Models and Theories in Human-Computer Interaction/Critique on Diffusion of Innovation model

From Wikibooks, open books for an open world
Jump to navigation Jump to search

Diffusion of Innovation theory is well articulated and extensively used in explaining phenomena. However, I find several factors that, in my opinion, diminish the power of the model in prediction of innovation diffusion and therefore make it harder to integrate in real-life environment.

First, the model defines groups of variables in diffusion research, including environmental factors and adopter’s characteristics that explain their influence on an actor's decision to adopt an innovation. However, it does not specify measures to identify which factors would dominate over others in certain conditions. For example, is political situation overtakes personal characteristics of innovators? Which channel of influence on the diffusion in organizational networks is the most effective as a dependent variable of organizational culture? Is vertical influence (flow of information from upper-level down to employees) more productive than horizontal (flow of information between cross-organizational upper-level) in traditional organizational culture?

In “A Prospective and Retrospective Look at the Diffusion Model” paper, Rogers wrote that the success of STOP AIDS program in San Francisco could not be replicated in other cities because of different environmental conditions. It would be very helpful to understand which diffusion methods should be changed in relation to new factors.

Second, the model does not attempt to explain the nature of S curve of diffusion. Which factors affect the length and the amplitude of the diffusion curve (given same phenomenon)? When it is a clear terminal condition? Using the same example of HIV prevention program in San Francisco, I was wondering when the diffusion of HIV was considered as finished, when number of HIV infections dropped to from 8000 (1983) to 650 (1987)? To 250 (In 2000)? What determines end condition, ineffectiveness of diffused innovation or saturation of the population?

If those and other questions were addressed, the model could be much more useful in prediction success and planning of future innovation diffusion.