13 (boring- interesting), 14 (exciting-unexciting), and 16 (mundane- fascinating) to reflect interest rather than importance or usefulness. Quantitative analysis by Machleit (1986) of student response data to the 20-item scale led to elimination of item 10 and to consideration of item 15 (appealing-unappealing) for inclusion. Machleit (1986) found Cronbach alpha scores for the resulting 4-item scale (see Table IV-Z) of .85 (for blue jeans), .88 (for soft 60 drinks), and .89 (for cookies). Machleit's results are confirmed by reliability scores obtained from the pretests performed for this dissertation. These scores equal .9118 (for computers), .9036 (for movie theater attendance), and .8137 (for restaurants). Machleit (1986) suggested that item 15 (appealing-unappealing) might not really fit well with any of the three possible dimensions. The pretest results obtained in this disserttation indicated that item 15 be retained. Table IV-2. Enduring Involvement Scale Below are sets of word pairs. Please circle the numbers that best reflect your FEELINGS ABOUT ........... 3 interesting 1. boring I 2 4 S 6 7 2. unexciting 1 2 3 4 5 6 7 exciting 3. appealing 1 2 3 4 5 6 7 .unappealing l 2 3 4 5 6 7 4. mundane fascinating SITUATIONAL INVOLVEMENT Perceived risk incorporates two main risk components: importance and uncertainty (Cunningham 1964). Of these two components, situational involvement reflects the importance component (Houston and Rothschild 1979). Importance refers to the perceived severity of making an incorrect innovation-decision. An early attempt to measure the importance component of perceived risk is provided by Arndt (1967) who asks: "How important is it to you that a brand 61 of this product you have never tried before is as good as your present brand?" The content of Arndt's question is retained in the measure of Situational involvement used in this dissertation. The measure is however structured to conform to Zaichowski's (1985) approach to the measurement of involvement. Zaichowski (1985) had suggested that her scale be used also to measure involvement with purchase situations. Rather than use all twenty items however, it was decided to reduce the number of items presented to subjects. Specifically, only items reflective of the importance domain sought for in the situational involvement construct were retained (see Table IV-3 below). Table IV-3. Pretest Situational Involvement Scale Next, using similar scales, we would like you to indicate how important it is to you that you do not make a mistake when choosing a movie to go see in a movie theater. 1. important 1 2 3 4 5 6 7 unimportant 2 of no concern 1 2 3 4 S 6 7 of concern to me 3 irrelevant l 2 3 4 5 6 7 relevant 4 means a lot 1 2 3 4 5 6 7 means nothing to me 5. trivial l 2 3 4 5 6 7 fundamental 6 matters to me 1 2 3 4 S 6 7 doesn't matter to me 7 significant 1 2 3 4 5 6 7 insignificant Pretesting of these seven items for each of the three product categories led to the elimination of items 2 (of no concern-of concern to me), 3 (irrelevant-relevant), and 5 62 (trivial-fundamental). The resulting 4-item situational involvement scale yielded reliability scores of .8741 (computers), .9495 (movie theater attendance), and .9525 (restaurants). As intended, mean situational involvement scores exhibited considerable variance between product categories. Mean situational involvement scores obtained during pretesting were 6.67 (standard deviation =. 54) for computers, 4.36 (standard deviation = 1.43) for movie theater attendance, and 5.71 (standard deviation = 1.26) for restaurants. OPINION LEADERSHIP Opinion leadership has long been a prominent construct in both marketing and diffusion theory (e.g., Robertson 1971; Rogers 1962; Rogers and Shoemaker 1971; Rogers 1983; Gatignon and Robertson 1985). Among several measures of opinion leadership, the most prominent is the self- designating method developed by King and Summers (1970). Previous research (see Riecken and Yavas 1983; Yavas and Riecken 1982) found Cronbach alpha scores ranging from .50 to .87 for the King-Summers scale. Childers (1986), in a recent effort to improve the scale's reliability and validity, revised it by rewording the questions to allow the range of response alternatives to include five response categories. This dissertation adopted Childer's refinement of the King-Summers scale and pretested the revised scale across the three product categories (see Table IV-4). The number 63 of response categories was however expanded to seven to maintain consistency throughout the questionnaire. Based on pretest results, item 5 was eliminated. Childers (1986) also suggested that item 5 be dropped. In addition, it was found that scale reliability improved greatly through elimination of item 7. Given the product category specific nature of opinion leadership, this was not surprising. Reliability scores for the resulting 5-item opinion leadership scale equaled .8744 (computers), .8857 (movie theater attendance), and .8944 (restaurants). EARLY ADOPTION Early adoption is operationalized in the personal computer product category on the basis of subjects' relative time of adoption from the introduction of the innovation. The purchase of a personal computer is likely to be a single and major purchase in that product category. Accordingly, it is not expected that respondents will encounter much difficulty in recalling the time of such a purchase, even if it occurred several years ago. Respondents who did not yet purchase a computer were aSked to indicate if such a purchase was anticipated and if so when (see Appendix D).‘ Previous research relying on anticipated purchase behavior to establish relative time of adoption includes studies by Jacoby (1971), King and Baumgarten (1970), and Summers (1971). For movie theater attendance, two alternative methods 64 Table IV-4. Pretest Opinion Leadership Scale We would like to ask a few questions about how you interact with friends and neighbours regarding......... 1. In general do you talk to your friends and neighbours about...? . Often I 2 3 4 5 6 7 Never 2. When you talk to your friends and neighbours about ... do you: give a great deal give very little of information 1 2 3 4 5 6 7 information ' 3. During the past six months, how many people have you told about a ...? told a number of people_ 1 2 3 4 5 6 7 told no one 4. Compared with your circle of friends how likely are you to be asked about ...? very likely not at all to be asked 1 2 3 4 5 6 7 likely to be asked 5. In a discussion of ... would you be most likely to: listen to your convince your friends friends' ideas 1 2 3 4 S 6 7 of your ideas 6. In discussions of ... which of the following happens most often? you tell your your friends tell you friends about ... l 2 3 4 5 6 7 about ... 7. Overall, in all of your discussions with friends and neighbours are you: often used as a not used as a source of advice 1 2 3 4 S 6 7 source of advice 65 for operationalization of early adoption behavior were used: the classical direct approach (Ryan and Gross 1943) and the cross-sectional method often used in marketing (e.g., Summers 1971, 1972; Darden and Reynolds 1974; Baumgarten 1975). Specifically, all respondents were asked to indicate which of ten major movies released during 1986 (see Appendix G) they attended and when. Then, following the direct approach, the mean time of attendance since release of the movie was computed to assess adoptive innovativeness. Next, following the cross-sectional method, the number of new movies attended was used to measure respondents' adoptive innovativeness. Pretests yielded a surprising -.16 correlation between the two measures. The precision of recall by respondents regarding the timing of movie attendance is subject to some doubt however since three respondents reported having seen movies prior to their release dates. Two additional respondents indicated they did not recall timing of attendance. Accordingly, on the basis of pretest results, it seems likely that the cross-sectional method may be more accurate and should be preferred. However, also on the basis of pretest results, it must be kept in mind that the cross-sectional method may measure a construct that is not necessarily identical to the early adoption construct from the traditional diffusion literature (Ryan and Gross 1943; Rogers 1983). It was decided to retain both approaches in 66 the final questionnaire to allow further examination of this issue, using a larger sample. Finally, early adoption behavior for restaurants was operationalized by asking respondents to indicate the number of restaurants visited for a first time during the six months prior to filling out the questionnaire. This approach is similar to the cross-sectional method dis- cussed above. In effect, the number of new adoption decisions made in the past is used to establish indirect- ly a likelihood that a particular respondent would be among the earlier adopters of any given new restaurant. The above has described and justified the selection of the sample used in this study, the selection of product categories investigated, and the operationalization of the variables. The balance of this chapter discusses the statistical methods used to test the hypotheses. Section II. Analysis HYPOTHESES H1 THROUGH H5 The analysis takes place in two parts. This part addresses hypotheses H1 through H5. The second part addresses hypotheses H6 and H7. Recall H1: Enduring involvement will have a positive impact on innovativeness Situational involvement will have a negative impact on innovativeness H2 67 H3: Heavy usage will have a positive impact on enduring involvement H4: Heavy usage will have a positive impact on early adoption H5: Innovativeness will have a positive impact on early adoption To test hypotheses H1 through H5, the causal model illustrated in Figure IV-l is specified and tested. The attendant structural and measurement model equations are provided in Table IV-5. The structural equation model (Tables IV-5.1 and IV- 5.2) consists of three equations. In the first equation, enduring involvement (eta-1) is a function of usage (ksi- 1). In the second equation, innovativeness (eta-2) is a function of enduring involvement (eta-1) and situational involvement (ksi-2). In the third equation, adoptive innovativeness (eta-3) is a function of innovativeness (eta-2) and usage (ksi-l). The measurement models are presented in Tables IV-5.3 and IV-5.4. Each of the five latent variables in the three structural equations is measured by one or more observable variables as discussed in the previous section on operationalization of the variables. By definition, these five latent variables do not have a definite scale, since they are unobserved. Hence, for each it is necessary to assign an origin and a unit of measurement (Joreskog and Sorbom 1984, p. I-7). The origins are assigned by constraining one lambda for each variable to 68 611 81 52 53 5|: 62 53 Sq 65 X1 Y1 Y2 Y3 YH X1 X2 X3 Xq T TH z 41.2 51 01 52 51.1 Endurin - t- 1 Usage 9 (1 Situa iona 4' I "”1“ me "i I nvol ve me nt Z3 :2 Earl g Adoption E10 65 E; 57 E. E, Figure IV-l. LISREL Model of Innovativeness 69 Table IV-S. Model Equations The structural equations are n; I 51 E1 "' (I IV-SJ n2 " [32,1711 + 52,2 52 + (2 'l3 ' [33.202 + 63.1 51 + (3 OP '1: O O O "I: 51.1 0 51 (1 lV-5.2 I); = sz 0 0 ll; + 0 52.2 E2 + z 2 '13 0 93.2 0 "3 53.1 0 :3 The measurement model equations are 'Y.