Journal of Management Information Systems / Winter 2007–8, Vol. 24, No. 3, pp. 45–77.
© 2008 M.E. Sharpe, Inc.
0742–1222 / 2008 $9.50 + 0.00.
A Design Science Research Methodology
for Information Systems Research
KEN PEFFERS, TUURE TUUNANEN, MARCUS A. ROTHENBERGER,
AND SAMIR CHATTERJEE
KEN PEFFERS is an Associate Professor and Chair of the Management Information
Systems Department at the University of Nevada, Las Vegas. He earned his Ph.D.
in Management Information Systems from Purdue University. His current research
focuses on making the right IS investments for the ﬁrm, on IS planning, and on require-
ments determination for new information systems. His research articles have been
published in journals such as Communications of the ACM, Journal of Management
Information Systems, Information Systems Research, IEEE Transactions on Engineer-
ing Management, Organization Science, Journal of Information Technology Theory
and Application, and Information & Management.
TUURE TUUNANEN is a Senior Lecturer in the Department of Information Systems and
Operations Management at the University of Auckland. He holds a D.Sc. (Econ) from
the Helsinki School of Economics. His current research interests lie in the areas of
IS development methods and processes, requirements engineering, risk management,
and convergence of IS and marketing disciplines, speciﬁcally in design of interactive
consumer services and products. His research has been published in Information &
Management, Journal of Database Management, Journal of Information Technology
Theory and Application, Journal of Management Information Systems, and Technology
Analysis and Strategic Management. In addition, his work has appeared in a number
of conference proceedings within his research interest areas such as European Con-
ference on Information Systems, Design Science Research in Information Systems
and Technology, Hawaii International Conference on System Sciences, Mobility
Roundtable and Requirements Engineering Conference. Dr. Tuunanen is a member
of the Association of Computing Machinery, the Association of Information Systems,
and the Institute of Electrical and Electronics Engineers.
MARCUS A. ROTHENBERGER is an Associate Professor of Management Information Sys-
tems at the University of Nevada, Las Vegas. He holds a Ph.D. in Information Systems
from Arizona State University. Dr. Rothenberger’s work includes theory testing, theory
development, and design science research in the areas of software process improve-
ment, software reusability, performance measurement, and the adoption of enterprise
resource planning systems. His work has appeared in major academic journals, such as
Decision Sciences, IEEE Transactions on Software Engineering, Communications of
the ACM, and Information & Management. Dr. Rothenberger is regularly involved in
major academic conferences, including the International Conference on Information
Systems and the Americas Conference on Information Systems. He is a member of the
Association for Information Systems and the Decision Sciences Institute.
46 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
SAMIR CHATTERJEE is a Professor in the School of Information Systems & Technol-
ogy and Founding Director of the Network Convergence Laboratory at Claremont
Graduate University, California. He holds a Ph.D. in Computer Science from the
University of Central Florida. He is widely recognized as an expert in the areas of
next-generation networking, voice- and video-over Internet protocol, and network
security. His current research is on designing IT solutions to problems in health care
and collaboration. He has published over 75 articles in respected scholarly journals
and refereed conferences including IEEE Network, IEEE Journal on Selected Areas
in Communications, Communications of the ACM, Computer Networks, International
Journal of Healthcare Technology & Management, Telemedicine & e-Health Journal,
Information Systems Frontiers, Computer Communication, IEEE IT Professional,
ACM Computer Communication Review, Communications of AIS, and Journal of In-
ternet Technology, among others. He is the principal investigator on several National
Science Foundation (NSF) grants and has received funding from numerous private
corporations for his research. He is on the editorial board of International Journal of
Business Data Communications & Networking. He is the founding Program Chair
for the First International Design Science Conference (DESRIST 2006) and General
Chair for DESRIST 2008. He has been an entrepreneur and successfully cofounded
a startup company, VoiceCore Technologies, Inc., in 2000.
ABSTRACT: The paper motivates, presents, demonstrates in use, and evaluates a meth-
odology for conducting design science (DS) research in information systems (IS). DS
is of importance in a discipline oriented to the creation of successful artifacts. Several
researchers have pioneered DS research in IS, yet over the past 15 years, little DS
research has been done within the discipline. The lack of a methodology to serve as a
commonly accepted framework for DS research and of a template for its presentation
may have contributed to its slow adoption. The design science research methodology
(DSRM) presented here incorporates principles, practices, and procedures required to
carry out such research and meets three objectives: it is consistent with prior literature,
it provides a nominal process model for doing DS research, and it provides a mental
model for presenting and evaluating DS research in IS. The DS process includes six
steps: problem identiﬁcation and motivation, deﬁnition of the objectives for a solu-
tion, design and development, demonstration, evaluation, and communication. We
demonstrate and evaluate the methodology by presenting four case studies in terms
of the DSRM, including cases that present the design of a database to support health
assessment methods, a software reuse measure, an Internet video telephony applica-
tion, and an IS planning method. The designed methodology effectively satisﬁes the
three objectives and has the potential to help aid the acceptance of DS research in
the IS discipline.
