Information Technology Reference
In-Depth Information
3
Data Collection and Analysis
3.1
Methods
While there are techniques to include the stakeholders involved in the studied
processes in the model building [6], [7], methods requiring less stakeholder involve-
ment were applied. The intention was to focus on more widely used methods such as
interviews and observation, in order to explore the extent to which they were suitable
to generate evidence for an agent simulation. Qualitative methods in general, and the
ones applied here in particular produce narratives, i.e. text based information about
human actions and mental concepts of their environs. Three such methods were com-
bined for this study:
1.
Participating observation was conducted by a researcher attending several meetings
of the regional assembly and recording the course of the discussions about the
submitted projects. The analysis focused mainly on the arguments used in the dis-
cussions and identifying the features of submitted projects that were regarded im-
portant by the discussants. The observation method is termed as 'participating',
because it seems fair to assume that the presence of the researcher cannot be unno-
ticed in a group of 20-25 persons. However, the researcher did not actively take
part in the discussions, in order to minimize the influence of the observer on the
processes taking place.
2.
A number of qualitative interviews with members of the regional assembly served
as the second empirical source. The interviewees were chosen equally from the
three stakeholder subgroups. The interviews focused on issues that were not appar-
ent from the observations of meetings, such as the arrangements of networks be-
tween the stakeholders (emergence and maintenance of relations), their ways of
collaboration in the assembly (course of the meetings, their personal appraisal, and
decision processes), and finally their own development goals for the region. For the
analysis, the audio recordings of the interviews were transcribed into written text.
3.
Because observations rely to a considerable extent on the interpretations of the re-
searcher, and the interviews focused on those of the assembly members, both
groups of findings were controlled and underpinned by analysis of the official do-
cumentation of the work of the regional assembly.
While observations served to grasp the regularities in the stakeholders' actions, the
interviews were capable of generating information about the motivations behind those
actions. Another advantage of this combination of methods is that it helps to identify
discrepancies between stakeholders' self reflections as expressed in the interviews
and their overt behavior. However, due to the object oriented programming of agent
models, these contradictions may well be a part of the resulting simulation model.
3.2
Observation Results Concerning Assembly Meetings
Using the observation protocols and partly also interview transcripts, the process
of discussions and decisions could be captured. Figure 1 illustrates the prevalent
