Information Technology Reference
In-Depth Information
network connectivity
Scenario 2 (S2)
Heterogeneous population
High network connectivity
(full connected network)
Scenario 1 (BS)
Homogeneous population
High network connectivity
(full connected network)
Pupulation heterogeneity
Scenario 3 (S3)
Homogeneous population
Low network connectivity
(small-world network)
Scenario 4 (S4)
Heterogeneous population
Low network connectivity
(small-world network)
Fig. 3.
Scenarios (Source: the authors)
Shape of the curves is similar between BS and S3, and between S2 and S4, and
variation into each one of those groups is related with the speed of diffusion. In this
sense, it can be argued that the network topology has implications for the speed at
which diffusion of two innovations occur: the small-world structure network causes
the diffusion occurs at a lower speed, taking longer to saturate the market.
This is because in a fully connected network, every adoption decision increases the
value of the innovation for every one of the individuals in the network, making easier
its own adoption decision; while, in a small-world network, adoptions increase the
innovation value only for those individuals connected with the adopter, so that this
information does not spread efficiently over the network.
Unlike the network topology, population heterogeneity change, not the diffusion
rate, but the shape of diffusion curves. Thus, greater heterogeneity causes an earlier
takeoff point in diffusion curves, so the earlier periods present adopters at higher rates
than those observed for homogeneous population in the same periods.
Further, it is possible to see a double-S shape curve for both, S2 and S4. According
the logic considered in the model, the saddle points appear as a result of three
characteristics: (1) population heterogeneity, (2) asymmetric information and (3)
satisfaction criterion for selection.
Unlike in homogeneous population, the presence of extreme points in the
thresholds of satisfaction of individuals in heterogeneous population ensures the
existence of individuals whose thresholds are below the valuations of innovations
they know.
Given that in the first periods there are many individuals who know only one of the
innovations, the threshold adjustment process is not necessary and, therefore,
adoptions make an early appearance. These adoption decisions increase the value of
innovation for other individuals, so the diffusion process is energized.
However, when the number of individuals who know only one of the innovations
decreases, the diffusion makes slower, because of individuals must now wait until
their thresholds of satisfaction reach suitable levels to make an adoption decision.
This deceleration takes the shape of saddle point, which is observed in the graphs.
Once satisfaction levels reach values around the valuations of the innovations,
diffusion is accelerated again until market gets saturated, so the double-S shape in the
curve is done.
