| To Main Proceedings Document
Structuring managerial problem situations
Assessing the suitability of different methodologies
Theo van Mullekom and Jac A.M. Vennix
Department of Methodology, Nijmegen University
P.O. Box 9104
6500 HE Nijmegen, The Netherlands
phone: +31-24-3615568
fax: +31-24-3612351
e-mail: T.vanMullekom@maw.kun.nl, ].V ennix@maw.kun.nl
Introduction
From its origin system dynamics has been employed as a tool to find leverage points to im-
prove the performance of systems. In a meta-analytic study Rouwette et al. (1999) found that
the number of case descriptions of system dynamics model building studies involving the
client in the process of model construction has increased exponentially over the last couple of
decades. Clearly, system dynamics is more and more being used not only as a research
method, but also as an intervention tool in organizations, and quite successful in a number of
cases. Seen from an intervention perspective, system dynamics is not the only methodology
that is used to solve managerial problem situations. There is a host of methods and tech-
niques, most notably those from the realm of soft operational research, also known as problem
structuring methods. All of these methods have proven that they can be successful in structur-
ing policy problems. However, it is not entirely clear if system dynamics and other problem
structuring methods are equally suitable for each type of managerial problem situation.
In the past some researchers have attempted to clarify the similarities and differences be-
tween soft OR and system dynamics and indicated how one could benefit from the other
(Lane, 1994). Some have even made an attempt to design a typology of problem situa-
tions/contexts and accompanying methods (see for example: Geurts et al., 1985; Flood and
Jackson, 1991; Mingers and Brocklesby, 1997). Unfortunately, these typologies are too gen-
eral in nature and at too abstract a level to be helpful as a guide in answering the question
when to apply which methodology.
This paper addresses the latter question in more detail. It takes three different, well estab-
lished methodologies which have been used widely as tools to support strategic decision-
making in organizations, i.e. Group Model Building, Soft Systems Methodology, and Strate-
gic Options Development and Analysis. In order to be able to answer the question when to
use which methodology, one needs to systematically compare their similarities and differ-
ences first. Criteria for comparison include such things as objectives of the method, origina-
tion, operating procedure, the role of models, et cetera. In this paper we will deal primarily
with this comparison. Our objective in the longer run is to specify criteria to make an appro-
priate selection of a method in a particular managerial situation as well. We will start with a
brief description of the three methodologies.
Group Model Building
Group Model Building is based on system dynamics. System dynamics was developed by Jay
Forrester in the second half of the 1950s as a critique to operations research (also called man-
agement science). Forrester regarded the latter paradigm as being too scientific and too
mathematically oriented to apply it to managerial problem situations. Moreover, operations
research had an open-loop approach towards problem solving and decision making. This
means that the results of decisions are seen as having no effect on new decisions (Forrester,
1990; Lane, 1994; Vennix, 1996). System dynamics has a closed-loop view by assuming that
the decision making process consists of information feedback systems. A decision in a social
system affects the environment, which, in turn, affects the decision. The feedback structure of
a system as a whole (instead of individual parts) determines its behavior (Forrester, 1970;
Forrester, 1973; Forrester, 1980; Forrester, 1990; Lane, 1994; Lane and Oliva, 1998; Richard-
son, 1991; Richardson and Pugh III, 1983; Vennix, 1996). Social systems are in general so
complex that it is impossible for the human mind to capture the dynamic consequences of
feedback systems and to allow an analytical solution of the model (Forrester, 1973; Forrester,
1980; Forrester, 1990; Lane and Oliva, 1998; Vennix, 1996; Wolstenholme, 1990). Forrester
(1970: 9) describes this as the counterintuitive character of social systems: "[...] corrective
action [... ] will often be ineffective or even adverse in its results. Very often one finds that the
policies that have been adopted for correcting a difficulty are actually intensifying it rather
than producing a solution. Choosing an ineffective or detrimental policy for coping with a
complex system is not a matter of random chance. The intuitive processes will select the
wrong solution much more often than not." Because of this complexity of social systems, sys-
tems dynamics makes use of simulation (Forrester, 1973; Forrester, 1980; Forrester, 1990;
Lane and Oliva, 1998; Vennix, 1996; Wolstenholme, 1990). The goal for building a simula-
tion model of a system is twofold. In the first place, one tries to gain a better understanding of
the structure and behavior of the system under study. Secondly, one is looking for robust poli-
cies to alleviate problems in the system and to improve its performance (Lane and Oliva,
1998; Vennix, 1996).