‘ rA... o o ‘ n. ”2.3 Y2 A23 0 0 '12 E 2 y:4 7‘11 0 0 “3 E 3 M53 Ya ML: 0 0 6 ti y5 = 0 A51 0 + E 5 Y: 0 K3,; 0 E 6 y? 0 >"Lz 0 £7 yo 0 Am: 0 e r y, 0 A12 0 e s \YIOJ t 0 0 No.3) H510. and (XIT rx” 0 ‘05] r61‘ X2 0 X22 52 52 N-5.4 X3 = O X32 + 63 M 0 Mg 6'1 70 equal 1 (see Joreskog and Sorbom 1984, p. I-7; see also Howell 1987). Since ksi-l (usage) and eta-3 (adoptive innovativeness) are each measured by a single observed variable, the above procedure also constrains error terms delta-1 and epsilon-3 (see Tables IV-5.3 and IV-5.4) to equal zero. Finally, it is assumed by convention that the error terms (vectors epsilon and delta) of the observed variables are independently distributed (Bagozzi 1980, p.89). Preliminary runs of the specified model indicate that the information matrix ("the probability limit of the matrix of second order derivatives of the fitting function used to estimate the model", see Joreskog and Sorbom 1984, p.I-24) is positive definite. That result indicates with near one hundred percent reliability that the model is identified (see Joreskog and Sorbom 1984, p. I-24). HYPOTHESES H6 AND H7 The final part of this chapter discusses the methodology for testing hypotheses H6 and H7. Recall H6: All innovators are likely to be opinion leaders H7: Not all opinion leaders are necessarily innovators In order to test these two hypotheses, it is necessary to assign respondents to four groups based on their scores on dichotomized innovativeness and opinion leadership scales. The dichotomization process will be discussed 71 shortly. These four groups (see Figure IV-2 below) include: (1) innovators who are also opinion leaders, (2) non-innovators who are opinion leaders, (3) innovators who are not opinion leaders, and (4) non-innovators who are also not opinion leaders. INNOVATORS NON-|NNOVATORS OPHNON LEADERS I 2 3 4 NON OPHNON LEADERS Figure IV-2. Cross-Tabulation of Innovators and Opinion Leaders Hypothesis H6 implies that cell 3 be empty, and hypothesis H7 implies that cell 2 may, but need not, be empty. That is, if respondents' scores on the dichotomized innovative- ness and opinion leadership scores do not overlap perfectly, then according to hypotheses H6 and H7, this should be attributable to subjects who belong in cell 2 (opinion leaders but not innovators), rather than to subjects who belong in cell 3 (innovators but not opinion leaders). The strongest possible test of these two hypotheses is 72 found in a simple inspection of the 2x2 matrix. Hypothesis H6 is falsified if even only one subject is found to occupy cell 3. Obviously, given the nature of research in social science, this purely falsificationist approach is unlikely to result in a finding of support for the stated hypotheses. Hence, a weaker test is proposed to examine if support is found in the data for the basic argument underlying hypotheses H6 and H7. Namely, the absence of overlap between innovativeness and opinion leadership is said to be attributable to a greater extent to opinion leaders who are not innovative (cell 2) than to innovators who are not opinion leaders (cell 3). Hence, it is expected that the frequency of observations in cell 3 be less than indicated by pure chance (given a certain degree of overlap of innovators and opinion leaders) and that the frequency of observations in cell 2 be greater than indicated by pure chance (again, given a certain degree of overlap of innovators and opinion leaders). Accordingly, to test hypotheses H6 and H7, first a simple inspection of frequencies in the 2x2 matrix is called for. Then a determination will be made of the frequencies in cells 2 and 3 that would obtain if a lack of overlap were equally attributable to non-innovative Opinion leaders and to innovators who are not opinion leaders. Subsequently, a simple binomial proportions test will be used to assess if the actual frequencies differ 73 Significantly in the proper direction from these expected frequencies. Finally, before proceeding to the analysis of the data, it is necessary to examine in this chapter the dichotomization of the innovativeness and opinion leadership constructs. Obviously, the selection of the actual cutoff points for dichotomization may affect the results. Hence, it is appropriate to consider this issue in some detail, both in the context of precedent in the literature and in the context of assumptions regarding the two constructs that have been made in this dissertation. Previous approaches in the marketing literature, conceiving of innovativeness as early adoption behavior, used a rather wide variety of cutoff points, including 10% (Robertson and Kennedy 1968), 16% (Uhl, Andrus, and Paulson 1970), and 33% (Ostlund 1972). The literature on opinion leadership exhibited greater consistency. Researchers used cutoff points of 27.7% (Baumgarten 1975), 28% (Summers 1970), and 23% (Katz and Lazarsfeld 1955). While any one of these cutoff points is necessarily arbitrary, consistency presumably offers the advantage of inter-study comparability. However, this dissertation suggests that innovativeness is product category specific and that proportions of innovators may vary across product categories. Similarly, it has been accepted in the literature that opinion leadership is product category specific (Katz and Lazarsfeld 1955; King and Summers 1970; 74 Montgomery and Silk 1971), and it is certainly likely that proportions of opinion leaders also vary across product categories. Theoretically, it is not impossible therefore that in some product categories a cutoff point on the basis of assumed proportions of innovators and opinion leaders would include subjects who on average disagreed with items designating them as being innovators or opinion leaders. Theoretical considerations therefore suggest that the dichotomization be based on a particular level of innovativeness and opinion leadership, independent of proportions of respondents who may or may not exceed such level. In this dissertation, it was decided to use an averaged score of 5.0 or higher on the 7-point innovativeness and opinion leadership scales for dichotomization. In other words, innovators and opinion leaders are operationally defined as subjects who on balance agree with items designating them as innovators or opinion leaders. 75 Chapter V Results INTRODUCTION This chapter contains three sections. The first provides a general overview of the results obtained in the survey. The second addresses hypotheses H1 through H5, using the causal model discussed in the previous chapter. The third section addresses hypotheses H6 and H7. Section I. Summary Results SAMPLE Three hundred and thirty-three questionnaires were distributed to administrative and clerical staff at the University of New Orleans. Of these, one hundred and fifty-seven were returned within a three week cutoff period. Subsequently, four questionnaires were eliminated because of severely incomplete responses across all three product categories. In addition, nine questionnaires were eliminated where the respondent was not a member of the intended sample population, including four faculty members and five students. Hence, one hundred and forty-four questionnaires (43.2%) were retained for analysis. This effective response rate of 43.2% within the three week cutoff period indicates that early concerns regarding the repetitive and burdensome nature of the response task were sufficiently addressed in the final questionnaire. 75 76 The mean age of the 144 respondents is 40.7 years, ranging from 20 to 64. They include 50 males and 94 females. For each product category, respondents were able to indicate that they "never or almost never" use a computer, go to a movie theater, or go to a restaurant. Respondents selecting that answer were designated non- users, and all others are designated users; 107 in the personal computer product category, 126 in the movie theater attendance category, and 141 in the restaurant product category. RELIABILITIES Given the changes in the layout and design of the final questionnaire after the pretests, it is appropriate to assess whether these changes impacted on the high degree of reliability attained during pretesting. The comparisons provided in Table V-1 clearly Show that such was not the case. Table V-l. Reliabilities Final Sample - Pretests COMPUTERS Enduring Involvement .8945 .9118 Situational Involvement .9732 .8741 Innovativeness .8343 .8414 Opinion Leadership .8584 .8744 MOVIE THEATER ATTENDANCE Enduring Involvement .9086 .9036 Situational Involvement .9669 .9495 Innovativeness .8808 .9444 Opinion Leadership .8993 .8577 RESTAURANTS Enduring Involvement .8562 .8137 Situational Involvement .9678 .9525 Innovativeness .8327 .8809 Opinion Leadership .8696 .8944 77 VALIDITY The validity of existing measures used in this dissertation has been addressed in depth by previous authors (Zaichowski 1985; Childers 1986; Rogers 1983). Nomological validation of the reconceptualized innovativeness construct is attempted in the context of hypotheses H1 through H7. In addition, an examination is provided below of certain assumptions made, critical to the theoretical framework underlying the hypotheses, regarding the reconceptualized innovativeness construct. Also, the variance of the situational involvement construct is examined with respect to Houston and Rothschild's (1974) prediction that situational involvement varies more between than within product categories. Finally, in the movie theater attendance category, the relationship between two alternative measures of early adoption is evaluated. First, regarding the reconceptualized innovativeness construct, the assumption was made that proportions of innovators vary among product categories. This assumption will be supported if the means of innovativeness scores differ between the three product categories. The results of the analysis of variance reported below in Table V-2 provide such support. 78 Table V-2. Analysis of Variance - Innovativeness by Product Category (3 groups) Sum of Mean F F Source D.F. Squares Squares Ratio Prob. Between Groups 2 568.45 284.23 158.62 (.001 Within Groups 425 761.55 1.79 427 1330.00 Furthermore, it was asserted that the reconceptualized innovativeness construct does not reflect a personality trait. Therefore, regardless of differences among innovativeness scores between different product categories, it is necessary that individual subjects' scores not correlate highly across product categories. Specifically, innovativeness scores should not be correlated to an extent where it would have to be concluded that a single underlying force (i.e., personality trait) gives rise to these scores. Confirmatory factor analysis is used to assess whether the three five-item innovativeness constructs are different across the three product categories. The LISREL model is used to perform the analysis. Table V-3 provides the standardized solution for the model. Depicted are the solutions for the lambda vectors (here factor loadings) and for the phi matrix (the true correlations between the constructs). The results reveal that the fifteen items load on three dimensions (the three product categories), that subjects' innovativeness scores are highly correlated within product categories, but not across product 79 Table V-3. Innovativeness - Confirmatory Factor Analysis Lambda x (factor loadings) Factor 1 Factor 2 Factor 3 Computers Movies Restaurants Computer . Innovativeness item 1 0.893 0.000 0.000 item 2 0.975 0.000 '0.000 item 3 0.576 0.000 0.000 item 4 0.515 0.000 0.000 item 5 0.449 0.000 0.000 Movies ‘ Innovativeness item 1 0.000 0.639 0.000 item 2 0.000 0.781 0.000 item 3 0.000 0.951 0.000 item 4 0.000 0.929 0.000 item 5 0.000 0.579 0.000 Restaurants Innovativeness item 1 0.000 0.000 0.747 item 2 0.000 0.000 0.777 item 3 0.000 0.000 0.837 item 4 0.000 0.000 0.845 item 5 0.000 0.000 0.434 Total coefficient of determination a .998 Phi matrix (True correlations between constructs) Factor 1 Factor 2 Factor 3 Computers Movies Restaurants Factor 1 Computers ‘ 1.000 Factor 2 Movies .093 1.000 Factor 3 Restaurants .106 .553 1.000 80 categories, and that the innovativeness construct therefore does not reflect some underlying personality trait. Next, some comments are in order regarding the situational involvement construct. Houston and Rothschild (1974) assert that situational involvement is more likely to vary between product categories than between subjects (within product