Beruflich Dokumente
Kultur Dokumente
The Usage of System Dynamics in Organizational Interventions
The Usage of System Dynamics in Organizational Interventions
The Usage of
System Dynamics
in Organizational
Interventions
A Participative Modeling Approach
Supporting Change Management Efforts
Deutscher Universitäts-Verlag
Bibliografische Information Der Deutschen Nationalbibliothek
Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der
Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über
<http://dnb.d-nb.de> abrufbar.
D 17
Peter Milling
Preface
The managers below the top executives in large corporate organizations are
often placed in the challenging situation of implementing other people’s
ideas. Top executives will frequently launch strategic initiatives, and expect
the managers at lower levels to act as change leaders even though they have
often played only a small or no part at all in the groping strategy forming
process where the strategic initiative has its origin. Consequently, there is a
need for learning processes that focus on the transfer of the insights and
reasoning behind the decision, as well as supporting the refining of
implementation plans. Furthermore, the processes should allow iterations
with top executives, with the dual objective of adjusting the strategic
initiative according to implementation issues and giving the managers
responsible for implementation true influence on the entire process. To a
great extent, this dissertation addresses the process of transfer of insights
and ownership as well as the operationalization of strategic initiatives and
other change projects. The main topic is the usage of system dynamics
modeling in organizational interventions in general, and specifically the use
of system dynamics modeling for the purposes of change management. The
first two chapters mainly discuss organizational interventions and the use of
modeling in decision-making and policy forming processes, which is the
predominant application of system dynamics. The last three chapters
concentrate on a rather specific application of system dynamics: modeling in
a change management context. Change management dedicated application of
system dynamics builds upon the theories and methods of system dynamics
in a decision-making and policy-forming context, but aims at the transfer of
insights and ownership from decision-makers to implementers, as well as
refining and aligning cross-organizational implementation plans.
Writing this doctoral dissertation has been an interesting journey for
me. I have enjoyed the opportunity to take the time to go into depth with the
literature especially from the disciplines of system dynamics and
organizational psychology. Coming directly from 10 years of management
consulting, the academic experience has been of great personal and
X Preface
educational value to me. I would therefore like to sincerely thank first of all
Professor Dr. Peter Milling for supporting me on my journey. He and
assistant Professor Dr. Andreas Größler have guided me with great patience
through the learning process – helping me to adjust my normative and
solution-oriented worldview from the consulting world to also embrace
scientific and academic viewpoints. I would also like to thank all my
doctoral colleagues at ‘Industrieseminars’ for the weekly discussions at the
doctoral seminars, and Markus Salge and Dr. Nadine Schieritz especially for
always volunteering to finding literature and discussing modeling issues.
Furthermore, I want to thank my good friend since early childhood, Kirstine
Munk, who has been struggling with her own dissertation at the same time as
me. Although our subjects are very different, we have had many and
interesting discussions on a wide variety of issues including theories of
science, social constructivism, aesthetic in consultations and workshops, and
using cognitive frameworks (being both models and horoscopes!) to reduce
personal barriers for involvement and honesty in discussions. Last but not
least, I want to thank my husband, Jim Hagemann Snabe. As well as
receiving personal support in many ways, I have also been very privileged to
be able to draw on his extensive business experience and conceptual
capabilities.
Birgitte Snabe
List of Contents
Preface .................................................................................................. IX
Appendices ...........................................................................................165
Bibliography.........................................................................................201
List of Figures
Figure D-1: Conceptual modeling steps in the case study .................. 128
Table C-2b: Main sources for evaluation of outcomes on group level…… 112
Table C-2e: Main sources for evaluation of the usage of system dynamics
in a change management context……………………………. 115
1
Forrester, Jay W.: Industrial Dynamics, Cambridge, 1961, p. 1.
2
See Kotter, John P: Leading Change, Boston, 1996, p. 18; Fine, Charles H.:
“Clockspeed-based strategies for Supply Chain Design”, Production and
Operation Management, Vol. 9, No. 3, 2000, p. 213; Brown, John Seely:
“Research That Reinvents the Corporation”, Harvard Business Review, August
2002, p. 105.
3
In de Geus, Arie P.: “Planning as Learning”, Harvard Business Review, Vol. 66,
No. 2, March-April 1988, p. 71, it is proposed that ”the ability to learn faster
than competitors may be the only sustainable competitive advantage.“
4
The Japanese concept of Kaizen is an example of a continued process
improvement focus, whereas the western world typically is more oriented
towards innovation- and result-oriented thinking, see Imai, Masaaki: Kaizen:
Der Schlüssel zum Erfolg der Japaner im Wettbewerb, German translation,
München, 1992, p. 15.
2 A. Organizational Intervention Skills as Corporate Competence
Managing Stratum
Goals
Interventions Feedback
Resource Achieved
Causal Stratum output
input
5
Takahara, Yasuhiko: “A Formal Model of Organization”, in Takahashi, Singo,
Kyoichi Kijima and Ryo Sato (eds.): Applied General Systems Research on
Organizations, Tokyo, 2004, p. 3.
6
Own translation of figure in Milling, Peter: Systemtheoretische Grundlagen zur
Plannung der Unternehmenspolitik, Berlin, 1981, p. 17. It should be noted that
the original figure uses the German term “Führungsstratum” (translated to
Managing Stratum) which is a broader term also encompassing the meaning of
leading, steering, controlling. This model is chosen due to its abstraction level
suitable to illustrate the concept of interventions. Takahara offers a more
detailed basic model of organizations decomposing the operational level (the
causal stratum), which inherent has a stronger focus on internal structures and
coordination challenges; see Takahara, Yasuhiko: A Formal Model of
Organization, in Takahashi, Singo, Kyoichi Kijima and Ryo Sato (eds.): Applied
General Systems Research on Organizations, Tokyo, 2004, pp. 10—13.
A. Organizational Intervention Skills as Corporate Competence 3
domain producing the output, and the higher-level managing stratum, which
includes information processing to goal compliant forming and controlling of the
causal stratum.7
In the social-psychological field of science, Argyris describes
organizational interventions from a task point of view stating: “the
interventionist’s primary tasks are to generate valid information, to help the client
system make informed and responsible choices, and to develop internal
commitment to these choices”.8 The terms ‘the interventionist’ and ‘the client’
are often used in intervention literature.9 Although disagreement exists with
regards to the importance of independence of the system and the intervener, the
contemporary literature focusing on organizational development typically sees
organizational interventions as embracing both change processes with and
without the use of external interventionists.10 The client system can, in the social-
psychological field of science, be an individual, a group of people or an
organization. At these three levels, behavioral changes are largely explained with
7
Milling, Peter: Systemtheoretische Grundlagen zur Plannung der Unternehmens-
politik, Berlin, 1981, p. 18. Milling later decomposes the managing stratum into
four levels: the normative level (formulation of long-term goals), the structuring
level (determination of the basic structures), the adaptive level (specification of
change programs) and the operative level (selection of actions), p. 20.
8
Argyris, Chris: Interventions Theory and Method – A Behavioural Science
View, Reading, Massachusetts, 1970, p. 21.
9
Cummings, Thomas G. and Christopher G. Worley: Organizational Development
and Change, Ohio, 2001, p. 142, describe the term intervention as “sequenced
planned actions or events intended to help an organization increase its
effectiveness. Interventions purposely disrupt the status quo; they are
deliberately attempts to change an organization or sub-unit towards a different
and more effective state.” Linguistic, the term ‘intervention’ indicates, that a
party is proactively doing something to change the system. This is also seen in
fields like economy (state interventions) and foreign affairs (armed conflicts).
10
In Argyris, Chris: Interventions Theory and Method – A behavioural Science
View, Reading, Massachusetts, 1970, p. 15, the importance of independency
between the client and the interventionist is stressed, whereas in Schein, Edgar
H.: Process Consultation, Boston, 2000, part II, p. 35, (collection of work first
published in the 1960’s), it is argued that both external consultants as well as
managers from within the company can serve the role of the interventionist.
Recent textbooks generally support the latter view.
4 A. Organizational Intervention Skills as Corporate Competence
11
See Chin, Robert and Kenneth D. Benne: “General Strategies for Effecting
Changes in Human Systems”, in Bennis, Warren G., Kenneth D. Benne and
Robert Chin: The Planning of Change, 4th edition, New York, 1985, p. 24; and
Bungard, Walter and Catrin Niethammer: “Psychologische Aspekte des Change
Management im interorganisationalen Kontext”, in Walter Bungard, Jürgen
Fleischer, Holger Nohr, Dieter Spath and Erich Zahn (eds.), Customer
Knowledge Management, Stuttgart, 2003, p. 109.
12
Schein, Edgar H.: Process Consultation, Boston, 2000, part I, pp. 9—12, and
part II, pp. 29—35. It should be noted, that this book mainly consists of reprints
from his work in the late 1960’s. Schein’s work in general addresses the last
intervention model type, the Process Consultation Model.
13
In Akkermanns, Henk: Modelling With Managers, Breda, The Netherlands,
1995, pp. 7—12, a literature overview of decision-making and problem solving
is found, covering Operation Management, System Dynamics, Strategic
Management, Operations Research/“Soft OR”, Group Decision Support Systems
and Organizational Psychology.
A. Organizational Intervention Skills as Corporate Competence 5
4
Action planning
1
Problem
formulation
3 Felt 2
need
Forecasting Producing
consequences, proposals
6 testing proposals for solutions 5
Evaluating Taking
outcomes action steps
14
See Schein, Edgar H.: Process Consultation, Boston, 2000, part I, p. 61. The
model is an elaboration of a model originally developed by Richard Wallen for
use in sensitivity training programs. The model has strong similarities with
Dörner’s “Steps in Planning and Action”, although Dörners model less sharp
separate in a planning and an implementation part, see Dörner, Dietrich: The
Logic of Failure, New York, 1996, p. 43. Also, the model has similarities with
the PDCA-cycle (Plan-Do-Check-Act) as seen in TQM. The PDCA-cycle is a
generic version of the Deming Cycle, see Imai, Masaaki: Kaizen: Der Schlüssel
zum Erfolg der Japaner im Wettbewerb, german translation, München, 1992,
p. 87.
6 A. Organizational Intervention Skills as Corporate Competence
4
Action planning
1
Problem
formulation
3 Felt 2
Forecasting need Producing
consequences, proposals
6 testing for solutions 5
Evaluating proposals Taking
outcomes action steps
4
Action planning
1
Problem
formulation
3 Felt 2
Forecasting need Producing
consequences, proposals
6 testing for solutions 5
Evaluating proposals Taking
outcomes action steps
two distinct areas: the area of strategy forming and the area of strategy
implementation.15 The latter is described in the change management focused
literature from the disciplines of organizational psychology and organizational
development (OD). Cycle I of interventions have their primary focus on
diagnostics and decision-making, which for strategic organizational interventions
can be understood as strategy forming. In the field of organizational psychology,
cycle I activities are normally labeled ‘organizational diagnostics’ which is
leading to the initiation of the planned change process.16 Cycle II interventions,
focusing on planning, carrying out and following up on implementation, comprise
what in the fields of organizational psychology and OD are typically categorized
as ‘planned change’ interventions.
The iterative and recursive nature of the entire problem solving process
(both cycle I and cycle II) should not be underestimated, as also discussed in the
problem solving system described by Flood, “Total System Intervention”.17 Total
System Intervention focuses on creative problem investigation and deliberate
selection of methods to solve problems, through an iterative and recursive
process of three phases, (1) creativity, (2) choice, (3) implementation. In this
context, strategy forming is influenced by implementation considerations and
experiences, and the strategy implementation constitutes in itself a new cycle
with the need for creative ideas on how best to implement the strategy and the
choice of the best methods to achieve the implementation.
Research within the area of strategy forming proposes that the way
corporations address strategic problems should be considered as groping,
interactive processes emphasizing learning, creativity, synthesis, and sharing of
mental models among decision-makers.18 In this regard, the understanding of
15
Huff, Anne S. and Rhonda Kay Reger: “A Review of Strategic Process
Research”, Journal of Management, Vol. 13, No. 2, 1987, p. 212. It should be
noted, that Huff and Reger use the term strategy formulation rather than strategy
forming.
16
The term ‘diagnostics’ is often used as heading for activities leading to the
planned change interventions, e.g. see the list of contents in Bennis, Warren G.,
Kenneth D. Benne and Robert Chin: The Planning of Change, 4 th edition, New
York, 1985 as well as in Cummings, Thomas G. and Christopher G. Worley:
Organizational Development and Change, Ohio, 2001.
17
Flood, Robert L.: Solving Problem Solving, Chichester, 1995, p. 32.
18
See Mintzberg, Henry: The Rise and Fall of Strategic Planning, New York,
1994, p. 77; de Geus, Arie P.: “Planning as Learning”, Harvard Business
8 A. Organizational Intervention Skills as Corporate Competence
22
See Schein, Edgar H.: Organisationspsykologi, Danish translation, Herning,
1990, p. 53; Kieser, Alfred: “Human Relations-Bewegung und Organisations-
psychologie”, in Kieser, Alfred (ed.): Organisationstheorien, 3 rd edition,
Stuttgart, 1999, pp. 101—131.
23
This viewpoint is discussed in Jöns, Ingela: Managementstrategien und
Organisationswandel, Mannheim, 1995, p. 156.
24
See Preface, Warren, Kim: Competitive Strategy Dynamics, Chichester, 2002.
For a theoretical discussion of the implementation problem, see also McPherson
III, L. Fillmore: “Organizational Change: An Industrial Dynamics Approach”, in
Edward B. Roberts (ed.): Managerial Applications of System Dynamics,
Cambridge, Massachusetts, 1978, pp. 447—449.
25
See Repenning, Nelson P. and John D. Sterman: “Nobody Ever Gets Credit for
Fixing Problems that Never Happened: Creating and Sustaining Process
10 A. Organizational Intervention Skills as Corporate Competence
application of a force, the social process will not change without an additional
force to break the habit, to unfreeze existing customs.31 Unfreezing existing
behavior or attitudes can take place through the mechanisms of weakening
existing behavior or attitudes, then the establishment of feelings of dislike
regarding the present situation, and also establishment of psychological feeling of
safeness in the change process.32 The second phase in the planned change process
is the actual change part with development of new attitudes and behavior based
on new information and cognitive and affective redefinitions. The last phase in
the planned change process (freeze) is concerned with how to change behavior in
a sustainable way, avoiding it sliding back to its old level in a short time.33 To
strive for sustainable and continuous benefit of new attitudes and behavior, useful
mechanisms could be testing the congruence between the change and the
individual’s own situation, team-building efforts and continuous support, or
recognition from both formal and informal leaders in the organization.34 Lewin’s
theories of planned change originally focus on cognitive and behavioral change
aspects related to a specific change situation, but researchers often also
emphasize the broader term organizational learning. Argyris and Schein both
place interventions as part of the continuous learning and forming of the
organization and its change readiness, Senge emphasizes the importance of
improving system thinking skills, and Sterman emphasizes the importance of
helping organizations to improve the critical thinking skills necessary to
challenge future mental models and biases, opposed to only helping to solve a
specific problem.35
31
Lewin, Kurt: “Group Decision and Social Change” (first published in Newcomb
and Hartley’s Readings in social psychology, 1948, pp. 330—341), in Gold,
Martin: The Complete Social Scientist – A Kurt Lewin Reader, Washington,
1999, p. 281.
32
See Schein, Edgar H.: Organisations Psykologi, Herning, 1990, pp. 254—255.
33
Lewin, Kurt: “Group Decision and Social Change”, in Gold, Martin: The
Complete Social Scientist – A Kurt Lewin Reader, Washington, 1999, p. 265.
The arguments include, that behavior observed in a training program is often not
continued when the person goes back to his normal routines.
34
See Schein, Edgar H.: Organisationspsykologi, Herning, 1990, pp. 256—257;
Cummings, Thomas G. and Christopher G. Worley: Organizational Development
and Change, Ohio, 2001, pp. 22—30.
35
Argyris, Chris: Interventions Theory and Method – A Behavioural Science
View, Reading, Ma., 1970, chapters 1 & 2; Schein, Edgar H.: Organisations-
12 A. Organizational Intervention Skills as Corporate Competence
psykologi, 1990, p. 40; Senge, Peter M.: The Fifth Discipline, New York, 1994,
pp. 57—67; Sterman, John D.: “All models are wrong: reflections on becoming a
systems scientist”, System Dynamics Review, Vol. 18, No. 4, Winther 2002,
p. 526.
36
Borum, Finn: Strategier for organisationsændringer, Copenhagen, 1995, p. 15.
37
Chin, Robert and Kenneth D. Benne: “General Strategies for Effecting Changes
in Human Systems”, in Bennis, Warren G., Kenneth D. Benne and Robert Chin:
The Planning of Change, 4 th edition, New York, 1985, pp. 22—45.
38
The school of scientific management dates back to the early 20th century with
Taylor’s work on rational optimization of work processes. The most well known
mechanism element might be the detailed time studies of work procedures,
although this should be seen in context with the underlying principles, including
focus on science development, scientific basis for selection and development of
workmen, and friendly cooperation between the management and the men, see
Taylor, Frederick W.: The Principles of Scientific Management, New York,
1967, (first published in 1911), pp. 129—130.
A. Organizational Intervention Skills as Corporate Competence 13
39
In Borum, Finn: Strategier for organisationsændringer, Copenhagen, 1995, the
change strategies are discussed thoroughly, and at p. 117, a schematic overview
can be found. Furthermore a fourth change strategy regarding network
organizations/communities is discussed.
40
Anderson, Linda A. and Dean Anderson: “Awake at the Wheel: Moving beyond
Change Management to Conscious Change Leadership”, OD Practitioner,
Vol. 33, No. 3, 2001, p. 44. The two mindsets seem to a high extend to
correspond with the traditional two views on Man-in-Organization: The Human
View vs. The Resource View, see Leavitt, Harold J., William R. Dill, and Henry
B. Eyring: The Organizational World – A systematic view of managers and
management, New York, 1973, pp. 122—123.
41
Overcoming human resistance to change is among the most discussed topics the
literature of strategy implementation and change management, see Cummings,
Thomas G. and Christopher G. Worley: Organizational Development and
Change, Ohio, 2001, pp. 154—173; Argyris, Chris: Interventions Theory and
Method – A behavioral Science View, Reading, Ma., 1970, p. 70.
14 A. Organizational Intervention Skills as Corporate Competence
42
Ajzen, Icek: Attitudes, Personality and Behavior, Chicago, 1988, pp. 20—131.
See also Rouwette, Etiënne: Group model building as mutual persuasion,
Nijmegen, 2003, pp. 104—111, for a discussion on Ajzen’s framework.
43
See Rouwette, Etiënne: Group model building as mutual persuasion, Nijmegen,
2003, p. 103.
44
See Schein, Edgar H.: Organisationspsykologi, Herning, 1990, p. 142 for a
literature overview of management traditions in regards to involvement of
subordinates in decisions.
45
Jöns, Ingela: Managementstrategien und Organisationswandel, Mannheim, 1995,
p. 157.
46
An extensive literature-based discussion of the relationship between practice and
theory in organizations theories in general (not specific related to action
research) can be found in Scherer, Andreas G.: “Kritik der Organisation oder
Organisation der Kritik? Wissenschaftstheoretische Bemerkungen zum Umgang
mit Organisationstheorien”, in Kieser, Alfred (ed.): Organisationstheorien,
3 rd edition, Stuttgart, 1999, pp. 1—37.
47
See Schein, Edgar H.: Organisationspsykologi, Herning, 1990, pp. 249—259;
Cummings, Thomas G. and Christopher G. Worley: Organizational Development
and Change, Ohio, 2001, pp. 22—30. Action research has some parallels to the
field of cybernetics, where focus is on behavior of systems (what does it do)
A. Organizational Intervention Skills as Corporate Competence 15
G, DP
m
D, P
ue P y
The system dynamics field has its origin as a primarily analytical and rational
oriented problem-investigating and policy-forming discipline.53 Forrester states
that the purpose of system dynamics in corporate environments is to aid in the
design of improved industrial and economical systems, and system dynamics has
over the years contributed significantly to create insight being used in strategic
planning and policy design.54 System dynamics offers a complementary
opportunity to analyze complex and dynamic problems, as most of the traditional
tools offered by the strategic planning field, are largely static, and thereby often
insufficient in our present-day environment of complexity and dynamics,
consequently resulting in actions often being made based on intuition and
experience.55
System dynamics addresses the need for decision makers to learn and
understand complex problems and situations. Since Descartes, cognitive science
has been interested in how humans learn.56 Human brain processes are event-
orientated, which – without long experience or effective learning - makes it
53
In Forrester, Jay W.: Industrial Dynamics, Cambridge, 1961, p. 56, objectives in
using mathematical models are described as follows: “A mathematical model of
an industrial enterprise should aid in understanding that enterprise. It should be
a useful guide to judgment and intuitive decisions. It should help establish
desirable policies.” Milling, Peter: “Leitmotive des System-Dynamics-
Ansatzes”, Wirtshaftswissenschaftliches Studium, Vol. 10, 1984, p. 508, also
supports this understanding of system dynamics: “System Dynamics verwendet
formale Modelle, um zu einem verbesserten Verständnis des zu studierende
Phänomens zu gelangen und um Eingriffe in das System auf ihre Konsequenzen
hin zu untersuchen.”
54
See Forrester, Jay W.: Industrial Dynamics, Cambridge, 1961, p. 115. For
examples of the practical usage of system dynamics, see the numerous cases
published in System Dynamics Review over the years.
55
See Lyneis, James M.: Corporate Planning and Policy Design: A System
Dynamics Approach, Massachusetts, 1980, p. 3; and Warren, Kim: Competitive
Strategy Dynamics, Chichester, 2002, preface; Mintzberg, Henry: The Rise and
Fall of Strategic Planning, New York, 1994, p. 319.
56
In René Descartes first major contribution, 1628, “Regulae ad directionem
ingenii,” regarding rules for the use of the human’s cognitive means, a method
for acquiring scientific or any other type of rational founded insight is described,
see Lübcke, Poul (ed.): Politikens filosofi leksikon, Copenhagen, 2001,
pp. 82—87. Wikipedia (accessed April 2006) describes the work as a method for
scientific and philosophical thinking and translates the title of the book into
“Rules for the Direction of the Mind.”
18 A. Organizational Intervention Skills as Corporate Competence
57
See Sterman, John D.: Business Dynamics – Systems Thinking and Modeling for
a Complex World, Boston, 2000, pp. 10—11; Kampmann, Christian P. E.:
Feedback complexity and market adjustment – An experimental approach,
Boston, 1992, p. 31; Bakken, Bent E.: Learning and Transfer of Understanding
in Dynamics Decision Environments, Boston, 1993, pp. 29—30; Dörner,
Dietrich: The Logic of Failure, New York, 1996, pp. 38—42.
58
Größler, Andreas: “A Content and Process View on Bounded Rationality in
System Dynamics”, Systems Research and Behavioral Science, Vol. 21, No. 4,
July/August, 2004, p. 320.
59
Bonabeau, Eric: “Don’t Trust Your Guts”, Harvard Business Review, May 2003,
pp. 118—119.
60
Forrester, Jay W.: “System Dynamics, System Thinking, and Soft OR”, System
Dynamics Review, Vol. 10, No. 2, Summer 1994, p. 249.
A. Organizational Intervention Skills as Corporate Competence 19
61
Lyneis, James M.: Corporate Planning and Policy Design – A System Dynamics
Approach, Massachusetts, 1980, p. 9 and p. 15; Lane, David C.: “Should System
Dynamics be Described as a ‘Hard’ or ‘Deterministic’ System Approach?”
Systems Research and Behavioral Science, Vol. 17, 2000, p. 4.
62
de Geus, Arie P.: “Planning as Learning”, Harvard Business Review, Vol. 66,
No. 2, March-April 1988, pp. 70—74.
63
Miller, George A.: “The Magical Number Seven, Plus Minus Two: Some Limits
on Our Capacity for Processing Information”, The Psychological Review,
Vol. 63, No. 2, March 1956, p. 95.
64
Lane, David C.: “Should System Dynamics be Described as a ‘Hard’ or
‘Deterministic’ System Approach?”, Systems Research and Behavioral Science,
Vol. 17, 2000, p. 4.
65
Milling, Peter: “Modeling Innovation Processes for Decision Support and
Management Simulation”, System Dynamics Review, Vol. 12, No. 3, 1996,
p. 227.
66
Bakken, Bent E.: Learning and Transfer of Understanding in Dynamics Decision
Environments, Boston, 1993, p. 31.
20 A. Organizational Intervention Skills as Corporate Competence
the problem parameters, and stimulates best practice discussions and brings to
surface misperceptions among key individuals.67
In the 1990’s a school within the field of system dynamics, oriented
towards participative modeling approaches, emerged with increased embracing of
softer aspects such as organizational learning, group processes, and the
importance of consensus and commitment.68 This development might follow from
the change in organizational structures in many organizations. Modern
organizations with a high degree of employee empowerment typically have a
need for a large number of people to have an understanding for the whole of the
organization and its strategy, including the dynamics and the interdependencies,
to be able to make the right decisions in their daily work as well as for
motivational factors. This to some degree substitutes the “old way” with a few
executives directing strategies and policies to be implemented downwards in the
organization. The challenge of interventions nowadays is therefore not only to
find good solutions to problems or new situations. The solution must also be
understood and find acceptance among the many stakeholders, and efforts of
establishing internal commitment in organizational interventions are often
centered on the creation of awareness, consensus, and confidence regarding the
goals and the change process.69
Although system dynamics projects (participative modeling, in particular)
are concerned with both cognitive and behavioral aspects relevant for
implementation, system dynamics modeling efforts are typically elements in
organizations’ strategy forming, with less focus on strategy implementation.
