Discontinuance and Diffusion: Examination of the Post Adoption Decision Process

William Black, University of Arizona
ABSTRACT - Behavioral studies concerning the adoption and diffusion of innovations became a primary research area in the 1960's and early 1970's, but have diminished in recent years. One possible explanation is the limited view taken whereby the post adoption decisions of adopters are neglected. This paper first undertakes an interdisciplinary review of past research of post adoption decisions, from which a series of propositions are developed. Then, a conceptual framework and research questions are given.
[ to cite ]:
William Black (1983) ,"Discontinuance and Diffusion: Examination of the Post Adoption Decision Process", in NA - Advances in Consumer Research Volume 10, eds. Richard P. Bagozzi and Alice M. Tybout, Ann Abor, MI : Association for Consumer Research, Pages: 356-361.

Advances in Consumer Research Volume 10, 1983      Pages 356-361

DISCONTINUANCE AND DIFFUSION: EXAMINATION OF THE POST ADOPTION DECISION PROCESS

William Black, University of Arizona

ABSTRACT -

Behavioral studies concerning the adoption and diffusion of innovations became a primary research area in the 1960's and early 1970's, but have diminished in recent years. One possible explanation is the limited view taken whereby the post adoption decisions of adopters are neglected. This paper first undertakes an interdisciplinary review of past research of post adoption decisions, from which a series of propositions are developed. Then, a conceptual framework and research questions are given.

INTRODUCTION

Faced with increased complexity and competitiveness in today's business environment, study of the consumer's behavior is being widely integrated into business planning and strategy. The impacts of social contacts, lifestyle orientations, varying decision processes and the cognitive faculties for choice have all been explored and developed to a degree suitable for use in strategy development. Likewise, consumer research has played a pivotal role in societal causes, especially as an input to public policy in its attempts to educate, inform and protect the consumer.

One area of consumer research which has found widespread application is the adoption and diffusion of innovations. In the business sector, a fundamental element of any growth strategy is the introduction of new products. The importance of increasing the efficiency and success of these product introductions is characterized by the high failure rates,estimated to be as high as 95 percent, and the nigh cost associated with product development. From the societal perspective, similar problems are encountered. A wide variety of "products" need to be introduced, ranging .rom health-care practices to legislative actions. The truly interdisciplinary nature or adoption research and the broad spectrum of applications led to a steadily increasing stream of research in this area in the 1960's and early 1970's (Rogers 1976). The latter part or the 1970's, however, has seen a relative decrease in this research as other areas of consumer research have emerged. A basic question involves identifying the reason(s) for this decline.

DECLINE IN THE STUDY OF THE DIFFUSION OF INNOVATIONS

Given the importance of the adoption and diffusion of innovations to both business and society, this decline can be both puzzling and troublesome. Several possible reasons can be advanced. First, the increased research activity coincided with periods of relative economic prosperity when growth was perhaps the key objective or business strategy. Since the mid-1970's, however, the business climate has changed and growth objectives have been tempered. The fallacy of this argument lies in the fact that the successful introduction of products becomes more, not less, critical as the outcomes or failure become more pronounced.

A second argument centers on the notion that the adoption decision is only a specialized decision process and its specific nature therefore limits its role in the broader scope or consumer research. However, the actual impact of innovativeness is perhaps more widespread than realized, as Hirschman (1980, p.983) contends:

Few concepts in the behavioral sciences have as much immediate relevance to consumer behavior as innovativeness, The propensities or consumers to adopt novel products, whether they are ideas, goods or services, can play an important role in the theories of brand loyalty, decision-making, preference and communication. If there were no such characteristics as innovativeness, consumer behavior would consist or a series of routine buying responses to a static set of products.

The conceptual basis of innovativeness thus seems to have potential for further development and integration into other areas of consumer- research. The stability of the basic theoretical framework found across disciplines also reinforces the validity of the basic constructs.

A third possible argument, at which this paper is directed, is the contention that not enough has been done to fully develop the paradigm. A major shortcoming of most past research is its neglect or a major portion of the conceptual framework, thereby limiting its potential range of academic study and managerial applications. It is the objective of this paper to examine this overlooked construct, the confirmation stage, and examine its relationships to the more widely researched elements or the adoption decision and the diffusion process. To address this question, an interdisciplinary review of research will be undertaken, from which a series or propositions are developed. Finally, a conceptual framework; for further research is proposed.