KEY WORDS AND PHRASES: case study, design science, design science research, design
theory, mental model, methodology, process model.
INFORMATION SYSTEMS (IS) IS AN APPLIED RESEARCH discipline, in the sense that we fre-
quently apply theory from other disciplines, such as economics, computer science, and
the social sciences, to solve problems at the intersection of information technology
(IT) and organizations. However, the dominant research paradigms that we use to
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 47
produce and publish research for our most respected research outlets largely continue
to be those of traditional descriptive research borrowed from the social and natural
sciences. We recently accepted the use of interpretive research paradigms, but the
resulting research output is still mostly explanatory and, it could be argued, not often
applicable to the solution of problems encountered in research and practice. While
design, the act of creating an explicitly applicable solution to a problem, is an accepted
research paradigm in other disciplines, such as engineering, it has been employed
in just a small minority of research papers published in our best journals to produce
artifacts that are applicable to research or practice.
Without a strong component that produces explicitly applicable research solutions,
IS research faces the potential of losing inﬂuence over research streams for which
such applicability is an important value. For example, we wonder whether the prefer-
ence for theory building and testing research may help to explain why the center of
gravity for research in systems analysis and design—arguably, IS research’s raison
d’être—appears to have moved to engineering, dominated by research streams such
as requirements engineering and software engineering. Engineering disciplines accept
design as a valid and valuable research methodology because the engineering research
culture places explicit value on incrementally effective applicable problem solutions.
Given the explicitly applied character of IS practice and the implicitly applied character
of IS research, as part of the business academe, we should do so as well.
In recent years, several researchers succeeded in bringing design research into the
IS research community, successfully making the case for the validity and value of
design science (DS) as an IS research paradigm [20, 31, 55] and actually integrating
design as a major component of research . In spite of these successful efforts to
deﬁne DS as a legitimate research paradigm, DS research has been slow to diffuse
into the mainstream of IS research in the past 15 years  and much of it has been
published in engineering journals.
An accepted common framework is necessary for DS research in IS and a mental
model [18, 45, 54] or template for readers and reviewers to recognize and evaluate
the results of such research. Every researcher trained in the culture of social science
research has mental models for empirical and theory building research that allow
the researcher to recognize and evaluate such work, and perhaps one for interpretive
research as well. Even if all of these mental models are not exactly the same, they
provide contexts in which researchers can understand and evaluate the work of others.
For example, if a researcher reviewed an empirical paper that failed to describe how
the data were gathered, he or she would probably always regard that as an omission
that required notice and correction. Because DS research is not part of the dominant
IS research culture, no such commonly understood mental model exists. Without one,
it may be difﬁcult for researchers to evaluate it or even to distinguish it from practice
activities, such as consulting.
A number of researchers, both in and outside of the IS discipline, have sought to
provide some guidance to deﬁne DS research . Work in engineering [2, 14, 16, 38],
computer science [37, 46], and IS [1, 10, 20, 31, 33, 40, 55, 56] has sought to collect
and disseminate the appropriate reference literature ; characterize its purposes;
48 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
differentiate it from theory building and testing research, in particular, and from other
research paradigms; explicate its essential elements; and claim its legitimacy. However,
so far this literature has not explicitly focused on the development of a methodology
for carrying out DS research and presenting it.
We propose and develop a design science research methodology (DSRM) for the
production and presentation of DS research in IS. This effort contributes to IS research
by providing a commonly accepted framework for successfully carrying out DS re-
search and a mental model for its presentation. It may also help with the recognition
and legitimization of DS research and its objectives, processes, and outputs, and it
should help researchers to present research with reference to a commonly understood
framework, rather than justifying the research paradigm on an ad hoc basis with each
Problem Identiﬁcation: Completing a DSRM for IS Research
WHEN IS RESEARCHERS STARTED TO DEVELOP an interest in DS research in the early 1990s,
there already was agreement in prior research about the basic difference between DS
and other paradigms, such as theory building and testing, and interpretive research:
“Whereas natural sciences and social sciences try to understand reality, design science
attempts to create things that serve human purposes” [43, p. 55]. Three papers from
the early 1990s [31, 33, 55] introduced DS research to the IS community. Nunamaker
et al.  advocated the integration of system development into the research process,
by proposing a multimethodological approach that would include theory building,
systems development, experimentation, and observations. Walls et al.  deﬁned
IS design theory as a class of research that would stand as an equal with traditional
social science–based theory building and testing. March and Smith  pointed out
that design research could contribute to the applicability of IS research by facilitating
its application to better address the kinds of problems faced by IS practitioners.