However, the value creation of a corporate modeling study is seldom “a new
67
See Warren, Kim: Competitive Strategy Dynamics, Chichester, 2002, p. 30;
Snabe, Birgitte and Andreas Größler: “Targeted Participative Modelling as
Organisational Intervention: Concept and Case Study”, Journal of Systems
Research and Behavioral Science, Vol. 23, No. 4, in print, 2006, p. 20.
68
Through the introduction of participative model-building methodologies and
“planning as learning”, focus has been put on creating conceptual insights,
changing mental models of decision-makers and creating consensus and
commitment; see Vennix, Jac A. M.: Group Model Building, Chichester, 1996,
p. 97; de Geus, Arie P.: “Planning as Learning”, Harvard Business Review,
March-April 1988, p. 70; Lane, David C.: “Modelling as Learning: A
consultancy methodology for enhancing learning in management teams”,
European Journal of Operational Research, Vol. 59, No.1, 1992, pp. 64—84.
69
See Akkermans, Henk: Modelling With Managers, Breda, p. 20.
A. Organizational Intervention Skills as Corporate Competence 21
70
See Lane, David C.: “Modelling as Learning: A consultancy methodology for
enhancing learning in management teams”, European Journal of Operational
Research, Vol. 59, No.1, 1992, p. 64; Vennix, Jac A. M.: Group Model
Building, Chichester, 1996, pp. 98—99.
71
See Akkermanns, Henk: Modelling With Managers, Breda, 1995, p. 17.
72
Sterman, John D.: Business Dynamics, Boston, 2000, p. 80 and p. 88; Roberts,
Edward B.: “Strategies for Effective Implementation of Complex Corporate
Models”, in Edward B. Roberts (ed.): Managerial Applications of System
Dynamics, Cambridge, 1978, pp. 79—84. A more critical view on system
dynamics efforts in organizational interventions can be found in Zock,
Alexander: “A critical review of the use of System Dynamics for organizational
consulting projects”, at CD-ROM of Proceedings, System Dynamics Conference,
System Dynamics Society, 2004, p. 7, where it is argued that not even the
participative modeling approaches are sufficient attentive to the overall
challenges of change processes.
22 A. Organizational Intervention Skills as Corporate Competence
73
The use of modeling-oriented simulations vs. gaming-oriented simulations
comes from the taxonomy proposed by Maier, Frank und Andreas Größler:
“What are we talking about? A taxonomy of Computer Simulations to Support
Learning”, System Dynamics Review, Vol. 16, No 2, 2000, p. 143. The term
‘modeling-oriented simulations’ does not refer to the context of the project in
regards to decision-making or implementation.
74
In this dissertation, strategic formulation is understood to include both strategic
planning and policy formulation. In the system dynamics society, the term
“policy formulation“ is often used as the aim of modeling projects with policies
being rules stating how the day-by-day operating decisions are made, see
Forrester, Jay W.: Industrial Dynamics, Cambridge, 1961, p. 93. Strategies are
constituted by both corporate goals and corporate policies, and strategic
planning is defined as the process of transforming corporate goals into policies,
see Lyneis, James M.: Corporate Planning and Policy Design: A System
Dynamics Approach, Boston, 1980, p. 19 and p. 3.
75
‘Exploratory modeling’ should not be mistaken with ‘exploratory models’ as
described in Homer, Jack B.: “Why we iterate: scientific modeling in theory and
practice”, System Dynamics Review, Vol. 12, No. 1, Spring 1996, p. 1. Homer
defines exploratory models as a kind of easy-made, draft models less occupied
with validation. In this dissertation ‘exploratory modeling’ refers to the purpose
of the project: to explore and understand a given problem, no matter if the model
used is less detailed or if it is highly developed and refined with scientific rigor.
A. Organizational Intervention Skills as Corporate Competence 23
communication plans. This is due to the fact, that for exploratory modeling
interventions, the organizational change process cannot be defined before the
outcome of the modeling process is (at least to some extent) clear.76 Often, a
strategy forming modeling intervention will result in changed mental models
among decision-makers; frequently, implementation will not take place in an
explicit, planned change manner.77 Exploratory modeling can take place as
‘participative modeling’ or as ‘expert modeling’, the difference primarily being
the way people are involved. In expert modeling, people – apart from the main
decision-makers and a few modelers - are primarily involved for information
collection purposes.78 In participative modeling, such as Group Model Building
and Modeling for Learning, people representing an extensive array of viewpoints
are involved in the modeling process itself, with strong focus on mental model
alignment and refinement.79
Transfer-oriented usage of system dynamics modeling belongs to the
planned change type of organizational interventions and has some common
characteristics with the field of action research, with its parallel focus on the
implementation of planned change as well as on continued knowledge
80
development. The modeling project supports implementation processes where a
76
In Vennix, Jac A. M.: Group Model Building, Chichester, 1996, p. 99, it is
argued that learning cannot be predicted in the outset of a project.
77
In Vennix, Jac A. M.: Group Model Building, Chichester, 1996, p. 97, it is
argued that insights are conceptual rather than instrumental, and although
stressing that the purpose of system dynamics is performance improvement, he
also states (p. 99) that “implementation becomes evasive.” Richardson, George
P. and Alexander L. Pugh: Introduction to System Dynamics Modeling with
DYNAMO, Cambridge, 1981, p. 355, write that “a modeling study usually
focuses on what policies will help, not on how those policies ought to be
introduced into the system.”
78
See Forrester, Jay W.: “Policies, decisions and information sources for
modeling”, European Journal of Operational Research, Vol. 59, No. 1, 1992,
pp. 59—60, and Forrester, Jay W.: Industrial Dynamics, Cambridge, 1961,
p. 364, where it is recommended to use industrial dynamics in a business
company by initiating in a small, exclusive group of people with the right
qualities to go in-depth with the dynamics of the company including the
“innermost secrets and hopes of the organization.”
79
Vennix, Jac A. M.: Group Model Building, Chichester, 1996, p. 112.
80
For definitions of action research, see Cummings, Thomas G. and Christopher G.
Worley: Organizational Development and Change, Ohio, 2001, p. 23.
24 A. Organizational Intervention Skills as Corporate Competence
strategy forming process has outlined business objectives and targets, but where
the optimization of the strategy or policy is left as a part of the implementation.
Therefore a balance is needed between the initial detail level of the objectives
81
and the degree of freedom to make decisions in the implementation process.
This could be called “framing the intervention”, giving participants
empowerment to explore, decide and act within a given ‘frame’ (how to do), but
not to explore, decide and act outside the given ‘frame’ (what to do). Transfer-
oriented usage of modeling could be called instrumental usage of modeling for
change management purposes. In respect to transfer of existing learning from one
group of people to another group of people, this type of modeling has similarities
with gaming-oriented simulations, such as flight simulators or educational games.
Gaming-oriented simulations make use of fixed models, with the purpose being
to transfer the understanding of the causal relations and the behavior of the
system. Compared with gaming-oriented simulations, transfer-oriented modeling
to a higher extends aims to transfer commitment in the search for sustainable
change. Through involvement and participation in modeling sessions,
implementers take part in the refining of the change program and the
operationalization of the implementation. In his description of the system
dynamics process, Forrester proposes a phase called “Educate and Debate”, in
which consensus for implementation is aimed for.82 The phase is placed after the
actual exploratory modeling, but is expected to raise questions resulting in
repeated analysis in the previous phases. If modeling were to be used in the
Educate and Debate phase, it would be an example of change management
oriented modeling. Although for transfer-oriented usage of modeling, it may or
may not be the case that a model has been developed in an earlier strategy
formulation phase.83 The modeling can also be based on a model especially
81
Borum, Finn: Strategier for organisationsændringer, Copenhagen, 1995, p. 58,
discuss the problem of “a free, informed choice as a condition for establishment
of commitment” in a change process planned and controlled by consultants.
82
Forrester, Jay W.: “System Dynamics, System Thinking, and Soft OR”, System
Dynamics Review, Vol. 10, No. 2, 1994, p. 247.
83
An interesting case, where a modeling project was continued into instrumental
implementation activities is the well-known “Maintenance Game” described in
Repenning, Nelson P. and John D. Sterman: “Nobody Ever Gets Credit for
Fixing Problems that Never Happened: Creating and Sustaining Process
Improvement”, California Management Review, Vol. 43, No. 4, Summer 2001,
pp. 64—88.
A. Organizational Intervention Skills as Corporate Competence 25
drafted for the project. Using such a preliminary model of the problem-system,
the process should allow for interactive refining and evaluation of the model
itself, and through both model adjustments, model enhancements and model
simulations, the modeling approach has the threefold aim of change program
refining, transfer of system understanding, and establishment of commitment.
The design of participative modeling interventions supporting change
management can draw from the normative, prescriptive management literature of
planned change with regards to activities such as intervention planning,
stakeholder management and implementation planning and review. Intervention
planning includes the definition of business objectives and targets, the framing of
the intervention, the identification of consultation relationships, roles and
responsibilities in the project organization, and time and budget planning.
Stakeholder management involves a thorough analysis of all the major interest
groups and individuals who have significant influence - directly or indirectly - on
the success of the intervention. Focus is on interests and power, importance for
solution design and implementation, and relevant means of involvement and
84
communication. Stakeholder analysis is a major input to intervention planning,
both to secure relevant parameters to be included in the process, and to secure
appropriate involvement and communication with stakeholders and other
85
employees. Implementation planning and review deals with the planning of the
implementation, including a communication plan and a clear assignment of
responsibilities. The communication plan develops over the course of the
intervention and includes elements such as motivating change and the
communication of visions, results, implementation plan and successes. Planning
activities for implementation should to be understood as something to be done
84
See Flood, Robert L. and Michael C. Jackson: Creative Problem Solving – Total
Systems Intervention, Chichester, 1991, p. 12; Argyris, Chris: Interventions
Theory and Method – A Behavioural Science View, Reading, Massachusetts,
1970, p. 81; and Borum, Finn: Strategier for organisationsændringer,
Copenhagen, 1995, pp. 77—89.
85
For discussions on “Employee Involvement”, see both Cummings, Thomas G.
and Christopher G. Worley: Organizational Development and Change, Ohio,
2001, p. 317; and Thun, Jörn-Henrik, Peter M. Milling, and Uwe Schwellbach:
“The Impact of Total Employee Involvement on Time-based Manufacturing”, in
Blackmon, Kate, Steve Brown, Paul Cousins, Andrew Graves, Christine
Harland, Richard Lamming, and Harvey Maylor (eds.): “What Really Matters in
Operations Management”, Bath, 2001, pp. 133—135.
26 A. Organizational Intervention Skills as Corporate Competence
only after a strategy or policy has been designed. Involving the right people in the
right way early in the process might be one of the most important criteria for
successful implementation later, together with communication strategies
including timely information and dialogues. Also, the iterative process continues
after implementation activities, as follow-up activities will create learning to be
used for further corrective actions.
The main differentiator of modeling supporting change management
compared to other approaches within the field of system dynamics, is the usage
and utility of participative modeling in a new context, namely in strategy
implementation rather than strategy formation. The main research objective is to
investigate whether change management-oriented participative modeling seems
to be an effective approach seeking sustainable change through:
• Transfer of insight from decision-makers to implementers in such a way
that not only the decisions but also the underlying arguments are
effectively transferred,
• Allow true involvement of implementers through participative strategy
refinement within a given decision ‘frame’ and strategic direction.
86
Sterman, John D.: Business Dynamics – Systems Thinking and Modeling for a
Complex World, Boston, 2000, pp. 4—5.
87
Martinez, Ignacio J and George P. Richardson: “An Expert View on the System
Dynamics Modeling Process: Concurrences and Divergences Searching for Best
Practices in System Dynamics Modeling”, at CD-ROM of Proceedings, System
Dynamics Conference, System Dynamics Society, 2002, p. 25.
28 B. Conceptual Foundation for the Usage of System Dynamics
handling conflicts among outcomes that occur at different times.88 The rationality
based expected-utility decision-making is often seen analyzed in the traditions of
operations research, decision trees, game theories, etc. 89
Whereas principles of rational choice are considered as reasonable in
abstract form, their implications are often violated in actual choices. In socio-
economic systems it is an illusion to assume perfect rational decision-making due
to complexity, uncertainty and human factors. The topic of complexity and
complex systems has been of great interest to scientists using terms such as
theories of holism, cybernetics, general system theory, chaos theories etc. since
World War I.90 Although not undertaking a formal definition of complexity,
Simon explains a complex system as “one made up of a large number of parts
that have many interactions.”91 In system theory traditions, Senge and Sterman
describe complexity as consisting of detail complexity and dynamic complexity.92
Milling further divides detail complexity into three sub-dimensions: number of
relevant elements (variety), number of connections between elements
(connectivity), and functional relationship between elements (functionality).93
88
Baron, Jonathan: Thinking and Deciding, 3 rd edition, Cambridge, UK, 2000,
pp. 223—243.
89
See Baron, Jonathan: Thinking and Deciding, Cambridge, UK, 2000, p. 227, for
a discussion on using game theory to examine expected-utility decision-making.
Expected-utility is also a cornerstone in the expected-monetary-value method;
see e.g. Tversky, Amos: “Additivity, utility and subjective probability”,
in Edwards, Ward and Amos Tversky (eds.): Decision Making, 1967,
pp. 208—238.
90
In Simon, Herbert A.: The Science of the Artificial, 3 rd edition, Cambridge,
1996, pp. 169—181, an overall discussion is offered on the major scientific
trends in this field.
91
Simon, Herbert A.: The Science of the Artificial, Cambridge, 1996,
pp. 183—184.
92
Senge, Peter M.: The Fifth Discipline, New York, 1994, p. 71; Sterman, John
D.: Business Dynamics – Systems Thinking and Modeling for a Complex World,
Boston, 2000, p. 21.
93
Milling, Peter: “Kybernetische Überlegungen beim Entscheiden in komplexen
Systemen”, in Entscheiden in komplexen Systemen, Wirtschaftskybernetik und
Systemanalyse, Band 20, Berlin, 2002, p. 12.
B. Conceptual Foundation for the Usage of System Dynamics 29
94
Herbert Simons view was expressed in the 1950’s, and discussed in Hogarth,
Robin: Judgement and Choice – The Psychology of Decision, 2nd edition,
Chicago, 1987, p. 63.
95
Simon, Herbert A.: The Science of the Artificial, 3 rd edition, Cambridge, 1996,
p. 39.
96
Miller, George A.: “The Magical Number Seven, Plus Minus Two: Some Limits
on Our Capacity for Processing Information”, The Psychological Review,
Vol. 63, No. 2, March 1956, p. 95; Vennix, Jac A. M.: Group Model Building,
Chichester, 1996, p. 27.
97
Forrester, Jay W.: “Counterintuitive Behavior of Social Systems”, in Collected
Papers of Jay W. Forrester, Cambridge, 1975, p. 216. In Sterman, John D.:
Business Dynamics – Systems Thinking and Modeling for a Complex World,
Boston, 2000, pp. 5—9, a larger number of examples of policy resistance are
described. In Dörner, Dietrich: The Logic of Failure, New York, 1996, a few,
but more detailed examples are discussed throughout the book; including
failures in Third World efforts and the Chernobyl disaster.
98
See Rouweette, Etiënne: Group model building as mutual persuasion, Nijmegen,
2003, pp. 19—29, for a description of descriptive and prescriptive view-points
in decision-making.
30 B. Conceptual Foundation for the Usage of System Dynamics
System
Dynamics
System Use of Formal,
Thinking Mathematical Models
(Simulation Models)
The three levels in figure B-1 are accumulative; i.e. in addition to the use
of formal mathematical models, system dynamics also comprises explicit,
conceptual feedback models as well as intuitive and experience-based models.
This is in accordance with Kampmann, who stresses that intuitive assumptions
underlie any type of model.99 The usage of qualitative and quantitative models
serve the purpose of changing the mental models of the decision makers, as
mental models are seen as a vehicle to change decisions and organizational
action.100 The difference between system thinking and system dynamics cannot be
99
Kampmann, Christian P. E.: Feedback complexity and market adjustment,
Boston, 1992, p. 28.
100
Keough, Mark and Andrew Doman: “The CEO as organization designer – An
interview with Professor Jay W. Forrester, the founder of system dynamics”, The
B. Conceptual Foundation for the Usage of System Dynamics 31
seen as reflecting the differences between hard and soft modeling approaches, as
system dynamics is located somewhere between the two extremes.101 Hard
modeling is a term used for single objective optimization, typically without
taking people and organization into account.102 Although system dynamics uses
mathematical formulas and relatively rigid model structures, it also encompasses
soft modeling fundamentals like focus on generating debate and new insights
about the problem at hand.103
The field of system dynamics often has vital debates concerning the
advantages of qualitative modeling as seen in the system thinking area vs. the
advantages of quantitative modeling as seen in system dynamics. The opinions
differ from the one extreme, that only if a model is quantified and simulated, a
study can be said to be complete, to the other extreme, that for a complex system
with many soft relationships, quantification itself can be damaging.104
Some of the better known qualitative modeling approaches include
Checkland’s Soft System Methodology (SSM), which is a “process of enquiry”,
Eden’s Strategic Option Development and Analysis (SODA), using cognitive
mapping for strategic options development and Senge’s use of Cause-Loop-
Diagrams (CLD) in building learning organizations.105 The goals of qualitative
110
Sterman, John D.: “Misperceptions of Feedback in Dynamic Decision Making”,
in Milling, Peter M. and Erich O.K. Zahn (eds.): Computer-Based Management
of Complex Systems, Proceedings of the 1989 International Conference of the
System Dynamics Society, 1989, p. 30.
111
In Mintzberg, Henry: The Rise and Fall of Strategic Planning, New York, 1994,
pp. 298—299 and pp. 326—328, system dynamics is criticized for being shallow
in depth and not embracing creativity and intuition, although on pp. 376—378 in
the same book, credits are given to a number of system dynamics case stories.
112
The whole article of Lane, David C.: “Should System Dynamics be Described as
a ‘Hard’ or ‘Deterministic’ System Approach?”, Systems Research and
Behavioral Science, Vol. 17, 2000, pp. 3—22, is a discussion of the
misinterpretations of system dynamics, and also holds the quote “It may seem
paradoxical but the results of a quantitative system dynamics study are
qualitative insights” (p. 17).
34 B. Conceptual Foundation for the Usage of System Dynamics
113
De Geus, Arie P.: The Living Company, Boston, 1997, p. 71.
114
Luna-Reyes, L.F. and D. L. Andersen: “Collecting and analyzing qualitative data
for system dynamics: methods and models”, System Dynamics Review, Vol. 19,
No. 4, 2003, pp. 271—296 give an overview of many qualitative data collections
methods to be used not only in the beginning of a modeling process but also in
the later stages. Furthermore, Hodgson, A. M.: “Hexagons for system thinking”,
European Journal of Operational Research, Vol. 59, 1992, pp. 123—136,
introduces a soft modeling technique that is incorporated in many modeling
approaches; e.g. in Group Model Building.
115
De Geus, Arie P.: The Living Company, Boston, 1997, p. 72.
116
See Bakken, Bent E.: Learning and Transfer of Understanding in Dynamics
Decision Environments, Boston, 1993, p. 31; Vennix, Jac A. M.: Group Model
Building, Chichester, 1996, pp. 97—99.
117
Akkermans, Henk: Modelling With Managers, Breda, 1995, p. 116.
B. Conceptual Foundation for the Usage of System Dynamics 35
118
These three levels for impacts of interventions are widely used in the system
dynamics literature, see Rouwette, Etiënne: Group model building as mutual
persuasion, Nijmegen, 2003, pp. 21—27 for a discussion on the three levels in
the literature of decision-making. In Argyris, Chris: Interventions Theory and
Method – A Behavioural Science View, Reading, Massachusetts, 1970, p. 38,
the three levels are listed together with an additional level; called intergroups
(formal and informal).
119
In Dörner, Dietrich: The Logic of Failure, New York, 1996, p. 6 cognitive
limitations in analytical, serial and visualized thinking are mentioned (as
opposed to female, “parallel” or non-western thinking); In De Bono, Edward:
Lateral Thinking for Management, England, 1971, pp. 4—9, it is argued that
linear vertical thinking being overly dominant in our education system. In
Miller, George A.: “The Magical Number Seven, Plus Minus Two: Some Limits
on Our Capacity for Processing Information”, The Psychological Review,
Vol. 63, No. 2, March 1956, p. 95, limitations on the amount of information
humans are able to receive, process and remember are discussed.
120
Bakken, Bent E.: Learning and Transfer of Understanding in Dynamics Decision
Environments, Boston, 1993, pp. 29—30; Kampmann, Christian P. E.: Feedback
36 B. Conceptual Foundation for the Usage of System Dynamics
Several studies within the field of system dynamics confirm these theories: based
on a study investigating a multiplier-accelerator model of capital investments,
Sterman concludes that many subjects fail to adequately account for the effects of
delays, and fail to understand the feedback between their own decisions and the
environment – even when provided with perfect information and knowledge of
the system structure.121 In gaming environments, Dörner has observed students
and professionals carrying out simulations of a variety of systems (small town,
eco-system etc.), with the majority of subjects failing to achieve “good results”;
failures he believes are primarily due to lack of system understanding and
tendencies to focus on short term, immediate effects rather than more long-term
and fundamental processes.122
E C
A => E => C => F
F
B
A
D -
System Perception
inferences.123 The underlying ideas and concepts can be traced further back in
psychology literature, including the work of Craik’s in the mid 1940’s proposing
that people construct internal symbolic representations or models of external
events.124 Definitions in the fields of psychology, communication and system
dynamics vary from seeing mental models as being stable internal models to
being reconstructed on an ad hoc basis; from being simple picture-like images to
being complex and intuitive theories; from considering people being aware of
their own mental models to being intuitive and inaccessible models.125
Doyle and Ford propose a definition of mental models suitable to the field
of system dynamics:
123
Johnson-Laird, P. N.: Mental Models, Cambridge, 1983, pp. 144—145. See also
Baron, Jonathan: Thinking and Deciding, 3 rd edition, Cambridge, UK, 2000,
p. 74.
124
In Doyle, James K. and David N. Ford: “Mental models concepts for system
dynamics research”, System Dynamics Review, Vol. 14, No. 1, Spring 1998,
p. 8, reference is made to Craik’s (1943) book “The Nature of Explanation”.
This reference is also made in Johnson-Laird, P. N.: Mental Models, Cambridge,
1983, p. 2.
125
Doyle, James K. and David N. Ford: “Mental models concepts for system
dynamics research”, System Dynamics Review, Vol. 14, No. 1, Spring 1998,
p. 4, p. 9 and p. 14.
126
Doyle, James K. and David N. Ford: “Mental models concepts revisited: some
clarifications and a reply to Lane”, System Dynamics Review, Vol. 15, No. 4,
Winter 1999, p. 414. The definition is a revised version from an earlier article,
based on comments in Lane, David C.: Friendly amendment: A commentary on
Doyle and Ford’s proposed re-definition of “mental model”, Systems Dynamics
Review, Vol. 15, No. 2, Summer 1999, pp. 185—194.
38 B. Conceptual Foundation for the Usage of System Dynamics
With Doyle and Ford’s definition the system dynamics society to some
degree has an aligned use of the term mental models. In line with this definition,
cause-loop-diagrams or stock-and-flow diagrams could be viewed as externalized
mental models.127 Lane recommends not using the term “cognitive map” for these
diagrams, as this is a well-established term in management science literature for
the representation used in the SODA methodology.128 Doyle and Ford have a
broader definition, and argue that cognitive map is a well suited term, due to its
intuitive appeal, and arguing that Eden’s use of the term is merely a particular
form of cognitive mapping.129
The conceptual application of mental models gives meaning in both
everyday life and in more complex decision-making situations. In a normal
discussion of a single topic, participants will implicitly employ different mental
models, with different underlying assumptions, and with different goals.
Underlying relationships, assumptions and goals are typically left unstated and
never brought into the open. With this in mind, Forrester argues that it is easy to
understand why compromises often takes long and often fail in their objectives or
produce new difficulties.130 A different view could be, that the “slack” in
understanding does not make compromising more difficult, but allows everyone
to have a different understanding of the compromise. For important and complex
decision-making it consequently makes sense to invest in the efforts of making
the models explicit. Forrester further argues for not only making mental models
explicit, but also using computer analyses to investigate the problem, as even
models correct in structure and assumptions due to the fact that limited
127
Lane, David C.: Friendly amendment: “A commentary on Doyle and Ford’s
proposed re-definition of ‘mental model’”, Systems Dynamics Review, Vol. 15,
No. 2, Summer 1999, p. 186.
128
Lane, David C.: “A commentary on Doyle and Ford’s proposed re-definition of
‘mental model’”, Systems Dynamics Review, Vol. 15, No. 2, Summer 1999,
p. 186.
129
Doyle, James K. and David N. Ford: “Mental models concepts revisited: some
clarifications and a reply to Lane”, System Dynamics Review, Vol. 14, No. 1,
Spring 1998, pp. 412—413.
130
Forrester, Jay W.: “Counterintuitive Behavior of Social Systems”, in Collected
Papers of Jay W. Forrester: Foreword by Gordon S. Brown, Cambridge, 1975,
p. 213.
B. Conceptual Foundation for the Usage of System Dynamics 39
131
Forrester, Jay W.: “Counterintuitive Behavior of Social Systems”, in Collected
Papers of Jay W. Forrester, Cambridge, 1975, p. 214.
132
Bakken, Bent E.: Learning and Transfer of Understanding in Dynamics Decision
Environments, Boston, 1993. pp. 29—30. An often mentioned example: Newell,
Allen and Herbert A. Simon: Human Problem Solving, Englewood Cliffs, New
Jersey, 1972, pp. 90—91, describe the well known ‘Nine Dot Problem’ (with
nine dots arranged in a 3 by 3 square array), where each subject is “directed to
draw four straight lines, without raising his pencil from the paper, that pass
through all nine dots.” Most subjects subconsciously assume, that the lines may
not continue outside the boundaries of the square, which make the problem
unsolvable.