THE ADOPTION AND DIFFUSION PROCESS

The two major components of innovation research have been the adoption decision and diffusion process. The individual's decision relating to the adoption or rejection of the innovation is one of the most studied areas of innovation research. The most widely accepted model of the individual's decision is termed the innovation decision model (Rogers and Shoemaker 1977). The model employs concepts from Learning theory, post-purchase dissonance and general decision-making processes to postulate four steps: 1) knowledge - the individual is exposed to the innovation by personal contact or social interaction; 2) persuasion - the individual forms a favorable or unfavorable attitude toward the innovation; 3) decision the innovation is adopted or rejected; and, 4) confirmation - the individual seeks reinforcement about the decision, with four possible outcomes. If adopted, the innovation may be used further or discontinued at a later point in time. If rejected, further consideration may result in continued rejection or later adoption.

Study of the individual involves the adoption decision, wherein adoption is defined as "the decision to make full use or a new idea as the best course of action available" (Rogers and Shoemaker 1977,p. 95). While generally used, the operationalization in most studies has only been one of purchase/nonpurchase or use/nonuse. The correct measurement, however, extends past the initial decision and necessitates adding the temporal dimension or usage. The dichotomous operationalization may be meaningful when discontinuous innovations are examined, but even in these cases, the evaluative processes or the confirmation stage and their potentially important effects on subsequent actions are neglected.

One of the unique aspects of innovation research is the dual nature of the process, whereby both relational (interpersonal and social interactions) and monadic (individual characteristics) concepts apply (Rogers 1976). From the monadic perspective, research has generally focused on the wide range of consumer characteristics associated with innovativeness. Profiles of adopter categories have been defined and discriminating variables identified (e.g., Robertson 1971, Donnelly and Ivancevich 1974, Midgley and Dowling 1978, Hirschman 1980).

The relational concept refers to the diffusion process, which involves the spread of the innovation through a social system and is the cumulative result of the individuals' adoption decisions. The classical definition is given by Katz, Levin and Hamilton (1967) as "the (1) acceptance, (2) over time, (3) of some specific item --an idea or practice, (4) by individuals, groups or other adopting units, linked (5) to specific channels of communication, (6) to a social structure, and (7) to a given system of-values, or culture." Past empirical studies have identified two empirical regularities particularly relevant to the study of diffusion processes in general. The first is the normality of the frequency distribution of adopters over time. This has been the basis for classifying adopters either as innovators, early adopters, early majority, late majority or laggards (Rogers 1962). The second is the s-shaped cumulative frequency distribution of adopters over time, indicative or the interpersonal communication process (Barnett 1953, Casetti 1969, Dodds 1955, Griliches 1957).

With past diffusion research extensively oriented toward problems dealing with the initial adoption decision, a result has been the neglect of two important issues. The first is that while most research has dealt with the individual's adoption in terms of only the initial acceptance or rejection, this is not the final decision for the individual. In the confirmation stage the individual may continue adoption or rejection, discontinue use or even adopt after previous rejection or discontinuance. The study of the complete decision process, therefore, necessitates an examination of both initial adoption decisions as well as subsequent decisions.

The second issue concerns the dynamic nature of diffusion. Participants are continually entering and leaving the process, with this continual change in number of users at any one time due to adoption and confirmation decisions that are continuously being made. In most cases, past studies have only dealt with initial adopters. Moreover, little attention has been directed towards monitoring the deletion of adopting units and measuring their impact.

THE POST ADOPTION PERIOD AND THE DISCONTINUANCE DECISION

Study of the confirmation stage can focus on the activities or adopters, rejectors, or both. One element common to any adopter is the discontinuance decision, which terminates use of the innovation. Relatively little research has been conducted concerning the continuance/ discontinuance decision and the resulting discontinuance process. .Moreover, the topic of discontinuance, when examined, has generally been an ancillary research topic. The necessity, however, of studying post adoption behavior is exemplified by Leuthold (1967):

The rate of continuance of a farm operator is often just as important as the rate of adoption in determining the "level of use".... The number of practices used at a particular time, theoretically, is a function or both adoption and discontinuance. Thus, high level or use is necessarily a result of high adoption and high continuance, while low level use may result from either low adoption or high discontinuance.

