DIFFUSION THEORIESinnovation. Implementation, the fourth stage, fol-
lows a decision to accept and involves putting the
innovation into some use (in either its accepted
form or some modified form). During the final
stage of confirmation, decision makers assess an
adopted innovation, gather information from sig-
nificant others, and choose to continue to use the
innovation as is, modify it (reinvention), or reject
it. While some have criticized the stage model as
too linear, Rogers (1983) has convincingly argued
that existing formulations afford a degree of inter-
pretative and predictive flexibility that averts his-
torical problems with stage models in social science.
Different innovations have different proba-
bilities of adoption and hence, different adoption
rates. That is, they travel through the innovation-
decision process at varying speeds. The literature
demonstrates that five characteristics of innova-
tions influence the adoption decision. Compatibili-
ty refers to the congruence between an innovation
and the prevailing norms, values, and perceived
needs of the potential adopter. Higher levels of
compatibility are associated with greater likeli-
hood of adoption. Innovation complexity, on the
other hand, is negatively associated with adoption.
The extent to which use of an innovation is visible
to the social group—called observability—is posi-
tively related to adoption. Relative advantage refers
to the extent to which an innovation is perceived
to be ‘‘better’’ than the idea, practice, or element
that it replaces. Higher relative advantage increas-
es the probability of adoption. Finally, trialability—
the extent to which an innovation may be experi-
mented with—also increases the probability of
adoption.
The third component of diffusion of innova-
tion theory addresses adopter characteristics.
Adopter categories are classifications of individu-
als by how readily they adopt an innovation. Rog-
ers (1983, p. 260) identifies nine socioeconomic
variables, twelve personality variables, and ten per-
sonal communication characteristics that have been
demonstrated to bear upon adoption choices. In
general, the literature holds that early adopters are
more likely to be characterized by high socioeco-
nomic status, high tolerance of uncertainty and
change, low levels of fatalism and dogmatism, high
integration into the social system, high exposure
to mass media and interpersonal communication
channels, and frequent engagement in informa-
tion seeking.
Identifying the characteristics of people who
adopt innovations raises the question of interper-
sonal influence. Three issues are addressed in the
development of propositions about the role of
interpersonal influence in the innovation decision
process: information flow, opinion leadership, and
diffusion networks. Over time, information flow
has been seen as a ‘‘hypodermic needle’’ model, a
two-step flow (to opinion leaders, then other
adopters), and a multi-step flow. Currently, infor-
mation flows are seen as multi-step in nature and
are described in terms of homophily and heterophily—
the degree to which pairs of interacting potential
adopters are similar or dissimilar. Opinion leader-
ship denotes the degree to which one member of a
social system can influence the attitude and behav-
ior of others. This concept is presently discussed
relative to spheres of influence, wherein a given
person may be a leader or follower depending
upon the part of the diffusion network being
referenced. The diffusion or communication net-
work is the structural stage upon which social
influence takes place. Considerable attention has
been devoted to developing analysis strategies and
tactics for such networks (Wigand 1988).COLLECTIVE BEHAVIORWhile diffusion is not a commonly used term
in collective behavior, processes of diffusion are
important in connection with understanding crowds,
fashion, and some aspects of disaster behavior.
In all cases, analytic concern centers on the dis-
semination of emotions, social practices, or physi-
cal elements through a collectivity. The study of
human behavior in disasters is recent and
multidisciplinary. In this field there has been a
concern with diffusion in the classic sense of track-
ing ideas and practices through networks. The
principle foci of research have been the adoption
of protective measures and the dissemination of
warning messages (Lindell and Perry 1992), with
the aim of research being both the development of
general theories of protective behaviors and more
effective protection of endangered populations.All three diffusion theory traditions converge
in the study of crowd behavior. In proposing
imitation as an explanatory mechanism for crowd
actions, Gabriel Tarde (1890, p. 45) drew upon
Edward Tylor’s concept of cultural diffusion. Sub-
sequently, Gustave LeBon (1895) and Gabriel Tarde