A Design Science Approach - Joan Van Aken PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 45

A Design Science Approach to Evidence-Based Management

Joan Ernst van Aken, Eindhoven University of Technology


Georges L. Romme, Eindhoven University of Technology
Abstract
In this chapter we adopt a Design Science approach to demonstrate how
Evidence-Based Management can extend experience-based management by
judiciously gathering, validating, selecting and using knowledge. The creative
processes in which valid knowledge is used in designing and creating preferred
futures are discussed, as well as the nature of the design knowledge needed for
this. This includes knowledge on possible interventions with expected outcomes
and on the mechanisms producing these outcomes.

Keywords: Design science


Collaboration between scientists and practitioners
Design propositions
Spin-offs
Preferred future
Professional practice
Technical rationality

140

Design is () the principal mark that distinguishes


the professions from the sciences. Schools of engineering,
as well as schools of architecture, business, education, law,
and medicine, are all centrally concerned with the process of design.
Simon (1996, p. 111)
This chapter builds on the idea proposed by Nobel laureate Herbert Simon that
organization and management research is a science of design. In keeping this idea,
Evidence-Based Management (EBMgt) is a design process individuals and teams
use to solve real-life organizational problems through the use of scientific
evidence and validated local facts.
EBMgt itself is a family of approaches to the practice of management, all
of which use relevant scientific evidence (Briner et al., 2009). In this chapter,
EBMgt is positioned in the context of the design science paradigm in management
and organization science (e.g. Bate, 2007; Boland & Collopy, 2004a; Jelinek,
Romme & Boland, 2008; Romme, 2003; Van Aken, 2004). As such, we argue
EBMgt is much more than rational decision-making; it is about changing the
actual into the preferred, with research-informed designing as the core activity.
This perspective draws on the conceptualization of EBMgt as research-informed,
organizational problem solving (Tranfield, Denyer & Smart, 2003) and a design
science approach (Romme, 2003; van Aken, 2004).
A Design Science approach to EBMgt has the following three key
features:

Design Science Research produces a critical part of the evidence to


be used in EBMgt.

141

This evidence is not to be used as a set of instructions or fixed


protocols, but as input to the creative and innovative process of
designing structures, processes or interventions.

Research-informed designing is the core activity in a complex


process of changing the actual into the preferred: EBMgt involves
acting, rather than (only) decision-making, on the basis of
evidence.

By positioning EBMgt in the context of Management as a design-oriented


discipline, this chapter draws on two seminal books, namely Herbert Simons
(1969, 1996) The Sciences of the Artificial and Donald Schns (1983) The
Reflective Practitioner, both building on the work of James (1907), Dewey (1929)
and Pierce (1955) on pragmatism. Schn analyzed the more science-based
professions like Medicine and Engineering, as well as the (somewhat) less
science-based professions like architecture, management and psychotherapy.
Following Schn, professional work in all these disciplines entails much more
than the model of pure technical rationality implies. In the latter model,
professional practice largely boils down to instrumental problem solving on the
basis of explicit knowledge, made rigorous by applying scientific theory and
technique. The classic example of pure technical rationality is the engineer who
uses a formula to calculate the maximum load for a buildings construction. By
contrast, professionals use rich repertoires of explicit and tacit knowledge in a
creative process of reflection-in-action. These repertoires contain tacit knowledge,

142

developed through personal experiential learning, as well as explicit knowledge,


derived from their academic discipline.
This chapter thus develops the following key ideas:

The core process in EBMgt is research-informed designing of


structures, processes or interventions, which entails a creative
process of developing situation-specific solutions to design issues,
combined with a more analytic process of evaluating these
solutions.

This process of designing on paper is an essential part of the


very complex and challenging process of changing the current
(organizational) world into a preferred one.

The design process draws on multiple forms of knowledge,


including tacit as well as explicit knowledge, through a creative
process of reflection-in-action.

Design science research produces a critical part of this explicit


knowledge.

Personal and interpersonal experiential learning are central to


developing both tacit and explicit knowledge for use in designing.

We first describe the design science perspective and then discuss a key
component in the design process: the role of reflection-in-action. The chapter then
describes the general process of changing the actual into the preferred, and
subsequently compares and contrasts designed-based changes in the material and
social worlds. We then explore the nature of generic evidence-for-design, and

143

contrast it with the nature of evidence in the social world, setting the stage for a
discussion of EBMgt from a design science perspective. The chapter subsequently
illustrates the EBMgt approach to design in a study of creating university
spinoffs. Finally, we give recommendations for both practitioners and
organizational researchers in advancing EBMgt as a design process.
The Design Science Perspective
Designing a future is fundamentally different from describing and explaining the
present. Simons (1969; 1996) The Sciences of the Artificial has played an
important role in recognizing the centrality of design in applied disciplines, by
demonstrating the fundamental differences between describing and explaining
what is and designing and evaluating what can be. Scholars often emphasize
describing and explaining the present as the core research mission of any
academic discipline. By contrast, designing what can be is a core activity in
architecture and the arts, and also in research-based engineering. Yet, an
increasing number of disciplines now resonate with Simons recognition that the
human world is artificial and that design activity is essential in making this world.
The primary mission of academic disciplines like Medicine and
Engineering is the creation of preferred futures -- for example, restoring health in
case of pneumonia or creating a more fuel-efficient car. This mission is realized
through knowledge produced by research as well as through the actions of the
professionals trained in these disciplines. As design plays a central role in the
creation of preferred futures, one may call disciplines for which design is central
design sciences, as opposed to explanatory sciences. The mission of an

144

explanatory science, like Physics and Sociology, is to describe and explain the
world as it is. Explanatory research is largely driven by a quest for knowledge as
an end in itself. The iconic research product here is the causal model. A causal
model is assessed in terms of descriptive validity, that is, how well it accounts for
the observed world.
Mainstream research in a design science is driven by field or real-world
problems. Field problems are situations in reality that, according to influential
stakeholders, can or should be improved. Knowledge development here is
instrumental in solving field problems; students are thus trained to become
professionals, using the knowledge of their discipline to explore, design and
create preferred futures. Whereas explanatory research typically uses an
independent observer perspective, Design Science Research draws on a
participant-observer perspective that requires knowledge on interventionoutcome-context combinations to be able to make judgments on the (likely)
outcomes of interventions in specific cases.
With disciplines like Medicine and Engineering as inspiration one can
define Design Science Research (DSR) as a family of approaches to research that
are driven by field problems, use a participant-observer instead of the independent
observer perspective, and pursue a solution orientation. This implies design
science researchers are not satisfied with describing field problems and analyzing
their causes, but also develop alternative general solution concepts for these field
problems. In Medicine and Engineering the participant-observer perspective does
not pose specific problems because most researchers in these disciplines are, or