Once this literature provided a conceptual and paradigmatic basis for DS research,
Walls et al.  expected its widespread adoption within IS, believing that this would
lead to IS research having more impact on practice through close ties between DS
research and practical applications. Despite the precedents of these early papers, Walls
et al.  observed that this rush to publish DS research did not occur and that the DS
research paradigm had only occasionally been used explicitly in the past ten years.
Given that many papers in reference disciplines, such as engineering and computer
science, use DS as a research approach and, in doing so, realize beneﬁts from the
practical applicability of research outcomes (e.g., [3, 19, 27, 28, 29]), it would seem
reasonable that it could also happen in IS.
Toward a DSRM
Some engineering literature (e.g., ) has pointed to a need for a common DSRM.
Archer’s  methodology focuses on one kind of DS research, which resulted in
building system instantiations as the research outcome, or “the purposeful seeking
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 49
of a solution” [32, p. 14] to a problem formulated from those desires . Archer 
believed that design could be codiﬁed, even the creative part of it. Archer’s industrial
engineering research outcomes reﬂect his views on research methodology. His work
included purpose-oriented designs for hospital beds and for mechanisms that prevented
ﬁre doors from being propped open. Through this work, he deﬁned six steps of DS
research: programming (to establish project objectives), data collection and analysis,
synthesis of the objectives and analysis results, development (to produce better design
proposals), prototyping, and documentation (to communicate the results). With these
steps, he asserted that designers can approach design problems “systematically,” by
looking at functional-level problems such as goals, requirements, and so on, and by
progressing toward more speciﬁc solutions .
A methodology is “a system of principles, practices, and procedures applied to a
speciﬁc branch of knowledge” . Such a methodology might help IS researchers
to produce and present high-quality DS research in IS that is accepted as valuable,
rigorous, and publishable in IS research outlets. For DS research, a methodology would
include three elements: conceptual principles to deﬁne what is meant by DS research,
practice rules, and a process for carrying out and presenting the research.
Principles: DS Research Deﬁned
With just a decade and a half of history, DS research in IS may still be evolving;
however, we now have a reasonably sound idea about what it is. “Design science . . .
creates and evaluates IT artifacts intended to solve identiﬁed organizational problems”
[20, p. 77]. It involves a rigorous process to design artifacts to solve observed prob-
lems, to make research contributions, to evaluate the designs, and to communicate the
results to appropriate audiences . Such artifacts may include constructs, models,
methods, and instantiations . They may also include social innovations  or
new properties of technical, social, or informational resources ; in short, this
deﬁnition includes any designed object with an embedded solution to an understood
Practice Rules for DS Research
Hevner et al.  provided us with practice rules for conducting DS research in the
IS discipline in the form of seven guidelines that describe characteristics of well
carried out research. The most important of these is that the research must produce
an “artifact created to address a problem” [20, p. 82]. Further, the artifact should be
relevant to the solution of a “heretofore unsolved and important business problem”
[20, p. 84]. Its “utility, quality, and efﬁcacy” [20, p. 85] must be rigorously evaluated.
The research should represent a veriﬁable contribution and rigor must be applied in
both the development of the artifact and its evaluation. The development of the artifact
should be a search process that draws from existing theories and knowledge to come
up with a solution to a deﬁned problem. Finally, the research must be effectively
communicated to appropriate audiences .
50 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
Procedures: A Process Model and Mental Model for
Prior research has introduced principles that deﬁne what DS research is  and
what goals it should pursue [16, 20], as well as practice rules that provide guidance
for conducting [2, 16, 20, 38] and justifying it [1, 33, 55]. Nevertheless, principles
and practice rules are only two out of the three characteristics of a methodology .
The missing part is a procedure that provides a generally accepted process for car-
rying it out.
Hitherto, IS researchers have not focused on the development of a consensus process
and mental model for DS research, such as that called for in engineering literature
[16, 38] and required by the IS research discipline. This lack of a consensus-based DS
research process model may help to explain why, despite many citations, the message
of DS research has not resulted in more research in IS that makes explicit use of the
paradigm . Instead, much of the DS research published by IS researchers has been
published in engineering journals, where DS behaviors are more the norm. Some of
that published in IS journals has required ad hoc arguments to support its validity [5,
9, 34, 36, 42]. For example, in Peffers and Tuunanen , the authors use information
theory to justify the use of an IS planning method, which, in reality, was a designed
method. In Peffers et al. , the researchers justify their work as a practical extension
of another methodology, rather than making explicit design claims. In Rothenberger
and Hershauer , the authors describe the development of a software reuse measure
in the context of a ﬁeld study and evaluate the artifact using one project of the ﬁeld
company that is treated as a case study.