133
Baron, Jonathan: Thinking and Deciding, 3 rd edition, Cambridge, UK, 2000,
p. 182.
134
The Rosenthal experiment was carried out in the 1960’s, and a description can
be found in Vennix, Jac A. M.: Group Model Building, Chichester, 1996, p. 20.
Other experiments showing conservatism in mental models include gaming
environments where participants have a tendency to stick to their initial
understandings of the system even when data would suggest otherwise, see
Dörner, Dietrich: The Logic of Failure, New York, 1996, p. 17.
40 B. Conceptual Foundation for the Usage of System Dynamics
Prior beliefs can also result in illusory correlations, e.g. if a teacher thinks a child
is intelligent, he or she will often tend to overestimate how well behaved the child
is.135
Before discussing how to bring about changes in individuals’ cognitive
structures and mental models, it is relevant to take a look at the theories of
judgment. Kahneman and Tversky argue, that human judgment relies on
heuristics rather than calculus of chance or statistical based prediction.136 Human
judgment is affected by a number of elements: contextual effects, the extent to
which cues are available, the order in which information is presented, whether
comparative judgments involve similar or dissimilar information, qualitative and
quantitative data, and other factors.137 Hogarth especially points out the
importance of availability of information in the forming of judgments.
Individuals tend to put more emphasis on clues available to them, and
furthermore, some information from the past has stuck better, and therefore
influences judgments to higher degree.138 The contextual effects influencing
perceptions include the absolute variation in size, as experiments have shown
distorted perceptions due to absolute sizes.139 Another contextual effect shows in
the study of scenario planning where scenarios that include both a possible cause
and an outcome seems to appear more probable than scenarios merely involving
the outcome.140
An interesting hypothesis stated by Enhager, is that the intensity of
emotions activated (being ether positive or negative) strongly influences how
well humans remember a given situation.141 He argues, that people typically
135
Baron, Jonathan: Thinking and Deciding, 3 rd edition, Cambridge, UK, 2000,
p. 183.
136
Kahnemann, Daniel and Amos Tversky: “On the Psychology of Prediction”,
Psychological Review, Vol. 80, No. 4, July 1973, p. 237.
137
Hogarth, Robin: Judgment and Choice: The Psychology of Decision, 2 nd edition,
Chicago, 1987, p. 55.
138
Hogarth: Judgment and Choice, Chicago, 1987, p. 53. Hogarth mentions work
by Tversky and Kahnemann in much of his argumentation.
139
Hogarth: Judgment and Choice, 1987, p. 52.
140
Hogarth: Judgment and Choice, 1987, p. 49.
141
Kjell Enhager is a known speaker on the psychology of sports, and he has used
his theories in the training of the female national golf team in Sweden, e.g. by
having the women forcing a positive feeling with a smile and a happy
exclamation for each good swing and trying to ignore the bad ones.
B. Conceptual Foundation for the Usage of System Dynamics 41
142
McKee, Robert: “Storytelling That Moves People”, Harvard Business Review,
June 2003, p. 52.
143
Baron, Jonathan: Thinking and Deciding, 3 rd edition, Cambridge, UK, 2000,
pp. 176—181.
144
See Hogarth, Robin: Judgment and Choice: The Psychology of Decision,
Chicago, 1987, p. 130; Baron, Jonathan: Thinking and Deciding, 3 rd edition,
Cambridge, UK, 2000, p. 182.
145
See discussions in Hogarth, Robin: Judgment and Choice, Chicago, p. 54;
Vennix, Jac A. M.: Group Model Building, Chichester, 1996, p. 28.
146
Vennix: Group Model Building, 1996, p. 28, offers a number of examples from
psychology literature.
147
Hogarth: Judgment and Choice, 1987, p. 109.
42 B. Conceptual Foundation for the Usage of System Dynamics
148
Dörner, Dietrich: The Logic of Failure, New York, 196, p. 8. An example of
personal motives and perceived behavioral control influencing behavior is given
in Berger, Ulrike and Isolde Bernhard-Mehlich: “Die Verhaltenswissen-
schaftliche Entscheidungstheorie”, in Kieser, Alfred (ed.): Organisations-
theorien, 3 rd edition, Stuttgart, 1999, p. 155. A practical example of how to
involve emotions in change processes is the use of storytelling, see McKee,
Robert: “Storytelling That Moves People”, Harvard Business Review, June
2003, p. 52.
149
Ajzen, Icek and Martin Fishbein: “The Prediction of Behavior from Attitudinal
and Normative Variables”, Journal of Experimental Social Psychology, No. 6,
1970, p. 483.
150
See Ajzen, Icek: Attitudes, Personality and Behavior, 1988, Chicago,
pp. 20—23, and pp. 120—121. Ajzen is using the three components of attitude
(cognitive, affective, and conative), arguing that this tripartite model of attitude
has served as the starting point of most behavioral analyses since the 1960’s.
151
See Kotler, Philip: Marketing Management, 7th edition, New Jersey, 1991,
pp. 573—575 where different buying situations are described in terms of models
embracing cognitive (learning), affective (feeling) and behavior (doing) stages.
The AIDA-model (awareness-interest-desire-action) is probably the most known
model. Kotler also argues, that high-involvement purchases with perceived high
differentiation calls for the “learn-feel-do“ sequence of marketing.
B. Conceptual Foundation for the Usage of System Dynamics 43
Attitude
towards the
behavior
Subjective
Intention Behavior
norm
Perceived
behavioral
control
152
Taken from the figure in Ajzen, Icek: Attitudes, Personality and Behavior,
Chicago, 1988, p. 133.
153
Ajzen, Icek: Attitudes, Personality and Behavior, Chicago, 1988, p. 46. See also
Rao, Abhijit: “Recognition of Conative and Affective Behavior in Web Learning
using Digital Gestures”, North America Web-Based Learning Conference,
Online Proceedings, New Brunswick, 2001, p. 1, on conative and affective
elements improving learning experiences.
44 B. Conceptual Foundation for the Usage of System Dynamics
empirically difficult to distinguish between cognition and affect, as they often are
highly correlated.154
The subjective norms are assumed to be a function of the person’s beliefs
that specific individuals or groups approve or disapprove of performing this
behavior, and deal with both social norms and group dynamics aspects.155 The
importance of subjective norms is often stressed since Elton Mayo in 1945,
published a set of social assumptions regarding human nature. This focuses on
the social need of employees and recognizes the strong influence high status
individuals have on peer employees independently of formal power.156 Schein
states that norms within a group are maybe the most important influential factors
for the everyday behavior of the employees in a company.157 A related view on
social norms is what Gladwell calls the Power of Context: how seemingly small
changes in context can influence behavior. A known example is the Broken
Window theory, proposing that if small context flaws are tolerated (e.g. broken
windows), soon larger context flaws will occur (e.g. more windows will be
broken, which will start a process towards more serous crimes).158
Together the attitudes towards behavior and the subjective norms
influence intended behavior in a way that can be considered as goals to be
pursued.159 The intention to pursue these goals are affected by the person’s
perceived control over a given behavior determined by internal factors including
information, skills, abilities and control of emotions and compulsion, and
external factors like opportunity and dependence on others.160 The broken arrow
154
See Ajzen, Icek: Attitudes, Personality and Behavior, 1988, p. 21.
155
Ajzen, Icek: Attitudes, Personality and Behavior, 1988, p. 121; Ajzen, Icek and
Martin Fishbein: “The Prediction of Behavior from Attitudinal and Normative
Variables”, Journal of Experimental Social Psychology, No. 6, 1970, p. 483.
156
See Schein, Edgar H.: Organisationspsykology, Herning, 1990, p. 67.
157
Schein, Edgar H.: Organisationspsykology, 1990, p. 37.
158
See Gladwell, Malcolm: The Tipping Point, paperback edition, New York, 2002,
p. 141, giving credit of the Broken Window theory to the criminologists James
Q. Wilson and George Kelling. At pp. 142—145 an example of interventions
partly based on this theory is described: the fight against crime in New York in
the 1990’s, where fight against graffiti and fare-beating in the subway was the
starting point
159
Ajzen: Attitudes, Personality and Behavior, Chicago, 1988, p. 128.
160
See Ajzen: Attitudes, Personality and Behavior, 1988, pp. 128—131.
B. Conceptual Foundation for the Usage of System Dynamics 45
in figure B-3 indicates that the influence will only emerge when there is some
agreement between a person’s actual control and the person’s perceived control
of the behavior.161 Furthermore, perceived control does not only influence a
person’s current behavior. It also has long-term perspectives, as there is a general
agreement that perceived control results in positive feelings of one’s own
competence and worth, and likewise that lack of or loss of perceived control
influences both the physical and psychological well-being negatively.162
Understanding mental models, judgment processes, and behavior theories
is essential in understanding how people learn and react. People learn primarily
on the basis of what they can observe.163 A fundamental attribute of system
dynamics offers the opportunity to model and simulate the behavior of dynamic
structures and through observing the behavior reaching an understanding of a
complex problems in a few days or weeks that would have taken years to
understand by “normal” experience. Simulations have shown to result in better
transfer than case studies and lectures and this may also be attributed to
motivational side-effects of the interactive pedagogy.164 It is often difficult to
change the truly fundamental assumptions and beliefs underlying one’s thoughts
and actions, and especially when they are vague and poorly understood are they
difficult to escape from.165 People are typically unaware of their own
inconsistencies leading to self-fulfilling prophecies, and self-sealing processes.166
Modeling and simulating system structures is a way to bring forward underlying
161
Ajzen, Icek: Attitudes, Personality and Behavior, 1988, p. 134.
162
See Bungard, Walter: “Zur Implementierungsproblematik bei Business-
Reengineering Projekten”, in Perlitz, Manfred, Andreas Offinger, Michael
Reinhardt and Klaus Schug (eds.): Reengineering zwischen Anspruch und
Wirklichkeit, Wiesbaden, 1996, pp. 260—261. Also Kieser, Alfred:
Organisationstheorien, 3 rd edition, Stuttgart, 1999, pp. 129—131, describes
“Humanisierung der Arbeit” and discusses the benefits of the individual’s
influencing own work situation.
163
Hogarth, Robin: Judgment and Choice – The Psychology of Decision, Chicago,
1987, p. 130.
164
Bakken, Bent E.: Learning and Transfer of Understanding in Dynamics Decision
Environments, Boston, 1993, p. 31.
165
Morgan, Gareth: Creative Organization Theory, Newbury Park, California, 1989,
p. 28.
166
See Argyris, Chris: Reasoning, Learning, and Action – Individual and
Organizational, San Francisco, 1982, p. 39.
46 B. Conceptual Foundation for the Usage of System Dynamics
167
In chapter A, learning was discussed in terms of single-loop and double-loop
learning. The discussions on double-loop learning and mental model refinement
are closely related. Both types of learning require impact of belief systems and
value systems, with beliefs being an understanding of causality and values being
a network where one value is supported by the other values. The belief system
especially impacts the internal process of “making sense” and defining the
situation, whereas the value system subsequent impacts the problem definition,
see Eden, Colin: “Cognitive mapping and problem structuring for system
dynamics model building”, System Dynamics Review, Vol. 10, Nos. 2-3,
Summer-Fall 1994, p. 263.
168
Doyle, James K. and David N. Ford: “Mental models concepts revisited: some
clarifications and a reply to Lane”, System Dynamics Review, Vol. 14, No. 1,
Spring 1998, p. 4.
169
Lewin, Kurt: “Group Decision and Social Change” (first published in Newcomb
and Hartley’s Readings in social psychology, 1948, pp. 330—341), in Gold,
Martin: The Complete Social Scientist – A Kurt Lewin Reader, Washington,
1999, pp. 276—279. Note that Lewin has a different terminology use compared
to most SD research, as he calls individual decision made in a group setting for
“group decisions“ (see p. 274).
B. Conceptual Foundation for the Usage of System Dynamics 47
individuals partly due to unwillingness of the individual to depart too much from
group standard of the group they belong to or wish to belong to.170
In organizational settings, where teams work together to make the business
run, and where decisions often are made in teams involving the main
stakeholders, the concept of mental model offers insight into many of the typical
problem solving challenges. Each individual has a limited linear understanding of
the main problems, and these will differ with regards to which causal
relationships are perceived as the most influential ones (see figure B-4).
E C
F
B
A A => E => C => -D A => B => D
-
D
System Perception
170
Lewin, Kurt: “Group Decision and Social Change”, in Gold, Martin: The
Complete Social Scientist – A Kurt Lewin Reader, Washington, 1999, p. 273
and p. 281.
48 B. Conceptual Foundation for the Usage of System Dynamics
although not only differing mental models, but also conflicting interests based on
conflicting goal-settings play a role.
Baron discusses the concept of cognitive dissonance resolution, as a way
humans eliminate conflicts among beliefs.171 An example is, that after a decision
is made, it seems that increased weight is put on the reasons in favor of the
decision, and less value is given to the arguments in favor for paths that were not
selected. In organizational settings, where decision-makers typically have had a
significant role in the forming of the present situation, and have often had
different viewpoints in earlier decision-making processes, cognitive dissonance
resolution must be expected to contribute to the creation of biases in their
individual mental models.
Janis has put forward the concept of Groupthink, which is partly related to
cognitive dissonance resolution on a group level, and the concept is based on
extensive case studies on group decision situations, including major US failures
such as the Bay of Pigs decision, the escalation of the Vietnam War and the
failure to predict the attack on Pearl Harbor.172 The Groupthink phenomenon
refers to the lack of critical thinking that can occur in groups characterized by
group cohesiveness and consensus. It is in line with social-psychological research
findings, showing powerful social pressures if a dissident in a cohesive group
reveals objections to group norms or consensus. Although the term Groupthink
primarily refers to self-censorship in the critical thought process to avoid group
disunity.
Dörner discuss the Chernobyl explosion in terms of Groupthink and
overconfidence, arguing that the team responsible for the accident felt they were
so experienced that they did not have to follow safety rules.173 Although the team
was in fact very experienced they made basic errors of interaction with dynamic
171
Baron, Jonathan: Thinking and Deciding, 3 rd edition, Cambridge, UK, 2000,
p. 208. Leavitt, Harold J.: Top Down – Why Hierarchies Are Here to Stay and
How to Manage Them More Effectively, Boston, 2005, p. 130, discuss the same
concept, stating that “we humans don’t just do what we believe. We also believe
what we do.”
172
Janis, Irving L.: “Groupthink: The Problems of Conformity” (original printed in
Psychology Today, Nov. 1971, pp. 271—279), in Morgan, Gareth: Creative
Organization Theory, Newbury Park, California, 1989, pp. 224—228.
173
Dörner, Dietrich: The Logic of Failure, New York, 1996, p. 34.
B. Conceptual Foundation for the Usage of System Dynamics 49
systems; they over-steered (regulation of the situation rather than the process),
they ignored side effects, and they interpreted system indicators erroneously.174
Groupthink is normally discussed is the context of smaller groups, often
being management teams, but some of the same mechanisms also apply to a
people’s philosophy of life as expressed in the below quote of Nietzsche:
174
Dörner, Dietrich: The Logic of Failure, New York, 1996, pp. 30—33; Salge,
Markus and Peter Milling: “Who is to blame, the operator or the designer? Two
stages of human failure in the Chernobyl accident”, System Dynamics Review,
Vol. 22, in print, 2006.
175
Quoted in Morgan, Gareth: Creative Organization Theory, Newbury Park,
California, 1989, p. 22.
176
For a definition on cognitive diversity, see Tilebein, Meike: “Eine struktur-
wissenschaftliche Betrachtung von Diversity Management”, Tagungsband,
GWS-Tagung, Greifswald, in print, 2006, p. 1.
50 B. Conceptual Foundation for the Usage of System Dynamics
177
Pearson, Alan: “You Drive for the Show but you Putt for the Dough”, appendix
in Beer, Stafford: Beyond Dispute – The Invention of Team Syntegrity, 1994,
p. 321.
178
See Beer, Stafford: Beyond Dispute – The Invention of Team Syntegrity, 1994,
p. 59 and p. 102.
179
Kanter, Rosabeth Moss: “Leadership and the Psychology of Turnarounds“,
Harvard Business Review, June 2003, p. 61.
180
Eden, Colin: “Cognitive mapping and problem structuring for system dynamics
model building”, System Dynamics Review, Vol. 10, Nos. 2-3, Summer-Fall
1994, p. 259; Rouwette, Etiënne: Group model building as mutual persuasion,
Nijmegen, 2003, p. 251. Also the alignment of mental models has some
similarities with building collective intuition, see Eisenhardt, Kathleen M.:
“Strategy as Strategic Decision Making”, Sloan Management Review, Spring
1999, pp. 66—67.
B. Conceptual Foundation for the Usage of System Dynamics 51
Figure B-5: Mental models as instruments between actual systems and formal models
181
Maier, Frank: Die Integration wissens- und modellbasierter Konzepte zur
Entscheidungsunterstützung im Innovationsmanagement, Berlin, 1995, p. 217.
182
In Kim, Daniel H. and Peter M. Senge: “Putting systems thinking into practice”,
System Dynamics Review, Vol. 10, Nos. 2-4, Summer-Fall 1997, p. 279, the
concept of mental models is called the transfer mechanism between individual
learning and organizational learning. In Senge, Peter M.: The Fifth Discipline,
New York, 1994, mental models and system thinking are together with team
learning, shared vision and personal mastery, described as the main elements in
organizational learning.
52 B. Conceptual Foundation for the Usage of System Dynamics
183
For literature overview and discussions of organizational learning, see Argyris,
Chris: On Organisational Learning, 2 nd edition, Oxford, 1999, pp. 7—14; and
Kieser, Alfred and Ulrich Koch: Organizational Learning through Rule
Adaptation: From the Behavioral Theory to Transactive Organizational
Learning, Mannheim, 2000, pp. 2—26.
184
Senge, Peter M.: The Fifth Discipline, New York, 1994, p. 69; Sterman, John
D.: “All models are wrong: reflections on becoming a systems scientist”, System
Dynamics Review, Vol. 18, No. 4, Winter 2002, p. 526.
185
Argyris, Chris and Donald Schön: “Organizational Learning: A theory of Action
Perspective” (first printed as part of book with the same title in 1978), reprint in
Morgan, Gareth: Creative Organization Theory, Newbury Park, California, 1989,
p. 140; Argyris, Chris: Reasoning, Learning, and Action – Individual and
Organizational, San Francisco, 1982, p. 160.
186
Argyris, Chris and Donald Schön: “Organizational Learning: A theory of Action
Perspective”, reprint in Morgan, Gareth: Creative Organization Theory,
Newbury Park, Ca., 1989, p. 142.
187
Argyris, Chris: On Organisational Learning, 2nd edition, Oxford, 1999,
pp. 230—238.
188
Schein, Edgar H.: Process Consultation, Boston, 2000, part I, p. 194.
B. Conceptual Foundation for the Usage of System Dynamics 53
189
De Geus Aire P: “Planning as Learning”, Harvard Business Review, March-
April 1988, p. 70.
190
Milling, Peter: “Organisationales Lernen und seine Unterstützung durch
Managementsimulatoren”, in: Zeitschrift fåür Bestriebswirtschaft, 65 Jg.
Lernende Unternehmen (Sonderausgabe), 1995, pp. 98—100.
191
Kim, Daniel H. and Peter M. Senge: “Putting systems thinking into practice”,
System Dynamics Review, Vol. 10, Nos. 2-4, Summer-Fall 1997, pp. 280—281.
This framework is more comprehensive, and includes a number of different types
of learning, but is as a consequence less intuitively understandable compared to
the framework of Milling.
54 B. Conceptual Foundation for the Usage of System Dynamics
Observed
Behavior
Comparison
Comparison
Statements
Comparison
Expectations
Comparison
Comparison Comparison
Comparison
Comparison Mental
Comparison Real World
Comparison
Models
Comparison
Comparison
Comparison
Reflection
Decision
Action
192
Own translation of figure in Milling, Peter: “Organisationales Lernen und seine
Unterstützung durch Managementsimulatoren”, in: Zeitschrift für
Betriebswirtschaft, Ergänzungsheft 3/95, Lernende Unternehmen (Sonder-
ausgabe), 1995, p. 100.
B. Conceptual Foundation for the Usage of System Dynamics 55
193
Kim, Daniel H. and Peter M. Senge: “Putting systems thinking into practice”,
System Dynamics Review, Vol. 10, Nos. 2-4, Summer-Fall 1997, pp. 278—279
and p. 286.
194
Senge, Peter M. and John D. Sterman: “Systems thinking and organizational
learning: Acting locally and thinking globally in the organization of the future”,
European Journal of Operational Research, Vol. 59, No. 1, 1992, pp. 146—148;
Repenning, Nelson P., Sterman, John D.: “Nobody Ever Gets Credit for Fixing
Problems that Never Happened: Creating and Sustaining Process Improvement”,
California Management Review, Vol. 43, No. 4, Summer 2001, pp. 64—88; and
Sterman, John D.: Business Dynamics – Systems Thinking and Modeling for a
Complex World, Boston, 2000, pp. 130—132.
56 B. Conceptual Foundation for the Usage of System Dynamics
Observed
Behavior
Comparison
Comparison
Statements
Comparison
Expectations
Comparison
Comparison Comparison
Comparison Formal
Comparison Mental
Comparison Real World
Comparison Model
Models
Comparison
Comparison
Comparison
Reflection
Decision
Action
195
Own translation of figure in Milling, Peter: “Organisationales Lernen und seine
Unterstützung durch Managementsimulatoren”, in: Zeitschrift für Betriebs-
wirtschaft, Ergänzungsheft 3/95, Lernende Unternehmen (Sonderausgabe), 1995,
p. 105.
196
In Luna-Reyes, Luis Felipe and Deborah Lines Andersen: “Collecting and
analysing qualitative data for system dynamics: methods and models”, System
Dynamics Review, Vol. 19, No. 4, 2003, pp. 274—279, primarily based on the
following mentioned sources: Randers, Richardson and Pugh, Roberts et al.,
Wolstenholme and Sterman.
B. Conceptual Foundation for the Usage of System Dynamics 57
197
See Vennix, Jac A. M.: Group Model Building, Chichester, 1996, chapter 2 and
3; and Vennix, Jac A. M; David F. Andersen; George P. Richardson and John
Rohrbaugh: “Model-building for group decision support: Issues and alternatives
in knowledge elicitation”, European Journal of Operational Research, Vol. 59,
1992, pp. 28—41.
198
Forrester, Jay W.: “System dynamics, system thinking, and soft OR”, System
Dynamics Review, Vol. 10, No. 2, 1994, p. 244.
199
For a description and discussion of management simulators, see Größler,
Andreas: Entwicklungsprozess und Evaluation von Unternehmenssimulation für
lernende Unternehmen, Frankfurt am Main, 2000.
58 B. Conceptual Foundation for the Usage of System Dynamics
200
See Sterman, John D.: Business Dynamics – Systems Thinking and Modeling for
a Complex World, Boston, 2000, pp. 87—88.
201
Homer, Jack B.: “Why we iterate: scientific modeling in theory and practice”,
System Dynamics Review, Vol. 12, No. 1, Spring 1996, p. 16.
202
Randers, Jørgen: “Guidelines for Model Conceptualization”, in Randers, Jørgen
(ed.): Elements of the System Dynamics Method, Cambridge, Connecticut, 1980,
p. 118.
203
In Forrester, Jay W.: “System dynamics, system thinking, and soft OR”, System
Dynamics Review, Vol. 10, No. 2, 1994, p. 253, it is stated that to describe the
system is “the most important and the least straightforward of the stages in
system improvement;” and Schein, Edgar H.: Process Consultation – MAOM
Capstone Course for the University of Phoenix, Boston, 2000, part I, p. 62 it
says: “In my own experience in solving problems and watching others solve
them, by far the most difficult step is the first one – defining the problem.”
B. Conceptual Foundation for the Usage of System Dynamics 59
204
Richardson, George P. and Alexander L. Pugh: Introduction to System Dynamics
Modeling with DYNAMO, Cambridge, 1981, p. 18, Sterman, John D.: Business
Dynamics – Systems Thinking and Modeling for a Complex World, Boston,
2000, p. 89. An exception from the problem-orientated viewpoint is the Strategy
Dynamics approach, where a company’s strategic architecture is modeled rather
than a problem, see Warren, Kim: Competitive Strategy Dynamics, Chichester,
2002, pp. 89—113.
205
Eden, Colin: “Cognitive mapping and problem structuring for system dynamics
model building”, System Dynamics Review, Vol. 10, Nos. 2-3, Summer-Fall
1994, p. 261.
206
Roberts, Edward B.: “Strategies for Effective Implementation of Complex
Corporate Models”, in Edward B. Roberts (ed.): Managerial Applications of
System Dynamics, Cambridge, Massachusetts, 1978, pp. 83—89; Sterman, John
D.: Business Dynamics – Systems Thinking and Modeling for a Complex World,
Boston, 2000, pp. 87—88.
207
Eden, Colin: “Cognitive mapping and problem structuring for system dynamics
model building”, System Dynamics Review, Vol. 10, Nos. 2-3, Summer-Fall
1994, pp. 261—262. Eden furthermore argues for the usage of cognitive
mapping in the problem definition phase, to overcome the problems of handling
complexity and politics (pp. 263—268).
208
Sterman, John D.: Business Dynamics – Systems Thinking and Modeling for a
Complex World, Boston, 2000, p. 90; Randers, Jørgen: “Guidelines for Model
Conceptualization”, in Randers, Jørgen (ed.): Elements of the System Dynamics
Method, Cambridge, Connecticut, 1980, pp. 121—122.
60 B. Conceptual Foundation for the Usage of System Dynamics
(referred to) in many stages of the project, and timescale should be appropriate to
the time frame where significant changes and feedbacks are likely to occur.209 The
literature in general furthermore recommends formulating a dynamic hypothesis
characterizing the problem in terms of underlying feedback, stock and flow
structures of the system.210 The dynamic hypothesis guides modeling efforts by
focusing on certain system structures.