145

have been, practicing professionals. In Management research this is typically not


the case. Therefore DSR in this discipline often draws on research strategies like
collaborative or interventionist research (e.g. Romme and Endenburg, 2006).
The iconic research product of DSR is a well-tested solution concept, i.e. a
generic intervention to solve a generic field problem, tested in the laboratory and
in the field of its intended use. In Medicine, for example, solution concepts for a
certain dysfunction (the field problem) can include types of drugs or surgery. In
Engineering, solution concepts to design a particular electronic equipment can
include types of electrical circuits or programmable logic for their operation. In
Management, for example, solution concepts for creating cooperative
arrangements can include types of contracts and procedures for assessing the fit
between the prospective parties. A solution concept is assessed in terms of its
real-world practical use, that is, pragmatic validity (Worren, N., Moore, K. &
Elliott, R., 2002). For example, solution concepts for organization design
problems are: hierarchy as an unambiguous sequence of accountability levels
and circularity for organizing the flow of power and information (Romme &
Endenburg, 2006).
The logic of applying the solution concept is the design proposition (also
called technological rule). The design proposition puts the solution concept into
its application context. It runs like: if you want to achieve a given outcome for
this generic problem-in-context, then use this generic intervention. For example:
if you want to achieve a successful entry in a rather inaccessible foreign market
by way of a cooperative arrangement with a local company, then use this

146

particular type of contract. The most powerful design proposition is the fieldtested and grounded one: the intervention is tested in its intended field of
application and is grounded in an understanding of the generic mechanisms -- that
is the cause-effect relations -- that produce the outcome (van Aken, 2004).
Therefore, in the example of the proposition regarding the type of contract (given
above), one would like to have it tested through actual applications and grounded
in a theoretical explanation of why this type of contract is superior in this
particular context. The logic of the field-tested and grounded solution concept is
called the CIMO-logic (Denyer, D., Tranfield, D. & Van Aken, J.E., 2008), a
combination of problem-in-Context, Intervention, generative Mechanisms
producing the outcome, and Outcome.
A discipline is called a Design Science if DSR-approaches are firmly
positioned in its mainstream. This is the case in fields like Medicine, Architecture
and Engineering. But the importance of design has also been acknowledged in
other fields, for example, IT (March & Smith, 1995; Hevner, A.R., March, S.T. &
Park, J., 2004), accounting (Kasanen, E, Lukka, K. & Siitonen, A., 1993; Labro &
Tuomela, 2003) and education (Kelly, 2003; Collins, A., Joseph, D. & Bielaczyc,
K., 2004). Moreover, there is a growing interest in developing knowledge-fordesign among management researchers. For example, inspired by architect Frank
Gehrys approach to design, Boland & Collopy (2004a) brought together a group
of people from very different backgrounds to explore the potential of design for
management studies. The design science perspective to organization and
management studies was initially developed by, among others, Romme (2003)

147

and Van Aken (2004) and later also in special issues of the Journal of Applied
Behaviorial Science (Bate, 2007) and Organization Studies (Jelinek, Romme &
Boland, 2008).
Like EBMgt, Design Science Research can thus be regarded as a family of
approaches: driven by field problems, using a participant-observer perspective,
and pursuing a solution orientation.
Research-Based Designing and Reflection-in-Action
The core activity of designing consists of synthesis-evaluation iterations. A
possible solution to the design problem is created or synthesized, typically by
contextualizing an appropriate solution concept. The designed solution is then
evaluated on paper to determine how well it solves the problem. Evaluation is
the basis for choosing among alternative solutions to a design issue; evaluation
on paper means judging the extent to which a design meets its desired
specifications, before that design is actually implemented. Evaluation may also
take place by engaging in small experiments or pilot studies to try out one or more
alternative designs. If the result is not satisfactory, the solution is re-designed and
evaluated again, an iterative process that continues until a satisfactory design is
obtained.
Engineers, psychotherapists, managers and other design professionals
draw on their creativity, skills and repertoires of tacit and explicit knowledge.
They engage with a specific situation and treat their case as a unique one. At the
same time, in a creative process of reflection-in-action they draw on an extensive
repertoire of, among others, examples, models, theories and solution concepts in

148

order to make sense of their case and to design alternative solutions (Schn,
1983). In this respect, the designer draws on a repertoire as a kind of grab bag
of knowledge, sometimes consciously but more often largely unconsciously, to
synthesize and evaluate alternative designs.
The repertoires of junior professionals are largely the result of their initial
formal training. These initial repertoires consist of internalized formal theory and
tacit clinical experience, obtained through personal experiential learning.
Subsequently, professionals continually enrich their repertoires with further
personal experiential learning, usually combined with efforts to keep their formal
disciplinary knowledge up-to-date.
The use of explicit disciplinary knowledge demands creativity and
considerable expertise: general knowledge has to be translated towards the
specific context in question. DSR is not really intended for use by the layperson,
but rather by experienced and well-educated professionals. That is, design science
results are best used by professionals, having mastered the body-of-knowledge of
their discipline, having the ability to locate and obtain (new) knowledge that is
relevant in their work setting, the ability to contextualize explicit knowledge, and
the ability to develop intimate knowledge on the case under consideration and its
context. For example, physicians adapt their use of patient care protocols to
individual circumstances and needs, and professional experience and judgment
are key elements in designing context-specific solutions.
The degree to which explicit knowledge can support the design and
realization of preferred realities varies by profession. In Engineering, this role is

149

so inherent and self-evident that evidence-based engineering is a non-issue.


Engineers do not want to reinvent the wheel. In Medicine, the role of explicit
knowledge is also important, but compared to Engineering more professional
judgment is needed to contextualize generic knowledge in designing and choosing
interventions. In some cases this need for contextualization is fairly low, allowing
for a crisp definition of best practices. In other instances, this need is so high that
it is impossible to define unambiguous best practices, and the role of the
practitioners judgment thus becomes more pervasive (Groopman, 2010). Both
Engineering and Medicine (largely) operate in the material world, although laws
in physics are more general and less contingent than those in biology (Mitchell,
2000). This allows Engineering to evaluate designs more unambiguously than
Medicine (e.g. evaluating the maximum load of alternative constructions in
designing a building versus evaluating alternative interventions in case of a brain
tumor).
Next to the generic knowledge of the professional repertoire, the core
process of designing requires case-specific inputs: the formulation of the design
problem, the specifications the design has to satisfy, (likely) root causes of the
problem, and an analysis of the problem context. But designs cannot be derived
from these inputs in a number of logical steps. Designing typically entails a
creative leap from inputs to design; this maybe a small leap, but more often it is a
large one (cf. logic of abduction; Peirce, 1955). For example, new product
development teams draw extensively on past knowledge, but only as a jumping
board for creating radically new products. Designing is thus creatively exploring