Deﬁning Objectives of a Solution: Process and Mental Models
Consistent with Prior Research
OUR OVERALL OBJECTIVE FOR THE PAPER IS THE DEVELOPMENT of a methodology for DS
research in IS. We do this by introducing a DS process model, which, together with
prior research on DS, provides DS research with a complete methodology. The design
of this conceptual process will seek to meet three objectives: it will (1) provide a
nominal process for the conduct of DS research, (2) build upon prior literature about
DS in IS and reference disciplines, and (3) provide researchers with a mental model
or template for a structure for research outputs.
A Nominal Process
Such a process could accomplish two things for DS research in IS. It would help provide
a road map for researchers who want to use design as a research mechanism for IS
research. Such a process would not be the only way that DS research could be done, but
it would suggest a good way to do it. It could also help researchers by legitimizing such
research, just as researchers understand the essential elements of empirical IS research
and accept research that is well done using understood and accepted processes.
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 51
Building on Prior Research
There is a substantial body of research, both within the IS literature and in reference
disciplines, that provides us with a tradition to support such a process. A process for
DS research should build on this work while integrating its principles into a compre-
hensive methodology for conducting DS research. There are two sets of applicable
literature. One revolves around issues of actually doing academic design work—that
is, design research. The second set addresses the meta level of conducting research
at a higher level of abstraction—research about design research. Below, we discuss
the differences and how both contribute to meeting this objective.
The design research literature contains a large number of references to processes
that are described incidentally to the production of research-based designs. Many of
these descriptions are speciﬁc to research contexts and to the practical needs of design
practitioners. In engineering, for example, there have been a number of design research
efforts in which the focus has been on processes targeting the production of artifacts
. Evbuonwan et al.  mention 14 such process models. Many, such as Cooper’s
StageGate [11, 12], are clearly intended as design or development methodologies, rather
than research methodologies or processes, such as the one we are seeking to develop
for DS research in IS. Likewise, in computer science, Maguire’s  human-centered
design cycle addresses the speciﬁc problems of requirements engineering methods
for different situations and, in IS, Hickey and Davis  addressed the issue from
a functional view. Iivari et al.  considered the differences between IS develop-
ment methods and methodologies and the needs that arise for method development.
Processes described in this literature are of interest, but, because they vary widely
and are generally context speciﬁc, they cannot necessarily be directly applied to the
development of a general process for DS research.
The research about design research literature is rich with ideas about how to conduct
research. This literature, while not providing process models that can be applied directly
to the problem of DS research, provides concepts from which we can infer processes.
In IS, Nunamaker et al.  provided an abstract model connecting aspects of design
research, but leave the actual process for conducting it to the researcher’s inference.
Walls et al.’s [55, 56] IS design theory provides theory at a high level of abstraction
from which we can infer a process. Hevner et al.’s  and March and Smith’s 
guidelines for DS research inﬂuence methodological choices within the DS research
process. In the computer science domain, Preston and Mehandjiev  and Takeda
et al.  proposed a “design cycle” for intelligent design systems.
In engineering, Archer  and Eekels and Roozenburg  presented design process
models that could be incorporated into a consensus process. Adams and Courtney
 proposed an extension of Nunamaker et al.’s  system development research
methodology via inclusion of action research or grounded theory approach as ways
to conduct research. Cole et al.  and Rossi and Sein  proposed basic steps to
integrate DS research and action research. So far, no complete, generalizable process
model exists for DS research in IS; however, if we develop such a process model, it
should build upon the strengths of these prior efforts.
52 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
A Mental Model
The ﬁnal objective of a DSRM process is to provide a mental model for the character-
istics of research outputs. A mental model is a “small-scale [model] of reality . . . [that]
can be constructed from perception, imagination, or the comprehension of discourse.
[Mental models] are akin to architects’ models or to physicists’ diagrams in that their
structure is analogous to the structure of the situation that they represent, unlike, say,
the structure of logical forms used in formal rule theories” . Outcomes from DS
research are clearly expected to differ from those of theory testing or interpretative
research. A process model should provide us with some guidance, as reviewers, edi-
tors, and consumers, about what to expect from DS research outputs. March and Smith
 contributed to this expectation with their ideas about research outputs. Hevner et
al.  further elaborated on this expectation by describing DS research’s essential
elements. A mental model for the conduct and presentation of DS research will help
researchers to conduct it effectively.
