CHANG, HSIU-JUNG; PHD
 
                         STATE UNIVERSITY OF NEW YORK AT BUFFALO, 1990
 
                         MASS COMMUNICATIONS (0708); INFORMATION SCIENCE (0723); SOCIOLOGY, THEORY AND METHODS (0344)
 

                         The diffusion of innovations is one of the central areas of study in communication science. Typically, the
                         diffusion process has been described by an S-shaped curve in which the cumulative numbers of
                         adopters is plotted with respect to the time of an innovation's adoption. Since this research examines the
                         process of how new ideas or practices diffuse, pro-innovation bias becomes inevitable. The
                         pro-innovation bias implies that an innovation should be diffused and adopted by all members of a social
                         system. It should be neither reinvented nor rejected. One manifestation of this bias is the focus of
                         diffusion research exclusively on adoption to the neglect of disadoption. There has been relatively little
                         research designed to investigate the nature of discontinuance, and as a result little is known about this
                         aspect of the diffusion behavior. One reason for this is the lack of mathematical models to describe this
                         process. This dissertation proposes a mathematical model which describes the patterns of
                         adoption/discontinuance in this dissertation. This model depicts the trend of disadoption and is not
                         inherently pro-innovation. To examine the robustness of the proposed model, five data sets are
                         employed to test the goodness-of-fit. Also, detailed comparisons are made between the proposed
                         model and Barnett, Fink and Debus's (1989) model. The contributions of this dissertation are several.
                         First, the pro-innovation bias in diffusion research can be corrected. Second, an elegant and powerful
                         mathematical model is proposed to depict the process of adoption/discontinuance. Third, the
                         robustness of the mathematical model is tested and critically reviewed. Fourth, different disadoption
                         curves and patterns are examined. Finally, a new model, which is a combination of the logistic,
                         exponential decay and the assymptote, is proposed to describe the process of
                         adoption/discontinuance. Consequently, the process of mathematical modelling can be documented
                         and elaborated to explain variations in the rates at which various innovations are adopted and
                         disadopted.

Social Systems Simulation Group P.O. Box 6904 San Diego, CA  92166-0904 Roland Werner, Principal Phone/FAX  (619) 660-1603