The following section will briefly examine the various research efforts which have dealt with the discontinuance decision, grouped by discipline. From these studies, a series of propositions can be formed. While space prohibits a comprehensive review of each study, Table 1 will summarize the empirical and/or conceptual support of each study for the propositions.

Rural Sociology

The field of rural sociology was one of the first disciplines to systematically examine the diffusion of innovations. Such research originated with the studies of Ryans and Gross (1943) detailing the adoption and diffusion of hybrid seed corn. A limited number of studies have examined the aggregate process of discontinuance or characteristics of classes of discontinuing adopters. Those studies examining the aggregate process have focused primarily on three topics: 1) the relationship between the rate of adoption and the rate of discontinuance; ?) the impact of discontinuance on the overall level of use; and, 3)characteristics of the aggregate process, focusing on the role of interpersonal communication and the resulting "negative" diffusion process.

Several specific findings concerning the aggregate process merit mention. First, the character of the process as one typified by a diffusion process and being interrelated with tlhe adoption process has been supported by the consistently significant inverse relationships across studies between the rates of adoption and discontinuance. This is an indication of the impact of negative information and performance. A second general finding pertains to the cross-sectional method or assessing innovativeness and the potential misrepresentations resulting from its use. An examination of the number of adoptions and discontinuances led Johnson and van den Ban (1959, p.,) to conclude:

The writers had assumed, and it had been indicated in some studies, that low adopters are low [in innovation usage] because they do not adopt, or are slow to adopt recommended farm practices. The evidence here indicates that they are low because they discontinue adopted practices. [added by author].

Examination of the individual's decision process dealt in three areas: 1) analysis of discontinuance behavior across adopter categories; 2) specification of information; and, 3) an assessment of factors related to the discontinuance decision. In terms of the variation of discontinuance behavior across categories, both Silverman and Bailey (1961) and Bishop and Coughenour (1964) found a decrease in the proportion of innovations discontinued when comparing those adopters of a larger number of innovations to those adopting a lesser amount. The role of interpersonal communication was also found to be potentially substantial. Leuthold (1967, p. 105) found what he termed a "negative interaction effect" and stated:

...the influence of those who discontinue an improved practice in curtailing adoption of the practice may be greater than the influence of those who continue in promoting adoption. Primary evidence here is the .-act that practices exhibiting high discontinuance rates nearly always exhibit low adoption rates, but practices exhibiting low discontinuance rates less frequently exhibit high adoption rates.

Finally, the general trait of innovativeness seems associated with continued use when the characteristics of continuing versus discontinuing adopters are compared. Leuthold and Wilkening(1965), Leuthold(1967) and Jorissen (1969) all suggested the profile of the individual with a lower rate of discontinuance as being similar to those of higher innovativeness.

Political Science

Political scientists have utilized diffusion theory in studying the diffusion of public policies, generally among states (Gray 1973, Walker 1969). A limited number have examined further the impact of the discontinuance process on the overall level of adoption.

Eyestone (1977) examined six public policies (e.g., fair employment policies, desegregation, labor laws) to study and categorize the patterns of diffusion. In the course or the study, the impact and pattern of discontinuance concerning the minimum wage law was detailed. Eyestone reflects on the patterns in the following passage:

Should we now speak of the "extinction" of innovations or the "recission" of innovations by outside forces? It seems that some kind of disinnovation may be as contagious as the initial pattern of adoption. (p.445)

Waves of adoption, readoption, and discontinuance were identified. The patterns of adoption and readoption generally coincided, while the peak periods of discontinuance fell between those of adoption/readoption, possibly characterizing the discontinuance of those who adopted hastily or quickly became disinterested.

Education

Research efforts within the field of education have been both empirical and conceptual. The empirical studies employed the school rather than the teacher as the unit of analysis, masking individual behavior (Aslin and DeArmen 1976, Cawelti 1967, Eicholz 1963). Focus was on the innovation's characteristics and how this affected adoption and continued use. The conceptual developments concentrated in two areas. First, Hall, Loucks, Rutherford and Newlove (1979) developed a schema based on two dimensions -- eight levels of use ranging from non-use to seeking out new innovations or modifying existing ones and seven functions which define the sophistication and intensity of use, allowing a more detailed analysis considering past usage and performance. Eicholz and Rogers (1964) developed a "Rejection-Adoption" model which paralleled the stages of adoption and attempted to describe the process of rejection (non-adoption) or discontinuance in terms similar to those used.in the initial adoption decision. A limited empirical study supported the model and identified a wide range of reasons for rejection, ranging from ignorance, lack of interest or maintenance of the status quo to social progress, interpersonal relations, "mistakes" and experience.