Design: Development of the Methodology
DEVELOPMENT OF THE METHODOLOGY REQUIRED the design of a DSRM process. To ac-
complish this, we looked to inﬂuential prior research and current thought to determine
the appropriate elements, seeking to build upon what researchers said in key prior
literature about what DS researchers did or should do. Our aim here was to design
a methodology that would serve as a commonly accepted framework for carrying
out research based on DS research principles outlined above. Rather than focusing
on nuanced differences in views about DS among various researchers, we sought to
use a consensus-building approach to produce the design. Consensus building was
important to ensure that we based the DSRM on well-accepted elements.
A number of researchers in IS and other disciplines have contributed ideas for
process elements. Table 1 presents process elements, stated or implied, from seven
representative papers and presentations and our synthesis: the components of the
DSRM process. The authors agree substantially on common elements. The result of
our synthesis is a process model consisting of six activities in a nominal sequence,
which we justify and describe here and graphically in Figure 1.
All seven papers include some component in the initial stages of research to deﬁne a
research problem. Nunamaker et al.  and Walls et al.  emphasized theoretical
bases, whereas engineering researchers [2, 14] focused more on applied problems.
Takeda et al.  suggested the need for problem enumeration, whereas Rossi and
Sein  advocated need identiﬁcation. Hevner et al.  asserted that DS research
should address important and relevant problems.
Activity 1: Problem identiﬁcation and motivation. Deﬁne the speciﬁc research
problem and justify the value of a solution. Because the problem deﬁnition will
be used to develop an artifact that can effectively provide a solution, it may be
useful to atomize the problem conceptually so that the solution can capture its
complexity. Justifying the value of a solution accomplishes two things: it moti-
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 53
Table 1. Design and DS Process Elements from IS and Other Disciplines and Synthesis Elements for a DSRM in IS
et al. 
Common design Takeda Eekels and Nunamaker Walls Rossi and Hevner
process elements Archer  et al.  Roozenburg  et al.  et al.  Sein  et al. 
Problem identiﬁcation Programming, Problem Analysis Construct a Meta- Identify a Important
and motivation data collection enumeration conceptual requirements, need and relevant
framework kernel problems
Objectives of a Requirements Implicit in
Design and Analysis, Suggestion, Synthesis, Develop a Design Build Iterative
development synthesis, development tentative system method, search
development design architecture, meta design process,
proposals analyze and artifact
Demonstration Simulation, Experiment,
conditional observe, and
prediction evaluate the
Evaluation Conﬁrmatory Evaluation, Testable Evaluate Evaluate
evaluation decision, design
deﬁnite design process/product
Communication Communication Communication
54 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
Figure 1. DSRM Process Model
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 55
vates the researcher and the audience of the research to pursue the solution and
to accept the results and it helps to understand the reasoning associated with the
researcher’s understanding of the problem. Resources required for this activity in-
clude knowledge of the state of the problem and the importance of its solution.
Some of the researchers explicitly incorporate efforts to transform the problem into
system objectives, also called metarequirements  or requirements , whereas
for the others, these efforts are implicit as part of programming and data collection
 or implicit in the search for a relevant and important problem. Identiﬁed problems
do not necessarily translate directly into objectives for the artifact because the pro-
cess of design is necessarily one of partial and incremental solutions. Consequently,
after the problem is identiﬁed, there remains the step of determining the performance
objectives for a solution.
Activity 2: Deﬁne the objectives for a solution. Infer the objectives of a solution
from the problem deﬁnition and knowledge of what is possible and feasible. The
objectives can be quantitative, such as terms in which a desirable solution would
be better than current ones, or qualitative, such as a description of how a new
artifact is expected to support solutions to problems not hitherto addressed. The
objectives should be inferred rationally from the problem speciﬁcation. Resources
required for this include knowledge of the state of problems and current solutions,
if any, and their efﬁcacy.
All of the researchers focus on the core of DS across disciplines—design and devel-
opment. In some of the research (e.g., [14, 33]), the design and development activities
are further subdivided into more discrete activities whereas other researchers focus
more on the nature of the iterative search process .
Activity 3: Design and development. Create the artifact. Such artifacts are poten-
tially constructs, models, methods, or instantiations (each deﬁned broadly)  or
“new properties of technical, social, and/or informational resources” [24, p. 49].
Conceptually, a design research artifact can be any designed object in which a
research contribution is embedded in the design. This activity includes determining
the artifact’s desired functionality and its architecture and then creating the actual
artifact. Resources required for moving from objectives to design and development
include knowledge of theory that can be brought to bear in a solution.