150

possible futures. It deploys accumulated knowledge and experience, without


becoming locked into the past.
Changing the Actual into the Preferred in the Material World
The core process of designing, as previously described, is a key part of the overall
process of changing the actual into the desired. This section explores this process
in the material world, using engineering design as an exemplar. As such, this
serves to identify key elements of design and change processes, whereas this
discussion will later also be used to understand how design processes in the social
world differ from those in the material world.
Prior to starting the core process of designing, designers engage with
principals and users to formulate the design problem and its specifications. These
specifications can change during the design process, as the understanding of the
problem and its setting deepens; moreover, time and cost constraints may make it
impossible to meet initial specifications (which may then be adjusted downward).
Alternatively, developments in technology or competitive conditions may lead to
specifications being adjusted upward. If the design problem and specifications can
be well defined, the development of specifications can be readily separated from
the core process of designing. If design problem and specifications cannot be
defined unambiguously, both problem formulation and design specification are
strongly intertwined with designing; designing then involves exploring alternative
futures in close collaboration with various stakeholders. For instance, many
people find it rather difficult to unambiguously specify the home furnishing they

151

want, so interior decorators have to design interiors in close interaction with their
principals.
An intense dialogue may be needed with those that will produce and
implement the design, to produce designs that are easy or inexpensive to realize.
In a wide range of routine settings (e.g. designing and producing shoes), designfor-manufacturing can be readily accomplished using explicit knowledge about
manufacturing requirements. Otherwise, a focused dialogue between designers
and producers is necessary.
Changing the actual into the preferred involves creating an action net
(Lindberg & Czarniawska, 2006), a network of actors -- individuals, groups,
organizations -- working together to create something new. This action net can be
emergent in nature, occurring in an ad hoc, as-needed fashion. However, if an
action net does not emerge spontaneously, design professionals need to
deliberately develop it.
A design can be defined as a model to be realized. In engineering design,
evaluating alternative designs on paper is typically done with help of
mathematics (e.g. analytically or through simulation modeling). As discussed
below, using mathematical models for evaluation purposes is enabled by the fact
that key mechanisms in the material world can be described in terms of universal
and invariant laws. However, even in the material world it is not always possible
to develop models of future entities that can be analyzed mathematically. If the
entity being designed is too complex to be adequately modeled in mathematical
terms, case-based reasoning is useful (e.g. Leake, 1996; Watson, 1997). In case-

152

based reasoning, the design is evaluated by comparing it with similar, previously


realized designs -- just as lawyers assess their cases on the basis of case law. If the
entity involves such a high level of complexity that even case-based reasoning
cannot be applied, small-scale pilot experiments with the alternative designs
might be used to test and evaluate them.
Evidence-for-Design in the Social World
A design science approach is a pragmatic one. It is not about developing true
propositions about reality. Rather, it develops propositions that inform people
about how to create preferred realities. This does not imply that design science
approaches merely aim at instrumentalistic propositions (Archer, 1995: 153),
informing agents about interventions producing preferred realities without
informing them about why these interventions would work. On the contrary, the
answer to the why question is essential for effective and thoughtful usage of a
design proposition. As discussed, in the implementation stage a design
proposition always has to be adapted to the specific context of application; in
designing the adapted version, one has to know why the intervention works in
order not to lose its power in the adaptation process. Nevertheless, the pragmatism
of design science deeply influences the kind of evidence used for designing in the
social world: evidence is needed on how, when and why the particular
intervention works.
Design Science Research produces knowledge on solution concepts for
solving field problems. The key scientific claim with respect to a design
proposition is that it predicts the outcomes of the use of the solution concept in

153

question (in the given context). In applied research in the material world, this
prediction is enabled by universal and invariant mechanisms governing the
behavior of matter -- even if these are contingent on this world (Mitchell, 2000).
An electron does not have the freedom today to behave differently from
yesterday, nor can it act differently in Amsterdam than in New York. A machine
tested in Barcelona will perform similarly in Buenos Aires -- assuming that the
human operator cannot significantly influence the machines performance.
Because of these universal and invariant mechanisms, responses identified in
past/current tests can reliably be used to predict responses in the future.
Human agency creates the social world and is subject to interactions with
the behavior of others. People do have the freedom to act on the basis of
anticipated effects of their actions. Thus, no universal and invariant mechanisms
exist in the social world. However, human behavior has patterns and regularities
as a result of human nature and nurture. These patterns and regularities can be
uncovered and then used in predicting -- within certain ranges -- the outcomes of
interventions in the social world. In fact, predicting human behavior is an almost
universal human competence (almost universal: autistic people lack this
competence, showing how important and rather generic it is). Starting from the
day we are born, we learn over the years what we can do to get what we want
from others: experiential learning (see Kolb, 1984).
The repertoires of professionals (both in the material and in the social
world) contain the explicit knowledge of their discipline and the results of their
personal experiential learning in real world settings. We argue that for applied

154

disciplines in the social world, given the absence of universal and invariant laws,
knowledge production should be largely based on the strategy of experiential
learning. However, knowledge production in this case should result from
objectified experiential learning, based among other things on controlled
observations (following protocols and using methods like triangulation),
comparative case analyses, and validation of designed interventions by alpha-and
beta-testing (van Aken, 2004) in the intended field of application. Objectified
experiential learning in academic research can, for instance, be done through the
comparative case study, resulting in a rich understanding of context, intervention
and outcome. But other approaches to objectified experiential learning can also be
of value.
According to Bhaskar (1998) it is the nature of the object that determines
the form of its possible science. In engineering design, the available science is
such that one often can evaluate alternative designs by means of calculations of
their expected performance. But in social system design, the nature of the
designed system implies that the evaluation of alternative designs proceeds
through experience-based judgment of expected performance. To support this
judgment, objectified experiential learning is a powerful knowledge production
strategy.
Problems arising in organizations are not limited to purely social
dimensions. They can also refer to socio-technical systems, as in operations
management. The smaller the social component in such a system, the stronger are
the universal and invariant mechanisms governing system behavior, and thus the

155

stronger the potential for adequate mathematical modeling of the system (of an
existing as well as a newly designed one). Designing socio-technical systems with
a small social component is more similar to engineering design than to social
system design, making math-based evaluation tools more useful. For instance, the
design and evaluation of a scheduling or inventory control system can be largely
based on mathematical modeling and analysis if the human operator can be
expected to quite closely follow the instructions from the information system.
However, if the operator must make frequent, significant adaptations to the
schedules calculated by the system (e.g. because the dynamic complexity of the
context leads the operator to make frequent exceptions to the built-in rules), any
evaluation effort has to draw on experiential learning as well as mathematical
modeling. Thus, if the social component of the system is large, one has to largely
rely on experience-based prediction of the performance of the designed system; if
the social component is small, one may fairly predict this performance on the
basis of calculations.
EBMgt in Action: Operating in the Swamp of Practice
Schn (1983) argued there is the high ground of theory and there is the lowly
swamp of practice. The ambition EBMgt is to provide some firm ground in the
swamp through the use of valid evidence. This will not drain the swamp: The use
of this evidence still needs a substantial level of competence, experience and
creativity.
In the present section we discuss EBMgt in the swamp of organizational
change, particularly in design-based organizational change. This discussion will