Sociology

The diffusion research tradition is rich within the field of sociology, but the studies of discontinuance are quite limited. An empirical study by DeFleur (1966) was an attempt to develop a general theory of social change that essentially encompassed all four stages of the innovation decision model. Emphasis was placed on specifically relating discontinuance, termed obsolescence in the study, to the process of adoption and defining it as part of one decision process. In addition to relating factors that influence adoption to those affecting discontinuance, he also examined the diffusion aspects in relation to adoption and concluded:

Obsolescence should show a kind of "reverse" diffusion curve. Its form should be opposite to the familiar S-shaped curve describing adoption. There should be a "curve of abandonment" for once-institutionalized behavior forms that are dropping out of the social or cultural system of a given group or society. (p. 318)

Empirical results concerning use of four types or mass media supported this contention by finding diffusion curves or discontinuance in the hypothesized pattern.

Medicine

The study by Coe and Barnhill (1967) focused on the discontinuance of an operationally successful innovation in a setting generally conducive to change. The innovation, a new method of drug prescription and administration, was studied for its acceptance by nursing personnel in several units of a hospital. While empirical results w-ere obtained, the conceptual developments are of particular interest. As implied in its title, emphasis was placed on factors pertaining to both the individual and relationships between individuals. The general hypothesis was that factors which had a positive effect on adoption (e.g., source of innovation, perception of characteristic and impact on organization) also affected the continued usage. The empirical portion of the study showed limited support.

The results indicated that those factors which positively influenced adoption also were necessary for continued use. Although the magnitudes of the relationships were not stated, all had positive relationships with continued usage. The absence of these factors was hypothesized to have led to discontinuance of the innovation.

Marketing

The research tradition in marketing has been toward the initial adoption and diffusion processes. Several studies, however, have examined aspects or the post adoption period. Ahl (1978) examined repeat purchase rates among classes of individuals analogous to adopter categories. Results indicated a greater degree of continued use among earlier adopters, who also exhibit a higher level of total usage. Midgley (1976), in the development of innovative behavior, performed the most rigorous examination of the relationship between adoption and discontinuance and focused on the pivotal role played by interpersonal communication. Of particular interest was the class of adopters (active rejectors) and their impact through negative information. Midgley proposed that active rejectors had an indirect effect on total usage, since they only s low adoption or foster rejection while active adopters transmitting favorable information are responsible for the contagion effect. Madden and Little (1978) formulated the discontinuance decision in an expectancy-disconfirmation framework and Offutt (1981) examined the impact of integration of the innovation with the lifestyle and the effect on continued use.

PROPOSITIONS FROM PAST RESEARCH

A series of propositions based on the research reviewed can be developed concerning both the aggregate discontinuance process as well as the individual's decision. The following section will detail each proposition, with Table 1 indicating the empirical or conceptual support for each proposition among each study.

1. The total number of adopters at a given time period is a function of both the number or initial adopters in that time period and the number of continuing adopters from the previous period. The total number or users, even .n periods of high initial adoption, may be stable or even declining due to the decreasing number of continuing adopters. Studies have shown that the condition of a stable number of users may actually be the result or considerable number or offsetting adoption and discontinuance dec is ions .

2. The discontinuance by adopters has a negative impact on the subsequent number of initial adopters. An inverse relationship exists between the number or discontinuing adopters and the expected number or initial adopters in subsequent time periods. The adoption curve may still increase, but at a slower rate in the presence or discontinuing adopters.

3. Earlier adopters tend to discontinue a lower percent of their total adoptions than do later adopters. Studies point out the consistent differences between early adopters' versus laggards' discontinuance rates, which depict the proportion of total innovations continued. Continuance, shown by repeat purchase rates, has also been shown to be higher for early adopters compared to others.

4. Innovations which have high rates of adoption exhibit lower rates of discontinuance. This proposition relates to two aspects of the discontinuance process. First, research has shown that the shorter the time for complete adoption, the longer will be the period of discontinuance. Second, research also indicates that innovations with high rates of adoption generally have lower rates of discontinuance.