Next, the solutions vary from a single act of demonstration  to prove that the idea
works, to a more formal evaluation [14, 20, 33, 40, 51] of the developed artifact. Eekels
and Roozenburg  and Nunamaker et al.  included both of these phases.
Activity 4: Demonstration. Demonstrate the use of the artifact to solve one or
more instances of the problem. This could involve its use in experimentation,
simulation, case study, proof, or other appropriate activity. Resources required
for the demonstration include effective knowledge of how to use the artifact to
solve the problem.
56 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
Activity 5: Evaluation. Observe and measure how well the artifact supports a
solution to the problem. This activity involves comparing the objectives of a
solution to actual observed results from use of the artifact in the demonstration.
It requires knowledge of relevant metrics and analysis techniques. Depending
on the nature of the problem venue and the artifact, evaluation could take many
forms. It could include items such as a comparison of the artifact’s functionality
with the solution objectives from activity 2, objective quantitative performance
measures such as budgets or items produced, the results of satisfaction surveys,
client feedback, or simulations. It could include quantiﬁable measures of system
performance, such as response time or availability. Conceptually, such evaluation
could include any appropriate empirical evidence or logical proof. At the end of
this activity the researchers can decide whether to iterate back to activity 3 to try
to improve the effectiveness of the artifact or to continue on to communication
and leave further improvement to subsequent projects. The nature of the research
venue may dictate whether such iteration is feasible or not.
Finally, Archer  and Hevner et al.  proposed the need for communication to
diffuse the resulting knowledge.
Activity 6. Communication. Communicate the problem and its importance, the
artifact, its utility and novelty, the rigor of its design, and its effectiveness to
researchers and other relevant audiences such as practicing professionals, when
appropriate. In scholarly research publications, researchers might use the structure
of this process to structure the paper, just as the nominal structure of an empiri-
cal research process (problem deﬁnition, literature review, hypothesis develop-
ment, data collection, analysis, results, discussion, and conclusion) is a common
structure for empirical research papers. Communication requires knowledge of
the disciplinary culture.
This process is structured in a nominally sequential order; however, there is no
expectation that researchers would always proceed in sequential order from activity
1 through activity 6. In reality, they may actually start at almost any step and move
outward. A problem-centered approach is the basis of the nominal sequence, starting
with activity 1. Researchers might proceed in this sequence if the idea for the research
resulted from observation of the problem or from suggested future research in a paper
from a prior project. An objective-centered solution, starting with activity 2, could
be triggered by an industry or research need that can be addressed by developing an
artifact. A design- and development-centered approach would start with activity 3. It
would result from the existence of an artifact that has not yet been formally thought
through as a solution for the explicit problem domain in which it will be used. Such
an artifact might have come from another research domain, it might have already
been used to solve a different problem, or it might have appeared as an analogical
idea. Finally, a client-/context-initiated solution may be based on observing a practical
solution that worked; it starts with activity 4, resulting in a DS solution if researchers
work backward to apply rigor to the process retroactively. This could be the by-product
of a consulting experience.
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 57
Demonstration in Four Case Studies
TO DEMONSTRATE THE USE OF THE DSRM, WE APPLY it retroactively to four already published
IS research projects. In the ﬁrst, researchers design and develop a data warehousing
solution to support data gathering and analysis necessary for public health policy
[4, 5]. The second explicates the design of a software reuse measure that was used
in subsequent case study research [41, 42]. The third reports on the design of an ap-
plication and middleware for the Internet2 environment that provides telephony and
video functionalities [9, 17]. Finally, the fourth depicts the development of a method,
critical success chains (CSC) [34, 36], for use in generating a portfolio of new ideas
for mobile ﬁnancial services applications.
In each case, we show how the process of motivating, developing, designing, dem-
onstrating, evaluating, and communicating the artifact is consistent with the DSRM.
In none of the cases were the publication outputs explicitly described and presented
as using a DS research process, because a designed methodology had not been hith-
erto available. In the summaries that follow, we used the language of the DSRM to
interpret the research processes actually used by the researchers to determine how
well the DSRM ﬁts with the research processes used.
Case 1: The CATCH Data Warehouse for
Health Status Assessments
The comprehensive assessment for tracking community health (CATCH) methods
was published  and successfully used in multiple counties in the United States.
The methodology requires data to be gathered from multiple sources, including
hospitals, health agencies, health-care groups, and surveys. CATCH organizes over
250 health-care indicators into 10 categories that represent a variety of health-care
issues. The output of the CATCH methodology is a prioritized listing of community
health-care challenges. In this work Berndt and colleagues [4, 5, 7] automated the
use of CATCH by developing a data warehouse that implements the methodology.
Figure 2 summarizes how the DSRM applies to the steps undertaken as part of this
DS research effort [4, 5, 7].