categories). The selection of the three product categories was partly motivated by that consideration. That is, they were selected in the expectation that they be dissimilar with respect to situational involvement. Situational involvement was measured following Zaichowski's methodology for measuring involvement and based on the content of Arndt‘s (1967) early single-item measure of the same construct. The validity of this measure may be assessed based on the extent to which it performs according to theoretical expectations. The results of analysis of variance of situational involvement by the three product categories, reported below in Table V- 4, confirm that situational involvement varies more between than within the three product categories selected. Table V-4. Analysis of Variance - Situational Involvement by Product Category (3 groups) Smnof than F F Source D.F. Squares Squares Ratio Prob. Between Groups 2 120.17 60.09 24.06 (.001 Within Groups 423 1056.42 2.50 Total 425 1176.59 81 Finally, the relationship between the two operationali- zations of early adoption in the movie theater attendance category is re-examined. The larger sample resulted in a correlation between the two measures of .72 (p<.001, n=9l). Non-response was substantial on the relative time of adoption measure. In addition, again, several respondents indicated having seen movies prior to their release date. These responses were excluded from the analysis. Yet the overall result seems to confirm the validity of the cross- sectional approach (simply counting the number of new products adopted) in measuring early adoption. Section II. Hypotheses H1 through H5 RESULTS This section reports the findings for hypotheses H1 through H5. Complete results are provided consecutively for each individual product category. Then, hypotheses H1 and H2 are also examined concurrently across the three product categories. A combined summary for all findings relating to hypotheses H1 through H5 can be found in Table V-10. The results for the personal computer product category are provided below in Figure V-l and Table V-5. For the purpose of clarity of presentation, the results for the measurement models (vectors lambda x and y) are provided in Table V-5 only. Multiple item measures are depicted in single boxes in Figure V-l. This too was done for clarity of 82 5 1 E i- E '1 62,- 55 X1 Y,_.' X 2-5 Ax A 1-'-l.l H ”'2 .. ‘164 H “(z-5.2 £1 6 a 1 , C 2 Usage "‘ "' 7° Endurino Z1 Situational ., I nvolve me nt | nvolve me nt 132.1: .174" 62.2 = -.126 C3 '12 £2 Earl u I nnovative ness Adoption "’ 7“110.25 NJ s-s.2 r Y1o Y5-9 E10 E5-9 * significant at p .-.- .IO .— Figure V-l. Structural Model Results for Personal Computers 83 Table V-S. Parameter Estimates for Personal Computers Standardized LISREL 2 Parameters Estimates x lambda 1 1.000 lambda 2 .911 lambda 3 .877 lambda 4 .656 lambda 5 .922 lambda 6 .907 lambda 7 '.931 lambda 8 .977 lambda 9 ..992 lambda 10 1.000 lambda 11 .875 lambda 12 ..984 lambda 13 .590 lambda 14 .510 lambda 15 .502 beta 2 l .174 2.74* gamma 1 l .178 2.99* gamma 2 2 -.126 1.49 gamma 3 l .080 .63 phi l 2 .164 -Chi-square 8 178.36 R-square (eta 1)=.032 GFI=.826 p < .001 R-square (eta 2)=.045 AGFI=.7S7 d.f. = 86 R-square (eta 3)=.022 RMSR=.107 * Significant at p=.10 84 presentation. This convention will be followed throughout this chapter. Table V-S reveals that, for computers, the overall model does not fit the data very well. The structural parameters are all quite small and the relationship between innovativeness and early adoption Shows a Sign opposite to that hypothesized. Tests are conducted on the individual parameters by fixing each structural parameter in turn, re-estimating the model, and examining the differences in Chi-square with one degree of.freedom (see Howell 1987; Steiger, Shapiro and Brown 1985). These individual parameter tests suggest that only the relationships between usage and enduring involvement and between enduring involvement and innovativeness approach statistical significance (see Figure V-2 and Table V-S). A complete discussion of these results will take place in the context of the results for the other two product categories. In the movie theater attendance category, adoptive innovativeness was measured both by using the cross- sectional approach and on the basis of relative time of adoption. First, the results from using the cross- sectional approach are presented in Figure V-2 and Table V-6. Although the overall model again fits poorly with the data, the tests of the individual parameters reveal that this may be primarily attributed to the absence of a 85 6 1 E 1" 5 '-I 52" 55 x1 YI-‘I Xz-s RXIJ TAM-8.1 91.2 = .182 Ax2-s.2 {MINI-i} n‘ E 2 61k .408 Enduri n9 (1 Situational ., lnvolveme nt Involvement 63.1=.706**** 92.1: 263*" x 62.2 = -.2oe** + + - D3 £2 :3 33.2 a -.toe “2 Earl u I nnovative ness Adoption + k 95110.3 7‘9 s-s.2 l on YS‘S 810 55-9 '5'! significant at p I .05 *4” significant at p I .01 *‘fl'i' significant at p = .001 Figure V—2. Cross-Sectional Approach: Structural Model Results for Movie Theater Attendance 86 Table V-6. Cross-Sectional Approach: Parameter Estimates for Movie Theater Attendance Standardized LISREL 2 Parameters Estimates x lambda 1 1.000 lambda 2 .855 lambda 3 .966 lambda 4 .669 lambda 5 .858 lambda 6 .886 lambda 7 .959 lambda 8 .940 lambda 9 . .963 lambda 10 1.000 lambda 11 .707 lambda 12 . .780 lambda 13 .960 lambda 14 .937 lambda 15 .642 beta 2 l .263 7.36*** beta 3 2 -.108 2.36 gamma 1 1 .408 19.88**** game 2 2 -.208 4.83" gamma 3 l .706 76.61**** phi l 2 .182 Chi-square = 172.41 R-square (eta-l) I .117 GFI I .777 p < .001 R-sguare (eta-2) I .075 AGFI - .689 d.f. = 86 R-square (eta-3) I .234 RMSR - .131 ** significant at p80.05 *** significant at p-0.01 *‘** significant at p80.001 87 relationship between innovativeness and early adoption. All other relationships hold quite strongly in accordance with the hypotheses, although the coefficients are admittedly not very high. The model was re-estimated using the direct relative time of adoption approach to the measurement of early adoption behavior. The pattern of the results (see Figure V-3 and Table V-7) is entirely similar, but statistical significance is reduced. One reason for this might be the much smaller number of subjects (N=91) for whom complete data were available for the relative time of adoption measure of early adoption behavior. Next, the results in the restaurant product category setting are provided in Figure V-4 and Table V-8. The significant relationships in this product category are found between usage and early adoption and between situational involvement and innovativeness. None of the other relationships turns out to be significantly different from zero. Finally, the relationships between enduring involvement, situational involvement, and innovativeness are examined across the three product categories combined. The results are reported in Figure V-5 and Table V-9. Both the relationship between enduring involvement and innovativeness and between situational involvement and innovativeness are found to be highly significant. 88 6 1 E 1" E ‘l 52' 5 s XXIJ PHI-".1 91.42: 3.063 AX2-s.2 *‘I'i'lt n' . E 2 61.1=.343 Enduri no (I Situati o nal Involvement Involvement 63.1=.464"" 62.2 = -.I72 Z3 . ”2 £2 Ea rl g I n novative ness Adoption 7"310.3 7‘9 s-s.2 Ym YS-S 510 65-9 **** significant at p I .001 Figure V-3. Direct Approach: Structural Model Results for Movie Theater Attendance 89 Table V-7. Direct Approach: Parameter Estimates for Movie Theater Attendance Standardized LISREL 2 Parameters Estimates x lambda 1 1.000 lambda 2 .881 lambda 3 .970 lambda 4 .698 lambda 5 .847 lambda 6 .890 lambda 7 .988 lambda 8 .960 lambda 9 .967 lambda 10 . 1.000 lambda 11 .719 lambda 12 .830 lambda 13 .935 lambda 14 .936 lambda 15 .653 beta 2 1 .223 2.63 gamma 1 1 .343 8.02 **** gamma 3 l ' .464 16.39 **** phi 1 2 .063 **** significant at p-.001 9O 5 1 E 1" E H 52' 55 X1 Y1-“ XZ‘S k 1-‘i.1 x” 01.2 - .034 1 g Xx?“ 711 E 2 611- .1 25 Enduring (1 Situational + l nvolvo mo nt l nvolvo mo nt 63.1-.491"" . -. 1321 098 62.2 a “03...... + + - U3 :2 Z3 53.2 =-.07a "12 Eo r1 9 I nnovotive ness Adoption " >“$10.3: 7‘9 s-s.2 E10 E5-9 **** significant at p I .001 Figure V-4. Structural Model Results for Restaurants 91 Table V-8. Parameter Estimates for Restaurants Standardized LISREL 2 Parameters Estimates x lambda 1 1.000 lambda 2 .740 lambda 3 .972 lambda 4 .543 lambda 5 .802 lambda 6 .920 lambda 7 .960 lambda 8 .962 lambda 9 .966 lambda 10 1.000 lambda 11 ' .752 lambda 12 .769 lambda 13 .863 lambda 14 .849 lambda 15 .435 beta 2 1 .098 1.07 beta 3 2 -.078 .89 gamma 1 l .125 1.92 gamma 2 2 -.303 10.13**** gamma 3 l ' .491 35.48**** phi 1 2 .034 Chi-square a 246.84 R-square (eta-1) a .016 GFI = .793 p < .001 R-square (eta-2) a .101 AGFI = .711 d.f. = 86 R-square (eta-3) I .247 RSMR . .100 **** significant at p-.001 92 i’12 = .234 Tlt E 2 Enduring Si t uoti onal Involvement Involvement 61.1=.146 **** rl2 l nnovotive ness *i-‘H’ significant at p = .001 Figure V-S. Structural Model Results: Hypotheses H1 and H2 Tested Across the Three Product Categories Table V-9. Parameter Estimates for Hypotheses H1 and H2 Tested Across the Three Product Categories Standardized LISREL 2 Parameters Estimates x lambda 1 .841 lambda 2 .937 lambda 3 .631 lambda 4 .850 lambda 5 .919 lambda 6 .958 lambda 7 .961 lambda 8 .974 lambda 9 .906 lambda 10 - .940 lambda 11 .807 lambda 12 .762 lambda 13 .683 gamma 1 1 .146 47.26 **** gm 2 1 -0442 100.10 *.** phi 1 2 .234 Shi-sguare = 299.76 R-square (eta 1) = .186 p < 0.001 d.f. 3 52 **** significant at p=.001 93 A summary of the results reported in Figures V-2 through V- 6 and Tables V-5 through V-9 is provided below in Table V— 10. The table reports the significance levels found for each of the hypotheses H1 through H5 in the three product- market settings separately and for hypotheses H1 and H2 across the three product categories combined. Table V-lO. Combined Results: Hypotheses Hl-HS Hypotheses Direction Product Categories IV Hl : Enduring Involvement I 11‘ *IIb .III — and Innovativeness + .10 .01 .001 HZ: Situational Involvement and Innovativeness - * .05 * .001 .001 H3: Usage and Enduring Involvement + .10 .001 .001 a at H4: Usage and Early Adoption + * .001 .001 .001 ** H5: Innovativeness and Early Adoption + t t t a to * Not significant ** Not applicable I Computers IIa Movies: using the cross-sectional approach IIb Movies: using the relative time of adoption approach 111 Restaurants IV Computers, Movies (Ila), and Restaurants DISCUSSION Two situational variables were hypothesized to underlie the innovativeness construct (viz., enduring involvement and situational involvement). Viewed in the aggregate, it may be concluded that both impact significantly on innovativeness. Admittedly, situational involvement was not significantly related to innovativeness for computers, while enduring involvement was not significantly related to ...:- 94 innovativeness for restaurants. However, the analysis across all three product categories, increasing both the number of observations and variance in the situational involvement and innovativeness constructs, yielded highly signmificant results for the relationships of both enduring and situational involvement with innovativeness. Inspection of the means and standard deviations for the three constructs in the different product-market settings (Table V-ll) shows why this result could be expected. Clearly, in comparison with the other two categories, a high level of situational involvement for personal computers is attended by a low level of innovativeness. Yet, within the personal computer product category no significant relationship between situational involvement and innovativeness was found (Figure V-l and Table V-S). Table V-ll. Means and (Standard Deviations) Computers Movies Restaurants Innovativeness 2.30 (1.29) 5.02 (1.43) 4.32 (1.30) Enduring Involvement 5.29 (1.46) 5.25 (1.32) 5.18 (1.10) Situational Involvement 6.01 (1.56) 4.71 (1.73) 5.34 (1.45) Hence, while the relationships are not very strong within the personal computer product category or within the 95 restaurant product category, they become very strong when viewed across the three product categories. Possibly, the homogeneity of the sample also created homogeneity with respect to their involvement with these product-market settings, reducing variance within the product categories. The model also included early adoption and usage. In the literature, usage is one of the few constructs consistently shown to be strongly related to early adoption. The results in this dissertation corroborate these previous findings for movies and restaurants. For computers no relationship was found. Note that for computers usage bears no relationship to frequency of purchase. Therefore, the often positive relationship between usage and early adoption may simply be due to a greater frequency of purchase occasions. As hypothesized, usage was also found to have a positive impact on enduring involvement, both for computers and for movies, but not for restaurants. Possibly, heavy users of restaurants lose some of the excitement that may be felt by those who go less often. Finally, no relationship was found between early adoption and innovativeness in any product category. This lack of a relationship between the traditional innovative- ness construct and the reconceptualized innovativeness construct is further addressed in the final chapter. Hypotheses H6 and H7 are examined next in the final section of this chapter. 