In the system conceptualization, the first explicit structures of the model
will typically take form. Decisions will be made with regards to what parameters
and casual relations to include in the model, focus on feedback loops etc.211 In
general, the system dynamics literature agree on the value of softer techniques in
problem identification, whereas a vital discussion exists with regards to the usage
of softer techniques in the starting point of the conceptualization and model
formulation. The discussions especially focus on the advantages of causal-loop-
diagrams vs. stock-and flow diagrams in the search for identifying the most
important structures of the system. Sterman, Richardson, Pugh and de Geus often
recommend the combination of use of quantitative modeling and qualitative
modeling, and having the qualitative modeling serve the purpose of providing
good understanding of the problem, it’s main parameters, and it’s boundaries
before actually outlining the stock and flow structures in the quantitative
model.212 The discussion on whether to include soft modeling techniques in the
209
Warren, Kim: Competitive Strategy Dynamics, Chichester, 2002, p. 27.
210
Forrester, Jay W.: “System dynamics, system thinking, and soft OR”, System
Dynamics Review, Vol. 10, No. 2, 1994, p. 246; Sterman, John D.: Business
Dynamics – Systems Thinking and Modeling for a Complex World, Boston,
2000, p. 95; Luna-Reyes, L.F. and D. L. Andersen, “Collecting and analyzing
qualitative data for system dynamics: methods and models”, System Dynamics
Review, Vol. 19, No. 4, 2003, p. 275.
211
In Sterman, John D.: Business Dynamics, Boston, 2000, p. 97, it is
recommended to do a formal model boundary chart explicitly grouping
endogenous, exogenous and excluded parameters.
212
Sterman, John D.: Business Dynamics, Boston, 2000, p. 102; Richardson,
George P. and Alexander L. Pugh: Introduction to System Dynamics Modeling
with DYNAMO, Cambridge, 1981, pp. 25—26; de Geus, Arie P.: The Living
Company, Boston, 1997, pp. 70—73; Luna-Reyes, L.F. and D. L. Andersen:
“Collecting and analyzing qualitative data for system dynamics: methods and
models”, System Dynamics Review, Vol. 19, No. 4, 2003, pp. 271—296;
Hodgson, A. M.: “Hexagons for system thinking”, European Journal of
Operational Research, Vol. 59, 1992, pp. 123—136.
B. Conceptual Foundation for the Usage of System Dynamics 61
213
Forrester, Jay W.: “System dynamics, system thinking, and soft OR”, System
Dynamics Review, Vol. 10, No. 2, 1994, p. 252.
214
Warren, Kim: Competitive Strategy Dynamics, Chichester, 2002.
215
Wolstenholme, Eric F.: “The definition and application of a stepwise approach
to model conceptualisation and analysis”, European Journal of Operational
Research, Vol. 59, 1992, p. 128. Wolstenholme describes the approach as a
combination between a feedback loop approach to model construction and a
modular approach to model construction (p. 136).
216
Jac Vennix, Professor at Nijmegen University, in a Course in Group Model
Buillding, Nijmegen, May, 2004.
62 B. Conceptual Foundation for the Usage of System Dynamics
217
Flood, Robert L. and Michael C. Jackson: Creative Problem Solving – Total
Systems Intervention, Chichester, 1991, p. 74.
218
Forrester, Jay W.: Principles of Systems, Cambridge, 1968.
219
Sterman, John D.: Business Dynamics – Systems Thinking and Modeling for a
Complex World, Boston, 2000.
220
Richardson, George P. and Alexander L. Pugh: Introduction to System Dynamics
Modeling with DYNAMO, Cambridge, 1981, p. 63.
221
Sterman, John D.: Business Dynamics, Boston, 2000, p. 95, argues: “system
dynamic models seek endogenous explanations for phenomena.”
222
Luna-Reyes, L.F. and D. L. Andersen, “Collecting and analyzing qualitative data
for system dynamics: methods and models”, System Dynamics Review, Vol. 19,
No. 4, 2003, p. 285.
B. Conceptual Foundation for the Usage of System Dynamics 63
Striving for high quality and efficiency in model formulation, the field of
system dynamics has dedicated much effort to the development of generic models
and model elements, which can serve as a starting point in a model building
process. Sterman offers a significant number of ‘example-models’ in his
extensive textbook, and the literature often discusses generic models, archetypes,
as well as microstructures to be used as building blocks in model formulation.223
In the field of group model building, Anderson and Richardson discuss scripts
supporting the entire modeling process.224
An important step in the modeling process, and closely linked to the model
formulation, is the model testing. Forrester and Senge identify three categories of
tests: tests of model structure, tests of model behavior, and tests of policy
implications.225 Tests of model structure include structure and parameter
verification (in comparison with knowledge of the real system), tests of extreme
situations, boundary-adequacy tests, and test of dimensions.226 Tests of model
behavior include behavior reproduction based on historical data, behavior
prediction tests, behavior-anomaly tests (study of anomaly features of the model
behavior), family-member test (comparisons to related models), surprise-
behavior tests (behavior which has gone unrecognized in the real system),
223
Sterman, John D.: Business Dynamics – Systems Thinking and Modeling for a
Complex World, Boston, 2000; Wolstenholme, Eric: “Using generic system
archetypes to support thinking and modelling”, System Dynamics Review,
Vol. 20, No. 4, Winter 2004, pp. 341—356.
224
Andersen, David F and George P. Richardson: “Scripts for group model
building”, System Dynamics Review, Vol. 13, No. 2, 1997, pp. 107—129.
225
Forrester, Jay W. and Peter Senge: “Tests for Building Confidence in System
Dynamics Models”, in Legasto, Augusto A., Jay W. Forrester und James M.
Lyneis (eds.), TIMS Studies in the Management Sciences, Vol. 14, Amsterdam,
1980, pp. 211—226. A discussion of many of these tests (although differently
labeled), as well as additional tests, can be found following a schematically
overview of formal model validation tests in Barlas, Yaman: “Formal aspects of
model validity and validation in system dynamics”, System Dynamics Review,
Vol. 12, No. 3, 1996, p. 189. Also Sterman, John D.: Business Dynamics,
Boston, 2000, pp. 845—901, offers an extensive discussion on validation of
models.
226
Forrester, Jay W. and Peter Senge: “Tests for Building Confidence in System
Dynamics Models”, in Legasto, Augusto A., Jay W. Forrester und James M.
Lyneis (eds.), TIMS Studies in the Management Sciences, Vol. 14, Amsterdam,
1980, pp. 211—216.
64 B. Conceptual Foundation for the Usage of System Dynamics
227
Forrester, Jay W. and Peter Senge: “Tests for Building Confidence in System
Dynamics Models”, in Legasto, Augusto A., Jay W. Forrester und James M.
Lyneis (eds.), TIMS Studies in the Management Sciences, Vol. 14, Amsterdam,
1980, pp. 217—223.
228
Forrester and Senge: “Tests for Building Confidence in System Dynamics
Models”, in Legasto, Forrester, und Lyneis (eds.), TIMS Studies in the
Management Sciences, Vol. 14, Amsterdam, 1980, pp. 224—225.
229
See Barlas, Yaman: “Formal aspects of model validity and validation in system
dynamics”, System Dynamics Review, Vol. 12, No. 3, 1996, p. 188; Forrester
and Senge: “Tests for Building Confidence in System Dynamics Models”, in
Legasto, Forrester, und Lyneis (eds.), TIMS Studies in the Management
Sciences, Vol. 14, Amsterdam, 1980, p. 210.
230
Barlas: “Formal aspects of model validity and validation in system dynamics”,
System Dynamics Review, Vol. 12, No. 3, 1996, p. 184.
231
Forrester, Jay W.: “System dynamics, system thinking, and soft OR”, System
Dynamics Review, Vol. 10, No. 2, 1994, p. 226.
232
See Sterman, John D.: “All models are wrong: reflections on becoming a systems
scientist”, System Dynamics Review, Vol. 18, No. 4, Winter 2002, p. 522;
B. Conceptual Foundation for the Usage of System Dynamics 65
bounded rationality in the process view and in the content view.233 In the system
dynamics modeling part of an organizational intervention, this can mainly be
understood as limitations in the modeling process skills and in problem content
skills.
Modeling process skills regards such elements as knowledge acquisition,
system modeling, methods and tools, testing principles, model analysis,
challenging assumptions, fostering discussions and creativity, and theories of
human learning. Also general intervention design skills such as stakeholder
involvement, ensuring agreement in the process, placing the intervention in a
broader organizational development view, and focus on implementation will
influence the rationality in the process view.
Problem content skills regarding elements as the understanding of the
problem at hand, understanding of goal-structures, identification and estimation
of the important parameters, establishment of cause-effect-relationships, etc.
If both modeling skills and problem content skills are high (i.e. relatively
low degree of bounded rationality in both views), it should be expected that the
developing process should create fruitful discussions and that the quality and
utility of the model should be high. If both modeling skills and problem content
skills are low, the risk should be high for both off-track discussions and wrong
models. On the other hand, if modeling skills are high but problem content skills
are low, the modeling process should hopefully open discussions that challenge
some of the wrong perceptions, but the end quality of the model could be
expected to be low. And lastly if problem content skills are high, whereas
modeling skills are low, then it is likely, that the problem would have been solved
better using alternative way of addressing it.
Barlas discusses practical differences for model validity with respect to
different uses of models, although the same general philosophy applies for the
Sterman, John D.: Business Dynamics – Systems Thinking and Modeling for a
Complex World, Boston, 2000, p. 846; Barlas, Yaman: “Formal aspects of
model validity and validation in system dynamics”, System Dynamics Review,
Vol. 12, No. 3, 1996, p. 187.
233
Größler, Andreas: “A Content and Process View on Bounded Rationality in
System Dynamics”, Systems Research and Behavioral Science, Vol. 21, No. 4,
July/August, 2004, pp. 319—330.
66 B. Conceptual Foundation for the Usage of System Dynamics
234
Barlas, Yaman: “Formal aspects of model validity and validation in system
dynamics”, System Dynamics Review, Vol. 12, No. 3, 1996, pp. 199—200.
235
Barlas, Yaman: “Formal aspects of model validity and validation in system
dynamics”, 1996, pp. 200—201.
236
Ashby, W. Ross: An Introduction to Cybernetics, paperback version, London,
1964, pp. 206—207. A regulator of the air traffic flows around New York is
given as example. See also Conant, Roger C. and W. Ross Ashby: “Every Good
Regulator of a System Must Be a Model of that System”, International Journal of
System Sciences, Vol. 1, No. 2, 1970, pp. 89—97.
237
Sterman, John D.: Business Dynamics – Systems Thinking and Modeling for a
Complex World, Boston, 2000, p. 104.
238
Forrester, Jay W.: “System dynamics, system thinking, and soft OR”, System
Dynamics Review, Vol. 10, No. 2, 1994, pp. 245—247.
B. Conceptual Foundation for the Usage of System Dynamics 67
239
Bakken, Bent E.: Learning and Transfer of Understanding in Dynamics Decision
Environments, Boston, 1993, p. 31; Sterman, John D.: “All models are wrong:
reflections on becoming a systems scientist”, System Dynamics Review, Vol. 18,
No. 4, Winter 2002, p. 522.
240
See Vennix, Jac A. M.: Group Model Building, Chichester, 1996, p. 97.
Although, the importance for designing the modeling process from the very start
with implementation in view is also emphasized in more traditional system
dynamics articles, see Roberts, Edward B.: “Strategies for Effective
Implementation of Complex Corporate Models”, in Edward B. Roberts (ed.):
Managerial Applications of System Dynamics, Cambridge, Massachusetts, 1978,
p. 83
68 B. Conceptual Foundation for the Usage of System Dynamics
241
Weil, Henry B.: “The Evolution of an Approach for Achieving Implemented
Results from System Dynamics Projects”, in Jørgen Randers (ed.): Elements of
the System Dynamics Method, Cambridge, Connecticut, 1980, p. 290.
242
Roberts, Edward B.: “Strategies for Effective Implementation of Complex
Corporate Models”, in Edward B. Roberts (ed.): Managerial Applications of
System Dynamics, Cambridge, Massachusetts, 1978, p. 77.
243
See Bakken, Bent E.: Learning and Transfer of Understanding in Dynamics
Decision Environments, Boston, 1993, p. 31; Vennix, Jac A. M.: Group Model
Building, Chichester, 1996, pp. 97—99.
B. Conceptual Foundation for the Usage of System Dynamics 69
244
See Vennix, Jac A. M.: Group Model Building, Chichester, 1996, pp. 97—99.
70 B. Conceptual Foundation for the Usage of System Dynamics
Conceptualization
(intellectual
learning)
Reflection Experimentation
(emotional (action
learning) learning)
245
Taken from the figure in Maani, Kambiz E. and Robert Y. Cavana: Systems
Thinking and Modeling – Understanding Change and Complexity, Auckland,
2000, p. 112.
246
In Maani and Cavana: Systems Thinking and Modeling – Understanding Change
and Complexity, 2000, pp. 112—113, the three elements of the learning cycle
are discussed.
B. Conceptual Foundation for the Usage of System Dynamics 71
Mental Database
Observation
Experience
Written Database
Numerical Database
247
Forrester, Jay W.: “Policies, decisions and information sources for modeling”,
European Journal of Operational Research, Vol. 59, No. 1, 1992, pp. 55—58.
248
Slight modification of figure in Forrester, Jay W.: “Policies, decisions and
information sources for modeling”, European Journal of Operational Research,
Vol. 59, No. 1, 1992, p. 56. Forrester’s figure is also depicted in Luna-Reyes,
L.F. and D. L. Andersen: “Collecting and analysing qualitative data for system
dynamics: methods and models”, System Dynamics Review, Vol. 19, No. 4,
2003, p. 280, along with a discussion on the three types of databases.
249
Forrester, Jay W.: “Policies, decisions and information sources for modeling”,
European Journal of Operational Research, Vol. 59, No. 1, 1992, p. 56.
72 B. Conceptual Foundation for the Usage of System Dynamics
250
Forrester.: “Policies, decisions and information sources for modeling”, European
Journal of Operational Research, Vol. 59, No. 1, 1992, p. 57.
251
Own notes from Kim Warren’s presentation at the System Dynamics Conference
in Oxford in 2004.
252
Luna-Reyes, L.F. and D. L. Andersen: “Collecting and analysing qualitative data
for system dynamics: methods and models”, System Dynamics Review, Vol. 19,
No. 4, 2003, pp. 271—296. Besides the qualitative methods often mentioned in
the system dynamics literature like interviews, focus and Delphi groups,
observations, etc., the article also include analysis methods like hermeneutics,
discourse analysis, grounded theory, ethnographic decision models and content
analysis.
253
Fey, Willard and John Trimble: “The Evaluation and Development of Knowledge
Acquisition in System Dynamics Studies”, in Proceedings, System Dynamics
Conference, System Dynamics Society, 1992, pp. 173—182; Vennix, Jac A. M.:
Group Model Building, Chichester, 1996, pp. 175 and pp. 187—188.
B. Conceptual Foundation for the Usage of System Dynamics 73
than leaving out an important parameter in a model.254 Warren argues that in the
everyday business, decisions are made implicit based on judgment and
assumptions of the parameters, where data is missing.255 Asking for qualified
guesses is just a way to make such judgments and assumptions explicit.
254
Milling, Peter: “Organisationales Lernen und seine Unterstützung durch
Managementsimulatoren”, in: Zeitschrift für Betriebswirtschaft, Ergänzungsheft
3/95, Lernende Unternehmen (Sonderausgabe), 1995, p. 104.
255
Warren, Kim: Competitive Strategy Dynamics, Chichester, 2002, p. 25.
256
Weil, Henry B. and Kenneth P. Veit: “Corporate Strategic Thinking: The role of
System Dynamics” in Peter M. Milling and Erich O.K. Zahn (Eds.): Computer-
Based Management of Complex Systems, Proceedings of the 1989 International
Conference on the System Dynamics Society, Stuttgart, 1989, p. 67, discuss how
system dynamics is often part of a larger package addressing a problem, e.g. in a
strategy consulting project.
257
See Vennix, Jac A. M.: Group Model Building, Chichester, 1996, p. 4; Lane,
p. 70; Richmond, Barry: “The Strategic Forum: aligning objectives, strategy and
process”, System Dynamics Review, Vol. 13, No. 2, 1997, p. 131; Morecroft, J.
D. W.: “Executive knowledge, models and learning”, European Journal of
Operational Research, Vol. 59, 1992, p. 13.
258
See Vennix, Jac A. M.: Group Model Building, Chichester, 1996, p. 6;
Andersen, David F., George P. Richardson and Jac A. M. Vennix: “Group model
building: adding more science to the craft”, System Dynamics Review, Vol. 13,
No. 2, 1997, p. 191.
74 B. Conceptual Foundation for the Usage of System Dynamics
Learning, Lane strongly stresses the learning objectives, such as changing mental
models, improving intuition, allowing risk-free experimentation and revealing
systemic complexity; as vehicles to business improvement.259 Morecroft discusses
the usage of system dynamics in the context of executive dialogue and debate:
Concepts and
theory plus
facilitation
action plans and change
facilitation MAPS
words, diagrams
FRAMEWORK
friendly algebra
259
Lane, David C.: “Modelling as Learning: A consultancy methodology for
enhancing learning in management teams”, European Journal of Operational
Research, Vol. 59, No.1, 1992, p. 72.
260
Taken from the figure in Morecroft, J. D. W.: “Executive knowledge, models
and learning”, European Journal of Operational Research, Vol. 59, 1992, p. 14.
The full article is also printed in the book: Morecroft, John D. W. and John D.
Sterman (eds.): Modeling for Learning Organizations, Portland, Oregon, 1994,
pp. 3—28 (the book is in principle a reprint of the special issue of European
Journal of Operational Research, Vol. 59, No. 1, 1992).
B. Conceptual Foundation for the Usage of System Dynamics 75
261
Morecroft, J. D. W.: “Executive knowledge, models and learning”, European
Journal of Operational Research, Vol. 59, No. 1, 1992, p. 13.
262
See Vennix, Jac A. M.: Group Model Building, Chichester, 1996, pp. 97—99.
263
Argyris, Chris: Interventions Theory and Method – A Behavioural Science
View, Reading, Massachusetts, 1970, p. 83
264
Schein, Edgar H.: Organisationspsykologi, Danish translation, Herning, 1990,
p. 40; Argyris, Chris: Interventions Theory and Method – A Behavioural Science
View, Reading, Massachusetts, 1970, chapter 1 and 2.
76 B. Conceptual Foundation for the Usage of System Dynamics
number of people, using a setup with steering committee, a core group, reference
groups etc. 265 The involvement will often be designed with the dual purpose of
acquiring information and ideas, balancing differing perspectives and interests,
and establishing internal commitment.
A way to coordinate the stream of information in a group model building
process is the usage of a Workbook. Vennix describes a Workbook as a
document consisting of text and diagrams, and presenting material as well as
asking questions.266 The Workbook can be seen as an instrument to summarizing
workshops and preparing for the next workshops and is a way to formalize the
information stream in a modeling project. Akkermans discusses the concept of
The Workshop-Workbook Cycle, being a structured, iterative process involving a
number of in-house discussions (among consultants) and client consultations, as
well as Workbook distribution and feedback, between all workshops.267
When designing a modeling intervention, one of the decisions is whether
or not to employ a preliminary model. The advantages include speeding up the
process and reducing participants time investment, but the major risk is to
infringe the degree of the group’s sense of model ownership, and therefore
Vennix stresses the importance of not being defensive about a preliminary model
when discussing and refining the model. 268 Vennix furthermore offers a
schematic overview of factors indicating in which situations it is appropriate to
use a preliminary model, as well as other design decisions such as
communication pattern, usage of soft methods to initiate the process, and
facilitator preparation.269
265
See Akkermans, Henk: Modelling With Managers, Breda, The Netherlands,
1995, p. 87; Vennix, Jac A. M.: Group Model Building, Chichester, 1996,
p. 114.
266
Vennix, Jac A. M.: Group Model Building, Chichester, 1996, p. 128.
267
Akkermans, Henk: Modelling With Managers, Breda, 1995, p. 89.
268
Vennix, Jac A. M.: Group Model Building, 1996, p. 113.
269
Vennix: Group Model Building, 1996, p. 132.
B. Conceptual Foundation for the Usage of System Dynamics 77
270
Morecroft, J. D. W.: “Executive knowledge, models and learning”, European
Journal of Operational Research, Vol. 59, No. 1, 1992, p. 15.
271
See Andersen, David F and George P. Richardson: “Scripts for group model
building”, System Dynamics Review, Vol. 13, No. 2, 1997, p. 108. See also
Andersen, David F., George P. Richardson and Jac A. M. Vennix: “Group model
building: adding more science to the craft”, System Dynamics Review, Vol. 13,
No. 2, 1997, p. 195, where furthermore The Gifted Practitioner hypothesis is
stated, indicating that some facilitators are skilled to a degree where choice of
intervention method is less important – and also the other way around - if
facilitation skills are lacking, no methods and tools can assure an effective
intervention.
272
Vennix, Jac A. M.: Group Model Building, Chichester, 1996, p. 134.
273
Richardson, George P. and David F. Andersen: “Teamwork in group model
building”, System Dynamics Review, Vol. 11, No. 2, 1995, pp. 114—115.
78 B. Conceptual Foundation for the Usage of System Dynamics
274
Argyris, Chris: Interventions Theory and Method – A Behavioural Science
View, Reading, Massachusetts, 1970, p. 83.
275
Vennix, Jac A. M.: Group Model Building, Chichester, 1996, p. 6.
276
Scheper, Willem J.: Group Decision Support Systems, Tilburg/Utrecht, 1991,
p. 23.
277
Scheper.: Group Decision Support Systems, 1991, p. 143.
B. Conceptual Foundation for the Usage of System Dynamics 79
278
Slight simplification of the model in Eco, Umberto: A Theory of Semiotics,
Bloomington, 1976, p. 141. The model has originally also focus on sub-codes,
which is especially relevant in the discussion of undercoding and overcoding.
The simplification is mainly inspired by the use in Scheper, Willem J.: Group
Decision Support Systems, Dissertation, Tilburg/Utrecht, 1991, p. 143.
279
Miller, George A.: “The Magical Number Seven, Plus Minus Two: Some Limits
on Our Capacity for Processing Information”, The Psychological Review,
Vol. 63, No. 2, March 1956, p. 93.
280
Morecroft, J. D. W.: “Executive knowledge, models and learning”, European
Journal of Operational Research, 59, 1992, p. 18.
80 B. Conceptual Foundation for the Usage of System Dynamics
281
For further descriptions on facilitation of group model sessions, see Vennix, Jac
A. M.: Group Model Building, Chichester, 1996, pp. 140—171.
282
See Akkermans, Henk: Modelling With Managers, Breda, The Netherlands,
1995, pp. 91—92, where credit for the term The Wallow Curve is given to
McKinsey.
283
Vennix, Jac A. M.: Group Model Building, 1996, p. 154.
284
Numerious examples are seen in the art and music industries.
285
Decision quality as a function of conflict level is often seen depicted as a
reversed U-curve.
B. Conceptual Foundation for the Usage of System Dynamics 81
problem
understanding
irritatoin &
conflict level
286
Taken from the figure in Akkermans, Henk: Modelling With Managers, Breda,
1995, p. 91.
C. A Case Study Using Participative System Dynamics
Modeling in the Implementation of a Sensitive Change
Project
287
An alternative scientific approach could have been to investigate the challenges
of bridging the two problem-solving cycles in organizational change projects,
and to evaluate and compare the usage of a number of different methods and
tools in such a setting. The arguments for the applied scientific approach include
the decision to make one in-depth case study in action research tradition, rather
than to spread focus and efforts over a number of field studies.
288
For definitions of action research, see Cummings, Thomas G. and Christopher G.
Worley: Organizational Development and Change, Ohio, 2001, p. 23.
289
Voss, Chris, Nikos Tsikriktsis, and Mark Frohlich: “Case research in operations
management”, International Journal of Operations & Production Management,
Vol. 22, No. 2, 2002, p. 198.
84 C. Case Study
the research approach involving a single site case study does not have the design
characteristics necessary to qualify for testing a theory. 290 Discussing whether to
use one or more case studies in a research project, Stake argues for the benefit of
maximizing knowledge through focusing on a single, in-depth case study, rather
than to disperse focus over a number of case studies.291 Other researchers,
however, champion the viewpoint that theory building should preferably be based
on insights from a larger number of case studies.292 The main argument for the
decision to use only one case study in the present research is the iterative seek-
and-learn relationship that has taken place between the case study and the
theoretical discussion on how best to apply modeling in change management.293
Furthermore, a survey approach was not considered a possible alternative, due to
lack of known cases utilizing system dynamics modeling in change management
context. As an alternative to a multiple case study approach or a survey approach,
the single-site case study approach puts the spotlight on one instance to be
investigated in more detail and thereby concentrates efforts rather than trying to
290
This fits well with the quote from Andersen, David F.; George P. Richardson
and Jac A. M. Vennix: “Group model building: adding more science to the
craft”, System Dynamics Review, Vol. 13, No. 2, 1997, p. 196: “case studies are
only suitable to generate hypotheses, not to test them rigorously.”
291
Stake, Robert E.: The Art of Case Study Research, Thousand Oaks, 1995,
pp. 4—5, argues that a good instrumental case study not necessarily needs to
examine a typical case, as an unusual case helps illustrate matters often
overlooked in typical cases. Furthermore, he argues that even when designing a
collection of cases, representation is often difficult to defend.
