Information Technology Reference
In-Depth Information
2
Connecting the Empirical Data with the Consumat
Framework
In the following Figure 1 the basic architecture of the consumat framework is pre-
sented. The core of the model is the consumer decision-making. The type of decision-
making an agent engages in depends on its satisfaction and uncertainty level. For
example, an agent that is satisfied with a fuel car (an opportunity), and not feeling
uncertain about this because everybody else also uses a fuel car will engage in repeti-
tion. In this situation this implies continue using the same car. No new information
about the attributes of fuel or electric cars will be collected, hence the memory will
not be updated. When an agent becomes dissatisfied with the current car, for example
because the variable costs (fuel) rise rapidly, and also uncertain because they increa-
singly see agents with electric cars and are more often confronted with information
about electric cars, the agent may switch towards inquiring, which implies the consul-
tation of other agents concerning the attributes of different cars, and decide on the
optimal car. Agents differ with respect to their uncertainty tolerance and ambition-
level, which affects their tendency to use certain strategies. For example, a highly
ambitious agent that has a high uncertainty tolerance is more likely to engage in opti-
mizing. This causes this agent to be the first to find and adopt attractive alternatives.
As such the consumat approach allows for “growing” phenomena that have been
described in the innovation diffusion theory of Rogers (1993).
The aggregated choice of the agents has an impact on the environment, but also on
culture. In particular it results in a social environment of the agent of other agents that
adopted or not an electric car. Especially when an agent has a strong social need for
conformity, this may result in a low social need satisfaction, stimulating the agent to
consider an alternative.
In simulating the diffusion of electric cars we have to realize that we are entering a
complicated market with many brands, models and fuel types. In the questionnaire
respondents indicate what specific type of car (brand and model) they are interested
in, and the questions address this specific type of car. More specifically, the respon-
dents have to compare a fuel and electric version of a chosen car. This provides us
with a direct comparison between electric and fuel versions of cars, which allows for
the formalization of two opportunities in the simulation model: a fuel car and an iden-
tical electric car. Also respondents have indicated when they expect to purchase a new
car, which also is relevant. The following Table 1 provides an overview of a number
of key demographic variables, showing that the sample of 3000 respondents is repre-
sentative for the Dutch population.
In the following sections we provide an overview of the architecture of the consu-
mat, and describe how the data from the questionnaire can be used in parameterizing
the different elements. Because of the many variables only in a few cases we will go
into detail to give an example of the approach.