96 Section III. Hypotheses H6 and H7 RESULTS As noted in the previous chapter, subjects are classified as innovators and/or opinion leaders when their average scores on the innovativeness and/or opinion leadership .scales exceed 5.0, in other words, if on average they agree with scale items designating them as innovators and/or opinion leaders. Figure V-6 shows a cross-tabulation of innovators and opinion leaders for each of the three product categories. Simple inspection of Figure V-6 shows that H6 (All innovators are opinion leaders) does not hold. In each of the three product categories, subjects are found who are innovators, but who are not opinion leaders. H7 (All opinion leaders are not necessarily innovators) is confirmed, replicating previous findings of imperfect overlap of the two constructs. As discussed in the previous chapter, a less stringent test of the general argument underlying hypotheses H6 and H7 examines whether, given the lack of overlap, this occurs more often due to opinion leaders who are not innovators than to innovators who are not opinion leaders. Thus, the null hypothesis is that the number of subjects in cell 2 (non-innovative opinion leaders) is less than or equal to the'number of subjects in cell 3 (innovators who are not opinion leaders), the attendant research hypothesis being that the number of subjects in cell 2 is greater than the 97 LEADERS I R -7 0MPUT R INNOVATO Rs NONI NNOVATO RS I I OPINION I LEADERS 0 : 31 V-7.l i I I OPINION I LEADERS I I movlgg INNOVATORS NONINNOVATORS ! OPINION I LEADERS 35 : 3 I I v-7.2 : i NON 51 E 49 OPINION I LEADERS I I TA R INNOVATORS NONINNOVATORS i OPINION I LEADERS 22 : 34 I I v-7.3 g I I NON I OPINION 23 E 64 I I I 98 number of subjects in cell 3. A one-tailed binomial test of proportions (viz., testing whether the proportion of subjects in cell 2 among the total of subjects in cell 2 and 3 combined is greater than .5) allows rejection of the null hypothesis for the computer product category (31/37 > .50, n=37, p < .01). The results are exactly opposite however with respect to movies (8/59 < .5, see Figure V-6). For restaurants, the results are again as hypothesized (34/57 > .50, n=57, p < .01, see Figure V-6). These conflicting results prohibit a definitive conclusion at this point. In addition, it should be pointed out that only a minor degree of overlap between the two constructs was found for movies (r=.175) and restaurants (r=.068). A slightly negative relationship existed for computers (r=-.118). This result indicates that opinion leaders are no more likely to be independent innovation-decision makers than other members of a target market. That is, they are distinct as purveyors, not creators, of opinion. Chapter VI Conclusion INTRODUCTION This chapter contains four parts. The first part examines the dissertation's achievement of intended theoretical contributions. Additional theoretical considerations and implications are discussed in the second part. Next, future research directions are explored. The final part examines managerial implications. THEORETICAL CONTRIBUTIONS The first major intended contribution of this disserta- tion, as stated in Chapter I, was to reconceptualize the innovativeness construct in response to criticisms and concerns in the literature. Recall that such concerns and criticisms of the state of diffusion research in gen- eral and the traditional approach to the innovativeness construct in particular were summarized in Chapter I as follows: 1. Methodological issues raised by Rogers (1976). These include the absence of consideration of causality, the pro-innovation bias found in most studies, and the general lack of a process ori- entation in diffusion research. 2. The inconsistency and/or weakness of empirical findings in the literature on innovativeness and innovators. 99 :31 100 3. The lack of integration of the behavioral dif- fusion literature with the theory and practice of new product management and new product (con- cept) testing. 4. The lack of integration of the behavioral and modeling diffusion literatures. 5. The need for advances in consumer diffusion theory beyond merely applying concepts from the general diffusion literature. The reconceptualization of the innovativeness construct in this dissertation took each of these issues into ac- count. First, considerations of causality, taking a teleological (idealist) metatheoretical perspective, were introduced through specification of situational variables that lead individuals to engage in innovative behavior. The pro-innovation bias was eliminated by cutting the link between innovativeness and early adoption. The former was taken to represent the making of innovation-decisions independently, while the latter represents one of many possible outcomes of that decision process. A process orientation was included by linking innovativeness to communication behavior. Specifically, the absence of a need for interpersonal communication in the completion of the innovation-decision making process was used to dis- tinguish innovators from non-innovators. Second, explanations in the literature for the in- consistency and/or weakness of empirical findings led to the adoption of a product category specific perspective of innovativeness. This is also the approach taken in the diffusion modeling literature. 101 Third, the resulting reconceptualization has impli- cations for the theory and practice of new product manage- ment and new product (concept) testing. These are dis- cussed in the third and fourth parts of this chapter. Fourth, the new construct is essentially that of the diffusion modeling literature. The diffusion modeling literature implicitly considers innovativeness to be product category specific and links it to the (absence of a) need for communication. The diffusion modeling literature however focuses on adoption, contrary to the approach taken in this dissertation. Finally, by explaining innovativeness on the basis of consumer behavior constructs (enduring and situational involvement) and by integrating existing attempts in the consumer behavior literature toward reconceptualization of the innovativeness construct, consumer diffusion theory is advanced beyond the confines of traditional diffusion research. The second major intended contribution of this disserta- tion was to build and test a causal model for empirical validation of innovativeness as reconceptualized. The empirical results verified that a situational and product category specific approach to innovativeness is valid, and that approaches relying on personality traits to explain innovativeness are not. Specifically, it was shown that subjects' innovativeness scores need not be correlated across product categories. 102 Early adoption and usage were also included in the model, linking the alternative approach of this disser- tation to traditional diffusion research. However, the reconceptualized innovativeness construct showed no rela- tionship with early adoption. That finding is likely attributable to variables intervening between favorable innovation-decisions and subsequent adoption behavior. These intervening variables, based on the empirical results, seem to nullify the impact of favorable innovation-decisions having been made earlier. Earlier adoption instead may depend more on the particular quality of communication received and favorability of the situation (Midgley and Dowling 1978) or "priority acquisition patterns" (Gatignon and Robertson 1985) extant. In addition, earlier adoption does not occur when the innovator rejects the innovation altogether. Finally, as its third major intended contribution, the dissertation endeavored to refine the opinion leadership construct by re-examining the overlap between opinion leadership and innovativeness. As discussed in the pre- vious chapter, conflicting empirical results did not allow for any definitive conclusion. It was also found however that opinion leaders are no more likely to be independent innovation-decision makers than other members of a target market. 103 ADDITIONAL THEORETICAL IMPLICATIONS First, the results clearly indicate that much of consumer behavior at the individual level is product category specific. That is, relationships between innovativeness, usage, early adoption, and enduring and situational involvement generally hold across different product categories. However, the individual consumer who is an innovator, heavy user, early adopter, or enduringly or situationally involved in one product category need not behave similarly in another product category. Hence, previous diffusion research, focusing on personality traits to explain and predict adoption behavior across product categories, will likely remain unsuccessful. Second, the product category specific nature of inno- vative behavior suggests that variables describing aspects of the person-product dyad are likely conducive to expla- nation of such behavior. Such variables include usage, enduring and situational involvement, innovativeness, etc. These are situational variables, describing aspects of a situation made up of an individual and his or her relationship with a product category. Third, the idealist metatheoretical perspective, recognizing human action as consciously motivated in the pursuit of certain goals, was successfully adopted as the foundation for the theoretical framework of this disser- tation. Much of marketing theory takes a managerial perspective, aiming to improve managerial (rather than 104 consumer) decision making. It may be argued that the idealist perspective is well suited to militate against biases (in casu, the pro-innovation bias) arising from researchers' identification with one of the parties to processes studied. That is, the idealist perspective forces explicit recognition of consumers as active participants in the exchange process rather than as passive recipients of marketers‘ manipulations. “is. . FUTURE RESEARCH DIRECTIONS First, research results validated the reconceptualized innovativeness construct in reference to enduring and situational involvement. The results also indicate however that additional explanatory variables relating to innovativeness may be successfully included in the model. Given the confirmed product category specific nature of the innovativeness construct, such