A systems dynamics model consists of levels, flows, decision functions, and information
channels. Levels are the accumulators within the system. They represent a state or condition
of a system. Forrester (1980; 1990) makes a distinction between six kinds of levels: materials,
orders, money, personnel, capital equipment, and information. Flows, the second element of a
system dynamics model, transport the content of one level to another. They represent an activ-
ity, while the levels show the resulting state of this activity. Decision functions are the actual
policies taken in order to move the system in a desired direction. Finally, information chan-
nels form the link between levels and decisions. They give the decision-maker the necessary
information to take a decision (Forrester, 1970; Forester, 1973; Forrester, 1980; Forrester,
1990; Richardson, 1991; Richardson and Pugh III, 1983).
The model building process starts with the identification of the problem. This problematic
behavior of the system is depicted in a so-called reference mode of behavior. Next, the system
under study and the problem it contains are transformed into a conceptual model. Depending
on the (problem) situation and the wishes of the client organization, this conceptual model can
be quantified. The model that results, either qualitative or quantitative, is being tested and
validated. The intervention ends with formulating policy measures and their implementation
(Forrester, 1980; Lane and Oliva, 1998; Richardson and Pugh III, 1983; Vennix, 1996).
However, it is not always enough to stop at the end of the process of building a model of a
system and performing simulations. Practice shows that people in organizations often all have
their own view on policy problems. They often disagree if there is a problem at all or not, and
if they agree that there is a problem they frequently disagree on what the problem is precisely.
People construct their own reality by selecting and interpreting the information they receive.
This results in a situation of multiple realities and no viewpoint can be regarded as being the
best or most valid. This makes it difficult to talk about ‘objective’ problems. The problem is
that such situations of multiple realities hamper the effective and efficient functioning of an
organization seriously. These problem situations are often referred to as messy problems
(Vennix, 1996). Involvement of the client in the model building process may overcome, or at
least help to deal with these different and antagonistic viewpoints. Then building a system
dynamics model becomes not a goal in itself, but rather a means to enable problem owners to
learn about the system and to come to consensus among them and commitment with a pro-
posed policy. These participative variants of systems dynamics have become known as Group
Model Building (Andersen, Richardson, and Vennix, 1997; Rouwette et al., 1999; Vennix,
1996). Andersen and Richardson (1997) describe a host of more ore less standardized proce-
dures, techniques, and tools that can be helpful in Group Model Building interventions.
A facilitator guides the Group Model Building intervention. His task is twofold. First, as
for content, he tries to bring the individual mental models of the team members together in a
system dynamics model. Second, as for the group process, he tries to make the individual
group members cooperate so as to actually work as a group (Richardson and Andersen, 1995;
Vennix, 1996). It is best for the facilitator to be a person from outside the organization. Oth-
erwise, he easily gets involved in the content of the model building process. The danger is that
the facilitator becomes a member of the group, which makes it difficult for him to act as a
neutral person (V ennix, 1996).
Group Model Building separates fact from value. As Vennix (1996: 3) puts it: "[...] the
way team members think a system works is separated from how they would like a system to
work". The reason for this separation is that the group process is somehow less politicized.
To keep the group process manageable, Group Model Building implies working with
small groups (V ennix, 1996).