The lack of automated support made the data gathering for the CATCH methodology
labor intensive and slow; thus, extended trend analyses were cost prohibitive for most
communities. The need for a more efﬁcient automated data access for CATCH health
assessments triggered the development of the CATCH data warehouse.
Problem Identiﬁcation and Motivation
The United States has the highest health-care spending of any nation in the world, both
as a percentage of gross domestic product (GDP) and per capita. Nevertheless, the
United States does not rank among the countries with the healthiest populations. Thus,
58 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
Figure 2. DSRM Process for the CATCH Project
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 59
there was a need to assess the country’s health status in order to assist communities
to develop comprehensive health strategies, leading to better resource allocation for
prevention and treatment. The formulation of such strategy had to be based on local
health data. The availability and quality of health data was low, which is why health
data rarely were the basis for decision making on health policies. Although CATCH
was an available assessment method at the time, the labor intensity of nonautomated
data gathering limited its adoption.
Objective of the Solution
The objective was to develop a data warehouse solution for the automated support
of the CATCH methods that enables users to run cost-effective analyses. The major
challenges included the diversity of the data sources, the diversity of target groups for
which reports were generated, and the need to conform to the public policy formulation
process. The data warehouse was to provide a rich environment that would enable an
improvement of research capabilities on critical health-care issues with the long-term
goal of centering the role of public health agencies around monitoring and improving
the health status of the population using this technology.
Design and Development
The artifact is the data warehouse that supports and automates CATCH. The researchers
drew from data warehousing research to develop the CATCH data warehouse with data
arranged in a star schema. The design includes three levels of granularity: the report
structures, aggregate dimensional structures, and ﬁne-grained and transaction-oriented
dimensional structures. Staging and quality assurance methods were established to
enable a successful use of the data warehouse and performance issues were addressed.
The design and related methods have been and continue to be reﬁned based on emerg-
ing performance needs.
After developing proof-of-concept-level prototypes, the artifact was extensively adapted
to production use by user organizations. The researchers point to the application of the
CATCH data warehouse in multiple counties and provide screenshots of several output
screens in their articles. In related research, it was also demonstrated that the CATCH
data warehouse could be used to conduct bioterrorism surveillance: a similar data
warehouse approach was used in a demonstration surveillance system in Florida.
The original CATCH methods have been used and reﬁned for more than 10 years in
more than 20 U.S. counties. The researchers implemented the CATCH data warehouse
as a fully functional version in Florida’s Miami-Dade County. The veriﬁcation of the
accuracy of the automated generation of the report through a comprehensive manual
60 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
check identiﬁed only minor problems in use. The data warehouse was found to be
ﬂexible and effective in this ﬁeld application.
Manuscripts relating to the CATCH data warehouse have been published in academic
journals, academic conference proceedings, and professional outlets. The development
of the health-care data warehouse was presented in Decision Support Systems  and
Upgrade, a professional online magazine . The challenges of quality assurance in
the CATCH data warehouse were discussed in IEEE Computer . Further, the use
of data warehousing technology and CATCH in the context of bioterrorism has been
explored in proceedings of the IEEE International Conference on Intelligence and
Security Informatics [6, 8]. In addition, this research effort received attention from
various newspapers in Florida.
The CATCH data warehouse research resulted in architecture and applications. This
artifact was used effectively to collect data in a consistent and automated fashion
from disparate local health-care organizations, which, among themselves, had no
consistent IS or data collection infrastructure. The immediate contribution of this
research to public health policy was the ability to collect data that could be effectively
used to formulate such policy within Florida, where it was implemented. In a broader
context, the artifact could serve as a template for the implementation of such systems
elsewhere. Furthermore, the architecture and applications could serve as a model for
the development of similar systems, such as one that was developed for bioterrorism
surveillance, to serve other public or business needs.
Case 2: A Software Reuse Measure Developed at
MBA Technologies was a medium-sized Phoenix-based software developer that spe-
cialized in the development of business process and accounting systems; the company
obtained high reuse in its software development by leveraging of existing compo-
nents that were mapped to an enterprise-level model. The model and its components
represented generic business solutions that could be customized to a speciﬁc set of
requirements. The objective of this work by Rothenberger and Hershauer  was
to develop a generic reuse metric for such an enterprise-level model-based software
development environment and to apply the generic measure to the speciﬁcs of the
organization. Figure 3 provides a summary of the research steps discussed below.