additional variables should likely describe aspects of the person-product dyad; for example, a person's information searching behavior with respect to a product category, or perhaps additional dimensions of involvement (e.g., utility). Note that usage also is such a variable and was linked indirectly to innovativeness by way of enduring involvement. Second, no relationship was found between innovative- ness and early adoption. Early adoption however was strongly related to usage. It may be suggested therefore ‘ that early adoption is mostly a function of purchase occasions arising more frequently. If a new product is 105 perceived as superior to existing alternatives, purchases will occur, earlier and later, as purchase occasions arise. The important question in explaining successful dif- fusion of innovations may therefore not be when (earlier or later in the diffusion process) an individual purchases a new product. Instead, attention should be focused on why an individual purchases some new products and rejects many others. In other words, an understanding of the diffusion process and characteristics of adopters at different stages of that process is not sufficient to explain why the diffusion process is set in motion for some products, but not for many others. To explain new product failure, one needs to identify and understand the motivations of early rejectors. Current diffusion research can only shed light on succesful diffusion, not on the more common occurrence of failure. Yet, to achieve success and avoid failure, an understanding of both is needed. Third, consideration of theoretical and empirical contributions from the literature on adoption and diffu- sion of innovations is virtually absent in new product (concept) testing. However, if imitation guides non- innovative adoption behavior, then it is incorrect to rely on non-innovators' evaluations in new product (concept) testing. Such evaluations might not be predictive of their actual behavior in the marketplace. Future research in new product (concept) testing should investigate differences among innovators' and non- 106 innovators' evaluations predicting products' ultimate market acceptance. Similar suggestions to differentiate among potential customers' evaluations of a new product have been made by Taylor (1977), who advocates a focus on heavy users, and Kleyngeld (1974) who advocates a focus on early adopters in a product category. Finally, additional research should be undertaken relating alternative new product entry strategies to differing configurations of situational and enduring involvement found for different product categories. The following part addresses managerial implications and provides tentative recommendations pertaining to this issue. MANAGERIAL IMPLICATIONS First, it was determined that proportions of innovators vary between product categories. Managers should accordingly recognize that sales for new products may be much slower to develop in some categories than in others. Moreover, a new product that does not readily fit into any existing category, for example Pampers when introduced by Procter and Gamble, may lack a constituency of innovators to pass judgment positively or negatively. It took several years before Pamper sales finally took off. Test markets however are normally limited to periods of fairly short duration. When enduring involvement is low and/or situational involvement is high, management should take longer before conceding defeat. 107 Second, a lack of innovators may be caused by low enduring involvement, high situational involvement, or both. Marketers' actions should accordingly be tailored to address a lack of interest, a fear to make incorrect inno- vation decisions, or both. For example, promotion of a new very expensive audio system may be aimed primarily at very high income groups (addressing situational involvement) or primarily at audiophiles (capitalizing on enduring involve- ment). Research prior to introduction can map the target market on a two-dimensional, enduring and situational involvement plane (see Figure VI-l). If subjects are found grouped in quadrant A (Figure VI-l), a high proportion HIGH — di rection of i ncreaai no innovative neaa EIDURI MG INVO LVEHENT — '— LOW 41 LOV HIGH SITUATIONAL INVOLVEI‘IEIIT Figure VI-l. Enduring and Situational Involvement 108 of innovators exists, and success or failure of new products should be quickly determined. If subjects are found generally in quadrant B, marketers may want to focus on strategies enhancing trialability. For quadrant C, a choice exists to increase trialability or to enhance interest in the product category. For quadrant D, interest in the product category would need to be enhanced. In any of these alternative situations however marketers may accept the existing distribution as given and work to identify and inform subjects found in quadrant A. Specifically, smaller numbers of innovators, properly identified, can often be reached quite economically. In such situations a proper entry strategy may be one that spends few resources at any point in time, but is committed to doing so for a long time. Metaphorically, the marketer may choose to patiently teach a targeted social system at the pace at which it is naturally inclined to learn. CONCLUSION The dissertation extended recent attempts in the consumer diffusion literature toward reconceptualization of innovativeness. A new definition of innovativeness was proposed and empirically validated. The new construct was shown to be promising for additional theoretical and empirical work in several areas relating to the diffusion of innovations. The new construct and the theoretical ' considerations on which it is founded were shown to entail 109 significant managerial implications. More importantly, however, they offer new directions from which to approach what may be the primary challenge in marketing, the successful introduction of new products. APPENDI CES Appendix A Pretest Questionnaire: Computers Instructions: Please answer the following questions. Mark your answers on the survey and return the questionnaire in the accanpanying envelope. First we would like to know how often you use a computer. Throughout this questionnaire, when we speak about computers we mean desk top or hone canputers. Please circle the nunber that approximates how often you use a computer: at bone or at work. 1. Never or almost never. 2. Less than 3 times a year 3. Fran3to6timesayear 4. Fran 6 to 12 times a year 5. More than once a month 6. More than twice a month 7. mce a week or more Now we would like you to circle, on a- scale from 1 to 7. the nulber that best describes your agreement or disagreement with the following statements. strongly . strongly agree disagree I know more about new can- puters than most of the people that I talk to. l 2 3 4 5 6 7 I like to read about new com- puters even if I have no inten- tion to go and buy one. 1 2 3 4 5 6 7 IthinkIlookatadsaboutnew canputers more than most people. 1 2 3 4 5 6 7 110 111. Do you yourself own a personal computer? (Please circle correct answer): Yes. No. If 'Yes': How long have you owned it? (Please circle correct number) 1. Less than a year 2. l to 3 years 3. 4 to 5 years 4. More than 5 years If 'No': When do you think you might purchase one in the future? (Please circle correct number) 1. I am in the process of purchasing one 2. Within one year 3. l to 2 years from now 4. More than 2 years from now 5. Never . Below are sets of word pairs. Please circle, on the scales from 1 to 7 provided, the numbers that best reflect your feelings about computers, according to the word pairs below. 1. boring 1 2 3 4 5 o 7 interesting 2. unexciting l 2 3 4 ' 5 6 7 exciting 3. appealing l 2 3 '4. 5 6 7 unappealing 4. mundane l 2 3 . 4 5 6 7 fascinating Next, using similar scales, we would like you to indicate how important it is to you that you do not make a mistake when choosing a computer to buy: 1. important 1 2 3 4 5 6 7 unimportant 2. of no concern 1 2 3 4 5 6 7 of concern to me 3. irrelevant 1 2 ‘ 3 4 s 6 7 relevant 4. means a lot 1 2 3 4 5 6 7 means nothing to me 5. trivial l 2 3 f 4 5 6 7 _ fundamental 6. matters to me 1 2 3 4 5 6 7 doesn't matter to me 7. significant 1 2 ' 3 4 5 6 7 insignificant 1112 Next. we would like to ask a few questions about how you interact with friends and neighbors regarding computers. 1. In general. do you talk to your friends and neighbors about computers? Often Never 1 2 3 4 5 6 7 2. When you talk to your friends and neighbors about computers do you: give a great deal give very little of information ’ information 1 2 3 4 5 6 7 3. During the past six months, how many people have you told about a computer? . told a number told of people no one 1 2 3 4 5 6 7 4. Compared with your circle of friends, how likely are you to be asked about computers? very likely ' not at all to be asked likely to be asked 1 2 3 ‘ 4 5 o 7 5. In a discussion of computers would you be most likely to: listen to your convince your friends friends' ideas of your ideas 1 2 3 4 - 5 6 7 6. In discussions of computers which of the following happens most often? you tell your - your friends tell friends about computers - you about computers 1 2 3 4 ‘ 5 6 7 7. Overall in all of your discussions with friends and neighbors are you: often used as a not used as a source of advice . source of advice 1 2 3 4 5 6' 7 113 Now we have some additional questions on how you would decide on a new canputer to buy. Please circle. on a scale frcm l to 7. the nuIber that best describes your agreement or disagreement with the statenents below. strongly strongly agree disagree 1. I would decide to buy a new computer without asking for advice fran people who had previously bought it. 1 2 3 4 5 6 7 I 2. I 910016 talk with others who had bought 'a new canputer before I would decide whether tobuyit. 1. 2 3 4 5 6. 7 3. I would decide to buy a new canputer based on the opinions of friends who had alreadybought it. 1 2 3 4 5 6 7 4. I would decide whether to buy a new computer before I knew what friends who had bought it thought. 1 2‘ 3 4 5 6 7 5. I would seek advice fron other people who have tried a new cau- puter before I would buy it. 1 2 3 4 5 6 7 6. I vwould get advice fron others who have tried a new canputer when making up my mind about whether to buy it. 1 2 3 4 5 6 7 7. I find it hard to decide whether to buy a new canputer before I learn the opinions of those who have already bought it. 1 2 3 4 5 6 7 8. Before I would buy a new computer I would try to find out what friends who have already pur- chased it think. 9. I would talk with other people who have purchased a new com- puter before I would buy it. 102 I would wait to buy a new computer until I knew whether friends who had bought it thought it was "0k”. 11. I would listen to friends who had bought a new computer before I would buy it. 12.1 would talk with people I know who had bought a new computer before I would decide whether to buy it. 13.1 am one of those people who would decide on buying a new com- puter without consulting others who had previously purchased one. THANK YOU. 1Jl4 Strongly Agree 1 2 1 2 1 2 l 2 1 2 1 2 Strongly Disagree Appendix B Pretest Questionnaire: Restaurants The questionnaire is designed to examine consuners' behavior with respect to dining out in restaurants or having a lunch in a nice restaurant. Therefore, throughout this questionnaire WHEN WE SPEAK ABGJT RESTAURANTS marmrsmrmmsrmmmmm. Instructions: Please answer the following questions. Mark your answers on the survey and return the questionnaire in the enclosed envelope. First we would like to know how often.you go to a restaurant. Please circle the nunber that approxinotes how often you eat in a res- taurant. 