Page 266 -

12 Participatory Approaches 265

agents implement a deep connection between participants and the social simulation
model. Information conveyed in the interaction is relative to the model assumptions,
as well as to the model content.
Ramanath and Gilbert have reviewed a number of software engineering tech-
niques which may be coupled to participatory approaches (Ramanath and Gilbert
2004). This union between software design and participatory approaches is based
on joint production not only between developers but also with end users. Not
only interaction with stakeholders contributes to better software ergonomics—
the computer-supported cooperative work (CSCW) workshop series being an
example—but their participation tends to improve their acceptation and further
appropriation of the mode.
The implementation of interactive techniques may take place at all stages of a
software development process. In early stages, joint application design (Wood and
Silver 1995) allows issues raised to be dealt with during the software development
phase, attributing a champion to each issue. It is also concerned with technical
issues. This protocol might involve other developers, as well as potential users. It
may also increase the computing literacy of the participants involved in the process.
This process is based on the implementation of rather well-framed workshops.
Joint application design is supported by using prototypes. It is here we find
a link with a second technique: prototyping. This technique can be used all the
way through a software development cycle. It is based around providing rough
versions or parts of the targeted product. For example, it allows the pre-product
to be criticised, respecified or the interface improved. Quite close to prototyping,
in the final stages of the process, user panels can be used to involve end users in
assessment of the product. These panels are based on a demonstration or a test of
the targeted product.
In these cases, control of the process is dependent on the hiring of a skilful
facilitator. Otherwise, control of the process may become rather implicit. The
content of the interaction is rather technical, which makes it potentially unbalanced
according to participants’ literacy in computer science. An assessment of 37 joint
application design experiments has shown that the participation of users during the
process is actually rather poor, notably due to the technical nature of debates, which
is hardly compatible with the time allocated to a joint application design process by
users, compared to the time allocated by developers (Davidson 1999). Interaction
is rather superficial and needs translation. However, identification of a champion of
specific tasks gives a little bit more control to participants, as does involvement in
the content of pieces of the tool being developed.
Besides these approaches originating from software engineering, people working
in thematic fields such as the environmental sciences propose co-design workshops
that focus on the development of simulation models. Such workshops are a type of
focus group, organised around the identification of actors, resources, dynamics and
interactions, suitable for a set of stakeholders to represent from a socioecological
system on which they express their own point of view (Étienne 2006). This
approach, which occurs at the design stage of the modelling process, is supposed
to lead participants to design the simulation model by themselves, by formalising