156

highlight several important aspects of EBMgt that draws on a Design Science


perspective.
EBMgt: Dealing with Major Managerial Challenges
Managers deal with most problems on the fly. However, in this chapter we
discuss ways of dealing with major managerial challenges in a more or less
organized way, allowing for the deliberate use of EBMgt. An organized way
implies giving sufficient attention to following various process steps, to relevant
stakeholders, as well as to the information needed and its quality. This can result
in a formally organized project with a principal, a project team, a project leader, a
project assignment and a project plan. Major challenges can also be engaged
within the framework of a series of regular management meetings in which the
incumbent issue is part of the normal agenda -- possibly supplemented by
assignments on aspects of these issues completed between meetings or by breakout sessions for more focused discussions.
In the following we give a generic process model of formal or informal
solution design projects (for more details: Van Aken, J. E., Van der Bij, H. &
Berends, H., 2007). These projects deal with a significant issue, like a revision of
strategy, a merger, a reorganization, or starting a large, high-risk product
development project. Typically, the project is preceded by a fuzzy front end, in
which some stakeholders start to recognize the issue and try to obtain support for
addressing it. Weick (2004) characterizes the experience of entering and engaging
a new project as being thrown into a continuously evolving and ambiguous
context. This fuzzy stage may result in an explicit acknowledgement of the issue

157

by major stakeholders and the decision to start a major effort to deal with it. If one
decides to adopt an organized process, then a project definition step is initiated.
The products of this step should include:

The problem definition

The project assignment, among other things giving the


specifications for the intended outcomes of the project

The project plan that outlines the approach to analysis, design and
realization, and provides a time line for these activities

The project team, project leader, and reporting structure

If a less organized approach is chosen, a sound problem definition is


critical and its importance cannot be overstated. The rest of the project brief may
be less detailed and formalized, but agreement is still needed on what will be
done, by whom and when. If using EBMgt, an essential part of these agreements
refers to the efforts to be spent on collecting validated local knowledge as well as
explicit scientific knowledge on the issue in question through systematic review
and research synthesis.
Another key issue is the intended realization. If it is a significant issue,
already at the project definition step one needs to develop a rough idea of who the
possible change recipients are as well as their involvement in the design and
realization processes. The entire process should lead to the development of a
strong action net (Lindberg & Czarniawska, 2006), as previously discussed, to
realize the designed solution.
The process steps following project definition generally are:

158

Problem analysis (elaborating the initial problem analysis, made at


the problem definition step), which involves a process of naming
and framing (Schn, 1983); this naming and framing will give
pointers to relevant scientific literatures; the analysis will include a
diagnosis of the causes of the problem and the background of the
issue

Context analysis, involving both the internal and external


organizational context

The core process of designing the solution and the change plan

Realization, that is, the actual changes in roles, role structures and
activities

Learning for performance by the change recipients

Interpretation of and reflection on the results, design process and


change process -- with the intent to capture the main lessons
learned, to be used next time

Typically, the project will not follow such an undisturbed process, because
of delays or accelerations in the process, resulting from the interactions with the
daily operations in the organization. Other urgent issues may demand attention, or
the issue itself suddenly may increase in urgency.
The Core Process: Designing
The basic process steps for solution design and change planning are:

Sketching, by informally and creatively exploring possible


solutions

159

Actually designing alternative solutions through a process of


synthesis-evaluation iterations, resulting in the outline design and
in a change plan to realize it

Detailing the outline design and change plan

Usually the actual design process is not a linear one. Rather, the design
process will use iterations and explorations. In an iteration one may return to a
previous step, because new insight reveals that more information is needed. There
may even be a need to change the project objectives at such an iteration. In an
exploration, one jumps to a step further in the process, for example to understand
what information is needed to design and evaluate alternative solutions.
In organizational solution design, many people may be involved. There
may be a lead designer or main sponsor (e.g. the CEO or his/her delegate), but in
all cases a group of people will work together on the issue. The composition of
this group may change over time. In the remainder of this section, we will call this
group of tightly or loosely coupled people the design group.
For major issues, this design group generally should be multidisciplinary.
The quality of the designed solutions depends on the extent to which this group
uses its members expertise effectively and efficiently. The quality of the
designed solution strongly depends on the performance of the Transactive
Memory System (TSM) of this group (Wegner, 1986; Peltokorpi, 2008). That is,
in terms of Schns repertoire notion: the quality of the design depends to a large
extent on the quality of the design groups shared repertoire.

160

With respect to the actual designing (i.e., the synthesis-evaluation


iterations), we distinguish between a decision mode and a design mode (cf.
Boland & Collopy, 2004b). In the more common decision mode the emphasis is
on the evaluation part of the iterations, fairly tangible, allowing the use of familiar
analytic scientific methods. In the design mode, on the other hand, the creative but
much less tangible synthesis part is emphasized.
Boland and Collopy give an interesting illustration of the working in
design mode (Boland & Collopy, 2004b). This illustration concerns the design of
the Lewis building for the Weatherhead Business School, by the architectural firm
of Frank Gehry. Toward the end of the design process the floor space had to be
reduced by 4,500 square feet, a really major change. At that point, the project
architect Matt Fineout took two days to discuss this design problem with a
representative of the client, going into many details and into many alternative
detail solutions, making a lot of drawings. But to the complete astonishment of
the client, at the end of the second day, as the client thought the problem had
largely been solved, the architect tore these drawings up and threw them in the
trashcan. The client had expected that the design would emerge through a process
of successive detailing, each intermediate design coming closer to the eventual
design. But the architects behavior in this case demonstrated two key principles
of designing:

Designing is playing with alternatives (dont marry the first design


idea)

161

Assign the main effort to the outline design (start the time-and
resource-consuming detailing only after the key design dilemmas
are solved)

The architect needed discussions with the client to become familiar with
the design dilemmas he faced. But he was not yet ready to resolve these dilemmas
in an outline design, being still in the phase of creatively synthesizing alternative
solutions. In organizational problem solving in the design mode, participants
therefore need to take time to play with alternatives, and put a lot of effort into
making a sound outline design. In managerial settings, too often the design group
becomes entrapped in the first feasible solution found, and then shifts all its effort
in detailing it.
As said previously, one cannot deduce the design from the inputs; there
always is a creative leap from input to design. But, of course, the inputs to the
design process are of critical importance to the designs quality. The explicit
inputs include:

The problem analysis and diagnosis

The context analysis

Solution concepts and ways to evaluate them

In the synthesis steps, the design group merges these explicit inputs with
their tacit and explicit knowledge, to subsequently synthesize alternative solutions
to design issues in a process of reflection-in-action. In the same way, tacit and
explicit knowledge is used to evaluate the alternative designs against the
specifications, the most important being that the solution should solve the

162

problem. Like in very complex situations in engineering design that are rather
difficult to model, case-based reasoning is an important alternative approach to
evaluate organizational designs on paper.
Design-Based Change in the Social World
The overall process of changing the actual into the preferred in organizations has
similarities to engineering design. Much of the process is comparable: problem
analysis, development of specifications, interactions with the various
stakeholders, building of an action net to realize the solution. Fundamental
differences also exist.
Foremost among the differences is the need to contextualize generic
knowledge. How teams, meetings, or policies are thought of and operate in one
firm differs from those in another. Thick, rich and detailed descriptions are
therefore needed for the various solution concepts, the contexts in which these
concepts have been tested, and the nature of the (intended and unintended)
outcomes. Furthermore, the evaluation of alternative designs is done differently.
Many engineering designs can be adequately modeled, so engineers can use their
impressive array of engineering mathematics to evaluate their designs on paper.
This typically is not the case in the social world. An experiential learning
approach, like case-based reasoning, therefore has to be used to evaluate designs.
For instance, to evaluate the design for an account management system in a sales
department, the design group can draw on data on the performance of
implemented account management systems in other sales departments (within or
outside the same company).

163

Finally, a fundamental difference exists in realizing the design. In


Engineering, the design determines the realized entity and thus also its behavior.
In management and organizational settings, realizing any design always involves
the ultimate redesign and adaptation of the initial design by the change recipients:
their behavior depends on how they interpret the design. Their collective
interpretation can produce social realities at odds with the design groups
intentions, and thus different outcomes. A nursing group implementing a new
self-managing team approach, developed at the level of the hospital at large, may
distrust the given protocol and still appoint a group leader. In engineering design,
this dilemma is solved in the design-for-manufacturing approach. In management
and organization, however, the design group needs to facilitate the interpretation
and enactment of their design by frequently interacting with change recipients
during the design process as well as by staying involved in the entire subsequent
process of change and learning for performance.
In this process of participation by the change recipients, EBMgt can make
another contribution by informing the design group as well as the broader
stakeholder audience of the design and change -- thus supporting interpretation,
enhancing credibility, and generating confidence in the endeavors future
outcomes.
Other Contexts of EBMgt
Above we discussed EBMgt in design-based organizational change. EBMgt can
also be of great value in issues not necessarily involving major organizational
change, like in decision-making on large investments or technology-driven

164

innovations. Moreover, in organizational change there are important alternatives


to design-based change. In some situations, one cannot design a preferred end
state, for instance because the available knowledge is insufficient or because
organizational politics make it ill-advised. In such cases one may prefer to adopt a
step by step approach, each step based on the results of the previous one. In other
cases, change objectives are difficult to realize by designing preferred end states
(e.g. in cultural change). In each of these alternative settings, however, there still
is a lot of design to do -- for example, defining the objectives and set-up of the
next step of the change process or the type of intervention to be used.
The Contribution of Evidence-Based Management
EBMgt intends to extend and build on intuitive, experience-based managerial
problem solving. It does so by:

Giving attention to the informational inputs to the problem-solving


process

Enhancing the shared repertoire (or Transactive Memory System)


of the design group

Providing significant content to the design process.

In the course of problem and context analysis, the design group gathers a
lot of case-specific information, with respect to both the organization and its
environment. EBMgt promotes a rigorous approach to collecting information,
calling attention to scientific evidence as well as local facts and experiences, and
adopting a critical attitude toward the information collected. EBMgt thus

165

enhances the quality of the inputs to the process of designing solutions to


organizational problems.
An EBMgt approach emphasizes not just the solution to a single problem,
but also enhances the organizations readiness to address issues in future. Thus, it
also aims to make full use of the explicit knowledge of the members of the design
group and at making their tacit knowledge as explicit as possible (e.g. see the case
in the next section), thus creating a shared repertoire for the design group. This is
an important issue, especially because design groups for significant organizational
issues often represent many different disciplines. Making knowledge explicit and
critically assessing its quality can significantly enhance the design groups
performance by promoting knowledge sharing and deeper understanding.
The primary contribution of EBMgt, however, is to produce valid generic
knowledge on a particular issue, based on rigorous scientific research, including
the outcomes of Design Science Research. This explicit knowledge can support
problem analysis and diagnosis, context analysis and especially design, both for
synthesis and for evaluation. Naming and framing at the problem definition step
gives pointers to the literatures to be searched. The synthesis step of the design
process proper can be supported by identifying a range of generic solution
concepts along with their indications and contra-indications. Evidence from the
literature arising from testing the solution concepts in various settings can support
the subsequent evaluation step. As in engineering design, the evaluation and
justification of designs in management and organization may draw on case-based
reasoning.

166

The process of producing this valid generic knowledge is systematic


research synthesis (Denyer & Tranfield, 2006; Denyer et al., 2008; Briner &
Denyer, this Handbook). Quantitative meta-analysis may be a powerful way to
synthesize quantitative research outcomes in the material world. In the social
world, however, a highly effective knowledge production strategy is to synthesize
research outcomes on the basis of an objectified experiential learning approach.
At each step of the literature review, the basic question is: what have we learned
until now to understand our problem and to synthesize alternative solutions, and
what evidence have we generated to enable us to judge what the outcomes of
these solutions would be in our specific context?
As discussed in the section on evidence-for-design in the social world, if
the system to be designed has a significant material component, designing is more
like engineering design. In that case the knowledge production strategy can make
more use of quantitative approaches to research synthesis.
The results of the systematic research synthesis are shared within the
design group, thus further adding to its shared repertoire. These results can also be
shared with the prospective change recipients. This serves to create an even
broader shared repertoire, which can be of great value in realizing the design -- as
will be discussed below.
EBMgt in Developing University Spin-offs
To illustrate evidence-based solution design in the context of organization and
management, we draw on a study of university spin-off creation by Van Burg,
Romme, Gilsing & Reymen (2008). This study provides an example of designing

167

and developing policies, informed by codifying and synthesizing practitioners


experiences and the academic body of knowledge inferred from an extensive
literature review.
The problem was how to create university spin-offs, that is, business startups that commercialize technologies developed at the incumbent university. Its
particular context is the common case where the focal university resists any
deviation from its core teaching and research processes. The assignment to
develop evidence-based guidelines for spin-off creation was given to a research
team by a Dutch university -- Eindhoven University of Technology (TU/e). The
spin-off creation project was already underway when this assignment was given.
Three years earlier, TU/e had started the development of a new venturing
incubation unit. In a quest to realize some early results, this was done without
giving much attention to the literature on the subject. After some time the
University Board and the incubator director became interested in how their
incubator system compared to those at other (leading) universities and in what the
scientific literature ad to say on it. In other words, the incubator was set-up
without using EBMgt, but after some time the key players became interested in it.
This way of starting EBMgt meant that the research team (in which one of the
authors of this chapter participated) could benefit from the initial practitioners
learning outcomes and experiences (i.e. reflection-in-action). On the other hand,
arriving in the middle of the project meant that the design group -- the group
actually involved in designing -- consisted not only of the research team, but also
included the university staff members already engaged in developing university