1. Never or almost never 2. Less than 3 times a year 3. Frau 3 to 6 times a year 4. Fran 6 to 12 times a year 5. Momeummemealmmuh 6. Mmmrflmntwnnaammmh 7. Once a week or more Now we would like you to circle, on a scale fran l to 7. the nunber that best describes your agreenent or disagreanent with the statenents below. strongly strongly I know more about restaurants agree disagree than most of the people that I talk to. _ l 2 3 4 5 6 7 I like to read about restaurants even if I have no intention to go and eat there. 1 2 3 4 5 6 7 I think I look at ads for restau- rants more than most people. 1 2 3 4 5 6 7 115 Now. during the past six months how many restaurants do you recall having gone to that you had not visited before? (Please circle correct number). How many of these restaurants do you recall having opened for business only fairly recently? (Please circle correct number). Below are sets of word pairs. 1. boring 2. unexciting 3. ' appealing 4. mundane Next. 1.‘ 2. 3. 4. 5. 6. 7. important of no concern irrelevant means a lot trivial matters to me significant 1 1 1 1 l l 1 1 1 l l 2 2 2 2 NNNNNNN 3 3 3 3 using similar scales, UUUUUUU 1. 1. 116 2. 2. 3. 3. 4. 5. 6.. more than 6. 4. 5. 6. more than 6. Please circle on the scales from 1 to 7 provided. the numbers that best reflect your feelings about restaurants. 4 5 4 5 4 5 4 5 6 6 6 6 7 7 7. 7. interesting exciting unappealing fascinating we would like you to indicate how important it is to you that you do not make a mistake when you select a restaurant. . . b . . h I. . UIUIU'IUIUIUIUI OOOOOGO‘ I l ddfilfldflfl unimportant of concern to me relevant means nothing to me fundamental doesn't matter to me insignificant 117 Now we would like to ask a few questions about how you interact with friends and neighbors regarding restaurants. 1. _In general. do you talk to your friends and neighbors about restaurants? Often ' Never 1 2 3 4 5 6 - 7 2. When you talk to your friends and neighbors about restaurants do you: give a great deal give very little of information information 1 2 3 4 5 6 7 3. During the past six months. how many people have you told about a restaurant? told a number told of people no one 1 2 3 4 5 6 7 4. Compared with your circle of friends. how likely are you to be asked about restaurants? very likely not at all to be asked likely to be asked 1 2 3 4 5 6 7 5. In a discussion of restaurants would you be most likely to: listen to your I convince your friends friends' ideas of your ideas 1 2 3 4 5 6 7 6. In discussions of restaurants. which of the following happens most often? you tell your your friends tell friends about movies you about movies 1 2 3 4 5 6 7 7. Overall in all of your discussions with friends and neighbors are you: often used as a not used as a source of advice . source of advice 1 2 3 4 s 6.7 118 Now we have some additional questions for you on how you choose the res- taurants that you go to. Please circle the number that indicates, on a scale of l to 7. how often the following statements hold true for you. very Some- very Never Seldom Seldom times Often Often Always l. I make decisions to visit new restaurants without asking for advice from people who have pre- viously gone there. 1 2 3 4 5_ 6 7 2. I talk with others who have tried new restaurants before I decide whether to go there. 1 2 3 4 5 6 7 3. I decide to visit new restaurants based on the opinions of friends who have already tried them. 1 2 3 4 5 6 7 4. I decide whether to visit new restaurants before I know what friends who have . ‘ tried them think. 1 2 3 4 5 6 7 5. I seek advice from other people who have tried new restaurants before I visit them. 1 2 3 4 5 6 7 6. I get advice from others who have tried new restaurants when making up my mind about whether to go there. 1 2 3 4 5 6 7 7. I find it hard to decide whether to visit new restaurants before I learn the opinions of those who have already . tried them. 1 2 3 4 5 6 7 8. Before I go to a new restaurant I try to find out what friends who have already tried it think. 9. I talk with other people who have tried new restaurants before visiting them. 10.1 wait to visit new restaurants until I _ know whether friends who have tried them think they are 'ok?. 11.1 listen to friends who have tried new restaurants before I go there. 12.1 talk with people I know who have tried new restaurants before I decide whether to go there. 13.1 am one of those people who makes decisions on going to new restaurants without consulting others who have previously gone there. Finally, please fill in the following classification Sex: 1. Male 2. Female Age: THANK YOU VBR! MUCH 1119 Ver Y Never Seldom Seldom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 Some- Very times Often Often Always 4 Position at ONO: 5 6 7 5 6 7 5 6 7 5 6 7 5 6 7 5 6 7 data: Clerical/technical Supervisory Maintenance Administrative Other (specify) Appendix C - Pretest Questionnaire: Movies Instructions: Please answer the following questions. Mark your answers on the survey and return the questionnaire in the enclosed envelope. First we would like to know how often you go to a movie theater. Please circle the appropriate number to indicate approximately how often you go to see a movie in a theater. . Never or almost never . Less than 3 times a year . From 3 to 6 times a year . From 6 to 12 times a year . More than once a month . More than twice a month . Once a week or more NOUbUNI-I Now, please circle the appropriate numbers to indicate which of the following movies, if any, you went to see in a movie theater, and indicate approximately when you went to see them. Date: 1. Top Gun 2. Crocodile Dundee 3. Karate Kid II 4. Back to School 5. Star Trek IV 6. Aliens 7. Ruthless People I l 8. Ferris Bueller's Day Off 9. Down and Out in Beverly Hills 10. Golden Child 120 121 Now we have some additional questions for you on how you choose movies that you go to. Please circle the number that indicates, on a scale of 1 to 7, how often the following statements hold true for you. Very Some- Very Never Seldom Seldom times Often Often Always l. I make decisions to see new movies without asking for advice from people who have pre- viously seen them. 1 2 3 4 5 6 7 2. I talk with others who have seen new movies before I decide whether to go see them. 1 2 3 4 5 6 7 3. I decide to see a new movie based on the opinions of friends who have already seen it. 1 2 3 4 5 6 7 4. I decide whether to see a new movie before I know what friends who have seen it think. 1 2 ' 3 4 5 6 7 5. I seek advice from other people who have seen a new movie before ‘ . I go see it. 1 2 3 4 5 6“ 7 6. I get advice from others who have seen a new movie when making up my mind about whether to go see it. 1 2 3 4 5 6 7 _7. I find it hard to decide whether to go see a new movie before I learn the opinions of those who have already -seen it. 1 2 3 4 5 6 7 8. Before I see a new movie I try to find out what friends who have already tried it think. 9. I talk with other people who have seen a new movie before going to see it. 10. I wait to see a new movie until I know whether friends who have tried them think they are "ok". 11.I listen to friends who have seen a new movie before I go to see it. 12. I talk with people I know who have seen a new movie before I decide whether to go see it. 13. I am one of those people who makes decisions on whether to see a new movie without consulting others who have previously seen it. 122 Very‘ Some- Very Never Seldom Seldom times Often Often Always 1 2 3 4 5 6 7 l 2 3 4 5 6 7 1 2 3 4 5 6 7 l 2 3 4 5 6 7 l 2 3 4 5 6 7 1 2 3 4 5 6 7 123 Now we would like you to circle, on a scale from 1 to 7, the number that best describes your agreement or disagreement with the following statements. agree disagree strongly strongly I know more about recent movies than most of the people that I . talk to. 1 2 3 4 5 6 7 I like to read about movies even if I have no intention to go and see them. 1 2 3 4 5 6 7 I think I look at ads for movies more than most people. 1 2 3 4 5 6 7 Among my friends, I am usually one of the first to have seen a particular new movie 1 2 3 4 5 6 7 I like to see a new movie as soon as possible 1 2 3 4 5 6 7 I don't like going to the theater to see a movie that has been around for several years 1 2 3 4 5 6 7 Next, we would like to ask a few questions about how you interact with friends and neighbors regarding movies. 1. In general, do you talk to your friends and neighbors about movies? Often I ' Never“ 1 2 3 4 5 6 7 2. When you talk to your friends and neighbors about movies do you: give a great deal ’give very little of information information 1 2 3 4 5 6 7 1J24 3. During the past six months, how many people have you told about a-movie? told a number ' told of people no one 1 2 3 4 5 6 7 4. Compared with your circle of friends, how likely are you to be asked about movies? very likely not at all to be asked likely to be asked 1 2 3 4 5 6 7 5. In a discussion of movies would you be most likely to: listen to your . convince your friends friends' ideas of your ideas 1 2 3 4 5 6 7 6. In discussions of movies, which of the following happens most often? ' you tell your your friends tell friends about movies you about movies 1 2 3 4 5 6 7 7. Overall in all of your discussions with friends and neighbors are you: often used as a I not used as a source of advice source of-advice l 2 3 4 5 6 7 Below are sets of word pairs. Please circle on the scales provided the numbers, from 1 to 7, that best reflect your feelings about movies according to the word pairs below. 1. boring I 2 3 4 5 6 7 interesting 2. unexciting l 2 3 4 5 6 7 exciting ‘3. appealing l 2 3 4 5 6 7 unappealing 4: mundane 1 2 3 4 5 6 7 fascinating 2125 Next, using similar scales, we would like you to indicate how important it is to you that you do not make a mistake when choosing a movie to go see in a movie theater. 6. matters to me 1 doesn't matter to me 1. important 1 2 3 4 5 6 7 unimportant 2. of no concern 1 2 3 4 5 6 7 of concern to me 3. irrelevant 1 2 3 4 5 6 7 relevant 4. means a lot 1 2 3 4 5 6 7 means nothing to me 5. trivial l 2 3 4 5 6 fundamental 2 3 4 5 6 2 3 4 5 6 NV“ 7. significant 1 insignificant Appendix D Questionnaire First we would like to know how often you use a computer. When we speak about computers wo mean desk top or home computers. Please circle the number that approximates how often you use a computer, at home or at work. 1. Never or almost never 2. Less than 3 times a year 3. From 3 to 6 times a year 4. From 6 to 12 times a year 5. More than once a month 6. More than twice a month 7. Once a week or more Now we would like you to circle, on a scale from 1 to 7, the number that best describes your agreement or disagreement with the following statements. strongly strongly agree disagree I know more about new com- puters than most of the people that I talk to. l 2 3 4 5 6 7 - I like to read about new com- puters even if I have no inten- tion to go and buy one. 1 2 3 4 ‘ 5 6 7 I think I look at ads about new computers more than most people. 1 2 3 4 5 6 7 Do you yourself own a personal computer? (Please circle correct answer): 1. Yes 2. No If 'Yes': How long have you owned it? (Please circle correct number) 1. Less than a year 2. l to 3 years 3. 4 to 5 years 4 . More than 5 years If 'No': when do you think you might purchase one in the future? (Please circle correct number) . I am in the process of 1 purchasing one 2. Within one year 3. 1 to 2 years from now 4. More than 2 years from now 5. Never 126 127 Below are sets of word pairs. Please circle the numbers that best reflect your PEELIKGS ABOUT COMPUTERS. l. boring I 2 3 4 5 6 7 interesting 2. unexciting l 2 3 4 S 6 7 exciting 3. appealing l 2 3 4 S 6 7 unappealing 4. mundane l 2 3 4 5 6 7 fascinating Similarly, please indicate how IMPORTART I? IS TO YOU THAT YOU DO NOT MAKE A MISTAKE “BEN CBOOSIIG A COMPUTER TO BUY: 1. important 1 2 3 4 5 6 7 unimportant 2. means‘a lot 1 2 3 4 5 6 7 means nothing to me 3. matters to me I 2 3 4 5 6 7 doesn't matter to me 4. significant 1 2 3 4 5 6 7 insignificant Next, we would like to ask a few questions about how you interact with friends and neighbors regarding computers. I. In general, do you talk to your friends and neighbors about computers? Often lever l 2 3 4 S 6 7 2. When you talk to your friends and neighbors about computers do you: give a great deal give very little of information information 1 2 3 4 5 6 7 3. During the past six months, how many people have you told about a computer? told a number , told of people no one 1 2 3 4 S 6 7 4. Compared with your circle of friends, how likely are you to be asked about computers? very likely not at all to be asked likely to be asked 1 2 3 4 5 6 7 128 S.