Soft Systems Methodology
Soft Systems Methodology (SSM), developed by Peter Checkland, belongs to the realm of
soft operations research methodologies. SSM is based on systems engineering. Systems engi-
neering is a problem solving approach for dealing with major and complex technological
problems. In this approach one selects an appropriate means to achieve an end which is de-
fined and thereafter taken as given. Systems engineering is similar to approaches such as sys-
tems analysis or operations research. These approaches are in general referred to as hard sys-
tems thinking approaches, to set them apart from soft systems thinking, which grew out of it
(Checkland, 1989; Checkland and Haynes, 1994). The hard approaches, however, turned out
to be less suitable for policy problems, which tend to be messy, ill-defined and changing
(Checkland, 1989). SSM, like other soft operations research methodologies, leaves from such
soft problem situations. This means that SSM has a subjective and interpretive way of looking
at social settings. Individuals are said to have their own perception and interpretation of the
world around them in general and of the problem situations they encounter in particular.
Checkland (1989) calls this the Weltanschauung of people. Both what to do about the prob-
lem and how to do it are problematical. There is no objective view on the problem. There is
no rational, calculable way from goals to means and from means to solutions. People have
different interests at stake, which makes goals part of the problem (Checkland, 1989). At the
same time, however, people involved have to act together in order to take action to alleviate
the problem situation. The multi-perspectiveness makes it a difficult and complex process,
though (Checkland, 1989). Therefore, a neutral consultant is introduced into the problem
structuring process. His task is to guide the process by helping the group to structure the prob-
lem themselves instead of acting as an external expert (Checkland, 1989).
So SSM is not concemed with an objective world, but with the way people try to make
sense of the world in which they live (Checkland, 1989; Lane and Oliva, 1998). To prevent
individual views from clashing, people are in a constant process of negotiation. The goal is to
come to an accommodation (Checkland, 1989; Checkland and Haynes, 1994) between differ-
ent, and often conflicting, views and interests of participants in a process of debate and learn-
ing instead of optimization. It is crucial that actions both lead to improvement of the problem
situation and motivate people to implement defined changes. To realize this, learning and ac-
tion need to alternate in a cyclic process. Since it is not the consultant who has to learn but the
client and to prevent the development of inappropriate models because of miscommunications
between the consultant and the client, SSM should be participative (Checkland, 1989; Check-
land and Haynes, 1994; Patching, 1990).
The SSM process is as follows. First, an organized finding out about the problem situation
takes place. This contains making a model of every viewpoint on the problem situation and
how it should be dealt with. These models are called root definitions. It is important to note
that these models are not meant as a representation of the 'real world’. Their function is to
structure debate about how reality is perceived by individual participants. In the next stage,
the root definitions are transformed into qualitative, conceptual models of human activity sys-
tems. The conceptual models are compared with reality or, perhaps more correctly, with real-
ity as people involved perceive it. This comparison provides clues for improving a problem-
atical situation. Changes are then proposed and implemented. However, there are no perma-
nent solutions. So SSM is a process that is in fact never ending (Checkland, 1989; Checkland
and Haynes, 1994).
Strategic Options Development and Analysis
Strategic Options Development and Analysis (SODA) is a problem structuring methodology
that is focused on structuring complex and messy problems (Eden, 1989). It was developed by
Colin Eden. SODA tries to reach consensus and commitment to action. Eden (1989) states
that a SODA intervention is successful only when consensus rather than compromise, and
commitment rather than agreement have been reached. The client group needs to have the
feeling it ‘owns' the outcomes. This makes SODA not so much concerned with finding the
‘right answer’, but reaching a point where those involved feel confident to take action (Eden,
1989).
Like SSM, SODA can be classified as a soft operations research methodology. In SODA
there is no such thing as an objective world. SODA takes the individual perspective as its
starting point. Every individual problem owner has a personal view on the problem situation.
Each view is equally valid and equally necessary to come to consensus and commitment to
action. By forming a personal view on a problem situation, people try to make sense of their
world in order to manage and control it (Eden, 1989). This process of sensemaking leads to a
view of an organization as a changing set of coalitions in which forces of politics and power
determine decision-making to a large extent. To use Eden's words: "[...] organizations are a
negotiated enterprise whose participants are continuously negotiating and renegotiating their
roles within it" (Eden, 1989: 25-27).