168

spin-offs. Thus the research team faced the challenge of transforming an ongoing,
largely emergent design process into a more deliberate one. A further challenge
was, as always in evidence-based solution design, the need to contextualize any
generic findings of a systematic literature review to take into account the features
specific to TU/e.
In view of these challenges, a design science-driven EBMgt approach was
adopted in which two key notions connected professional practices and research
findings (cf. Romme & Endenburg, 2006): design propositions and solution
concepts (as discussed in the section on the design science perspective). Design
propositions can be based on research, but can also be derived from successful
professional practice. The research-based propositions serve as tangible artifacts
that allow different groups of people to focus upon and create shared
understandings (Romme & Endenburg, 2006). That is, these explicit propositions
allowed participants from diverse backgrounds to focus on a common set of issues
in the design process (cf. boundary objects).
The research team developed design propositions for university spin-off
creation by separately developing propositions based on practice (practice-based
propositions) and propositions based on scholarly knowledge (research-based
propositions). The synthesis of these propositions subsequently resulted in design
propositions, which thus provides a body of knowledge that is grounded in
research as well as tested in practice (cf. Van Aken, 2004). Thus, the key steps in
developing design propositions were as follows (Van Burg et al., 2008):

169

a) So-called practice-based propositions were developed by converting the


largely tacit knowledge of key actors in university spin-off creation into explicit
propositions;
b) Propositions were derived from a review of the literature; these researchbased propositions then served to understand (and possibly improve) practices and
solutions already in place as well as create entirely new solutions;
c) Finally, the research team synthesized the practice-based and researchbased propositions in a set of design propositions defined as propositions that
are tested in practice as well as grounded in the existing body of research.
The practice-based propositions were derived from the data by means of a
careful coding and reduction process (Strauss & Corbin, 1990). First, the research
team coded all different practices and experiences reported by initial design team
members and support advisors as well as described in key documents. Next, the
coded practices were clustered and reduced to a small number of categories. For
each category, crucial elements of the solutions and any common denominators
were identified. Finally, for each practice-based proposition the different
experiences of support staff and entrepreneurs were listed. For example, some of
the practice-based propositions identified by Van Burg et al. (2008: 122) were as
follows:
Create arrangements for starters to use university labs and other resources;
() enable starters to use the academic network of the university; establish a
network of investors, industry contacts and financers around the support
organization.

170

The research team derived research-based propositions by means of a


systematic literature review using a qualitative meta-synthesis approach (Denyer
& Tranfield, 2006). The domain of this review was defined in terms of all
research in the area of university spin-offs. The purpose of the review was to
derive normative (generic) propositions rather than to provide a comprehensive
overview. The findings from the review were synthesized in a number of key
concepts and a preliminary set of propositions. Subsequently, this set of researchbased propositions was linked to general theories, to explain the key mechanisms
addressed by these propositions (according to the CIMO logic, advocated by
Denyer, Tranfield and Van Aken, 2008).
Finally, Van Burg et al. (2008) developed a set of design propositions by
confronting and comparing the list of practice-based propositions with the list of
research-based propositions. This resulted in the following set of design
propositions for building and increasing capacity for creating spin-offs; in this
respect, a (European) university should design and implement practices that:
1. Create university-wide awareness of entrepreneurship opportunities,
stimulate the development of entrepreneurial ideas, and subsequently screen
entrepreneurs and ideas by programs targeted at students and academic staff.
2. Support start-up teams in composing and learning the right mix of
venturing skills and knowledge by providing access to advice, coaching and
training.

171

3. Help starters in obtaining access to resources and developing their social


capital by creating a collaborative network organization of investors, managers
and advisors.
4. Set clear and supportive rules and procedures that regulate the university
spin-off process, enhance fair treatment of the parties involved, and separate spinoff processes from academic research and teaching.
5. Shape a university culture that reinforces academic entrepreneurship by
creating norms and exemplars that motivate entrepreneurial behavior (Van Burg
et al., 2008: 123).
This set of emerging design propositions was shared with the design group
as a whole. They were discussed repeatedly with startup advisors, the incubation
unit manager, entrepreneurship professors and other TU/e stakeholders. This
ongoing dialogue also served to reposition and fine tune two specific TU/e
incubation practices. The first involved a program in which MSc students work in
teams to develop value propositions and business models for a particular
university-developed technology. The second practice was a regional incubator
network in which investors, banks, incubators of local companies (including
Philips Electronics), applied research institutes and regional developments
agencies collaborate with the startup advisors of the TU/e to provide access to
resources, network contacts and other facilities to entrepreneurs engaging in spinoff projects. The design propositions developed by Van Burg et al. (2008) served
to create awareness and understanding of each of the elements of the TU/e
approach to spin-off creation, by placing these in a broader framework.

172

The set of design propositions exposed blind spots and major opportunities
for improvement and development. Moreover, they helped to create a shared
repertoire of knowledge on university spin-offs for the various parties concerned.
Thus, the EBMgt project by Van Burg and co-authors served to define the lack of
an entrepreneurial university culture (cf. design proposition 5) at the TU/e as the
main weakness of its spin-off creation capability. Throughout the university, the
awareness of this deficiency grew and several new initiatives and projects were
used to expose the TU/e community to role models (e.g. successful entrepreneurs
among the TU/e alumni) and to motivate entrepreneurial behavior (e.g. by
creating attractive financial benefits for scholars whose patented technologies are
used in successful spin-offs). The universitys senior management and other key
stakeholders were well aware of the long-term effort required here, because
universities tend to be rather conservative in view of their long-standing
commitment to academic research and education.
This evidence-based project on spin-off creation, set up in 2005, is still
ongoing at this writing. The preliminary results of the project are as follows. First,
the TU/e increased its spin-off rate from zero in the late 1990s and about 5 per
year around 2005, to the current rate (in 2010-11) of about 15 new firms
exploiting IP of the university per year. The design framework developed by Van
Burg et al. (2008), in combination with this boost in the spin-off rate, motivated
the universitys top management to increase its ambitions -- to double the spin-off
rate in the next 5 years. Of course, it is not possible to determine precisely to what
extent the EBMgt project of Van Burg et al. (2008) contributed to the increase in