292
See Eisenhardt, Kathleen M.: “Building Theories from Case Study Research”,
Academy of Management Review, Vol. 14, No. 4, 1989, pp. 532—550; and
Leonard-Barton, Dorothy: “A Dual Methodology for Case Studies: Synergistic
use of a Longitudinal Single Site with Replicated Multiple Sites”,
Organizational Science, Vol. 1, No. 3, 1990, pp. 248—266. It should be noted,
that even though the present dissertation includes only one case study, it
indirectly benefits from a larger number of case studies from the group model
building literature, as many of these have elements of a change management
focus.
293
See also discussion on exploratory case studies, Yin, Robert K.: Case study
Research, 3 rd edition, Thousand Oaks, 2003, p. 23.
C. Case Study 85
294
Denscombe, Martyn: The Good Research Guide for small-scale social research
projects, Philadelphia, 2 nd edition, 2003, p. 30.
295
The external facilitator (the author of this dissertation) has most of her
management consulting experience from IBM Management Consulting and
Deloitte & Touche Consulting Group (now split into Deloitte Consulting and
Braxton).
296
Yin, Robert K.: Case study Research, 3 rd edition, Thousand Oaks, 2003, p. 28,
emphasizes the importance of theory development as a part of case study design,
regardless of whether the purpose of the case study is to develop or test theories.
86 C. Case Study
297
Schein, Edgar H.: Organisationspsykologi, Danish translation, Herning, 1990,
p. 253.
298
Anderson, Linda A. and Dean Anderson: “Awake at the Wheel: Moving beyond
Change Management to Conscious Change Leadership”, OD Practitioner,
Vol. 33, No. 3, 2001, p. 45. See also Roberto, Michael and Lynne Levesque:
“The Art of Making Change Stick”, MIT Sloan Management Review, Summer
2005, Vol. 46, No. 4, Summer 2005, p. 56, where modeling is relevant with
regards to what they call chartering and learning.
299
Intervention planning, stakeholder management and implementation planning
and review are three conceptually different, but also strongly interdependent
tasks.
C. Case Study 87
time and budget planning.300 Thus, this activity - as the other two
activities discussed next - contains both system dynamics and non-
system dynamics parts, for instance the coordination with related
projects and activities.
• Stakeholder management involves a thorough analysis of all the major
interest groups and individuals who have significant influence—directly
or indirectly—on the success of the intervention. Focus is on the
stakeholders’ interests and power, their importance for solution design
and implementation, and on relevant means of involvement and
communication.301 Stakeholder analysis is vital to intervention planning,
both to secure relevant parameters to be included in the process, and to
secure appropriate involvement of and communication with
stakeholders, including employees.302
• Implementation planning and review deals with the planning of the
implementation, including an action plan, communication plan and a
clear delegation of responsibilities.303 The communication plan develops
300
Roberto, Michael and Lynne Levesque: “The Art of Making Change Stick”, MIT
Sloan Management Review, Summer 2005, Vol. 46, No. 4, Summer 2005,
pp. 55—56, describe intervention planning (although calling it chartering). In
Koningswieset, Roswita and Alexander Exner: Systemische Intervention,
Stuttgart, 1998, it is throughout the book discussed how intervention planning
can be seen as creating an overall architecture and design of the intervention,
including selection of relevant techniques.
301
Flood, Robert L.: Solving Problem Solving, 1995, p. 20, places ‘organizational
politics’ as one of the four key dimensions to take into consideration in a whole
organizational system view, the three others being organizational processes,
design and culture. Borum, Finn: Strategier for organisationsændringer,
Copenhagen, 1995, pp. 79—92, offers a discussion on politics in change
processes.
302
For discussions on “Employee Involvement”, see both Cummings, Thomas G.
and Christopher G. Worley: Organizational Development and Change, Ohio,
2001, p. 317; and Thun, Jörn-Henrik, Peter M. Milling, and Uwe Schwellbach:
“The Impact of Total Employee Involvement on Time-based Manufacturing“, in
“What Really Matters in Operations Management“, Proceedings of the European
Operations Management Association, 8th International Annual Conference,
2001, pp. 133—135.
303
For implementation actions, focus is typical on the practical adjustments of
business processes and organizational design with single-loop orientation (result
88 C. Case Study
The case study addresses problems and issues in bridging between the top-
management launching a strategic initiative and the sustainable rollout of the
organizational change. The main arguments for selecting this specific case
include: (1) that the change initiative represented important and classical change
management challenges such as significant change resistance, strong emotional
attitudes, and that the decision to launch the initiative was made without
involving the key managers responsible for the implementation, and (2) the
change project had characteristics making it likely to gain from the insights from
a system dynamics model, due to the importance of understanding the influence
of delays in an aging chain. Furthermore, the writer of this dissertation had
extensive knowledge of the company and easy access to the decision-makers, due
to previous engagements with the company as well as due to personal
relationships. The existence of personal relationships allowed for increased top
executive involvement in the design of the modeling process, and also allowed
for extensive theoretical discussions with the project owners on the practical
implications of the usage of system dynamics modeling in corporate settings.304
However, personal relationships also create biases in terms of design
opportunities as well as participant perceptions of the researcher’s objectivity,
which has to be taken into account especially in the evaluation of the case study.
As a conclusion on the research considerations: the main research idea has
been to investigate the usefulness of participative system dynamics modeling in
the bridging of the two problem-solving cycles: the cycle of diagnostics and
decision-making, and the cycle of change management.305 In other words, the
dissertation investigates the utilization of proven, successful system dynamics
methods and tools in a new, specific context. An alternative scientific approach
could have been to investigate the challenges of bridging the two problem-
solving cycles in organizational change projects, and to evaluate and compare the
usage of a number of different methods and tools in such a setting. The
arguments for the applied scientific approach include the decision to make one
in-depth case study in action research tradition, rather than to spread focus and
efforts over a number of field studies. A survey approach was not considered a
possible alternative, due to lack of known cases utilizing system dynamics
modeling in change management context. Regarding the use of the case study, it
is stressed that the purpose primarily is to generate inspiration to the focus of the
research due to the seek-and-learn nature of the case study, seeking for theory
building in the action research tradition. Additionally, the case study serves as a
practical illustration in chapter D, in the discussion of the usage of system
dynamics modeling in the implementation of strategic initiatives. The next
subchapter describes the case itself: the problem context, the intervention
planning, the model, and its results. Following the case description comes a
subchapter outlining the evaluation framework for the case study as well as the
actual evaluation.
305
For a discussion on the two cycles of the problem-solving process,
see chapter A.I.
90 C. Case Study
organization. The case study was carried out in one of the major R&D divisions,
consisting of a number of rather different R&D business units.
Strategically, the company works towards creation of a truly global
company in terms of the entire value chain distribution, and as part of these
efforts the board of directors launched a strategic initiative regarding a balanced
location strategy.306 For R&D, specific targets were decided pertaining to the
distribution of the number of R&D employees placed in high-cost countries (e.g.
the USA and Western Europe) vs. the number of R&D employees placed in low-
cost countries (e.g. India, China, Eastern Europe). At the time of the launch of
the initiative, fall 2004, the company had significantly more R&D employees in
high-cost countries with the consequence of relatively high development costs.
The cost of an R&D employee in a high-cost country was at this time
approximately four times the cost of an R&D employee in a low-cost country. It
was a sensitive issue due to the fear among employees in high-cost countries that
the future could bring reductions of staff in high-cost locations as seen in many
other companies in the USA and Western Europe. The situation at the case study
company, however, included strong growth expectations and the company did not
intend to weaken existing high-cost locations in terms of the absolute numbers of
employees. The expansion of low-cost R&D locations should reflect a worldwide
growth of the R&D to gain speed in time-to-market and an extended product
portfolio.307 The motivation was not only to increase capacity and cost-efficiency,
but also to strengthen the local presence in growing markets such as China and
India with a view to increased future sales. New employees in low-cost locations
were primarily supposed to take over existing tasks from high-cost locations.
306
Roberto, Michael and Lynne Levesque: “The Art of Making Change Stick”, MIT
Sloan Management Review, Summer 2005, Vol. 46, No. 4, Summer 2005, p. 53,
define strategic initiatives as “corporate programs aimed at creating new
business processes or transforming existing ones to accomplish major goals,
such as enhancing productivity or improving customer service.” The new
location strategy has major impact on both R&D processes, delivery processes,
and sales support processes as well as the company’s risk profile, currency
spread, and organizational culture elements, which is why it was perceived as a
major strategic initiative in the company.
307
Perlitz, Manfred: Internationales Management, 5 th edition, Stuttgart, 2004, p. 72,
discusses factors for trade between high-cost and low-cost countries. These
factors are also relevant for decisions on distribution of employment.
C. Case Study 91
This way the company wanted to free capacity of experienced senior R&D
employees in high-cost countries to be used in new, challenging R&D projects.
The objective of the organizational intervention consisted of both ‘harder’
and ‘softer’ elements. The softer elements included the creation of acceptance
and commitment to the launched strategic initiative among key implementers, and
the harder elements included understanding the most influential parameters in
building up capacity in low-cost locations, with special focus on productivity and
costs. The process should seek to define the ‘ideal’ implementation approach,
balancing board expectations regarding a reduced cost/capacity ratio with an
effective and realistic implementation plan. The strategic initiative was intended
to stimulate a reinforcing growth loop as depicted in Figure C-1. As such, the
project context was not to explore and identify potential new strategies or
policies, but to implement a given decision in the best possible way and to gain
commitment and understanding from the key implementers. Thus, the
intervention was more of a change management project than a policy formulation
project.
strengthening low-cost
locations without reducing
high-cost locations
+ + +
+ maintained
company improved reduced
time-to- unit-price motivation in
growth high-cost
market of R&D
+ locations
+
improved
competitiveness
+
+
Figure C-1: The reinforcing growth loop underlying the intervention 308
308
Causal-loop diagrams are normally constructed in such a way that the variable
names do not indicate the direction of change. In Figure C-1, the variable names
include the direction, e.g. company growth, reduced unit-price etc. This was
done for communication purposes, emphasizing how the reinforcing growth loop
was intended to work.
92 C. Case Study
309
Snabe, Birgitte, Andreas Größler, and Peter M. Milling: “Policies and Politics of
Establishing R&D Capacity in Low-Cost Locations”, Tagungsband, GWS-
Tagung, Greifswald, in print, 2006, focus especially on the political challenges
in the case study.
C. Case Study 93
2. Intervention Process
The location strategy project was first initiated with a project team in each of the
five business units in the division, but due to lack of consistency and efficiency,
and lack of structure in the coordination and communication between the teams,
it was decided to develop a shared, formal model on an abstract and highly
aggregated level, aiming towards:
• creating a structured and ‘objective’ frame for the rather emotional and
diverse discussions;
• establishing a forum for exchange of experiences and best practices;
• refining and making operational the strategy outlined by the board;
• improving the change process effectiveness.
The change imperative was stated as: “right now is the right time to hire
people in the low-cost locations, because due to company growth, it can be done
right now without staff reduction at high-cost locations, and the expected results
are improved competitiveness and further company growth, also securing jobs at
high-cost locations in the future.” This was a difficult message to communicate
because of the fear of jobs moving from high-cost to low-cost locations.
An external facilitator from Mannheim University (the author of this
dissertation) was brought in as process coach and modeling facilitator, based on a
participative modeling approach. For system dynamics modeling, the Vensim
modeling environment was used.
Targeting and planning the intervention initially involved a discussion
with the project owners about the problem, the business objectives, the
intervention objectives, the dynamic hypotheses, and the suitability of applying
system dynamics to the problem. A preliminary model was built to frame the
modeling project, and also for the project owners to feel confident that major
results from the model were consistent with the directions of the strategy
expressed in the objectives and targets. This preliminary model was the basis for
the decision to move forward with modeling.
94 C. Case Study
DIVISION LEVEL
3.
3. Consolidate,
Consolidate, Balance
Balance &
&
1.
1. Vision
Vision & Modeling
Meta Modeling Coordinate
Agree on objectives & guiding principles Reduce limitations/barriers
Agreement on shared,
- formal model Set division targets
Set and agree critical parameters Balance between areas
Define ideal strategy Identify critical success factors
Agree on communication strategy
BUS. AREA LEVEL
4.
4. Detailed
Detailed Planning
Planning
2.
2. Area
Area Strategy
Strategy & Execution
Execution
Evaluate ideal strategy against reality Detailed planning
Define Area strategy based on ideal strategy Align with budget process 2005
Identify limitations/barriers Build 3 year plan (quarterly breakdown)
Business cases based on agreed parameters Detailed communication plan
Identify short term wins
The modeling and simulation process with the core project team served as
a cognitive framework for objective discussions of the problem – challenging
preconceived perceptions and aiming towards reducing the tendency often known
from budgeting and business planning processes, that each stakeholder to some
degree primarily looks after his or her business unit’s interests rather than
corporate objectives. Using a shared model on a highly aggregated level moved
the focus and discussion towards a holistic view. In this process some important
new aspects – including one additional stock and a number of additional
parameters – were added to the preliminary model. Also some parameters and
relationships, which through simulations showed only little impact, were removed
in order to simplify the model. For communication purposes, the modeling
process was focused on developing a relatively simple model that could provide a
picture of the overall behavior of the problem-system without including too many
details, as overview was considered more important than detailed correctness.
The focus on keeping the model simple was also due to the role of the system
dynamics model role as a ‘meta-model’, creating cross-unit overview, and being
complemented by a detailed excel-model in each business unit with the format
needed in the budgeting and business planning process.
The parameter setting was a cornerstone in the change-process, as these
agreed parameters formed the basis for each business unit’s business case in
phase 2 of the intervention process. The discussion of the parameters often
resulted in a discussion on how the strategy could and should be executed, as the
parameter setting reflected implementation decisions; e.g. the logistics and cost
model of traveling, how to structure knowledge transfer, etc.
Model testing was done partly behind the scenes by the modeling
facilitator using some of the most respected sources as guides and checklists”,310
partly during the modeling and simulation efforts, as “validation is the process of
establishing confidence in the soundness and usefulness of a model“.311 Also
310
Barlas, Yaman: “Formal aspects of model validity and validation in system
dynamics”, System Dynamics Review, Vol. 12, No. 3, 1996, pp. 183—210; and
Forrester, Jay W. and Peter Senge: “Tests for Building Confidence in System
Dynamics Models”, in Legasto, Augusto A., Jay W. Forrester und James M.
Lyneis (eds.), TIMS Studies in the Management Sciences, Vol. 14, Amsterdam,
1980, pp. 209—228.
311
Forrester, Jay W. and Peter Senge: “Tests for Building Confidence in System
Dynamics Models, 1980, p. 210.
C. Case Study 97
312
Forrester, Jay W.: Principles of Systems, Cambridge, 1968, chapter 3, p. 4.
313
Rockart, John F.: “Chief executives define their own data needs”, Harvard
Business Review, March-April 1979, pp. 81—93. For a discussion on the usage
of critical success factors, see also Snabe, Birgitte and Jakob F. Ølgaard:
Informationsstrategi, Master Thesis at The Institute of Mathematical Statistics
and Operations Research, Technical University of Denmark, Lyngby, 1992,
pp. 25—30.
98 C. Case Study
314
See Sterman, John D.: Business Dynamics, Boston, 2000, p. 491. Sterman
operates with only two levels in his promotion chain, with employees leaving
both levels. The location strategy model operates with 3 and 4 levels, with
employees only leaving the latest stage as this reflects the historical data well.
Other examples of system dynamics models of aging chains include Martinez,
Ignacio J. and Luis F Luna: “The Dynamics of Best Practices: A Structural
Approach”, at CD-ROM of Proceedings, System Dynamics Conference, System
Dynamics Society, 2001, which operates with three levels of practitioners:
junior, intermediate, and advanced.
315
In the model as well as the case description, the abbreviation FTE (Full Time
Equivalents) is used to indicate number of full-time employees. The abbreviation
is an established term in the case study company, used to adjust budget numbers
for employees working part-time.
handover capacity
reduction
<HC hire> <HC quit>
C. Case Study
HC CAPACITY USE
ON HAND-OVER
LC PRODUCTIVITY
LC replacing HC quit REDUCTION
LC ROOKIE HC ROOKIE
PRODUCTIVITY production per month PRODUCTIVITY
ADDITIONAL
GROWTH REPLACEMENT
IN HC VS. LC
LC QUIT HC QUIT
FRACTION FRACTION HC ROOKIE HC TRAINING <HC
TIME TIME quit>
LC new hire LC LC HC
LC H-O-R LC 1c quit quit
New hired 1a 1b Rookie Productive Productive Rookie New hired
FTE LC
FTE LC FTE LC FTE LC FTE HC HC FTE HC HC FTE HC HC
LC repl. hire LC 2a LC job-trained trained hire
2b
LC flow input
LC HAND-OVER TRAVEL COST
<H-O-R FTE LC>
handover
travel costs
over tasks from high-cost countries, which will be the case for all new employees
in low-cost countries who are not merely replacing people who have left a
position at a low-cost location. This stock is called H-O-R FTE (Hand-Over-
Rookies). These employees have zero productivity and, as they are spending time
physically with those HC employees, whose tasks they are taking over, they
furthermore tax time from fully productive employees in high-cost countries in
the process of knowledge transfer.
The two main decision points influencing the changes in stock levels in the
model are:
1. The rate “HC hire”, where only a share of the employees leaving high-
cost locations will be replaced at a high-cost location, depending on the
factor called REPLACEMENT OF HC VS. LC in the model. Those
headcounts not being replaced in HC will be replaced in LC.316
2. The rate “LC new hire”, consisting partly of replacement of employees
leaving the stock of productive LC employees, partly of headcounts not
being replaced in HC, and partly of the additional new employees
joining the division (called ADDITIONAL GROWTH in the model). All
additional new employees are allocated to low-cost countries, based on
a growth factor relative to the stock of experienced employees in low-
cost. 317
It is important to note that tasks will be moved from high-cost to low-cost
locations in bulk. Employees in high-cost locations, who have transferred their
tasks to low-cost locations, will either take over tasks from a person leaving the
division or take part in new R&D projects within the division. Tasks moved to
low-cost locations enable new and innovative tasks in high-cost locations; i.e.
does not lower productivity in high-cost locations.
316
Due to the non-replacements, the number of employees can be reduced in some
locations. However, the total number of employees in high-cost locations will
stay roughly stable due to necessary hiring in other functions in the division.
317
In Sterman, John D.: Business Dynamics, Boston, 2000, p. 491, the “two-level
promotion chain” has a growth factor related to the total number of employees,
but in the location strategy model it makes more sense to base the growth on
fully productive FTE’s, due to ramp-up limitations (ratio between experienced
staff and new staff).
C. Case Study 101
318
Examples of removed parameters: the number of productive days is higher in LC
is than in HC, but these parameters were removed from the model (and added in
a comment instead), as it was decided to let the higher productivity in HC equals
out with additional coordination overhead. Also the cost of hire was removed
from the model, as they proved to be insignificant in simulations.
319
In this case, it should be used in combination with a MIN-function in the LC
new hire-rate to secure that the number of newly hired employees did not exceed
102 C. Case Study
This would have made the policy of how many additional employees to hire
endogenous.320 However, in order to keep the model simple and with focus on the
few, most important parameters, the growth rate is simply implemented as a
constant (fixed for the first period, then gradually decreasing to zero after the 36th
month). For the same reasons, as well as to allow some degree of cost overrun, it
was decided not to handle the cost restrictions endogenously, but to incorporate
cost as an explicit auxiliary.
In Sterman’s Two-Level Promotion Chain, the rate of employees moving
from a ‘rookie’ state to an ‘experienced’ state was modeled as a fraction of the
number of Rookies321. In the location strategy model the rates between stages are
calculated as delay-functions of the inflow-rates. For the training period, a high-
order delay was used to imitate a pipeline delay, as this is a fixed period of time
for each employee. For the period of being a rookie, a lower order delay was
used, to reflect the variability in the learning curve for individuals. The
variability in the difficulties of the task areas is not explicit in the model, but is
considered in stipulating the average time for employees being rookies. Having
outflow-rates as a function of inflow-rates (rate-on-rate modeling) places the case
study model in a slight conflict with the tendency among system dynamic
practitioners to avoid rate-on-rate modeling, and rather calculate outflow-rates as
a fraction of stock-level. The reason for the tendency to avoid rate-on-rate
modeling can perhaps be seen partly as motivated by prevailing attention to
macro-trends of systems with continuous parameter development (rather than
systems with steps in inputs), and partly motivated by a historical tradition based
on Forrester’s “Principles of System”, which can be dated back to a time where
delay-functions constituted a seriously challenging workload for computers.322
For the location strategy model, the new strategy with a step in the hiring rate at
the very beginning is being investigated, and the use of the delay-function better
reflects the true pattern of employee-flows in the start-up period. Due to the use
of rate-on-rate modeling, a fixed period of time will pass before the first
additional new employees flow into the rookie state (e.g. 3 months) and an even
longer time will pass (e.g. 9 months) before the number of experienced
employees is beginning to increase. This is intuitively more acceptable for group
modeling participants, as they are not trained in system dynamics modeling, and
therefore could have trouble abstracting from the number of experienced
employees beginning to increase as early as the very first month. Interestingly
enough, though, the major trends and learning – even regarding year 1 - are the
same with both ways of modeling the delays, see appendix E.
The figures C-4, C-5 and C-6 show the behavior of selected variables for four
simulation runs. Naturally, a large number of simulations took place in the
workshops to investigate the importance of the influence of the different
parameter settings and the consequences of different scenarios, but the selected
simulation runs give an impression of some of the key learning from the model.
The selected simulation runs are labeled: “INI”, “Base run”, “40% HC
replacement”, and “Faster training and hand-over”. Below is a short description
of the parameter setting in each of the simulations:
The following three figures show major results of the four selected runs:
2 4 2 4 2 4 2 4 2
2 3 4 3 3 3 3
0.2 23 4 2 3 4
4
4 2 3
4 2 3
1 2 3 1 1 1 1 1 1 1 1 1 1
0
0 6 12 18 24 30 36 42 48 54 60
Months
INI 1 1 1 1 1 1 1 Dmnl
Base run 2 2 2 2 2 2 2 Dmnl
40% HC replacement 3 3 3 3 3 3 Dmnl
Faster training and hand-over 4 4 4 4 4 Dmnl
1.15 1
1
1
1
1
1.1 1
1
1 23
1 2 34
1 34
1.05 1 3 23 2 3 2 3 2 3 2 34 234 2
3 1 3 1 2 3 23 2
3 2 3 2 4 4 4 4 4 4
4 1 4 4 4
2
1
0 6 12 18 24 30 36 42 48 54 60
Months
INI 1 1 1 1 1 1 1 1 1 1 1 Dmnl
Base run 2 2 2 2 2 2 2 2 2 2 2 Dmnl
40% HC replacement 3 3 3 3 3 3 3 3 3 Dmnl
Faster training and hand-over 4 4 4 4 4 4 4 4 Dmnl
323
The “worse-before-better“ effect is widely recognized in change management
literature. The effect is also illustrated by the means of system dynamics models,
e.g. Repenning, Nelson P. and John D. Sterman: Nobody Ever Gets Credit for
Fixing Problems that Never Happened: Creating and Sustaining Process
Improvement”, California Management Review, Vol.43, No.4, Summer 2001,
p. 74.
C. Case Study 107
took place in a more indirect way due to discussions during model refinement
and parameter setting. Examples are:324
• The alignment of actual numbers to be used in the business cases (also
called the Excel-models). The business cases were often large ‘black-
box’ models. By having a system dynamics model on a highly
aggregated level, the most important parameters were made transparent.
An example of this is the employee turnover rate in low cost countries
that in the different business cases was set to be in the range of 10% to
25% based on personal assumptions, although historical figures showed
to be approximately 7%.
• The strategy could be accelerated by a shift in the entire company from
headcount orientation towards cost orientation.
• Identification of detailed transfer planning as a critical success factor.
The after-transfer solution for involved high-cost location employees
was perceived to be especially important in order to secure motivation
and morale. This included planning for replacements in high-cost
locations being made by colleagues who have handed over tasks to low-
cost locations.
• A number of best practice experiences were exchanged between the
business units because some units already had more experience in
building up capacity in low-cost locations. The modeling sessions also
served as a forum for the discussion of topics such as bridgeheads,
hiring strategies, and practical aspects of making new employees
productive as quickly as possible.
324
Further examples are discussed in chapter D.
108 C. Case Study
325
Huz, Steven, David F. Andersen, George P. Richardsen and Roger Boothroyd:
“A framework for evaluating systems thinking interventions: an experimental
approach to mental health system change”, System Dynamics Review, Vol. 13,
No. 2, 1997, pp. 149—169; Rouwette, Etiënne: Group model building as mutual
persuasion, The Netherlands, 2003, pp. 68—95. The framework of Rouwette
partly is based upon the work of Huz.
C. Case Study 109
with no test group - it will not give data to support or reject hypotheses about
relative effectiveness compared to alternative mechanisms. The value of the
evaluation is therefore strongly connected with the search for explanations and a
focus on the question of why certain outcomes might happen.
Data on the evaluation variables was collected from three sources: (1)
observations by the facilitator, (2) semi-structured interviews, and (3)
questionnaires among both core team members and non-core project participants.
The usefulness of the sources differs between the measurement variables, but
each source adds some information to all variables regarding outcome and
method. In the analysis, the appropriateness of the different sources for each
measure variable is taken into account. For observations and interviews, focus
was especially on potential manipulation and biases.326 Significant bias must be
expected to occur due to the mental model of the observer and interviewer (the
author of this dissertation).327 The usage of a structured evaluation framework
lower the bias, but the mental model will inevitable be a “filter” in the selection
and interpretation of observations and interview results. However, as the purpose
of the case study is theory building rather than theory testing, the bias is less
crucial for the research value.