The consultant is the instrument for facilitating this negotiation. His task is to bring to-
gether the different perspectives that exist among the group of problem owners, analyze the
implications of these perspectives, and guide the group in the process of formulating a com-
mon action strategy to which all members feel committed. The key notion is that the consult-
ant helps the group work with the problem situations they face, instead of acting as an expert.
The consultant has to pay equal attention to the content of the problem and the group process
of working on the problem together. To be able to achieve these goals the consultant must feel
confident to improvise and to work in a contingent and cyclic way (Eden, 1989). As the client
group must be considered a small number of significant people, instead of the organization as
a whole. In practice, a ‘small number’ comes down to three to ten people. With 'significant'
people is meant people who are in position to act as an intermediary between the client or-
ganization and the consultant and who have the power to bring about proposed changes result-
ing from the intervention (or at least to make them bring about). Moreover, a social factor is
also of great importance. Client and consultant should feel comfortable with each other and
with each other's way of working (Eden, 1989; Eden and Simpson, 1989).
SODA is based on the personal construct theory of George Kelly. This theory states that
people try to make sense of their world in order to be able to control it better. Kelly formu-
lates his theory around a fundamental postulate, which says that: "A person's processes are
psychologically channelized by the ways in which he anticipates events" (Kelly, 1955: 46).
The meaning of this postulate is that an individual is not so much an object, which is tempo-
rarily in a moving state but is himself a form of continuous motion. This makes it hard to an-
ticipate real events. Nonetheless, individuals seek prediction in order to be prepared for events
that may happen in the future. Therefore, an individual 'channels' his thoughts and actions
towards his goals. As Kelly (1955: 49) puts it: "Anticipation is both the push and the pull of
the psychology of personal constructs".
The individual viewpoints are the crucial input of the problem solving process in SODA.
Such a viewpoint is depicted in a model, which is called a cognitive map. This is a network of
ideas linked by arrows. The arrows are not causal relationships, but indicate the way in which
one idea may lead to, or have implications for, another (Eden, 1989). The SODA process be-
gins with making cognitive maps at an individual level. In this stage a cognitive map is made
of the point of view of every individual from the client group. Next, all individual cognitive
maps are merged into an aggregated map, called a strategic map. This strategic map is a rich
and qualitative description of the way in which the participants see the problem and it gives
the group an instrument to move to action. In forming a strategic map the consultant tries to
find a joint problem definition and consensus and commitment for an action strategy. The
consultant analyzes the strategic map, trying to identify so-called emerging themes and core
concepts. In the final stage, the SODA workshop, the goal is to enable the client group to de-
velop a mutual problem definition so that they can move to action (Eden, 1989).
Comparison
One can compare the three problem structuring methods on a whole lot of criteria. In our
view, the most important criterion is the general purpose the problem structuring methods
pursue. After all, when the methodologies do not strive for the same objective, it seems imag-
inable that the problem structuring processes resulting from each methodology are not the
same either. Other criteria for comparison, with which we deal at this moment, are worldview,
origination, and operating procedure.
General purpose
In what respect do the three problem structuring methods correspond to each other with re-
spect to their general purpose and on what point do they differ? A first point that stands out is
that the methodologies all have the same general goal. They all aim at structuring managerial
problem situations, which should enable the group to take well thought-out decisions. Another
important goal of the three problem structuring methods is that they urge participants to actu-
ally learn, which gives them more insight into the problem situation. In this way, the aim is to
realize a common vision on the problem situation. This, in its turn, enhances the chances that
participants commit themselves to decisions taken.
Taking a closer look, however, there seem to be some fundamental differences with regard
to the general goals of the various problem structuring methodologies. Group Model Building,
for example, seeks for robust strategies. This connotation, typical for the field of system dy-
namics, entails that policy measures do not break down with the slightest headwind. This of-
ten happens with strategies that are not robust, because, for example, they result in different
outcomes in different policy contexts, which makes decision-makers draw back the measures
in dissatisfaction. SSM and SODA on the other hand, do not strive for robust policies. These
two methodologies state that decisions are 'good' when participants in the problem structuring
process agree with them. In that sense every policy measure that is supported by the group is a
‘good' measure, even though it has little or no practical results. In looking for robust strate-
gies, Group Model Building sets itself somehow apart from SSM and SODA.