173

the spin-off rate at the TU/e. It is not unlikely that without this EBMgt project the
TU/e would also have increased its performance in creating spin-offs. Moreover,
any EBMgt project is embedded in a continuously evolving and ambiguous socioeconomic system. The spin-off creation project at the TU/e, however, does
illustrate that EBMgt serves to build a systematic, theory-driven understanding of
a rather complex managerial issue, which previously was dealt on the basis of
common wisdom and personal experiences.
Another, more academic, upshot of this project is that the design
propositions developed by Van Burg et al. (2008) served to reflect on the
comprehensiveness of (previous) research and theory development. In this
respect, Van Burg et al. (2008) observed that some of their design propositions
are not yet (firmly) incorporated in the university spin-off literature. For example,
the proposition referring to clear and supportive rules and procedures was not
previously identified, and as such not grounded in any theoretical frameworks. In
turn, this finding motivated a new study exploring the role of transparency and
fairness in university spin-off formation. Moreover, other research teams have
adopted and replicated the design framework developed by Van Burg et al.
(2008), for example Barr, Baker, Markham and Kingon (2009) in the context of
several US-based universities.
Discussion
The term Evidence-Based Management (EBMgt) may evoke a picture of rational
decision-making (like the Rational Manager of Keppner & Tregoe, 1965) with
academic research findings as its main input, replacing intuitive, experience-based

174

management. Instead, this chapter demonstrates and illustrates a design science


approach in which EBMgt involves extending intuitive, experience-based
management. This extension draws on scientific evidence as a source of design
propositions and for creating a design process that effectively makes use of a
broad array of knowledge and perspectives. EBMgt itself involves identifying and
using those business practices that work, according to the best available evidence.
EBMgt demands a significant effort in gathering valid information for
designing solutions for management and organization problems. As yet, researchbased information on many managerial and organizational issues is not readily
accessible; our academic community thus needs to engage in a systematic
literature review and synthesis on each significant managerial challenge
practitioners are facing. In Medicine, the Cochrane database (www.Cochrane.org)
provides professionals with the evidence to design and optimize their
interventions for a broad range of disorders. As long as there is no such equivalent
in the field of management and organization, EBMgt in this field needs to focus
on major issues and challenges for which it is feasible to invest in an extensive
systematic research synthesis.
Even so, practitioners need persuasion and incentives to engage in EBMgt,
since its added value appears less obvious than in some other disciplines. We
already observed that the value of valid, explicit knowledge is self-evident in
Engineering. Interestingly, in the social world we have a similar example: law
practice. Evidence-based law practice simply is a non-issue: lawyers and judges
have to extensively use the law, case-law and all available evidence to get their

175

job done. On the other hand, experienced medical doctors sometimes need to be
seduced to acquire and use state-of-the-art explicit knowledge; they may feel that
their initial training in the medical discipline, complemented with their rich tacit
knowledge gained through experiential learning, is quite sufficient to operate as a
medical practitioner. The challenge here is to demonstrate the added value of
recently developed evidence to medical professionals.
A similar problem occurs for experienced managers and management
consultants: the added value of research-based knowledge is not always selfevident to them (and experienced managers may also have different views on
what counts as evidence; see Green & Potwoworski, this Handbook). The
community of management and organization researchers and educators must take
charge of this challenge, by building a strong case for EBMgt, by creating
attractive conditions for practitioners to join collaborative EBMgt projects and
especially by giving much attention to EBMgt in management training.
The term evidence suggests a decision mode; that is, the idea that the
main challenge is to choose between known alternatives. This implies that the
main contribution of evidence-based practice is to produce the evidence for
making a rational choice between these alternatives. As previously argued,
EBMgt will typically address major managerial issues. For such issues, we
advocate a design mode that puts substantial effort in the design of alternative
solutions or arrangements of solutions, thus crafting a high-quality outline design
before going into details. Academic research can provide significant support for
this type of design effort by providing a range of well-tested solution concepts.

176

A final comment relates to the roots of the evidence-based practice


movement in Medicine -- being both an asset and a liability. As an asset, it
provides a link with a well-respected design science. It is also a liability that
evokes criticism, like by those perceiving EBMgt as a solely technical-rational
approach to managerial issues (e.g. Denzin, 2009). Or the criticisms leveled by
Morrell (2008) of, among other things, having a commitment to positivism (quod
non), having simplistic ideas on the nature of evidence and on accumulation and
progress in social science knowledge and of being troubled by physician envy
(as opposed to the physics envy in some other social science arenas) . This
chapter serves to argue that this is not necessarily the case. In this respect, a
design science approach suggests that effective EBMgt is complementary to
experience-based management, by giving research evidence a productive role in
management as a complex professional process.
One of the characteristics of a Design Science approach to EBMgt is that a
significant part of the general evidence to be used is produced by DSR, an
approach to research that has three defining characteristics: it is driven by field
problems, uses a participant-observer perspective, and adopts a solution
orientation. As such, DSR differs from the (in the social sciences) far more
common explanatory research. But DSR is not foreign to management research:
both observe reality, both analyze observations and both draw grounded
conclusions. Explanatory research studies the world as it is, DSR is interested in
what the world can be. But knowledge about what comes from experimenting
with these future states and observing and learning from these experiments. In its

177

methods DSR is thus not fundamentally different from explanatory research.


Furthermore, almost every actual DSR-project starts with an explanatory part,
describing and analyzing the type of field problem at hand and only then starts
with developing alternative solutions. If there is a substantial body of knowledge
about the field problem in question, the DSR part can be large; if little is known,
the explanatory part needs to be larger, leaving much work for DSR in follow-up
projects. Finally, DSR is not discipline-specific. In disciplines like Medicine and
Engineering it is an important part of mainstream research, but in other disciplines
one can do DSR equally well.
Implications for Practice and Research
Like Bismarck said: fools learn from their experience, I prefer to learn from the
experience of others. EBMgt does not replace intuitive, experience based
management, but extends it by judiciously gathering, validating, selecting and
using knowledge on the incumbent organization, its environment and the types of
issues at stake. In a design science approach to EBMgt, a critical part of the
general evidence to be used is produced by DSR. Moreover, one deals with these
significant issues in a design rather than decision mode: the emphasis is on
designing and creating a preferred future, rather than on using evidence to choose
from existing alternatives. EBMgt is not about finding exact interventions in the
literature, but about interrogating scientific evidence to obtain valid input for the
design of managerial initiatives, interventions and systems.
Next to practitioners, educators and researchers also have an important
role in advancing and applying EBMgt. MBA as well as DBA programs should

178

include rigorous training in EBMgt, including systematic review and research


synthesis, preferably also in real life settings. Researchers can support EBMgt by
including design science in their repertoire of research strategies, in order to
develop field-tested solution concepts for significant managerial issues. In the
medical discipline, the Cochrane database does not only give practitioners
diagnoses of dysfunctions, but also field-tested alternative treatments. Likewise,
EBMgt should (ideally) be able to draw on a body of valid knowledge on
alternative solutions to types of field problems.
EBMgt can and should become part of mainstream managerial thinking,
just like Total Quality Management became part of this mainstream after the hype
surrounding it had abated. For this to become reality a strong partnership between
interested practitioners, educators and researchers is needed.