The questionnaires are closely linked to the evaluation framework, and are
designed to be answered anonymously. They are furthermore short (1 page), in
order to increase response rate, and they aim to provide unambiguous and
objective questions about participants opinions regarding both intervention
output and mechanism on a 1 to 7 scale (from strongly disagree to strongly agree,
4 being neutral). The questions refer directly to the measure variables in the
evaluation framework, to reduce bias due to the researcher’s own pre-coded view
of the research. In addition, the use of checklists for how to use and design
questionnaires was applied in an attempt to reduced bias.328 The questionnaires
326
Yin, Robert K.: Case study Research, 3 rd edition, Thousand Oaks, 2003,
pp. 93—96, discusses the problems of participant observations. Kvale, Steinar:
InterViews: An Introduction to Qualitative Research Interviewing, Thousand
Oaks, 1996, pp. 235—252, discusses the reliability and validity of interviews.
327
For discussions on self-reinforcing of mental models, see Bakken, Bent E.:
Learning and Transfer of Understanding in Dynamics Decision Environments,
Boston, 1993. pp. 29—30; Argyris, Chris: Reasoning, Learning, and Action –
Individual and Organizational, San Francisco, 1982, p. 39.
328
Denscombe, Martyn: The Good Research Guide for small-scale social research
projects, Philadelphia, 2 nd edition, 2003, chapter 9.
110 C. Case Study
were pre-tested with both a project owner and a research colleague. The
questionnaire was given to the five core-team members, to the three most
involved steering committee members, and to three other participants, who were
not as involved in the project, but had only been exposed to the model in one or
two meetings. The latter group received a reduced version of the questionnaire,
as they did not have an overview of the complete project. The interviews are
structured around the same measure variables as the questionnaires, but with
open questions, and the interviews were conducted with the two project owners.
The facilitator’s own observations are also decoded in the same structure as the
questionnaires and the interviews.
The following tables (C-2a to C-2e) illustrate the applied framework’s sources
for data collection.
C. Case Study 111
Table C-2d: Main sources for evaluation of system dynamics compared to other approaches
C. Case Study 115
Table C-2e: Main sources for evaluation of the usage of system dynamics
in a change management context
Question n mean sd
I believe it was useful to include the model in the
8 6.00 0.76
project
I gained interesting learning from the
11 5.82 0.87
model
I agree with the recommendations derived from the
8 5.00 1.41
model and will act accordingly
The modeling affected some of my
8 4.63 1.85
decisions
The model was a useful framework facilitating
11 6.18 0.87
discussions
The modeling process helped building a shared
11 5.73 0.90
view of the location strategy
The modeling efforts helped creating a common
11 5.27 1.01
language for the location strategy
The model was a useful tool in the presentation of
11 5.09 1.14
the ideal location strategy
I believe the recommendations from the modeling
11 4.27 1.19
process will be implemented
I believe the recommendations will have positive
11 4.91 1.22
business impact
The use of modeling increased the efficiency of the
8 5.13 1.13
project process
Using modeling in this case was more efficient
8 5.13 1.13
compared to other approaches
Using modeling in this case created higher quality
8 5.38 0.92
results compared to other approaches
I intend to use modeling in other
8 5.00 0.93
change projects
I believe the model reflects the core
8 5.38 1.19
of the problem
It was useful to start the modeling with a
8 6.25 0.71
preliminary model to kick-start the process
329
In Warren, Kim: “Improving strategic management with the fundamental
principles of system dynamics”, System Dynamics Review, Vol. 21, No. 4,
Winter 2005, p. 329, differentiation of strategic management is made between
one-off challenges and the continuous direction of enterprise strategy.
330
The ‘context – process – outcome’ structure is inspired by the ‘context –
mechanism – outcome’ configurator as described Rouwette, Etiënne: Group
model building as mutual persuasion, Nijmegen, 2003, pp. 87—92; and also by
the ‘intervention – organization – effect’ model (a variant of a classical S-O-R
model) as described in Borum, Finn: Strategier for organisationsændringer,
Copenhagen, 1995, p. 56.
120 D. The Usage and Utility of Participative Modeling in Change Management
331
For a discussion on change strategy taxonomies, see chapter A.II.
332
Palmer, Ian and Richard Dunford: “Who says change can be managed?
Positions, perspectives and problematics”, Strategic Change, Volume 11, August
2002, p. 244.
333
Anderson, Linda A. and Dean Anderson: “Awake at the Wheel: Moving beyond
Change Management to Conscious Change Leadership”, OD Practitioner,
Vol. 33, No. 3, 2001, p. 45.
334
Leavitt, Harold J.: Top Down – Why Hierarchies Are Here to Stay and How to
Manage Them More Effectively, Boston, 2005, pp. 164—168.
335
According to Akkermans, Henk: Modelling With Managers, Breda, p. 20, the
efforts of establishing internal commitment in organizational interventions are
often centered on the creation of awareness, consensus, and confidence
regarding the goals and the change process
D. The Usage and Utility of Participative Modeling in Change Management 121
336
Although action research has it origin from micro-organizational issues with
much of the earlier literature primarily discussing changes of group behavior, the
field of organizational development has included a whole range of techniques to
adjust the theories of action research to be suitable for large-scale changes in
organizations. Practically all of the literature from Lewin, Argyris and Schein
used in this dissertation discuss mostly interventions on group level. For the
evolution of action research in the organizational development literature, see
Palmer, Ian and Richard Dunford: “Who says change can be managed?
Positions, perspectives and problematics”, Strategic Change, Volume 11, August
2002, p. 247.
337
Winch, Graham and Sonja Derrick: “Flexible Study Processes in ‘Knotty’ System
Dynamics Projects”, Journal of Systems Research and Behavioral Science,
Vol. 23, No. 4, in print, 2006, pp. 1—2.
122 D. The Usage and Utility of Participative Modeling in Change Management
338
Eskinasi, Martijn and Eppie Fokkema: “Bursting the myth of making easy
modeling money”, unpublished working paper (permission granted from the
authors), Proceedings, Second European System Dynamics Workshop,
Nijmegen, 2005, p. 63.
339
Example: implementation of a new business process might benefit from other
modeling mechanisms such as the ARIS toolkit as described in Scheer, August-
Wilhem: Business Process Reengineering – Reference Models for Industrial
Enterprises, 2nd edition, Berlin, 1994.
340
Flood, Robert L. and Michael C. Jackson: Creative Problem Solving – Total
Systems Intervention, Chichester, 1991, p. 12. For a discussion on power,
conflict and organizational differences, see also Argyris, Chris: Interventions
Theory and Method – A Behavioural Science View, Reading, Massachusetts,
1970, p. 81.
D. The Usage and Utility of Participative Modeling in Change Management 123
issues of interests, conflict and power (see table D-1). From a change
management viewpoint, modeling is primarily relevant in a pre-dominantly
pluralist situation. In a dominant unitary situation, change can be managed
without investing time on modeling. In a dominant coercive situation, it is
illusionary to expect the gain the necessary consensus and trust in the modeling
sessions.
341
The table is based on the table in Flood, Robert L. and Michael C. Jackson:
Creative Problem Solving – Total Systems Intervention, Chichester, 1991, p. 13.
The indication of modeling relevance is a modification of the table.
124 D. The Usage and Utility of Participative Modeling in Change Management
the people responsible for implementing the new structure and processes. In this
situation, the overall business strategy and objectives of the companies was used
to seek interest alignment, and the modeling process was used as the main
mechanism ensuring involvement and focusing the discussions on both the
usefulness and the practical implications of the decided change.342 The level of
conflict was relatively high, resulting in strong engagement in the discussions,
although not of a magnitude preventing a constructive atmosphere after the
initiation of a structured and framed process. Within the frame given by the
outlined targets, the project participants had a high degree of empowerment.
Even radical new ideas on how to conduct business were discussed, recognizing
the need for creativity to come from all levels in the organization, and not as a
top-down process.343
A modeling process will inevitably be part of the social constructivism in
the setting.344 In the social constructivism typology by Wenneberg, ‘the critical
perspective’ and ‘the ethical reflection’ provide particularly interesting
perspectives on modeling.345 In the case study, at a point in time the core team
discussed whether to include the parameter ‘motivation’: having capacity build-
up in low-cost locations resulting in reduced motivation among high-cost
employees and consequently reducing productivity in high-cost locations. From a
critical perspective, this might be an actually existing underlying social construct,
but the project owners did not want to legitimize such behavior, which can be
342
See Snabe, Birgitte and Andreas Größler: “System Dynamics Modelling for
Strategy Implementation: Case Study and Issues”, Journal of Systems Research
and Behavioral Science, Vol. 23, No. 4, in print, 2006, pp. 16—17, for a
discussion on ‘manipulation’ when refining and implementing a given decision.
343
See Hamel, Gary: Leading the Revolution, Boston, 2000, p. 280, opposing the
idea that new strategies, innovation and change should always start from the top.
Throughout the book it is discussed how innovation and radical ideas on how to
change the way a company does business should also come from ‘activists’ from
inside the company; with activists being people dedicated to rule-bursting and
daring unconventional business.
344
Kieser, Alfred: “Kontruktivistische Ansätze”, in Alfred Kieser (ed.):
Organisationstheorien, 3rd edition, Stuttgart, 1999, p. 288, discusses social
constructivism as how human communication and interaction produce a social
reality that appears as the objective reality.
345
Wenneberg, Søren B.: “Socialkonstruktivisme som videnskabsteori – Sisyfos’
videnskab”, Online Paper, Institut for Ledelse, Politik og Filosofi, Copenhagen,
2002, pp. 8—9.
D. The Usage and Utility of Participative Modeling in Change Management 125
346
Negative consequences on organizational trust among employees has been seen
to be avoided even in cases of both redeployment and relocation of employees,
which was interpreted as being due to factors such as acceptance of the
arguments for change, perceived organizational support, and fairness; see Ferres,
Natalie, Julia Connell, and Anthony Travaglione: “The effect of future
redeployment on organizational trust”, Strategic Change, Vol. 14, March-April
2005, pp. 87—88.
347
Van der Smagt, Ton: “Causation and Constitution in System Dynamics;
Modelling a Socially Constituted World”, Journal of Systems Research and
Behavioral Science, Vol. 23, No. 4, in print, 2006, pp. 13—14. Van de Smagt,
p. 1, furthermore argues: “causal models blind us for constitutional change.”
126 D. The Usage and Utility of Participative Modeling in Change Management
348
See Checkland, Peter: “Systems Thinking, Systems Practice”, Chichester, 1993,
p. 219, for a discussion on the importance of the concept of Weltanschauung in
both ‘hard’ and ‘soft’ system methodologies.
349
Van der Smagt, Ton: “Causation and Constitution in System Dynamics:
Modelling a Socially Constituted World”, 2006, p. 14, argues that causal models
do not even ‘invite’ to think about constitutional issues. This might very well be
true for the modeling workshops, but in the preparation of the modeling
processes (especially in drafting the preliminary model), model owners have a
clear opportunity to include structured commitment to constitution.
350
See Lyneis, James M.: Corporate Planning and Policy Design: A System
Dynamics Approach, Boston, 1980, pp. 6—9. Lyneis furthermore emphasize
corporate policies not being sufficiently robust to handle changing conditions.
351
In Thygesen, Henriette H.: System Dynamics in Action, Copenhagen, 2004,
p. 189, it is argued that the system dynamics modeling process assist in creating
“an atmosphere of shared reality” between project participants.
D. The Usage and Utility of Participative Modeling in Change Management 127
with little newness, due to the executives own extensive, intuitive understanding
of the overall system behavior.352
A final comment regarding context factors relevant to the appropriateness
of using system dynamics modeling concerns skills, attitudes and traditions
among project participants. It is relevant to consider whether the employees have
skills and attitudes suitable for the usage of a highly conceptual and mathematical
approach. Also, the process must correspond with how the organization normally
handles change, solves problems, and involves employees. In the case study,
most participants were mathematical skilled people; many were engineers. The
question is whether modeling is only relevant for this type of company. An
argument against this viewpoint is that in most companies, major change projects
make use of extensive Excel-models to plan, track and optimize. Thus, formal
models are part of strategic change initiatives anyway. Implementation of change
projects as a part of business planning must be able to give input to the budgets
for the coming quarters and years. Therefore, managers responsible for leading
change are generally used to models, used to mathematical approaches, and often
appreciate the advantages that system dynamics offers compared to black-box
models. Furthermore, people who want to hide weak arguments in huge
datasheets will have a problem. Through the system dynamics modeling process,
the numbers from the business cases in each business unit came out in the open,
and “the approach made it difficult for people to play politics”. The structure of
the Excel-based business cases often made them into ‘black box’ models, where
it was difficult for people other than the model owners to see the critical
parameter setting through the ‘jungle’ of detailed information and hidden
formulas. The system dynamics modeling process served as a ‘meta-model’,
aiming at having a model on a high aggregation level only including the most
significant parameters—creating transparency even beyond the people doing the
actual modeling.
352
A former Forrester student, who has worked five years with system dynamics
modeling in McKinsey, has stated this viewpoint. The CEO of the case study
company has also stated that viewpoint.
128 D. The Usage and Utility of Participative Modeling in Change Management
The design of the modeling process is a sub-set and closely interrelated to the
overall intervention design, with its stakeholder analysis, communication strategy
and plan, adjustment of business processes and measurement systems etc. In the
case study, although recognizing the interdependencies as well as the iterative
nature both internally between the steps and with regards to the overall
intervention design, eight conceptual modeling steps crystallize:
Step 1
Frame and Design Modeling Process
Step 2
Present Preliminary Model as Kick-off
Step 3
Adjust and Refine Model
Step 4
Discuss and Stipulate Parameters
Step 5
Analyze Sensitivity and Identify Critical Parameters
Step 6
Simulate Scenarios
Step 7
Plan for Improvements of Critical Parameters
Step 8
Provide Feedback to Initial Decision-Makers
The first step, Frame and Design Modeling Process, explicitly formulated
the given decisions (‘what is not for discussion’), respectively the degrees of
decision freedom in the implementation and modeling process (‘what is open for
discussion’). The preliminary model had a central role in step 1, both in the
investigation of the appropriateness of using a model and in the process of having
project owners feel comfortable with expected key learning from the model. The
business objectives and the targets in combination with the preliminary model
clearly framed the modeling process, and because of transparency of process
intentions, the project owners expected no perceptions of manipulation.
In the second step, the preliminary model was used to initiate the modeling
workshops with core project participants, who represented some of the key
managers responsible for implementation. Also, a number of other stakeholders
were introduced to the model in early versions, in order to allow for a sense of
involvement and for getting input, and to strive for both effectiveness and
efficiency in the process. The third step, Adjust and Refine Model, was intended
to serve the purpose of participants gaining trust in the model, as well as
improving the model. Allowing for model adjustments, together with later steps,
was seen as important for giving modeling participants true influence, which is an
important factor in getting commitment in the change process. In step four,
Discuss and Stipulate Parameters, the parameters were discussed in the group.
Some of the parameters were highly subjective, e.g. how long does it take for a
rookie employee to be fully productive, and it was perceived as important for
participant trust, that these estimations were made without interference from top
executives. The quantification process served as a means for aligning parameters
between business units, to be used in the different business units in the detailed
business cases. Also, the parameter estimations initiated discussion on best
practices across the business units.
The fifth step, Analyze Sensitivity and Identify Critical Parameters,
together with the sixth step, Simulate Scenarios, enabled the change leaders to
identify the crucial parameters, where improvement initiatives would have
significant impact.353 The main purpose was to focus the discussion on the critical
parameters in a way leading to focused implementation decisions. The facilitator
played an important role in turning barriers and issues into improvement
353
Example: the group realized the importance of reducing the elapsed time of
handing over tasks. If this had been a top-down request, it could very well have
caused significant change resistance.
130 D. The Usage and Utility of Participative Modeling in Change Management
354
The selection of four significant areas is made to focus the discussion and to
avoid the impossible mission of writing a ‘manual’ on how to apply system
dynamics modeling in change management context.
355
In Schneiderman, Arthur M.: “Setting Quality Goals: Use observed rates of
continuous improvement to position targets”, Quality Progress, April 1988,
p. 56, rational goal setting is discussed in terms of the individual company’s
situation. The term “half-life“ is introduced, indicating that a constant time
factor exist for reaching half of the theoretical possible process improvement at
any given time (comparable with the behavior of radioactive decay).
132 D. The Usage and Utility of Participative Modeling in Change Management
356
Akkermans, Henk: Modelling With Managers, Breda, 1995, pp. 116—117.
357
Akkermans: Modelling With Managers, 1995, p. 116.
D. The Usage and Utility of Participative Modeling in Change Management 133
strategy case study, one extreme viewpoint could be that no free exploration of
all possible scenarios took place, no free discussions over sound and just business
objectives were held, and only limited hierarchy-free conversations resulted.
Sterman discusses situations, where clients use models to support already reached
conclusions, recommending modelers to “speak truth to power” and get “a better
client,” and Borum discusses the importance of free, informed choices in
establishing commitment in change processes. 358 A different viewpoint on the
same matter could be that everyone in the process was informed about the
expected goals of the project, no-one was given the impression that he or she had
more influence on the direction of the strategy than was actually the case, and no-
one did generally criticize the task of implementing a strategy that was decided
by someone else. On the contrary, everyone agreed that by modeling, structure
was given to a usually chaotic and emotionally charged process and with the
system dynamics model a basis for discussion existed. Furthermore, free,
informed choices took place both with regards to the model refinement,
parameter settings, and the decisions on how to implement the strategy. This was
a deliberate part of enhancing participant ownership of the model.
It can be argued that the context in which the case study took place reflects
a rather typical and frequently occurring set-up in corporate environments with
which organizational members do not necessarily disagree. In the evaluation
interviews, there were some remarks concerning the fact that participants in the
process were quite informed about their options to change things. Most agreed
that their task was to refine and implement a given strategy, not to re-formulate
the strategy. However, for others the modeling process served as a disciplinary
tool, as the following quote from a participant shows: “A few participants did not
agree with the business objectives and for that reason they also did not agree with
the process, but nevertheless the process forced them in the decided direction,
and through the modeling they gained some of the insights motivating the
intervention in the first place”. However, this can be true for any intervention
approach and for many situations in organizations: people in business firms are
well aware of the fact that they sometimes have to implement things they would
358
Sterman, John D.: Business Dynamics – Systems Thinking and Modeling for a
Complex World, Boston, 2000, p. 85; Borum, Finn: Strategier for
organisationsændringer, Copenhagen, 1995, p. 58. It is important to note, that
Sterman’s recommendation is made in a modeling context different from change
management.
134 D. The Usage and Utility of Participative Modeling in Change Management
rather not. In interviews with the project owners, the issue of possible
‘manipulation’ was addressed explicitly. The project owners did not agree with
this viewpoint, as it is normal business that the top executives outline a strategy
and the level below is responsible for implementing it. In their view, top
executives as a part of their role and responsibility exercise legitimate use of
power, which has noting to do with manipulation because it is openly stated
upfront. Thus, there are frequently decisions in larger organizations that are not
open for discussion. It is difficult to say whether the problematic of a framed
intervention had a negative impact on the participants’ ownership and trust in the
model. The questionnaires do not explicitly include questions regarding this
possible impact of a preliminary model, due to the problem of a measure that
influences the system (in this case, creating negative attitudes by explicitly
raising the issue).
As a final remark on the role of business objectives and targets, it is
relevant to consider the common trends in the literature discussing change
management. In general, clear objectives and targets are seen as important and
necessary elements in creating planned change. Therefore, when modeling efforts
are used within the context of change management it could be seen as quite
natural that the process is framed by business objectives and targets.359
Furthermore, the literature also emphasizes the iterative nature of problem
solving and strategic processes, implying the need for implementation issues and
new insights gained in an implementation process feeding back to the strategy
forming or decision-making phases.360 In the case study, feedback of modeling
insights to the original decision-makers took place as formally stated critical
success factors, implying needed adjustments in the overall company governance
model.361
359
A parallel can be drawn to the importance of clear objectives in learning and
experimentation situations supported by simulators, see Größler, Andreas:
“Don’t Let History Repeat Itself: Methodological Issues Concerning the Use of
Simulators in Teaching and Experimentation”, System Dynamics Review,
Volume 20, Number 3, Fall 2004, pp. 268—269.
360
See discussion in chapter A on the diagnostics and decision-making cycle, and
the change management cycle, in problem-solving processes.
361
One example is the importance of head-count focus vs. cost focus in the
governance procedures, which is discussed elsewhere in this dissertation.
D. The Usage and Utility of Participative Modeling in Change Management 135
362
Argyris, Chris: Interventions Theory and Method – A behavioural Science View,
Reading, Massachusetts, 1970, p. 83.
363
This observation is based on a number of consulting projects, which the author
has carried out as consultant or engagement manager in IBM Management
Consulting and Deloitte Consulting Group between 1993 and 2002. Andersen
Consulting uses a similar approach (although it uses other terms), see
Lochmann, Hans-Dieter und Michaela Rüsch-Kornasoff: “Organization Change
Strategy – Ein wesentlicher Baustein des Reengineerings”, in Manfred Perlitz,
Andreas Offinger, Michael Reinhardt and Klaus Schug (eds.): Reengineering
zwischen Anspruch und Wirklichkeit, Wiesbaden, 1996, pp. 329—340.
364
Involvement of reference groups is often discussed in change management
literature with regards to the capture of important information and input, and in
order to ease a later implementation by avoiding negative attitudes to changes
only due to the fact that some people feel offended that they were not at all
involved in the project. In the approach “The Reference Group” by Jørgen
Randers, referees are involved primarily through interviews and workshops, see
Rouwette, Etiënne: Group model building as mutual persuasion, Nijmegen,
2003, p. 44. In group model building approaches emphasis is rather on direct
participation in modeling and simulations efforts.
365
Anderson, Linda A. and Dean Anderson: “Awake at the Wheel: Moving beyond
Change Management to Conscious Change Leadership”, OD Practitioner,
Vol. 33, No. 3, 2001, p. 45.
136 D. The Usage and Utility of Participative Modeling in Change Management
366
The structured involvement of change leaders is a way to exercise what Klein
calls ‘sympathetic understanding’ regarding the change resistance, seeking
valuable input about the nature of the system that is going to change, potentially
motivating a modification of the change itself or the change implementation
process, see Klein, Donald: “Some Notes on the Dynamics of Resistance to
Change: The Defender Role”, in Warren G. Bennis, Kenneth D. Benne and
Robert Chin: The Planning of Change, 4 th edition (first published in 1966), New
York, 1985, p. 103.
367
The five elements are taken from Ajzen’s framework for theories of planned
behavior; see discussion in chapter B.II.1. See also Schein, Edgar H.: Process
Consultation, Boston, 2000, part I, p. 68, emphasizing the importance that the
handover from the decision-makers to the implementers should be carefully
planned to avoid communication breakdown.
368
As seen in Business Reengineering projects, where individuals are taken out of
their previous environment, to use their skills and experiences in the creation of
reengineered business processes Bungard, Walter: “Zur Implementierungs-
problematik bei Business-Reengineering Projekten”, in Manfred Perlitz, Andreas
Offinger, Michael Reinhardt and Klaus Schug (eds.): Reengineering zwischen
Anspruch und Wirklichkeit, Wiesbaden, 1996, pp. 264—265.
D. The Usage and Utility of Participative Modeling in Change Management 137
369
Although it should be noted, that even though the change leaders are going to
implement an already outlined strategic initiative, the implementation in itself
also involves creative problem-solving. Projects involving system dynamics will
be likely to regard discontinuous change, requiring the change leaders to
navigate in a new landscape, where new and innovative solutions are required.
370
It should be noted that having change leaders also being informal change leaders
implies a broader definition of the term change leader compared to the use in
Anderson, Linda A. and Dean Anderson: “Awake at the Wheel: Moving beyond
Change Management to Conscious Change Leadership”, OD Practitioner,
Vol. 33, No. 3, 2001, pp. 40—48, where focus is mostly on top executives.
371
Schein, Edgar H.: Organisationspsykology, Danish translation, Herning, 1990,
p. 257.
138 D. The Usage and Utility of Participative Modeling in Change Management
372
Jöns, Ingela: Managementstrategien und Organisationswandel, Mannheim
University, 1995, p. 157.
373
Lewin, Kurt: “Group Decision and Social Change (first published in Newcomb
and Hartley’s Readings in social psychology, 1948, pp. 330—341), in Martin
Gold: The Complete Social Scientist – A Kurt Lewin Reader, Washington,
American Psychological Association, 1999, pp. 276—279.
374
Lewin, Kurt: “Group Decision and Social Change”, in Martin Gold: The
Complete Social Scientist – A Kurt Lewin Reader, Washington, 1999, p. 273.
375
Lewin, Kurt: “Group Decision and Social Change”, in Gold, Martin: The
Complete Social Scientist – A Kurt Lewin Reader, 1999, p. 281.
376
For a discussion on the importance of social norms and perceived behavioral
control in the changing of behavior, see Ajzen, Icek: Attitudes, Personality and
Behavior, Chicago, 1988, pp. 121—133.
377
Gladwell, Malcolm: The Tipping Point, paperback edition, New York, 2002,
p. 9.
D. The Usage and Utility of Participative Modeling in Change Management 139
epidemic will spread widely as a result of the behavior of the people transmitting
the virus, the characteristics of the virus, and environmental factors.378 Using this
analogy in a change management context gives rise to conscious reflections
regarding: who to involve, how to make the message and behavior ‘stick’ and
how to create settings and conditions that support the change process. In terms of
the use and selection of change leaders, Gladwell in particular has some
interesting thoughts regarding the first aspect: who to involve. He has formulated
the ‘Law of the Few,’ arguing that the effective spread of messages and behavior
often depends on relatively few individuals in a social system, namely individuals
with special skills regarding relations with others, regarding information
accumulation or regarding persuasion.379 The people with special skills regarding
relations with other people are called ‘Connectors.’ Compared to other
individuals, they know and interact with significantly more people. They are the
social glue of an organization, and they spread messages. The people who
accumulate information, the ‘Mavens’ are the one’s whose opinion most people
will take very seriously. The last group, the ‘Salesmen,’ consist of those who are
able to persuade unconvinced colleagues. Gladwell argues, that all three types of
personalities are critical to involve if a change process is to take advantages of
the ‘word-of-mouth’ phenomena.