5. Higher levels of adoption tend to be negatively related to rates of discontinuance in later time periods. The inverse relationship between the level of adoption and the rate of discontinuance indicates that innovations in the earlier stages of diffusion exhibit higher incidences of discontinuance. It should be noted that this differs from proposition four in that rates refer to measures of change between periods, while the levels referred to in this proposition refer only to the number of adoptions at a certain time period.

6. Adopters exhibiting a greater tendency toward continuance generally exhibit those characteristics associated with greater innovativeness Among those characteristics shown to be associated to continuance are: 1) exposure to more credible information sources; 9) more exposure to outside information sources; and, 3) Sigher opinion leadership. Those having a greater tendency toward discontinuance were shown to have lower opinion leadership status, less education and less communication exposure.

7. Characteristics or the innovation which facilitate adoption also positively influence continued use. Innovations which are less complex, have a lower cost and a greater ease or administration or use are more readily adopted and have a nigher level of continuance.

8. The impact of negative information can have a substantial impact on the continuance/discontinuance decision. Empirical support of the effect of positive information has been mixed. However, one common element in each case is that negative or unfavorable information, even by a

TABLE 1

EMPIRICAL AND CONCEPTUAL SUPPORT FOR DISCONTINUANCE GENERALIZATIONS

9. The cumulative discontinuance curve approximates the S-shaped diffusion curve. This finding supports the contention that the discontinuance may have underlying elements of "contagion" or other interpersonal processes.

10. The discontinuance decision can be described as a process or several stages. The discontinuance decision can be portrayed in a manner analogous to the adoption decision, based on usage patterns, or as a .unction of individual expectations and motivations.

DEVELOPING A CONCEPTUAL MODEL OF POST-ADOPTION BEHAVIOR

The development of the conceptual model of post adoption behavior must address four critical questions:

1. What process or trait is being examined?

2. What are the components of the post-adoption behavior?

3. What are the factors which could possible affect the continuance/discontinuance decision?

4. What are the interrelationships between the components of the process and the specified variables?

The first question primarily involves the role of innovativeness in discontinuance behavior. The trait ,f innovativeness is a concept basic to the first three stages of the innovation-decision process and acts as an agent for behavior. Innovativeness can be defined as "the degree to which an individual makes innovation decisions independently of the communicated experience of others." (Midgley 1977, p. ;9.) This definition implies a trait inherent in each individual to some degree which combines with communicated experience and the current situation to determine adoption behavior. The decision to adopt the innovation does not transform the trait of innovativeness into an irrelevant factor, but it does decrease it in relative importance given the nature of subsequent decisions in the post-adoption period. The decisions of the post adoption process involve not only those variables that affected the initial adoption decision, but also the factors of experience and subsequent communication of information. These two additional factors, independent of innovativeness, act as influences on subsequent decisions and/or influence the change of previous perceptions or beliefs. It does not seem reasonable, therefore, to define a separate trait relating to the tendency to discontinue, but instead to examine the actual decisions and assess the impact of these additional factors and any situational factors deemed relevant. Thus, this conceptualization will not introduce a new trait but will focus on the impact of additional variables on past decisions to determine future behavior.

A second question involves the definition of the components of the post adoption process. The process has two basic components: (1) the continuance/discontinuance decision; and (9) the length of continued use. The continuance/discontinuance decision is the result of the continuing evaluations evoked by the initial adoption decision. The length of continued use, while determined by the timing of adoption and discontinuance, is a separate component due to its unique impact on the continuance/discontinuance decision.

The third question deals with the factors which possibly have an impact on one or both components of the post-adoption process. The initial set of factors to consider are those that affected the adoption decision. These factors are important because they form the initial set of evaluative criteria utilized and determine the initial expectations as to the expected outcomes from the adoption decision. The five general types of variables are: personal characteristics; social system variables; perceived innovation attributes; communicated experience of others: and. situational considerations.

The continuance/discontinuance decision also is affected by additional sets of variables. These variables relate to changes that have occurred since the adoption decision. These changes may pertain to personal changes, such as communicated experience, personal experience, and post-adoption dissonance. They may also include exogenous events, such as the occurrence of new situational considerations. A second set of factors have an indirect effect on the continuance/discontinuance decision by affecting the "internal" components of the overall process. These factors include the level of diffusion, the length of continued use and the current rates of adoption and diffusion.