As became clear from the description of the methodologies, Group Model Building strives
(next to looking for robust policy) for learning, consensus, and commitment. In SSM, the fo-
cus is on an accommodation of interests. SODA is oriented towards consensus and commit-
ment. So on this point, the general goals of the three problem structuring methods seem to be
more or less similar: to align the points of view of problem owners, which might make it eas-
ier to move towards a common action strategy. The question is, however, whether this appar-
ent harmony is false. The crux is that Checkland states that SSM looks for an accommodation
of interests. Looking at the meaning of the word accommodation, one could get the impres-
sion that it is about a conciliation or adaptation of viewpoints. Participants in the problem
structuring process have to adjust their own point of view to that of the group. Participants
have to display, more or less, a certain kind of obligingness to make a common point of view
for the group possible. The question is whether this is the same or that it is fundamentally
different from the notion of consensus within Group Model Building and SODA. After all,
consensus encompasses a congruence of feelings, a general analogy of views. An accommo-
dation of interests (SSM) seems to lead to decisions that are compromises, while consensus
(Group Model Building and SODA) refers to true common decision-making to which all par-
ticipants in a group agree. However, things could even be more subtle than the above sug-
gests. Eden states that consensus is a general purpose of SODA. At the same time, however,
Eden sees organizations as a changing set of coalitions in which participants are continuously
negotiating and renegotiating their roles within it. Is it possible that the end product of such
negotiation processes can be labeled as something like consensus? Or is it, like in the case of
SSM, a compromise? To put it in other words, decisions in SSM and SODA might in fact be
compromises, while Group Model Building seems to strive for consensus in the true meaning
of the word.
In conclusion, the three problem structuring methods seem to have the same general goals,
but this accordance might not hold when taking a closer look. Group Model Building strives,
first, for robust strategies, and second, for learning, consensus, and commitment. Within SSM
and SODA the belief is that it is impossible to realize actual consensus. In this view, com-
promises are the best possible result.
Worldview
The three methods all leave from a world of multiple realities. Each participant in the problem
structuring process has his own view on the problem situation and each view is equally valid.
This means there is no vision that is 'right' or ‘wrong’. Consequently, it is not possible to ob-
jectively and indisputably determine what the problem is. The three problem structuring
methods have this subjectivity towards the problem situation in common.
Group Model Building is, out of these three methodologies, the only one to incorporate
the feedback concept. Systems are said to be steered by information feedback and solutions to
problem situations should be sought for in this direction. According to Group Model Build-
ing, however, the bottleneck in the group process remains that every participant has his own
view on the problem situation. In this way, Group Model Building combines the somewhat
objective notions of system dynamics with the strong subjective notions of SSM and SODA.
In short, all three problem structuring methods leave from a world of multiple realities. An
important difference is that Group Model Building, in contrast to the other two methodolo-
gies, describes and analyzes problem situations on the basis of feedback relations.
Origination
The origination of Group Model Building and SSM shows similarities. System dynamics, on
which Group Model Building is based, originated as a reaction to hard operations research.
Forrester stated that the mathematical and optimizing methods of the hard operations research
paradigm were incapable of handling strategic problem situations. SSM is based on systems
engineering, which can also be classified as a hard operations research paradigm. It is Check-
land's statement that these hard operations research approaches were not suitable for messy,
ill-defined, and changing problem situations. So both Group Model Building and SSM origi-
nated from a reaction to other methodologies that did not have an effective and convincing
answer to messy problem situations. However, the theoretical body of thought that resulted
out of these critiques differs fundamentally for Group Model Building and SSM. Group
Model Building takes as a point of departure individuals in organizations who can have quite
divergent points of view indeed, but in spite of this diversity in viewpoints it is possible to
come to consensus among the group of problem owners. The systems in which these indi-
viduals find themselves are characterized by feedback relations and can be steered. SSM
states that the differences in viewpoints among the problem owners are so divergent and so
opposite to each other that consensus is impossible. A compromise is the best one can reach.