Gladwell’s thoughts on creating change as initiating epidemics has some
similarities with Hamel’s 8-step process on how to “revolt” the business by
creating a movement within the organization, although Hamel’s process is a
bottom-up approach to be used by innovative employees rather than a top-down
change approach. The 8-step process includes “infecting” others with the idea,
creating coalitions, and working closely together with representatives from across
the organization.380
The structured development of change leaders should also be viewed in
terms of organizational learning. An organizational intervention cannot be seen
independently of the ongoing organizational development. The single
intervention is highly dependent on the flexibility and change readiness of the
individuals involved and the existing system structures. The intervention, on the
378
When an epidemic starts to spread widely, it has passed, what Gladwell calls
‘the tipping point, see Gladwell, Malcolm: The Tipping Point, paperback
edition, New York, 2002, p. 18.
379
Gladwell, Malcolm: The Tipping Point, New York, 2002, p. 19 and pp. 30—88.
380
Hamel, Gary: Leading the Revolution, Boston, 2000, pp. 187—206.
140 D. The Usage and Utility of Participative Modeling in Change Management
other hand, is also a part of the continuous learning and shaping of the
organization, its change readiness, and its change capabilities, why it seems
relevant to have the intervention planning including organizational development
considerations as well as considerations of each change leader’s individual
personal transition.381 The usage of modeling can help in the establishment of
understanding and commitment among change leaders regarding a change
process. But other elements are important for the effectiveness of change leaders,
mainly regarding general leadership skills such as interpersonal relations and
communication skills, e.g. presenting ideas in a way that activates feelings and
makes the message memorable.382
381
See Schein, Edgar H.: Organisations Psychology, Herning, Forlaget systime,
1990, p. 40; Argyris, Chris: Interventions Theory and Method – A Behavioural
Science View, Reading, Massachusetts, 1970, chapter 1 and 2.
382
McKee, Robert: “Storytelling That Moves People”, Harvard Business Review,
June 2003, p. 52.
383
Cummings, Thomas G. and Christopher G. Worley: Organizational Development
and Change, Ohio, 2001, p. 28. This framework fits also well with generic
consulting methodologies; e.g. a phase model of reengineering projects
described by Perlitz: (1) initializing, incl. project initialization and project
understanding, (2) problem analysis and redesign, (3) implementation, and (4)
anchoring and continuous development, see Perlitz, Manfred, Jürgen Bufka,
Andreas Offinger, Michael Reinhardt, und Klaus Schug: “Reengineering-
Projekte erfolgreich umsetzen – Ergebnisse einer Erfolgsfaktorenstudie”, in
Perlitz, Manfred, Andreas Offinger, Michael Reinhardt and Klaus Schug (eds.):
Reengineering zwischen Anspruch und Wirklichkeit, Wiesbaden, 1996, p. 186.
D. The Usage and Utility of Participative Modeling in Change Management 141
384
For further discussion on the unfreezing-movement-freezing process, see chapter
A.II as well as Schein, Edgar H.: Organisationspsykologi, Herning, 1990,
pp. 254—255.
385
See Kotter, John P: Leading Change, Boston, 1996, pp. 35—49; de Geus, Arie
P.: “Planning as Learning”, Harvard Business Review, Vol. 66, No. 2, March-
April 1988, pp. 70—74; and Dörner, Dietrich: The Logic of Failure, New York,
1996, pp. 49—54.
386
Chin, Robert and Kenneth D. Benne: “General Strategies for Effecting Changes
in Human Systems”, in Bennis, Warren G., Kenneth D. Benne and Robert Chin:
The Planning of Change, 4 th edition, New York, 1985, p. 22.
387
Doppler, Klaus and Christoph Lauterburg: Change Management – Den
Unternehmenswandel gestalten, 10 th edition, New York, 2002, p. 324.
142 D. The Usage and Utility of Participative Modeling in Change Management
thinking from the Board.”388 For the people expected to lead the change as well as
for the employees as such, the project owners found that all the three main causes
for change resistance listed by Doppler and Lauterburg were relevant when
designing the change process: addressing whether the reasons for change were
understood, addressing whether the reasons were believed, and addressing
whether people were willing and able to conform with the change.
Verbal Nonverbal
(Talk) (Behavior)
In the early years of the 21st century, normative management literature has
increased attention devoted to the affective side of planning and implementing
388
- The change imperative was stated as: “right now is the right time to hire people
in the low-cost locations, because due to company growth, it can be done right
now without staff reduction at high-cost locations, and the expected results are
improved competitiveness and further company growth, also securing jobs at
high-cost locations in the future.”
389
Own translation and modification from Doppler, Klaus and Christoph
Lauterburg: Change Management – Den Unternehmenswandel gestalten, 10th
edition, New York, 2002, p. 326.
D. The Usage and Utility of Participative Modeling in Change Management 143
change. Kotter and Cohen have introduced the term “see-feel-change” arguing
that, “people less change what they do because they are given analysis that shifts
their thinking than because they are shown a truth that influences their
feelings.”390 Roberto and Levesque use similar expressions, arguing for the “show
me” rather than the “tell me” approach.391 To make people “see” includes
communicating the visions and the need for change using emotionally engaging
approaches, such as storytelling, symbolic actions, a video of an angry customer
or an off-site event.392 The usage of participative system dynamics modeling fits
well with the affective oriented change approaches. The theories of system
dynamics are based upon not only a cognitive, but also an affective and ‘seeing is
believing’ learning approach, supporting the learning process by stimulating
experimentation and simulating experience. The importance of the latter is
expressed by Brown as follows: “It’s never enough to just tell people about some
new insight. Rather, you have to get them experience it in a way that evokes its
power and possibility”.393
A modeling process fosters involvement and participation, which is widely
recognized as some of the most effective strategies for overcoming change
resistance.394 Although dealing with change in large organizations, there is the
problem of scalability. For large-scale organizational change it seems rather
unrealistic to have a significant number of the implicated employees participate
390
See Kotter, John P. and Dan S. Cohen: The Heart of Change, Boston, 2002, p. 1.
As the book is practical rather than research oriented, there is no discussion of
the “see-feel-change” in terms of cognitive, affective and conative elements.
391
Roberto, Michael and Lynne Levesque: “The Art of Making Change Stick”,
MIT Sloan Management Review, Summer 2005, Vol. 46, No. 4, Summer 2005,
p. 56.
392
See Kotter and Cohen: The Heart of Change, 2002, p. 181; Roberto, Michael
and Lynne Levesque: “The Art of Making Change Stick”, MIT Sloan
Management Review, Summer 2005, Vol. 46, No. 4, Summer 2005, pp. 56—57.
Feelings facilitating change include faith, trust, optimism, urgency, reality based
pride, passion, excitement, hope, and enthusiasm, whereas feelings like anger,
false pride, pessimism, arrogance, cynicism, panic, exhaustion, insecurity, and
anxiety undermine change, see Kotter and Cohen: The Heart of Change, 2002,
p. 180.
393
Brown, John Seely: Research That Reinvents the Corporation, Harvard Business
Review, August 2002, p. 108.
394
Cummings, Thomas G. and Christopher G. Worley: Organizational Development
and Change, Ohio, 2001, p. 158.
144 D. The Usage and Utility of Participative Modeling in Change Management
in modeling efforts.395 In the case study, participative modeling did not support
the general learning process of the organization, but targeted the key managers
responsible for the implementation, by aiming to developing these individuals
into change leaders. The modeling efforts addressed the initial parts of the
change management process – or what could be called bridging the launch of the
strategic initiative and the implementation efforts. Roberto and Levesque have
proposed a change framework arguing that certain parts of the change processes
must be planned well before actually starting the main intervention; hence the
name Four Antecedent Processes (figure D-2).
CHARTERING
The process by which the organization defines the initiative’s purpose, its scope and the way
people will work with one another on the program.
• Boundary Setting – Definition of scope of initiative
• Team Design – Definition of roles, responsibilities, norms and ground rules for teamwork
How managers develop, test and refine ideas through experimentation before full-scale
LEARNING
rollout..
• Discovery – Data and information gathering to define goals of initiative and means of
achieving objectives
• Experimentation – Testing and refinement of initiative prior to full-scale rollout.
The use of symbolism, metaphors and compelling stories to engage hearts as well as
MOBILIZING
395
Although some of the same underlying mechanisms can be applied in larger
scale in the form of gaming environments, simulators, etc.
396
The figure is taken from Michael and Lynne Levesque: “The Art of Making
Change Stick”, MIT Sloan Management Review, Summer 2005, Vol. 46, No. 4,
D. The Usage and Utility of Participative Modeling in Change Management 145
Summer 2005, p. 56 and p. 60. The four processes are stated to be critical to
successful change; although the need for clear objectives, sound project
management, accountability and control systems are also stressed (p. 55).
397
An example of this kind of planning is described in the case study description of
the intervention design, see description chapter C.II.2.
398
The importance of leaders presenting ideas in a way that activate feelings and
make the message memorable is also discussed in McKee, Robert: “Storytelling
That Moves People”, Harvard Business Review, June 2003, p. 52.
146 D. The Usage and Utility of Participative Modeling in Change Management
seems fair to assume that at least one of the facilitators in the modeling projects
must be able to represent the viewpoints from the top executives. Furthermore,
traditional project steering mechanisms such as frequent check-points with a
steering committee would seem to be a likely approach in securing alignment
between modeling learning and change objectives. Although such aligning is a
balancing act: on the one hand to get unbiased input from and discussions within
the group, on the other hand to control the scope and orientation of the modeling
process.
For change projects dealing with more complex or cross-organizational
problems involving highly specialized managers, the cognitive capabilities of the
individuals must be expected to yield special challenges. Kieser and Koch
discuss the problem of shared knowledge when the individuals have difficulties
overlooking the possible combinations of knowledge, and they present case study
findings suggesting the relevance of thinking in terms of re-combining knowledge
rather than sharing knowledge.399 For organizational learning in organizations
with high specialization, Kieser and Koch also call for a knowledge integration
mechanism that does not rely on cross-learning, and discuss the benefits of
simulated prototyping, simulated experience, joint thought experiments et cetera,
and in their conclusions they emphasize the importance of teams and
communities in creative knowledge creation.400 It could be an area of further
research to investigate the possible role of system dynamics modeling in such a
context; maybe in a combination of a top executive “meta-modeling” process in
iterations with change leaders participating in modeling processes involving one
or more sub-models. For complex problems, the involvement of change leaders is
important in the search for sustainable change, but it must also be expected to be
crucial to involve the employees representing the critical specialized
knowledge.401
399
Kieser, Alfred and Ulrich Koch: Organizational Learning through Rule
Adaptation: From the Behavioral Theory to Transactive Organizational
Learning, Mannheim, 2000, p. 12 and pp. 15—16.
400
Kieser and Koch: Organizational Learning through Rule Adaptation, 2000,
pp. 18—27.
401
See Forrester, Jay W.: “Policies, decisions and information sources for
modeling”, European Journal of Operational Research, Vol. 59, No. 1, 1992,
pp. 42—63. See also Kieser and Koch: Organizational Learning through Rule
Adaptation, Mannheim, 2000, p. 19, discussing the transactive memory concept
D. The Usage and Utility of Participative Modeling in Change Management 147
In the modeling discussions and the change process around them, these
three questions are important when explicitly or implicitly addressing the change
resistance for each participant. The more open and honest the workshop
atmosphere is, the more directly the questions can be addressed. In general, the
atmosphere of the modeling sessions is important. Gladwell describes a number
of scientific research projects conveying that a positive atmosphere (e.g. a smile
from a charismatic person or physical head-nodding among the receivers)
influences the receiver’s likelihood to agreeing with a message.405 The
organizational situation and interpersonal factors influence the magnitude of the
challenge for the facilitator to establish a constructive atmosphere. If the change
is due to present and serious problems, for example, managers are likely to
exhibit behavior including hiding information, secrecy and denial, blame,
avoidance and/or passivity and helplessness.406 In such an atmosphere, the
404
Own translation from Doppler, Klaus and Christoph Lauterburg: Change
Management – Den Unternehmenswandelgestalten, 10th edition, New York,
2002, pp. 326—327.
405
Gladwell, Malcolm: The Tipping Point, paperback edition, New York, Back Bay
Books, 2002, pp. 74—87.
406
Kanter, Rosabeth Moss: “Leadership and the Psychology of Turnarounds”,
Harvard Business Review, June 2003, p. 61.
D. The Usage and Utility of Participative Modeling in Change Management 149
facilitator’s job of establishing trust in the process is challenging, but also crucial
for team members to be willing to contribute and cooperate.407
Although a good atmosphere improves the learning potential, an ‘extre-
mely’ good atmosphere also constitutes facilitation challenges. In groups
characterized by group cohesiveness and consensus, groupthink can occur with
lack of critical thinking, resulting in limiting discussions to few scenarios without
due consideration of alternative scenarios or possible alternative gains.408
Furthermore, groupthink often leads to not attaining sufficient information (even
from experts within the organization), ignoring facts not supporting the favorite
scenarios, dismissing feedback or new information that should lead to changing
earlier group decisions, as well as underestimating implementation challenges.409
In a change management context, the preliminary model and intervention
objectives and targets frame the modeling project, but still, within the framework,
such a project also involves innovative problem-solving and decision-making
regarding the implementation strategy and plan. O’Connor and McDermott stress
the importance of widening perspectives by stimulating creativity, having a
variety of different viewpoints represented, and getting different sorts of
feedback.410 The facilitator furthermore has the option to use different curiosity
stimulating mechanisms, as curiosity makes people question mental models and
generate new perspectives411 With regards to creativity and curiosity, creative
problem-solving literature often emphasizes the importance of “asking stupid
407
Henttonen, Kaisa and Kirsimarja Blomqvist: “Managing distance in a global
virtual team: the evolution of trust through technology-mediated relational
communication”, Strategic Change, Vol. 14, March-April 2005, p. 108. Here,
trust is discussed in respect to ‘normal’ teams in ‘normal’ projects, but
Henttonen and Kirsimarja later also discus evolution of trust in virtual teams
(which was actually one of the issues discussed among project participant the
case study outlining how to implement the location strategy).
408
Poor decision-making as result of groupthink is described in Janis, Irving L.:
“Groupthink: The Problems of Conformity” (original printed in Psychology
Today, Nov. 1971, pp. 271—279), in Morgan, Gareth: Creative Organization
Theory, Newbury Park, California, 1989, pp. 225—227.
409
Janis: “Groupthink: The Problems of Conformity”, in Morgan, Gareth: Creative
Organization Theory, Newbury Park, California, 1989, p. 227.
410
O’Connor, Joseph and Ian McDermott: The Art of System Thinking – Essential
skills for creativity and problem solving, London, 1997, p. 141.
411
O’Connor and McDermott: The Art of System Thinking – Essential skills for
creativity and problem solving, London, 1997, p. 141.
150 D. The Usage and Utility of Participative Modeling in Change Management
412
Leonard, Dorothy: Wellsprings of Knowledge – Building and Sustaining the
Sources of Innovation, Boston, 1998, pp. 74—75.
413
See discussion in chapter B.III.2.d.
414
Warren, Kim: Competitive Strategy Dynamics, Chichester, 2002, p. 264.
415
This has some similarities with how astrologers use horoscopes as a virtual
reality, where humans re-arrange their perception of their own life, see Munk,
Kirstine: “In the Airy Spaces of Our Minds…: Cosmology and ritual design in
modern, Western astrology”, in York, Michael (ed.): Nature, Religion, and
Culture, London, in print, 2006, p. 27. Munk refers to the Danish astrologer and
psychotherapist Pia Balk-Møller, who suggests that the use of a chart
(a horoscope) makes a discussion ‘safer’, as the secrets are already on the table,
and Munk furthermore stresses the importance of imaginative involvement in
discussions of charts, where insight “not only has to be understood
intellectually, but also has to be imagined and felt” (pp. 29—30).
D. The Usage and Utility of Participative Modeling in Change Management 151
teams. Distefano and Maznevski present research, showing that diverse teams
often perform either significant worse (more often) or significant better (less
often) compared to homogeneous teams.416 Managed well, diverse teams have
potential to outperform homogeneous teams, and modeling efforts might be a
relevant approach in such a setting due to possibilities of lowering personal
barriers as well as due to providing structure to otherwise diverse discussions.
The creation of an open and honest modeling atmosphere, which at the
same time has some degree of tension and investigative interest, has much to do
with creating a platform for learning and directing discussions, not with regards
to efficient model completion, but with regards to allowing exchange of
viewpoints and aligning mental models. In scenario simulations, focus should be
on understanding why the model produces the given behavior, rather than using
the model as “an evaluator” of different scenarios. The insights from discussion
of the model behavior can lead to new insights. For real life evaluations, Farson
and Keyes discuss the importance of moving beyond both success and failure,
analyzing the underlying reasons with less focus on evaluation and more focus on
interpretation.417 Being rewarded or complimented for success can actually be as
de-motivating as criticism, whereas most humans are motivated by getting a
deeper insight into their problems. In model understanding, not taking model
results at face value, but using them to generate more discussion on model
behavior encourages the long process of testing and gaining trust in the model,
and avoiding the risk of false trust in the model. It is important that there are
enough group activities to enable the necessary discussions and that each
individual has been sufficient involved in the model building to gain both a sense
of “ownership” and to gain trust in the model. Participant’s trust in the model is a
prerequisite for establishing learning, commitment, and alignment of mental
models. This is important when striving for sustainable change built upon a new
shared mental model.418
Effective facilitation depends not only on fostering discussions, but also
on making sure that the process relies on available facts. In the case study, the
416
Distefano, Joseph J. and Martha L. Maznevski: “Creating Value with Diverse
Teams in Global Management”, Organizational Dynamics, Vol. 29, No. 1, p. 46.
417
Quoted in Farson, Richard and Ralph Keyes: “The Failure-Tolerant Leader”,
Harvard Business Review, August 2002, pp. 66—67.
418
For a discussion of the importance of shared mental models in organizational
learning, see chapter B.II.3.
152 D. The Usage and Utility of Participative Modeling in Change Management
initial setting of each of the parameters involved issues that could have been
discussed for hours in both workshops and other related meetings. Opinions on
parameters were often very different within the core project team. As an example,
it was a widely accepted ‘fact’ among many of the project participants that
employees in low-cost countries very often stayed only 1–2 years, because as
soon as they attained experience in R&D, they could get a better-paid job in a
high-cost country. But with the ‘forced quantification process’ with parameter
stipulations, facts came on the table, as the data were actually available,
documenting a very low employee turnover in the low-cost countries.419 The
parameter stipulations furthermore served to set benchmarks between business
units, and this approach enabled cross-business-unit knowledge and experience
exchange. For example, one business unit already had high-scale experience with
building up resources in low-cost countries, reflecting very low hiring costs due
to procedures that the other business units decided to adopt. The modeling
approach this way served as a forum for transfer of best practices, which was one
of the objectives of the modeling project in the first place.
In the case study, the quantification of the model and the simulation of
scenarios helped to provide an understanding of which parameters most strongly
influenced the effectiveness of the strategy. Based on this understanding, the
facilitator directed the discussion on how the strategy could and should be
executed, as the parameter setting reflected implementation decisions. For
example, the discussion was focused on reducing the time spent on hand-over of
tasks, resulting in discussions on structuring knowledge transfer both during and
after the initial transfer period. The discussions also focused on how to reduce
both time and costs of classroom training. Without modeling, the intuitive choice
could have been to focus primarily on the costs, including the travel costs, but
due to the model-guided new understanding of the influence of the different
parameters, increased focus was placed on reducing the time spent on classroom
training. For system dynamics practitioners, the model seems very simple, but it
is interesting to notice that the project team first tried to handle the problem with
the use of a normal Excel-spreadsheet, which became a complicated “black box”,
419
Christensen, Søren and Jan Molin: Organisationskulturer, Copenhagen, 1987,
p. 27, discuss myths as inward, empty explanations used to legitimize certain
behavior. See also Ackoff’s morale: “There is nothing so deceptive as an
apparent truth“: Ackoff, Russel L.: The Art of Problem Solving – Accompanied
by Ackoff’s Fables, New York, 1978, p. 84.
D. The Usage and Utility of Participative Modeling in Change Management 153
420
Briggs, Robert, Gert-Jan de Vreede, and Jay F. Nunamaker Jr.: “Collaboration
Engineering with ThinkLets to Pursue Sustained Success with Group Systems”,
Journal of Management Information Systems, Vol. 19, No. 4, Spring 2003,
p. 32.
421
Briggs, de Vreede, and Nunamaker Jr.: “Collaboration Engineering with
ThinkLets to Pursue Sustained Success with Group Systems”, Journal of
Management Information Systems, Vol. 19, No. 4, Spring 2003, p. 45.
422
Briggs, de Vreede, and Nunamaker Jr.: “Collaboration Engineering with
ThinkLets to Pursue Sustained Success with Group Systems”, Journal of
Management Information Systems, Vol. 19, No. 4, Spring 2003, p. 46.
423
See table C3 in chapter C.III.2.
154 D. The Usage and Utility of Participative Modeling in Change Management
424
Salge, Markus and Peter Milling: “Who is to blame, the operator or the
designer? Two stages of human failure in the Chernobyl accident”, System
Dynamics Review, Vol. 22, in print, 2006, the figures 8 and 9.
425
See discussion in chapter B.III.2.d.
D. The Usage and Utility of Participative Modeling in Change Management 155
426
Leavitt, Harold J.: Top Down – Why Hierarchies Are Here to Stay and How to
Manage Them More Effectively, Boston, 2005, p. 98.
427
The discussion in chapter A.I on the two cycles of problem-solving, Diagnostics
and Decision-Making (cycle I) and Change Management (cycle II), is based on
the problem solving process described in Schein, Edgar H.: Process
Consultation, Boston, 2000, part I, p. 61.
428
Eisenhardt, Kathleen M.: “Strategy as Strategic Decision Making”, Sloan
Management Review, Spring 1999, pp. 66—67. Eisenhardt calls the mental
model alignment “building collective intuition.”
156 D. The Usage and Utility of Participative Modeling in Change Management
cycle. A modeling process can utilize a system dynamics model in bridging the
two cycles of the problem-solving process, transferring the main insights
motivating the change and illustrating the intended outcome of the change
project. Furthermore, by allowing for model refinement, parameter stipulation,
and scenario testing with a high degree of empowerment among the modeling
participants, the modeling process serves as an instrument ensuring the
implementers true involvement and influence in the change process. Involvement
and influence are often discussed as essential mechanisms in lowering change
resistance and ensuring the commitment among key employees necessary for
creating sustainable change.
A modeling process will impact the transfer of insights and ownership at
individual, group, and organizational level. On the individual level, learning of
causal relationships will influence cognitive structures, and if designed well, the
process in a group setting could be expected to also influence affective and
conative elements as well as social norms and perceived behavioral control for
the project participants.429 Consequently, based on the framework of Ajzen, one
could expect behavioral changes.430 In the case study, the participants’
questionnaire answers scored relatively high on the usefulness of including the
model in the project, its usefulness in facilitating discussions, the usefulness of
starting the project with a preliminary model, individual learning in general, and
the building of a shared view.431 From the project owners’ perspective one of the
main objectives of the modeling process was establishing consensus about the
change imperative, as significant change obstacles existed of both cognitive and
affective characteristics.432 A year and a half after the modeling process, it was
429
This conclusion is based on the discussions throughout the dissertation,
supported by both the organizational psychology literature and the normative,
prescriptive change management literature.
430
Ajzen, Icek: Attitudes, Personality and Behavior, Chicago, 1988, p. 133. See
also discussion in chapter B.II.1.
431
At the same time, the lowest score was given to expectations of actual
implementation of the recommendations. However, in follow-up discussions this
gave the impression of being due to pessimistic expectations to the required shift
from headcount-orientation to cost-orientation, see further description in
chapter C.
432
The short version of the change imperative being that in times of company
growth it is possible to build up capacity in low-cost locations without reducing
high-cost locations - and this is a necessary strategy for the continued
D. The Usage and Utility of Participative Modeling in Change Management 157
observed, that the recommendations derived from the modeling process were
implemented at an even higher speed than outlined in the implementation plan.
Although many influential factors exist, the project owners give significant credit
to the modeling project’s influence on the change leaders commitment to the
change. At group level, the modeling process allowed for discussion and
exchange of experiences, and the high-level aggregated model secured alignment
of mental models and codes among key implementers across the organization. At
organizational level, the change affected significantly more people than were
involved in the modeling project. However, indirect influence takes place
through the development of true change leaders.
In the system dynamics tradition, modeling projects should not only aim at
creating insights of relevance in a given intervention, but also strive for the
general improvement of system thinking skills among the modeling project
434
Snabe Birgitte: Brugtvognskoncept for Skandinavisk Motor Co., Thesis, Det
erhvervsøkonomiske diplomstudium, Copenhagen Business School,
Copenhagen, 1994, p. 9 and p. 89, discusses scenario evaluation and selection
among alternative scenarios in terms of strategic fit, profitability, risk,
competitor reactions and reversibility.
435
A discussion of the importance and the challenges of cross-organizational
alignment of robust policies can be found in Lyneis, James M.: Corporate
Planning and Policy Design: A System Dynamics Approach, Massachusetts,
1980, pp. 6—9.
D. The Usage and Utility of Participative Modeling in Change Management 159
436
For processes, where modeling has been used designing alternative policies and
structures, Forrester discusses the need for educating and debating with people
who will be involved the implementation. The challenge often includes changing
deeply embedded policies and emotional beliefs, where “it is not that people
disagree with the goals, but rather how to achieve them.“ See Forrester, Jay W.:
“System dynamics, system thinking, and soft OR”, System Dynamics Review,
Vol. 10, No. 2, 1994, p. 247.