The conceptual framework must also allow for defining the role of the post-adoption process of the individual in the total diffusion process, just as is done for adoption. This involves several aspects. The first pertains to the aggregate process of discontinuance and its characteristics. Past studies have indicated that the discontinuance process seemed similar to the diffusion process for initial adoptions. In addition to its particular form, attention must be given to the relationship between adoptions and discontinuances, where empirical analyses have indicated an inverse relationship exists. Finally, the impact of the post-adoption process, especially the decision to discontinue, on the total level of diffusion must be assessed. This involves relating the first two aspects to arrive at the expected effect on the future diffusion levels.

The final task involves the determination of the interrelationships between each of the components of the process and the sets of variables involved. The definition of these interrelationships allows for a conceptual model to be constructed from which some hierarchy of effects can be determined. The specification of the interrelationships resulted in the conceptual model presented in Figure 1. This model depicts the hypothesized relationships between the two components of the individual's decision and behavior -- the continuance/discontinuance decision and the length of continued use -- and the initial adoption decision, sets of influencing variables and the aggregate diffusion process.

FIGURE 1

A CONCEPTUAL MODEL OF THE POST ADOPTION PROCESS

IMPLICATIONS

The topic or diffusion has been extensively examined in several disciplines, but a fundamental component of the process has generally received little attention. This shortcoming has omitted a potentially significant factor from the analysis and limited the full range or applications available to this area or study. These limitations in scope have thus inhibited the integration of adoption and diffusion topics into other areas of consumer research as well as limited the managerial application.

Several avenues exist for continued research. One potentially significant extension is the use of the full innovation-decision model to add a behavioral perspective to the product life cycle. The refinement or marketing strategies would be possible if some systematic characteristics of the behavioral processes underlying this aggregate process can be identified. A second possible application is use of knowledge about the discontinuance process to foster rejection of certain behaviors, such as smoking or drinking. While the innovation-decision model is not the only appropriate means or analysis, hopefully increased attention focused on its unique contributions can continue to provide insights into the consumption process of individuals and organizations.

REFERENCES

Ahl, David H., (1970), "New Product Forecasting Using Consumer Panels," Journal of Marketing Research, 7, 160-7.

Aslin, Neil C. and John W.. DeArman, (1976), "Adoption and Abandonment of Innovative Practices in High School,'! Educational Leadership, (May), 601-606.

Barnett, Homer G., (1953), Innovation -- The Basis or Cultural Change (New York: McGraw-Hill).

Bishop, Rowland and C. Milton Coughenour, (1964), Discontinuance of Farm Innovations (Columbus: Departmental Series, Department of Agricultural Economics and Rural Sociology, Ohio State University, A.E. 361).

Casetti, Emilio, (1969), "Why Do Diffusion Processes Conform to Logistic Trends," Geographical Analysis, 1 101-105.

Cawelti, Gordon, (1967), "Innovative Practices in High Schools: Who Does What -- and Why -- and How," Nation's Schools 79, (April).

Coe, R. M. and E. A. Barnhill, (1967), "Social Dimensions of Failure in Innovation," Human Organization, 96, 149-58.

Coughenour, C. Milton, (1965), "The Problem or Reliability of Adoption Date in Survey Research," Rural Sociology, 30, 184-203.

Coughenour, C. M., (1961), "The Practice Use-Tree and the Adoption, Drop-out, and Non-Adoption of Recommended Farm Practices: A Progress Report," Paper presented at .he Rural Sociological Society Annual Meeting, Ames, Iowa.

DeFleur, Melvin L., (1966), "Mass Communication and Social Changes," Social Forces, 44, 314-96.

Deutschman, Paul J. and A. Eugene Havens, Discontinuance: A Relatively Uninvestigated Aspect or Diffusion (San Jose, Costa Rica: Programs Inter-Americano de Information Popular, 1964).

Dodd, Stuart, C., (1955), "Diffusion is Predictable: Testing Probability Models for Laws or Interaction," American Sociology Review, 90, 399-401.

Donnelly, J. H. and J. M. Ivancevich, (19/4), "A Methodology for Identifying Innovator Characteristics of New Brand Purchases," Journal or Marketing Research, 11, 331-4.

Eichholz, Gerald C., (1961), "Development or a Rejection Classification for Newer Educational Media," unpublished doctoral dissertation (Columbus: Ohio State University).