£ In discussions of computers which of the following happens most 0 ten? you tell your your friends tell friends about computers you about computers I 2 3 4 5 6 7 Now we have some additional questions on how you would decide on a new computer to buy. Please circle the numbers that best describe your agreement or disagreement with the statements below. strongly strongly l. I would talk with others agree disagree who had bought a new computer before I would decide whether to buy it. 1 2 3 4 5 6 7 2. I would seek advice from other peeple who have tried a new computer before I would buy it. 1 2 3 4 5 6 7 3. I find it hard to decide whether to buy a new computer before I learn the opinions of those who have already bought it. 1 2 3 4 S 6 7 4. I would wait to buy a new computer until I knew whether friends who had bought it thought it was 'ok'. 1 2 3 4 5 6 7 5. I am one of those people who would decide on buying a new computer without consulting others who had previously purchased one. I 2 3 4 5 6 7 Next, we want to investigate how people make decisions with respect to GOING TO MOVIE THEATERS: Please circle the apprOpriate number to indicate approximately how often you go to see a movie in a theater. I. Never or almost never 2. Less than 3 times a year 3. Prom 3 to 6 times a year 4. from 6 to 12 times a year 5. Hora than once a month 6. More than twice a month 7. Once a week or more 129 Now, please circle the appropriate numbers to indicate which of the following movies, if any, you went to see in a movie theater, and indicate approximately when you went to see them. Date: Top Gun Crocodile Dundee Karate Kid II 0 Back to School Stir Trek IV Aliens Ruthless People Perris Bueller's Day Off ooqmmeuuu 0 Down and Out in Beverly Hills 10. Golden Child how we would like you to circle, on a scale from I to 7, the number that best describes your agreement or disagreement with the following statements. agree disagree . strongly strongly I know more about recent movies than most of the people that I talk to. l .2 3 4 S 6 7 I like to read about movies even if I have no intention to go and see them. 1 2 3 4 S 6 7 I think I look at ads for movies more than most peOple. l 2 3 4 5 6 7 Next we want to investigate how you interact with friends and neighbors regarding movies. I. In general, do you talk to your friends and neighbors about movies? Often lever l 2 3 4 S 6 7 2. When you talk to your friends and neighbors about movies do you: give a great deal give very little of information information 1 2 3 4 5 6 7 130 3. During the past six months, how many people have you told about a movie? told a number . told of people no one 1 2 3 4 5 6 7 4. Compared with your circle of friends, how likely are you to be asked about movies? - very likely not at all to be asked likely to be asked 1 2 3 4 S 6 7 6. In discussions of movies, which of the following happens most often? you tell your your friends tell friends about movies you about movies 1 2 3 4 5 6 7 Below are sets of word pairs. Please circle the numbers that best reflect your PEBLIRGS ABOUT MOVIES. l. boring I 2 3 4 S 6 7 interesting 2. unexciting l 2 3 4 5 6 7 exciting 3. appealing l 2 3 4 S 6 7 unappealing 4; mundane l 2 3 4 5 6 7 fascinating Next, please indicate how IHPORIAIT IT Is TO YOU TBA? YOU DO IO? KARE A HISTAX! “HEN CHOOSING A HOVI! TO GO SE2 II A ROVIB THEATER. 1. important I 2 3 4 S. 6 7 unimportant 2. means a lot I 2 3 4 5 6 7 means nothing to me 3. matters to me I 2 3 4 S 6 7 doesn't matter to me 4. significant I 2 3 4 S 6 7 insignificant how some questions on how you choose movies to go to. Please indicate your agreement or disagreement with the following statements. Strongly Strongly 1. I talk with others who Agree Disagree have seen new movies before I decide whether to go see them. I 2 3 4 S 6 7 131 Strongly Strongly 2. I seek advice from other Agree Disagree people who have seen a new movie before I go see it. 1 2 3 4 5 6 7 3. I find it hard to decide whether to go see a new movie before I learn the opinions of those who have already seen it. I 2 3 4 S 6 7 4. I wait to see new movies until I know whether friends who have tried them think they are 'ok'. 1 2 3 4 5 6 7 5. I am one of those people who would decide to go see a new movie without consulting others - who had previously tried it. 1 2 3 4 5 6 7 Finally, we have some questions about restaurants, WHEN HE SPEAK ABOUT RESTAURANTS THAT DOES NOT INCLUDE FAST FOOD PLACES OR CAFETERIAS. Please circle the number that approximates how often you eat in a res- taurant. 1. Never or almost never 2. Less than 3 times a year 3. From 3 to 6 times a year 4. From 6 to 12 times a year 5. More than once a month 6. Nora than twice a month 7. Once a week or more Now, please circle the numbers that best describe your agreement or disagreement with the statements below. strongly strongly I know more about restaurants agree disagree than most of the people that I talk to. l 2 3 4 S 6 7 I like to read about restaurants even if I have no intention to go and eat there. 1 2 3 4 5 6 7 I think I look at ads for restau- rants more than most people. 1 2 3 4 5 6 7 During the past six months how many restaurants do you recall having gone to that you had not visited before? (Please circle correct number). 0. l. 2. 3. 4. 5. 6. or more 132 Please circle the numbers below that best reflect your FEELINGS ABOUT AESTAU RANTS . l. boring I 2 3 4 5 6 7 interesting 2. unexciting l 2 3 4 S 6 7 exciting 3. appealing l 2 3 4 S 6 7 unappealing 4. mundane l 2 3 4 S 6 7 fascinating Similarly, please indicate how IMPORTANT IT IS TO YOU TNAT YOU DO NOT HARE A MISTAKE Hill YOU SELECT A RESTAURANT. 1. important I 2 3 4 S 6 7 unimportant 2. means a lot I 2 3 4 5 6 7 means nothing to me 3. matters to me I 2 3 4 S 6 7 doesn't matter to me 4. significant 1 2 3 4 S 6 7 insignificant Then, 1. In general, do you talk to your friends and neighbors about restaurants? Often Never 1 2 3 4 5 6 7 2. Nhen you talk to your friends and neighbors about restaurants do you: . give a great deal give very little of information information I 2 3 4 5 6 7 3. During the past six months, how many people have you told about a restaurant? told a number told of people no one I 2 3 4 5 6 7 4. Compared with your circle of friends, how likely are you to be asked about restaurants? very likely not at all to be asked likely to be asked 1 2 3 4 5 6 7 133 5. In discussions of restaurants, which of the following happens most often? you tell your friends your friends tell about restaurants you about restaurants 1 2 3 4 S 6 7 Finally, a few more questions about decision making: Strongly Strongly Agree Disagree 1. I talk with others who have tried new restaurants before I decide whether‘to go there. 1 2 3 4 5 6 7 2. I seek advice from other people who have tried new res- taurants before I visit them. 1 2 3 4 S 6 7 3. I find it hard to decide whether to visit new restau- rants before I learn the opinions of those who have already tried them. I 2 3 4 S 6 7 4. I wait to visit new res- taurants until I know whether friends who have tried them think they are 'ok'. 1 2 3 4 S 6 7 S. I‘am one of those people who makes decisions on going to new restaurants without consulting others who have previously gone there. I 2 3 4 S 6 7 Finally, please fill in the following classification data (by cir- cling the appropriate numbers): Sex: 1. Male Position at ONO: l. Clerical/technical 2. Female 2. Supervisory 3. Naintenance Age: 4. Administrative 5. Other (specify) THANK YOU VERY HUCR Appendix E - Carlson-Grosbart Innovativeness Scale Very Some- Very Never Seldom Seldom times Often Often Always I make decisions to purchase new products and brands without asking for previous purchasers' advice I talk with others who have tried new products and brands before I decide whether to buy them I decide to buy new products and brands based on the opinions of friends who have already tried them I decide whether to buy new products and brands before I know what friends who have tried them think I seek advice from other people who have tried new products and brands before I buy them I get advice from others who have tried new products and brands when making up my mind about whether to buy them I find it harder to decide whether to pur- chase new products and brands before I learn the opinions of those who have already tried them 134 10e 11. 12. 13. Before I buy a new product or brand I try to find out what friends who have already tried it think I talk with other people who have tried new products and brands before purchasing them I wait to buy new products and brands until I know whether friends who have tried them think they are 4! 0k 4! I listen to friends who have tried new products and brands before I buy them I talk with people I know who have tried new products and brands before I decide whether to buy them I am one of those people who makes new product and brand purchase decisions without consulting others who have previously made the purchase 135 Very Some- Very Never Seldom Seldom times Often Often Always Appendix F Zaichowski's Involvement Scale (insert name of object to be judged) important : : : : : : unimportant* of no concern : : : : : : of concern to me irrelevant : : : : : : relevant means a lot to me : : : : : : means nothing to me* useless : : : : : :___useful valuab1e_:_:__:__:_:_:_worth1ess* trivial. : : : :___: :___fundamental beneficial : : :___:___: :___not beneficial* matters to me_:_:__:_:___:___:_doesn't matter* uninterested : : : : : :___interested significant : : : :___:___:___insignificant* vital_:_:_:_:_:___:_superf1uous* boring___:___:___: : : : interesting unexciting : : : :___: :___exciting appealing___:___:___:___:___:___:___unappealing* mundane : : : : : :___fascinating essential : : : : : : nonessentia1* undesirable_:_:__:_:__:_:_desirable wanted : : : : : :___unwanted* not needed : : : : : : needed *Indicates item is reverse scored. Items on the left are scored (1) low involvement to (7) high involvement on the right. Totaling the 20 items gives a score from a low of 20 to a high of 140. 136 Appendix G Major Movies of 1986 MOVIES AND RELEASE DATES MOVIE 1. Top Gun 5-20-86 2. Crocodile Dundee 9-26-86 3. Karate Kid II 6-20-86 4. Back to School 6-13-86 5. Star Trek IV 11-26-86 6. Aliens 7-18-86 7. Ruthless People 6-27-86 8. Ferris Bueller's Day Off 6-11-86 9. Down and Out in Beverly Hills 1—31-86 10. Golden Child 12-12-86 Note: Rank is based on box office receipts. 137 BIBLIOGRAPHY BIBLIOGRAPHY Anderson, P. F. (1983), "Marketing, Scientific Progress, and Scientific Method," Journal gg Marketing, 47 (Fall), 18-31. Arndt, J. (1976), "Role of Product Related Conversation in the Diffusion of a New Product," Journal 9; Marketing Research, 4 (August), 291-295. (1985), "On Making Marketing Science More Scientific: Role of Orientations, Paradigms, Metaphors, and Puzzle Solving," Journal 9f Marketing, 49 (Summer), 11- 23. Bagozzi, R. P. (1980), Causal Models ig Marketing. New York: John Wiley & Sons, Inc. Bass, F. M. (1969), "A New Product Growth Model for Consumer Durables," Management Science, 15 (January), 215-227. Baumgarten, S. A. (1975), "The Innovative Communicator in the Diffusion Process," Journal 9; Marketing Research, 12 (February), 12-18. Belk, R. W. (1975), "Situational Variables and Consumer Behavior," Journal 9; Consumer Research, 3 (December) 1975, 165-168. Bell, W. E., (1963), "Consumer Innovators: A Unique Market for Newness," in Proceedingg 9f Egg American Marketing Association, ed. S. Greyser, Chicago: American Marketing Association, 85-95. Blake, 8., R. Perloff, and R. Heslin (1970), "Dogmatism and Acceptance of New Food Products," Journal 9; Marketing Research, 7 (November), 483-486. Bloch, P. H. and M. L. Richins (1983), "A Theoretical Model for the Study of Product Importance Perceptions," Journal ‘gf,Marketing, 47 (Summer), 69-81. 