437
The two cycles of the problem-solving process are inspired from the problem-
solving process in Schein, Edgar H.: Process Consultation, Boston, 2000, part I,
p. 61.
162 E. Targeted Participative Modeling in Change Management
possible to know whether a different approach would have been more effective or
efficient, as the case study was a single-site study without any test group.
The usage of system dynamics modeling for change management purposes
could be called a ‘targeted participative modeling’ process, addressing how to
implement a strategic initiative with already established objectives and targets.
The managers responsible for an implementation process often played only a
small part, if any, in the decision-making process from which the strategic
initiative originated. Targeted participative modeling looks especially promising
with regards to:
(1) The usage of modeling and simulation as a tool to transfer insights and
ownership from decision-makers to implementers.438 In the pursuit of
sustainable change, iterations with top executives might be relevant,
with the dual objective of adjusting the strategic initiative according to
implementation issues and giving the managers responsible for
implementation true influence on the entire change process. 439
(2) The refining and aligning of cross-organizational implementation plans
through scenario simulation. Although simulations should not serve as
‘an evaluator’ of scenarios, they can aid in the investigation of expected
system behavior for alternative scenarios and, as such, support decision-
making at the operational level critical for successful implementation.
The positive indications of both the effectiveness and efficiency yielding
from the case study could give rise to expectations to potential significant usage
of targeted participative modeling, as larger organization often launch strategic
initiatives involving the need to bridge the decision-making process with the
implementation process. A question arising from the case study is whether a
modeling approach as used in the case study would also work if the problem at
hand had been significantly more complex. For some complex problems, a
business simulator might prove more suitable compared to a modeling process in
438
With regards to the transfer of insights, a targeted participative modeling process
has certain similarities with the usage of management simulators.
439
Communication skills, interpersonal skills and other personal leadership skills
influence effectiveness in the change process, together with company factors
including change readiness, change capabilities and culture, see Cummings,
Thomas G. and Christopher G. Worley: Organizational Development and
Change, Ohio, 2001, p. 144.
164 E. Targeted Participative Modeling in Change Management
440
Some of the same challenges are seen in the design and evaluation of business
simulators, see Größler, Andreas: Entwicklungsprozess und Evaluation von
Unternehmenssimulation für lernende Unternehmen, Frankfurt am Main, 2000,
p. 178.
441
See Van der Smagt, Ton: “Causation and Constitution in System Dynamics:
Modelling a Socially Constituted World”, Journal of Systems Research and
Behavioral Science, Vol. 23, No. 4, in print, 2006, pp. 13—14, as well as the
discussion in chapter D.I.
Appendices
HC TRAINING COST Basic new-hire training costs per person per month
(EUR/(Person*Month))
LC TRAINING TIME Training time for the basic training of newly hired
(Month) employees (classroom training)
LC new hire =
+ Productive FTE LC*ADDITIONAL GROWTH
+ LC replacing HC quit
Comment:
LC replacing HC quit = HC quit – HC hire
Main equations for production per month and cost per month
Starting below is a complete list of the model equations from the case study
model. According to the agreement with the case study company, the most
confidential numbers (such as average employee costs etc.) have been made
unrecognizable, through having ‘NN’ replacing the first digits. A number of the
parameters do not relate directly to the model, but serve the purpose of producing
nice output graphs.
Formulas Dimension
ADDITIONAL GROWTH=
(0.03*PULSE(0, 61)
+RAMP(-0.03/36, 12, 48)) Dmnl
LC new hire=
+(Productive FTE LC*ADDITIONAL GROWTH)
+LC replacing HC quit
+"HC micro-site optim."
+"LC micro- site optim." Person/Month
"HC job-trained"=
DELAY N(HC trained, HC ROOKIE TIME,
Rookie FTE HC/HC ROOKIE TIME, 3) Person/Month
LC 2a=
DELAY N(
("LC repl. hire"),
LC TRAINING TIME,
(New hired FTE LC/LC TRAINING TIME)
, 12) Person/Month
LC 1a=
DELAY N(
LC new hire,
LC TRAINING TIME,
0, 12) Person/Month
LC 2b=
DELAY N(LC 2a, LC TOTAL ROOKIE TIME,
Rookie FTE LC/LC TOTAL ROOKIE TIME, 3) Person/Month
LC 1b=
DELAY N(LC 1a, LC TOTAL ROOKIE TIME*
"HAND-OVER FRAKTION OF LC ROOKIE TIME", 0, 12) Person/Month
LC 1c=
DELAY N(LC 1b, LC TOTAL ROOKIE TIME*
(1-"HAND-OVER FRAKTION OF LC ROOKIE TIME"), 0, 3) Person/Month
LC QUIT FRACTION=
0.07/12 1/Month
176 Appendix B
adding=
production per month*1/12 Person/Month
summing=
cost per month EUR/Month
LC replacing HC quit=
HC quit-HC hire Person/Month
LC quit=
Productive FTE LC*LC QUIT FRACTION Person/Month
LC TRAINING COST=
NN00 EUR/(Person*Month)
LC PRODUCTIVITY REDUCTION=
0.2 +0*0.2*PULSE(0, 48) Dmnl
178 Appendix B
total FTE=
total FTE HC+total FTE LC Person
HC quit=
Productive FTE HC*HC QUIT FRACTION Person/Month
HC hire=
HC quit*"REPLACEMENT IN HC VS. LC" Person/Month
"MICRO-SITE OPTIMIZATION"=
0*(NN*PULSE(1, 1)+NN*PULSE(12, 1)
+NN*PULSE(24, 1)) Person/Month
HC FRACTION OF MICROSITES=
0.5 Dmnl
STY month=
1 Month
HC trained=
DELAY N(HC hire, HC TRAINING TIME,
New hired FTE HC/HC TRAINING TIME, 12) Person/Month
LC ROOKIE PRODUCTIVITY=
0.5 Dmnl
HC TRAINING COS
NN00 EUR/(Person*Month)
LC TRAINING TIME=
2 Month
HC ROOKIE PRODUCTIVITY=
0.5 Dmnl
HC TRAINING TIME=
3 Month
HC ROOKIE TIME=
6 Month
HC QUIT FRACTION=
0.0NN/12 1/Month
For both aging chains, the three stocks to be initialized are: Productive FTE, New
hire FTE, and Rookie FTE.
The model input parameters used to calculate the initial values are:
INI total FTE, QUIT FRACTION, TRAINING TIME, and ROOKIE TIME.
gives when New hire FTE + Rookie FTE from equation (2) is entered into equation (1):
<=> INI total FTE = Productive FTE + Productive FTE * QUIT FRACTION *(TRAINING TIME+ROOKIE TIME)
<=> Productive FTE = INI total FTE / (1+(TRAINING TIME+ROOKIE TIME)* QUIT FRACTION
and also, when the result of Productive FTE from equation (2) is entered to equation (1):
INI total FTE = ((New hire FTE+ Rookie FTE) / QUIT FRACTION * (TRAINING TIME+ROOKIE TIME))+New hire FTE + Rookie
FTE <=> INI total FTE = (New hire FTE + Rookie FTE) * (1+(1/ QUIT FRACTION * (TRAINING TIME+ROOKIE TIME) )
<=> (New hire FTE + Rookie FTE) = INI total FTE / (1+(1/ QUIT FRACTION * (TRAINING TIME+ROOKIE TIME) )
And the initial values of each of New hire FTE and Rookie FTE are then found by the equation (3):
New hire FTE = LC TRAINING TIME/(LC TRAINING TIME+LC ROOKIE TIME)**INI TOTAL FTE LC/(1+(1/((LC TRAINING
TIME+LC ROOKIE TIME)*LC QUIT FRACTION)))
1.15
1
1
1.1 1
1
3 2
1 3 2 4
1.05 3 3 3 4
1 2 2 2 2
32 4 4 4
4 4
41
1
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60
Months
INI 1 1 1 1 1 Dmnl
Base run 2 2 2 2 2 Dmnl
40% HC replacement 3 3 3 3 Dmnl
Faster training and hand-over 4 4 4 4 Dmnl
LC flow input
OF LC ROOKIE TIME
LC HAND-OVER TRAVEL COST
handover
<H-O-R FTE LC A> travel costs
Changes in behavior:
“In general the team members developed business cases Interview
in compliance with the modeling insights and results,
with only one exception”
192 Appendix F
Group communication:
The discussion seemed to be both very structured and Observation
very open and frank. The result-oriented process,
however, did not leave time to go into depth in all of the
relevant discussions, but due to the structure, most of the
time invested by the participants in discussions was used
very effectively.
Consensus:
Opinions on the importance of different causal- Observation
relationships differed initially, but through the model-
building process a more shared understanding of the
problem and its dynamics was created.
Common language:
Within the core project team, there was a tendency to Observation
increased alignment, but this was difficult to transfer to
non-core members in the relatively short meetings with
these people. Parts of the language did spread to some
extent, such as “one employee is one employee”
regardless of the type of location. The factor for reduced
productivity in low-cost countries only reflects a lower
average experience-level. A stronger outcome on this
dimension would have required a less result-oriented
process, with more time to in-depth discussion.
Transfer of insights:
The model clearly confirmed some viewpoints that the Observation
project team wanted to communicate to the board and the
corporate controlling. Whereas these insights did not
have much “newness” value, it was very valuable to have
a model that distinctly and clearly “proved” the matter.
These type of insight included the worse-before-better
effect, implying that the division even receiving a
relatively large number of additional head-counts in year
1, would have no additional productivity, but rather a
slightly reduced productivity. Also, the model showed
very clearly that even the relatively large growth in the
fraction of low-cost employees compared to the total
number of employees, does not result in a decreased cost
per produced development hour, as the inflation has
stronger influence than the benefits to be realized through
a location strategy of the discussed scope.
194 Appendix F
“the model made the strategy very transparent, with clear Interview
definitions – and was better than words for
communication”
System changes:
The board accepted the business cases developed in the Observation
project, and the hirings for 2005 went approx. as
planned. Also, the business case was implemented in the
3 years business plan, and the execution should follow
the plans.
The business unit that disagreed most with the project Observation
objectives has actually been “overperforming” year-end
2005 with regards to hirings in low-cost locations.
Results:
1½ years after the approval of the detailed Observation
implementation plans, the FTE ratio between low-cost (Q2 2006)
and high-cost locations has increased beyond the plan.
This must also be seen in relation to the fact, that due to
the success of the company, more hiring were needed and
approved than originally planned.
Efficiency:
The project kept established deadlines. Some disturbance Observation
and discussion took place due to the fact, that the
intervention also encompassed many elements not
included in the modeling.
Quality in results:
For SD practitioners, the model seems very simple, but it Observation
should be recognized that the project team first tried to
handle the problem using a normal Excel-model, which
became a complicated “black box”, where it was difficult
to see and understand how the different parameters
influenced the model.
Project framing:
The project owners had no intentions of starting a group Observation
model building process from clean sheets of paper with
the risk of losing control. This might be a general trend in
the corporate environment; that executives have a clear
view of the direction they want to drive a given change,
and that they will not take the risk that a model could
show contradicting results, which in their view could be
Appendix F 197
“We were open about the premises for the process, and Interview
participants should therefore not feel in the slightest way
manipulated” (This was the answer to a question, if the
use of a preliminary model and fixed business objectives
could have caused the participants to feel somewhat
manipulated)
Ajzen, Icek: Attitudes, Personality and Behavior, Chicago, (The Dorsey Press) 1988.
Ajzen, Icek and Martin Fishbein: “The Prediction of Behavior from Attitudinal and
Normative Variables”, Journal of Experimental Social Psychology, No. 6, 1970,
pp. 466—487.
Andersen, David F., George P. Richardson and Jac A. M. Vennix: “Group model
building: adding more science to the craft”, System Dynamics Review, Vol. 13,
No. 2, 1997, pp. 187—201.
Andersen, David F and George P. Richardson: “Scripts for group model building”,
System Dynamics Review, Vol. 13, No. 2, 1997, pp. 107—129.
Anderson, Linda A. and Dean Anderson : “Awake at the Wheel: Moving beyond
Change Management to Conscious Change Leadership”, OD Practitioner,
Vol. 33, No. 3, 2001, pp. 40—48.
Bennis, Warren G., Kenneth D. Benne and Robert Chin : The Planning of Change,
4 th edition, New York, (CBS College Publishing) 1985.
Bonabeau, Eric: “Don’t Trust Your Guts”, Harvard Business Review, May 2003,
pp. 116—123.
Brown, John Seely: Research That Reinvents the Corporation, Harvard Business
Review, August 2002, pp. 105—114.
Chin, Robert and Kenneth D. Benne: “General Strategies for Effecting Changes in
Human Systems”, in Warren G. Bennis, Kenneth D. Benne and Robert Chin: The
Planning of Change, 4 th edition, New York, (CBS College Publishing) 1985,
pp. 22—45.
Conant, Roger C. and W. Ross Ashby: “Every Good Regulator of a System Must Be
a Model of that System”, International Journal of System Sciences, Vol. 1,
No. 2, 1970, pp. 89—97.
Denscombe, Martyn : The Good Research Guide for small-scale social research
projects, 2 nd edition, Philadelphia, (Open University Press) 2003.
de Geus, Arie P.: “Planning as Learning”, Harvard Business Review, Vol. 66,
No. 2, March-April 1988, pp. 70—74.
de Geus, Arie P.: The Living Company, Boston, (Longview Publishing Limited)
1997.
Distefano, Joseph J. and Martha L. Maznevski : “Creating Value with Diverse Teams
in Global Management”, Organizational Dynamics, Vol. 29, No. 1,
pp. 45—63.
Doyle, James K. and David N. Ford : “Mental models concepts for system dynamics
research”, System Dynamics Review, Vol. 14, No. 1, Spring 1998,
pp. 3—29.
Doyle, James K. and David N. Ford : “Mental models concepts revisited: some
clarifications and a reply to Lane”, System Dynamics Review, Vol. 15, No. 4,
Winter 1999, pp. 411—415.
Dörner, Dietrich : The Logic of Failure, New York, (Metropolitan Books) 1996.
Eden, Colin : “Using cognitive mapping for strategic options development and
analysis”, in Jonathan Rosenhead (ed.): Rational Analysis for a Problematic
World, Chichester, (John Wiley & Sons) 1989, pp. 21—42.
Eden, Colin : “Cognitive mapping and problem structuring for system dynamics
model building”, System Dynamics Review, Vol. 10, Nos. 2-3, Summer-Fall
1994, pp. 257—276.
Eisenhardt, Kathleen M.: “Building Theories from Case Study Research”, Academy
of Management Review, Vol. 14, No. 4, 1989, pp. 532—550.
Eskinasi, Martijn and Eppie Fokkema: “Bursting the myth of making easy modeling
money”, unpublished working paper (permission to quote granted from the
authors), Proceedings, Second European System Dynamics Workshop,
Nijmegen, 2005, pp. 62—66.
Farson, Richard and Ralph Keyes: “The Failure-Tolerant Leader”, Harvard Business
Review, August 2002, pp. 64—71.
Ferres, Natalie, Julia Connell, and Anthony Travaglione: “The effect of future
redeployment on organizational trust”, Strategic Change, Vol. 14, March-April
2005, pp. 77—91.
Fey, Willard and John Trimble: “The Evaluation and Development of Knowledge
Acquisition in System Dynamics Studies”, in Proceedings, System Dynamics
Conference, System Dynamics Society, 1992, pp. 173—182.
Flood, Robert L. and Michael C. Jackson : Creative Problem Solving: Total Systems
Intervention, Chichester, (John Wiley & Sons) 1991.
Flood, Robert L.: Solving Problem Solving, Chichester, (John Wiley & Sons) 1995.
Forrester, Jay W. and Peter Senge: “Tests for Building Confidence in System
Dynamics Models”, in Augusto A. Legasto, Jay W. Forrester und James M.
Lyneis (eds.), TIMS Studies in the Management Sciences, Vol. 14, Amsterdam,
1980, pp. 209—228.
Forrester, Jay W.: “Policies, decisions and information sources for modeling”,
European Journal of Operational Research, Vol. 59, No. 1, 1992, pp. 42—63.
206 Bibliography
Forrester, Jay W.: “System dynamics, system thinking, and soft OR”, System
Dynamics Review, Vol. 10, No. 2, 1994, pp. 245—256.
Gladwell, Malcolm: The Tipping Point, paperback edition, New York, (Back Bay
Books) 2002.
Gold, Martin : The Complete Social Scientist: A Kurt Lewin Reader, Washington,
(American Psychological Association) 1999.
Hamel, Gary: Leading the Revolution, Boston, (Harvard Business School Press)
2000
Huff, Anne S. and Rhonda Kay Reger: “A Review of Strategic Process Research”,
Journal of Management, Vol. 13, No. 2, 1987, pp. 211—236.
Imai, Masaaki : Kaizen: Der Schlüssel zum Erfolg der Japaner im Wettbewerb,
german translation, München, (Wirtschaftsverlag Langen Müller Herbig) 1992.
Kieser, Alfred and Ulrich Koch : Organizational Learning through Rule Adaptation –
From the Behavioral Theory to Transactive Organizational Learning, Mannheim,
2000.
Kim, Daniel H. and Peter M. Senge: “Putting systems thinking into practice”,
System Dynamics Review, Vol. 10, Nos. 2-4, Summer-Fall 1997, pp. 277—290.
Kotter, John P : Leading Change, Boston, (Harvard Business School Press) 1996.
Kotter, John P. and Dan S. Cohen : The Heart of Change, Boston, (Harvard Business
School Press) 2002.
Leavitt, Harold J.: Top Down – Why Hierarchies Are Here to Stay and How to
Manage Them More Effectively, Boston, (Harvard Business School Press) 2005.
Leavitt, Harold J., William R. Dill, and Henry B Eyring: The Organizational World
– A systematic view of managers and management, New York, (Harcourt Brace
Jovanovich, Inc.) 1973.
Lewin, Kurt: “Group Decision and Social Change” (first published in Newcomb and
Hartley’s Readings in social psychology, 1948, pp. 330—341), in Martin Gold:
The Complete Social Scientist – A Kurt Lewin Reader, Washington, (American
Psychological Association) 1999, pp. 265—291.
Luna-Reyes, Luis Felipe and Deborah Lines Andersen : “Collecting and analysing
qualitative data for system dynamics: methods and models”, System Dynamics
Review, Vol. 19, No. 4, 2003, pp. 271—296.
Lyneis, James M.: Corporate Planning and Policy Design – A System Dynamics
Approach, Boston, (Pugh-Roberts Associates, Inc.) 1980.
210 Bibliography
Maier, Frank und Andreas Größler : ”What are we talking about? A taxonomy of
Computer Simulations to Support Learning”, System Dynamics Review, Vol. 16,
No. 2, 2000, pp. 135—148.
Martinez, Ignacio J and George P. Richardson: “An Expert View on the System
Dynamics Modeling Process: Concurrences and Divergences Searching for Best
Practices in System Dynamics Modeling”, at CD-ROM of Proceedings, System
Dynamics Conference, System Dynamics Society, 2002.
Martinez, Ignacio J. and Luis F Luna : “The Dynamics of Best Practices: A Struc-
tural Approach”, at CD-ROM of Proceedings, System Dynamics Conference,
System Dynamics Society, 2001.
McKee, Robert: “Storytelling That Moves People”, Harvard Business Review, June
2003, pp. 51—55.
Miller, George A.: “The Magical Number Seven, Plus Minus Two: Some Limits on
Our Capacity for Processing Information”, The Psychological Review, Vol. 63,
No. 2, March 1956, pp. 81—97.
Mintzberg, Henry: The Rise and Fall of Strategic Planning, New York, (The Free
Press) 1994.
Munk, Kirstine: “In the Airy Spaces of Our Minds…: Cosmology and ritual design
in modern, Western astrology”, in Michael York (ed.): Nature, Religion, and
Culture, in print, London, (Equinox) 2006.
Newell, Allen and Herbert A. Simon : Human Problem Solving, Englewood Cliffs,
New Jersey, (Prentice-Hall, Inc.) 1972.
O’Connor, Joseph and Ian McDermott : The Art of System Thinking – Essential
skills for creativity and problem solving, London, (Thorsons) 1997.
212 Bibliography
Palmer, Ian and Richard Dunford : “Who says change can be managed? Positions,
perspectives and problematics”, Strategic Change, Volume 11, August 2002,
pp. 243—251.
Pearson, Alan : “You Drive for the Show but you Putt for the Dough”, appendix in
Stafford Beer: Beyond Dispute: The Invention of Team Syntegrity, Chichester,
(John Wiley & Sons) 1994, pp. 311—322.
Perlitz, Manfred, Jürgen Bufka, Andreas Offinger, Michael Reinhardt, und Klaus
Schug : “Reengineering-Projekte erfolgreich umsetzen: Ergebnisse einer
Erfolgsfaktorenstudie”, in Manfred Perlitz, Andreas Offinger, Michael Reinhardt
and Klaus Schug (eds.), Reengineering zwischen Anspruch und Wirklichkeit,
Wiesbaden, (Gabler Verlag) 1996, pp. 181—208.
Repenning, Nelson P. and John D. Sterman : Nobody Ever Gets Credit for Fixing
Problems that Never Happened: Creating and Sustaining Process Improvement”,
California Management Review, Vol.43, No.4, Summer 2001, pp. 64—88.
Richmond, Barry: “The Strategic Forum: aligning objectives, strategy and process”,
System Dynamics Review, Vol. 13, No. 2, 1997, pp. 131—148.
Bibliography 213
Roberto, Michael and Lynne Levesque: “The Art of Making Change Stick”, MIT
Sloan Management Review, Summer 2005, Vol. 46, No. 4, Summer 2005,
pp. 53—60.
Rockart, John F .: “Chief executives define their own data needs”, Harvard Business
Review, March-April 1979, pp. 81—93.
Salge, Markus and Peter Milling : “Who is to blame, the operator or the designer?
Two stages of human failure in the Chernobyl accident”, System Dynamics
Review, Vol. 22, in print, 2006
Schein, Edgar H.: Process Consultation – MAOM Capstone Course for the
University of Phoenix, Boston, (Person Custom Publishing) 2000.
Schneiderman, Arthur M.: “Setting Quality Goals: Use observed rates of continuous
improvement to position targets”, Quality Progress, April 1988, pp. 51—57.
Scherer, Andreas G.: “Kritik der Organisation oder Organisation der Kritik?
Wissenschaftstheoretische Bemerkungen zum Umgang mit Organisations-
theorien”, in Alfred Kieser (ed.): Organisationstheorien, 3 rd edition, Stuttgart,
(Kohlhammer) 1999, pp. 1—37.
214 Bibliography
Senge, Peter M.: The Fifth Discipline, paperback edition, New York, (Currency
Doubleday) 1994.
Senge, Peter M. : The Dance of Change, New York, (Currency Doubleday) 1999.
Simon, Herbert A.: The Science of the Artificial, 3 rd edition, Cambridge, (The MIT
Press) 1996.
Snabe, Birgitte and Andreas Größler : “System Dynamics Modelling for Strategy
Implementation: Case Study and Issues”, Journal of Systems Research and
Behavioral Science, Vol. 23, No. 4, 2006, pp. 467—481.
Snabe, Birgitte, Andreas Größler, and Peter M. Milling : “Policies and Politics of
Establishing R&D Capacity in Low-Cost Locations”, Tagungsband, GWS-
Tagung 2005, Greifswald, in print, 2006.
Stake, Robert E.: The Art of Case Study Research, Thousand Oaks, (SAGE
Publications) 1995.
Sterman, John D.: Business Dynamics – Systems Thinking and Modeling for a
Complex World, Boston, (McGraw Hill Higher Education) 2000.
Taylor, Frederick W.: The Principles of Scientific Management, New York, first
time published in 1911, (Norton & Company) 1967.
Thun, Jörn-Henrik, Peter M. Milling, and Uwe Schwellbach : ”The Impact of Total
Employee Involvement on Time-based Manufacturing“, in Kate Blackmon, Steve
Brown, Paul Cousins, Andrew Graves, Christine Harland, Richard Lamming,
and Harvey Maylor (eds.): ”What Really Matters in Operations Management“,
Bath, 2001, pp. 131—143.
Van der Smagt, Ton : “Causation and Constitution in System Dynamics; Modelling a
Socially Constituted World”, Journal of Systems Research and Behavioral
Science, Vol. 23, No. 4, in print, 2006.
216 Bibliography
Vennix, Jac A. M. : Group Model Building, Chichester, (John Wiley & Sons) 1996.
Voss, Chris, Nikos Tsikriktsis, and Mark Frohlich : “Case research in operations
management”, International Journal of Operations & Production Management,
Vol. 22, No. 2, 2002, pp. 195—219.
Warren, Kim: Competitive Strategy Dynamics, Chichester, (John Wiley & Sons)
2002.
Weil, Henry B.: “The Evolution of an Approach for Achieving Implemented Results
from System Dynamics Projects”, in Jørgen Randers (ed.): Elements of the
System Dynamics Method, Cambridge, Connecticut, (Productivity Press) 1980,
pp. 269—291.
Weil, Henry B. and Kenneth P. Veit: “Corporate Strategic Thinking: The role of
System Dynamics” in Peter M. Milling and Erich O.K. Zahn (Eds.): Computer-
Based Management of Complex Systems, Proceedings of the 1989 International
Conference on the System Dynamics Society, Stuttgart, (Springer-Verlag) 1989,
pp. 66—75.
Winch, Graham and Sonja Derrick: “Flexible Study Processes in ‘Knotty’ System
Dynamics Projects”, Journal of Systems Research and Behavioral Science,
Vol. 23, No. 4, in print, 2006.
Bibliography 217
Yin, Robert K.: Case study Research, 3 rd edition, Thousand Oaks, (SAGE
Publications) 2003.