Eichholz, G., (1963), "Why Do Teachers Reject Change, Theory Into Practice, 9, 964-8.

Eichholz, Gerhard and Everett M. Rogers, (1964), "Resistance to the Adoption of Audio-Visual Aids by Elementary School Teachers -- Contrast and Similarities to Agricultural Innovation," in Matthew B. Miles, ed., Innovation is Education (New York: Columbia University, Teachers College Bureau or Publications).

Eyestone, Robert, (1977), "Confusion, Diffusion and Innovation," American Political Science Review, 71, 441-447.

Gray, Virginia, (1973), "Innovation in the States: A Diffusion Study," American Political Science Review, 67, 1174-85.

Hall, G. E., S. F. Loucks, W. L. Rutherford and B. W. Newlove, "Levels of Use of the Innovation: A Framework for Analyzing Innovation Adoption," Journal of Teacher Education, 96, 52-7.

Hirschman, Elizabeth, (1980), Innovativeness, Novelty Seeking and Consumer Creativity," Journal of Research, (Dec.), 283-85.

Johnson, D. E. and A. 1%. van den Ban, (1959), "The Dynamics or Farm Practice Change," paper presented to the Midwest Sociological Society, Lincoln, Nebraska.

Jorissen, Michael W., (1969), Discontinuance of Innovations by Farmers in Minas Gerais. Brazil, unpublished M.A. thesis (Lansing: Michigan State University).

Katz, Elihu. M. L. Levin and H. Hamilton, (1963), "Traditions of Research in the Diffusion of Innovations," American Sociological Review, 28, 237-253.

Leuthold, Frank O., (1965), Communication and Diffusion or Improved Farm Practices in Two Northern Saskatchewan Farm Communities (Saskatoon, Saskatchewan: University Campus, Center for Community Studies).

Leuthold, Frank O., (1967), Discontinuance of Improved Farm Innovations by Wisconsin Farm Operators, unpublished doctoral dissertation (Madison: University of Wisconsin).

Leuthold, Frank O. and Eugene A. Wilkening, (1965), "Measuring the Dimensions or the Adoption Process and Assessing the Factors Associated with the Acceptance and Continued Ese of New Farm Technology," paper presented at the Midwest Sociological Society, Minneapolis.

Madden, C. S. and E. L. Little, (1978), "Discontinuation Among Adopters of Innovation: An Exploratory Conceptualization," paper presented at the Midwest Marketing Association Meeting).

Midgley, David F., (1976), "A Simple Mathematical Theory or Innovative Behavior," Journal of Consumer Research, 3 (June) 31-41.

Midgley, David F. and Grahame R. Dowling, (1978), "Innovativeness: The Concept and its Measurement," Journal of Consumer Research, 4, 229-42.

Offutt, Nancy, (1981). Perception of Product Attributes and the Innovation Decision Process, unpublished doctoral dissertation (Columbus: The Ohio State University).

Robertson, Thomas S., (1971), Innovative Behavior and Communication (New York: Holt, Rinehart and Winston).

Rogers, Everett M. (1962), Diffusion or Innovations (New York: Free Press).

Rogers, Everett M., (1965), "New Product Adoption and Diffusion," Journal of Consumer Research, 9, 990-301.

Rogers, Everett M. and F. F. Shoemaker, (1977), Communication of Innovations: A Cross-Cultural Approach (New York: The Free Press).

Ryan, Bryce and Neal C. Gross, (1943), "The Diffusion o, Hybrid Seed Corn in Two Iowa Communities," Rural Sociology, 8, 15-24.

Silverman, L. J. and W. C. Bailey, (1961), Trends in the Adoption of Recommended Farm Practices, Alcorn County, Mississippi, 1954-1957 (Mississippi Agricultural Experiment Station Bulletin 617, April.

Walker, Jack L., (1966), "The Diffusion of Innovations Among the American States," American Political Science Review, 63, 880-899.

Wilkening, E. A. (1952) "Acceptance of Improved Farm Practices (Raleigh: North Carolina Agricultural Experiment Station Technical Bulletin 98).

Wilkening, Eugene A. and Frank A. Santopolo, (1952), The Diffusion of Improved Farm Practices From Unit Test- Demonstration Farms in the Tennessee Valley Countries of Northern Carolina (Raleigh: North Carolina Agricultural Experiment Station Mimeo Bulletin).

----------------------------------------