138 139 Bloch, P. H., D. L. Sherrell, and N. M. Ridgway (1986), "Consumer Search: An Extended Framework," Journal gf Consumer Research, 13 (June), 119-26. Calder, B. L., L. W. Phillips, and A. M. Tybout (1981), Designing Research for Application," Journal 9; Consumer Research, 8 (September), 192-207. Carlson, L. and S. L. Grosbart (1984), "Toward a Better Understanding of Inherent Innovativeness," in 1984 AMA Educators' Proceeding_, Belk et al. (eds. ), Chicago: American Marketing Association, 88- 91. Childers, T. L. (1986), "Assessment of the Psychometric Properties of an Opinion Leadership Scale," Journal 9; Marketing Research, 23 (May), 184-187. Clarke, K. and R. Belk (1979), "The Effects of Product Involvement and Task Definition on Anticipated Consumer Effort," in Advances in Consumer Research, Vol. 6, ed. William L. Wilkie, Ann Arbor, MI: Association for Consumer Research, 313-318. Cook, T., and Campbell, D. (1975), "The Design and Conduct of Experiments and Quasi-experiments in Field Settings," in Handbook of Industrial and Organizational ResearchI ed. Martin Dunnette, Chicago: Rand McNally & Co. Crawford, C. M. (1987), New Product Management, 2nd ed., Homewood, 111., Richard D. Irwin, Inc. Cunningham, S. M., 1966, "Perceived Risk as a Factor in the Diffusion of New Product Information." In Proceedings 9; the American Marketing Association, ed. Raymond M. Haas. Chicago: American Marketing Association, 698-721. Danko, w. D. and J. M. Maclachlan (1983), "Research to Accelerate the Diffusion of a New Invention," Journal 9; Advertising Research, 23 (June/July), 39-43. Darden, W. R. and F. D. Reynolds (1974), "Backwards Profiling of Male Innovators," Journal 9; Marketing Research, 47 (Fall), 79-85. Deshpande, R. (1983), "Paradigms Lost: On Theory and Method in Research and Marketing," Journal 9; Marketing, 47 (Fall), 101-110. Dickerson, M. D. and J. W. Gentry (1983), "Characteristics of Adopters and Non-Adopters of Home Computers," Journal 9; Consumer Research, 10 (2), 225-35. 140 Donnelly, J. H. and J. M. Ivancevich (1974), "A Methodology for Identifying Innovator Characteristics of New Brand Purchases," Journal 9; Marketing Research, 11, 331-4. Downs, G. W. and L. B. Mohr (1976), "Conceptual Issues in the Study of Innovations," Administrative Science anrterly, 21, 700-14. Engel, J. F. and R. D. Blackwell (1982), Consumer Behavior, 4th ed., Chicago: The Dryden Press. Engel, J. F., R. J. Kegereis, and R. D. Blackwell (1969), "Word-of-Mouth Communication by the Innovator," Journal 9; Marketing, 33 (July), 15-19. Engel, J. F., R. D. Blackwell, and R. J. Kegereis (1969), "How Information is Used to Adopt an Innovation," Journal 9; Advertising Research, 4 (December), 3-8. English, H. B. and A. C. English (1958), A Comprehensive Dictionary of Psychological and Psychoanalytical Terms, New York: Longmans, Green and Co. Eveland, J. D. (1979), "Issues in Using the Concept of 'Adoption of Innovation'," Baltimore, Maryland, Paper presented pg the American Society for Public Adm“nistration. Fourt, L. A. and J. W. Woodlock (1960), "Early Prediction of Market Success for New Grocery Products," Journal 9; Marketing, 25 (October), 31-38. Frank, R. E., W. F. Massy, and D. G. Morrison (1964), "The Determinants of Innovative Behavior with Respect to a Branded, Frequently Purchased Food Product," in Proceedingg 2f the American Marketing Associatigg, ed. L. G. Smith, Chicago: American Marketing Association, 312-323. Gatignon, H. J. and T. S. Robertson (1985), "A Propositional Inventory for New Diffusion Research," Journal 9; Consumer Research, 11 (March, 849-867. Hirschman, E. C. (1980), "Innovativeness, Novelty Seeking, and Consumer creativity," Journal 9; Consumer Research, 7 (December), 289-295. Houston, M. J. and M. L. Rothschild (1978), "Conceptual and Methodological Perspectives on Involvement, " in 1978 Educators' Proceedingg, S. C. Jain, ed., Chicago: American Marketing Association, 184-187. Howard, J. A. and J. N. Sheth (1969), The Theogy 9; Consumer. Behavior, New York: Wiley. 141 Howell, R. D. (1987), Covariance Structure Modeling and Measurement Issues: A Note on "Interrelations Among a Channel Entity's Power Sources", Journal 2: Marketing Research, 24 (February), 119-126. Hunt, S. D. (1982), "Are the Logical Empiricist Models of Explanation Dead?" in Marketing Theory: Philosgphy 9; Science Perspectives, R. F. Bush and S. D. Hunt, eds., Chicago: American Marketing, 7-10. Joreskog, K. G. (1982), "Recent Developments in Structural Equation Modeling," Journal 9: Marketing Research, 19 (November), 404-416. Joreskog, K. G. and D. Sorbom (1984), LISREL 2;, 4th ed., Chicago: International Educational Services. Katz, E. and P. F. Lazarsfeld (1955), Personal Influence, New York: The Free Press. King, C. W. (1963), "Fashion Adoption: A Rebuttal to the 'Trickle Down' Theory," in Proceedingg g; the American Marketing Association, ed., S. A. Greyser, Chicago: American Marketing Association, 108-125. King, C. W., Jr., 1965, "The Innovator in the Fashion Adoption Process." In Proceedings g; the American Marketing Association, ed. L. George Smith. Chicago: Amer1can Market1ng Association, 324-339. King, C. W. and G. B. Sproles (1973), "Predictive Efficacy of Psychopersonality Characteristics in Fashion Change- Agent Identification," in Proceedings 9; the American Psychological Association, Washington, DC: American Psychological Association. Kirkpatrick, J. (1982), "Theory and History in Marketing," in Marketing Theory: Philosophy 9; Science Perspectives, R. F. Bush and S. D. Hunt, eds., Chicago: American Marketing Association, 47-51. Kleyngeld, H. P., (1974), Adoption 9; New Food products, Tilburg, Holland: University Press. Kohn-Berning, C. A. and J. Jacoby (1974), "Patterns of Information Acquisition in New Product Purchases," Journal 9; Consumer Research, 2 (September), 18-22. Kohn, C. A. and J. Jacoby (1973), "Operationally Defining the Consumer Innovator," in Proceedings 9; Egg American Psychological Association, Washington, DC: American Psychological Association. 142 Lavidge, R. J., 1966, "The Cotton Candy Concept: Intra- Individual Variability," in Lee Adler and Irving Crespi, Attitude Research 35 ESE: Chicago: American Marketing Association, 39-50. Lazarsfeld, P. F., B. R. Berelson, and H. Gaudlet (1948), The People's Choice, New York: Columbia University Press. Leong, S. M. (1985), "Metatheory and Metamethodology in Marketing: A Lakatosian Reconstruction," Journal 9; Marketing, 49 (Fall), 23-40. Machleit, K. A. (1987), "An Investigation of Affective Res- ponses to Emotional Advertising Appeals", Diss. Michigan State, East Lansing, MI. Mahajan, V. and E. Muller (1979), "Innovation Diffusion and New Product Growth Models in Marketing," Journal 9; Marketing, 43 (Fall), 55-68. Mancuso, J. R. (1969), "Why Not Create Opinion Leaders for New Product Introductions?", Journal 9f Marketing, 33 (July), 20-25. Mansfield, E. (1961), "Technical Change and the Rate of Imitation," Econometrica, 29 (October), 741-766. Midgley, D. F. (1977), Innovation and New Product Marketing, New York: Halstedipress, JoHH’WIIey and Sons, Inc. Midgley, D. F. and G. R. Dowling (1978), "Innovativeness: The Concept and its Measurement," Journal 9; Consumer Research, 4 (March), 229-242. Montgomery, D. B. and A. J. Silk (1971), "Clusters of Consumer Interests and Opinion Leaders' Spheres of Influence," Journal 9; Marketing Research, 6 (August), 317- 321. Myers, J. H. and T. S. Robertson (1969), "Personality Correlates of Opinion Leadership and Innovative Buying Behavior," Journal 9: Marketing Research, 6 (May), 164-168. Ostlund, L. E. (1969), "The Role of Product Perceptions in Innovative Behavior," in Marketing Involvement $5 Society Egg Egg Economy, P. R. McDonald, ed., chicago: American Marketing Association, 259-266. (1972), "A Study of Innovativeness Overlap," Journal 9; Marketing Research, 9 (August), 341-343. 143 (1974), "Perceived Innovation Attributes as Predictors of Innovativeness," Journal 9; Consumer Research, 1 (Septem- ber), 23-29. Peat, N. C., J. W. Gentry, and T. L. Brown (1975), "A Comment on"Identifying Buyers of a Major Automotive Innovation,'" Journal 9; Marketing, 39, 61-2. Peter, J. P. (1982), "Current Issues in the Philosophy of Science: Implications for Marketing Theory - A Panel Discus- sion, " in Marketing Theory: Philosophy Lf Science Perspec- tives, R. F. Bush and S. D. Hunt, eds., Chicago: American Marketing, 11- 16. Peterson, D. R. (1973), "A Note on Optimal Adopter Category Determination," Journal Lf Marketing Research, 10, 325- 9. Rapoport, A. (1969), Operational Philosophy, New York: Har- per and Row, Inc. Reichardt, C. S. and T. D. Cook (1979), "Beyond Qualitative versus Quantitative Methods," in Qualitative Egg Quantitative Methods 12 Evaluation Research, T. D. Cook and C. S. Reichardt, eds., Beverly Hills, CA: Sage, 7-32. Riecken, G. and U. Yavas (1983), "Internal Consistency Reliability of King and Summers' Opinion Leadership Scale: Further Evidence," Journal 2; Marketing Research, 20 (August), 325-6. Robertson, T. S. (1971), Innovative Behavior and Com- munication, New York: Holt, Rinehart, and Winston. Robertson, T. S. and J. N. Kennedy (1968), "Prediction of Consumer Innovators: Application of Multiple Discriminant Analysis," Journal 2; Marketing Research 5 (February), 64- 69. Rogers, E. M. (1962), Diffusion 9; Innovation, Glencoe, IL: The Free Press. (1983), Diffusion g; Innovations, New York: The Free Press. (1976), "New Product Adoption and Diffusion," Journal Lf Consumer Research, 2 (March), 290-301. Rogers, E. M. and F. M. Shoemaker (1971), Communication Lf Innovations, New York: The Free Press. Rozeboom, W. W. (1966), Foundations Lf the meogy Lf Predic- tion, Homewood Ill: Dorsey Press. Runyon, K. E. (1980), Consumer Behavior, Columbus OH: Charles E. Merrill Publishing Co. 144 Ryan, B. and N. Gross (1943), "The Diffusion of Hybrid Seed Corn in Two Iowa Communities," Rural Sociology, 8 (March), 15-24. Steiger, J. H., A. Shapiro, and M. W. Browne (1985), "On thé Multivariate Asymptotic Distribution of Sequential Chi- Square Statistics," Psychometrika, 50 (September, 253-63. Summers, J. O. (1971), Generalized Change Agents and Inno- vativeness," Journal 9; Marketing Research, 8 (August), 313-316. (1972), "Media Exposure Patterns of Consumer Innova- tors," Journal 9; Marketing, 36, 43-9. Summers, J. O. and C. W. King (1969), "Interpersonal Com- munication and New Product Attitudes," in Marketing Involvement Lg Society Egg Egg Economy, Philip R. McDonald, ed., Chicago: American Marketing Association, 292-299. Taylor, J. H. (1977), "A Striking Characteristic of Innova- tors," Journal gfi Marketing Research, 14 (February), 104- 107. Uhl, K., R. Andrus, and L. Paulson (1970), "How are Lag- gards Different? An Empirical Inquiry," Journal 9; Marketing Research, 7 (February), 51-54. Urban, G. and J. R. Hauser (1980), Design and Marketing 9; New Products, Englewood Cliffs, N.J.: Prentice-Hall. Warren, W. E. (1985), "The Adoption and Diffusion of Alterna- tive Residential Long-Distance Telephone Services," Diss., Memphis State, Memphis, TN. Wolman, B. B. (1973), Dictionary 9; Behavioral Science, New York: Van Nostrand Reinhold Company. Yavas, U. and G. Riecken (1982), "Extensions of King and Summers' Opinion Leadership Scale: A Reliability Study," Journal 9; Marketing Research, 19 (February), 154-5. Zaichkowski, J. L. (1985), "Measuring the Involvement Con- struct," Journal g; Consumer Research, 12, (December), 341-52. 3 03082 3037 tum/um"11M:iiimanfliilfil'l’umlil'mmfim