Kelly's personal construct theory seems to share the subjective notion of individual problem
construction with Group Model Building and SSM.
Operating procedure
The operating procedures of the three methodologies have some characteristics in common. In
the first place, they are all participative problem structuring methodologies in which the group
process in particular is the steering factor. Participation of the client in the problem structur-
ing process is believed to be necessary for a valid representation of the problem situation and
for taken the most appropriate decisions, which have the client's support.
A second common characteristic is that Group Model Building, SSM, and SODA are, in
general, small group methodologies. Approximately up to ten people can participate in the
problem structuring process. If the group gets too large, this may hinder the progress of the
group process. One runs the risk then that there is not enough room for individuals to express
their own view on the problem situation, which, in tum, can be detrimental to learning and
commitment.
Third, the three problem structuring methods all have a neutral person who guides the
group process. Neutral in this sense means that the person is from outside the client organiza-
tion. This is to prevent the guide from becoming a participant in the group process, which
makes it difficult or even impossible for him to guide the group process objectively. In Group
Model Building this person is called the facilitator. SSM and SODA call this person the con-
sultant. Although the terms are different, the role of the person involved seems to be the same:
to try to enable the problem owners to learn and to come to a point where they can take deci-
sions.
However, there are also considerable differences in the processes that lead to making a
system dynamics model, root definitions, or a strategic map. SSM and SODA have a strong
subjectivist way of working by taking the individual as the point of departure. Models are
made of the point of view of each individual (root definitions and cognitive maps respec-
tively). Subsequently, SODA brings the individual models together in a group model, while
SSM does not result in a group model. The subjectivity is seen as the strength of the interven-
tion. Group Model Building also starts from looking at the individual as the building stone of
the group process. This methodology, however, states that the subjectivity hampers the group
process and joint decision-making. Therefore, Group Model Building tries to ‘objectivate’ the
group process by separating fact from value. In that sense, Group Model Building is diagnos-
tic in essence, while SSM and SODA depict situations how participants would like it to be.
Summary and concluding remarks
In this paper we have tried to make a systematic comparison between three problem structur-
ing methods: Group Model Building, SSM, and SODA. We have summarized this compari-
son in the following table:
Group Model Building |SSM SODA
General pur-| + robust policies * accommodation of|* consensus and com-
pose + leaming, consensus,) interests, through a/ mitment to action
and commitment cyclic process of
learning and action
Worldview * information feedback |* multiple realities, | * multiple realities,
* multiple realities,| strong subjectivism strong subjectivism
subjectivism
Origination | + system dynamics, | * hard systems thinking|* personal construct
operations research in general, systems| theory
engineering in par-
ticular
Operating * building a system] formulating root defi- | * making individual
procedure dynamics model with] nitions and subse-| cognitive maps and a
the client and imple-} quently conceptual} joint strategic map,
menting resulting pol-| models, proposing] enabling the client to
icy measures; separa-| and implementing] move to action
tion of fact and value measures * guided by a neutral
* guided by a neutral|* guided by a neutral} consultant
facilitator consultant
However, the answer to the question in what kind of managerial situation to apply which
methods has not been answered yet. We see the quest for this answer as an important chal-
lenge for both researchers and policy-makers.
References
Andersen, D.F., Richardson, G.P. (1997). Scripts for Group Model Building. System Dynam-
ics Review. 13(2): 107-129. Cambridge, MA: System Dynamics Society.
Andersen, D.F., Richardson, G.P., Vennix, J.A.M. (1997). Group Model Building. Adding
more science to the craft. System Dynamics Review. 13(2): 187-201. Cambridge, MA:
System Dynamics Society.