179

References
Archer, M.S. (1995). Realist social theory: The morphogenetic approach.
Cambridge: Cambridge University Press.
Barr, S.H., Baker, T., Markham, S.K., & Kingon, A.I. (2009). Bridging the Valley
of Death: Lessons learned from 14 years of commercialization of
technology education. Academy of Management Learning and Education,
8, 370-388.
Bate, P. (2007). Editorial to the special issue Bringing the design sciences to
organization development and change management. Journal of Applied
Behavioral Science, 43, 8-11.
Bhaskar, R. (1998). The possibility of naturalism: A philosophical critique of the
contemporary human sciences (3rd ed.). New York, NY: Routledge.
Boland, R. & Collopy, F. (2004a) (Eds.). Managing as designing. Stanford, CA:
Stanford University Press.
Boland, R. & Collopy, F. (2004b). Design matters for management. In Boland,
R. & Collopy, F. (2004a) (Eds.), Managing as designing. Stanford, CA:
Stanford University Press.
Briner, R.B., Denyer, D. & Rousseau, D.M. (2009). Evidence-based management:
concept cleanup time? Academy of Management Perspectives, 23, 19-32.
Collins, A., Joseph, D. & Bielaczyc, K. (2004). Design research: Theoretical and
methodological Issues. Journal of the Learning Sciences, 13, 15-42.
Denyer, D. & Tranfield, D. (2006). Using qualitative research synthesis to build
an actionable knowledge base. Management Decision, 42, 213- 227.

180

Denyer, D., Tranfield, D. & Van Aken, J.E. (2008). Developing design
propositions through research synthesis. Organization Studies, 29, 249269.
Denzin, N.K. (2009). The elephant in the living room; or extending the
conversation about the politics of evidence. Qualitative Research, 9, 139160.
Dewey, J. (1929). The quest for certainty: A study of the relation of knowledge
and action. New York, NY: Minton, Balch and Company.
Goodman, J. (2010). Healthcare: Who knows best? The New York Times Review
of Books, 57(2), February 11.
Hevner, A.R., March, S.T. & Park, J. (2004). Design science in information
system research. MIS Quarterly, 28, 75-105.
James, W. (1907). Pragmatism, a new name for some old ways of thinking. New
York, NY: Longmans, Green and Company.
Jelinek, M., Romme, A.G.L & Boland, R.J. (2008). Introduction to the special
issue Organization Studies as a Science for Design: Creating
collaborative artifacts and research. Organization Studies, 29, 317-329.
Kasanen, E, Lukka, K. & Siitonen, A. (1993). The constructive approach in
management accounting research. Journal of Management Accounting
Research, 5, 243-263.
Kelly, A.E. (2003). Research as design. Educational Researcher, 32, 3-4.
Keppner, C.H. & Tregoe, B.J. (1965). The rational manager. New York, NY:
McGraw Hill.

181

Kolb, D.A. (1984) Experiential learning: Experience as the source of learning


and development. Englewood Cliffs, NJ: Prentice-Hall.
Labro, E. & Tuomela, T-S. (2003). On bringing more action into management
accounting research. European Accounting Review, 12, 409-442.
Leake, D.B. (1966). Case-based reasoning: ExSperiences, lessons and future
directions. Menlo Park, CA: American Association for Artificial
Intelligence.
Lindberg, K. & Czarniawska, B. (2006). Knotting the action net, or organizing
between organizations. Scandinavian Journal of Management, 22, 292306.
March, F.T. & Smith, G.F. (1995). Design and natural science research on
information technology. Decision Support Systems, 15, 251-266.
Mitchell, S.D. (2000). Dimensions of scientific law. Philosophy of Science, 67,
242-265.
Mohrman, S.A. (2007). Having relevance and impact: The benefits of integrating
the perspectives of design science and organizational development.
Journal of Applied Behavioral Science, 43, 12-22.
Morrell, K. (2008). The narrative of evidence based management: A polemic.
Journal of Management Studies, 45, 613-635.
Peirce, C.S. (1955). Philosophical writings of Pierce. Selected and edited by J.
Buchler. New York, NY: Dover publications.
Peltokorpi, V. (2008). Transactive memory systems. Review of general
Psychology, 12, 378-394.

182

Romme, A.G.L. (2003). Making a difference: Organization as design.


Organization Science, 14, 558-573.
Romme, A.G.L. & Endenburg, G. (2006). Construction principles and design
rules in the case of circular design. Organization Science, 17, 287-297.
Rousseau, D. M. (2006). Is there such a thing as evidence-based management?
Academy of Management Learning and Education, 31, 256-269.
Schn, D.A. (1983). The reflective practitioner. London, UK: Temple Smith.
Simon, H.A. (1996). The sciences of the artificial (3rd ed.). Cambridge,MA: the
MIT Press (1st edition 1969).
Tranfield, D., Denyer, D., & Smart, P. (2003). Towards a methodology for
developing evidence-informed management knowledge by means of
systematic review. British Journal of Management, 14, 207-222.
Van Aken, J.E. (2004). Management research based on the paradigm of the design
sciences: The quest for field-tested and grounded technological rules.
Journal of Management Studies, 41, 219-246.
Van Aken, J. E., Van der Bij, H. & Berends, H. (2007). Problem solving in
organizations: A methodological handbook for business students.
Cambridge, UK: Cambridge University Press.
Van Aken, J.E. (2010). Donald Schns legacy to address the great divide
between theory and practice. Planning Theory and Practice, 11, 609-613.
Van Burg, E., Romme, A.G.L., Gilsing, V.A. & Reymen, I.M.M.J. (2008).
Creating university spin-offs: A science-based design perspective. Journal
of Product Innovation Management, 25, 114-128.

183

Watson, I. (1997). Applying case-based reasoning: Techniques for enterprise


systems. San Francisco, CA: Morgan Kaufman Publishers.
Wegner, D.M. (1986). Transactive memory: A contemporary analysis of the
group mind. In Mullen, B. & Goethals, G.R. (Eds.).

Theories of Group

Behavior. New York, NY: Springer Verlag, 185-208.


Weick, K.E. (2004). Designing for thrownness. In Boland, R. & Collopy, F.
(Eds.), Managing as Designing (pp74-78). Stanford, CA: Stanford
University Press.
Worren, N., Moore, K. & Elliott, R. (2002). When theories become tools: Toward
a framework for pragmatic validity. Human Relations, 55, 1227-1250.

184

You might also like