Checkland, P. (1989). Soft Systems Methodology. In: Rosenhead, J. (ed.). Rational analysis
for a problematic world. Problem structuring methods for complexity, uncertainty and
conflict, pp. 71-100, Chichester [etc.]: John Wiley & Sons.
Checkland, P.B., Haynes, M.G. (1994). Varieties of systems thinking. The case of Soft Sys-
tems Methodology. System Dynamics Review. 10(2-3): 189-197. Cambridge, MA: System
Dynamics Society.
Eden, C. (1989). Using cognitive mapping for strategic options development and analysis
(SODA). In: Rosenhead, J. (ed.). Rational analysis for a problematic world. Problem
structuring methods for complexity, uncertainty and conflict, pp. 21-42, Chichester [etc.]:
John Wiley & Sons.
Eden, C., Simpson, P. (1989). SODA and cognitive mapping in practice. In: Rosenhead, J.
(ed.). Rational analysis for a problematic world. Problem structuring methods for com-
plexity, uncertainty and conflict, pp. 43-70, Chichester [etc.]: John Wiley & Sons.
Flood, R.L., Jackson, M.C. (1991). Creative problem solving. Total systems intervention,
Chichester [etc.]: John Wiley & Sons.
Forrester, J.W. (1970). Urban dynamics. Third printing. Cambridge, Massachusetts: M.I.T.
Press. [1969].
Forrester, J.W. (1973). World Dynamics. Second edition. Cambridge, Massachusetts: Wright-
Allen Press. [1971].
Forrester, J.W. (1980). Industrial dynamics. Tenth printing. Cambridge, Massachusetts:
M.LT. Press. [1961].
Forrester, J.W. (1990). Principles of systems. Portland: Productivity Press. [1971].
Geurts, J.L.A., Hart, S.L., Caplan, N.S. (1985). Decision techniques and social research. A
contingency framework for problem solving. Human Systems Management. 5: 333-347.
Amsterdam: Elsevier Science Publishers.
Kelly, G.A. (1955). The psychology of personal constructs. Volume one: A theory of person-
ality. New Y ork: W.W. Norton & Company.
Lane, D.C. (1994). With a little help from our friends. How system dynamics and soft OR can
learn from each other. System Dynamics Review. 10(2-3): 101-134. Cambridge, MA: Sys-
tem Dynamics Society.
Lane, D.C., Oliva, R. (1998). The greater whole. Towards a synthesis of system dynamics and
soft systems methodology. European Journal of Operational Research. 107: 214-235.
Amsterdam [etc.]: Elsevier.
Mingers, J., Brocklesby, J. (1997). Multimethodology. Towards a framework for mixing
methodologies. Omega. The International Journal of Management Science. 25(5): 489-
509. Oxford [etc.]: Pergamon Press.
Patching, D. (1990). Practical Soft Systems Analysis. London: Pitman.
Richardson, G.P. (1991). Feedback thought in social science and systems theory. Philadel-
phia: University of Pennsylvania Press.
Richardson, G.P., Andersen, D.F. (1995). Teamwork in Group Model Building. System Dy-
namics Review. 11(2): 113-137. Cambridge, MA: System Dynamics Society.
Richardson, G.P, Pugh III, A.L. (1983). Introduction to system dynamics modeling with
DYNAMO. Second printing. Cambridge, MA [etc.]: M.I.T. Press [1981].
Rouwette, E.AJ.A, Vennix, J.A.M., Mullekom, T. van. (1999). Group Model Building. A
review of assessment studies. In: Cavana, R.Y., Vennix, J.A.M., Rouwette, E.AJ.A., Ste-
venson-Wright, M., Candlish, J. Systems thinking for the next millennium. Proceedings of
the 17th International Conference of the System Dynamics Society and the 5th Australian
& New Zealand Systems Conference, 20-23 July, Wellington, New Zealand.
Vennix, J.A.M. (1996). Group Model Building. Facilitating team learning using system dy-
namics. Chichester [etc.]: John Wiley & Sons.
Wolstenholme, E.F. (1990). System enquiry. A system dynamics approach, Chichester [etc.]:
John Wiley & Sons.