In spring 1997, Rothenberger and Hershauer  wanted to conduct an in-depth case
study on the reuse efforts at MBA Technologies that required the assessment of the
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 61
Figure 3. DSRM Process for the MBA Technologies Study
62 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
reuse rates the company obtained in its projects. Existing reuse measures available in
the IS and computer science literatures were not suitable to assess the reuse rate in the
enterprise-level model context; existing measures were only deﬁned on a high level
and did not deﬁne speciﬁcs required for an application to actual projects. To use the
underlying principles of a high-level measure in the ﬁeld setting, decisions had to be
made about how to assess and count modiﬁed component reuse, partial component
reuse, generated code, and multiple layers of abstraction.
Problem Identiﬁcation and Motivation
Most software development companies do not assess their success at reuse, even if they
are actively pursuing an increase in the reuse of software artifacts through a formal
reuse program. Thus, many software developers invest in corporate reuse programs
without being able to evaluate whether their programs lead to an increase of reuse.
Also, without a formal reuse measure, they are not able to identify differences in reuse
success among projects. The development and subsequent dissemination of a reuse
measure that can be applied to enterprise-level model-based reuse efforts would enable
the researchers to conduct an in-depth analysis of MBA Technologies’ reuse success
across multiple completed projects. Further, a measure would provide the means for
continued monitoring of reuse success in software projects.
Objective of the Solution
The objective was to develop a reuse rate measure that allowed the researchers to assess
the reuse rate, or reuse percentage, of the participating organization for subsequent
case study research. Such a measure would represent the development effort that was
reused from existing code as a percentage of the total project development effort. The
measure was to be developed in a generic fashion that would ensure its applicability
to settings other than the participating organization as well.
Design and Development
The software measurement literature was used to evaluate the suitability of potential
size or complexity measures. The concept of the reuse rate was obtained from software
reuse literature, which served as the theoretical foundation for the development of the
reuse metric. The result of the design effort was a generic reuse measure that could be
applied to any enterprise-level model-based reuse setting and that was customized to
the speciﬁc organizational setting at MBA Technologies. The reuse rate was deﬁned
as the reused development effort divided by the total development effort of the proj-
ect. The metric artifact operationalized this high-level deﬁnition by formalizing how
to count reused development effort and total development effort in the context of an
enterprise-level model-based reuse setting. This operationalization required making
decisions on how to count duplicate use of code stubs, modiﬁed reused components,
and other special cases. These decisions and assessments were made based on prior
ﬁndings in the software reuse literature.
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 63
Assessing and reporting the reuse rate for a project in the participating organization
demonstrated the measure’s feasibility and efﬁcacy. Details about the company’s
development environment, including a classiﬁcation of code into three levels of ab-
straction, the use of generated code, speciﬁcs about the component design, and the
classiﬁcation of certain code stubs, were obtained through structured interviews. Size
measures in thousands of lines of code (KLOC) and the classiﬁcation of code stubs at
the lowest level of abstraction were obtained directly from source code. The measure
yielded separate reuse percentages for code on three layers of abstraction, according to
the organization’s classiﬁcation, as well as a weighted total reuse percentage. Further,
reused generated code was reported separately.
In the subsequent case study, the measure was used to assess the reuse rates of ﬁve
projects at MBA Technologies, with sizes varying from 57 KLOC to 143 KLOC. The
assessed total project reuse rate for nongenerated code ranged from 50.5 percent to
76.0 percent. In structured interviews, developers were asked to assess the projects’
reuse rates without prior knowledge of the measured results. The relative assessments
were consistent with the actual measurements.
The contributions of this effort were disseminated in peer-reviewed scholarly publi-
cations. The development of the reuse rate measure was published in Information &
Management . Further, the measure was used to assess the projects of the software
development organizations in a subsequent case study that appeared in Decision
The research artifacts resulting from this study included a designed and evaluated
formal measure and metric for software reuse rates. These artifacts provide a valid
and effective measure for use in development practice at the organizational and project
level for evaluation and assessment of the effectiveness and performance of software
reuse efforts. They could be valuable measures for use in research where measures
of software reuse are required.
Case 3: SIP-Based Voice- and Video-Over IP Software
The session initiation protocol (SIP) is an Internet Engineering Task Force (IETF)
standard for Internet protocol (IP) telephony that was developed for voice-over Internet
communication. Researchers at the Network Convergence Lab (NCL) at Claremont
Graduate University were involved since early 2000 in the standardization process
64 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
for SIP-based voice communication. In early 2001, the Internet2 Consortium wanted
to explore video-over IP applications as an emerging architecture for IP networks in
cooperation with the NCL. This led to a research effort by Chatterjee , Gemmill et
al. , and Tulu et al. [47, 48] focusing on the extension of the SIP standard [9, 17,
47, 48], which is summarized in Figure 4.
A Design- and Development-Centered Approach
Building on the SIP-based voice communication standard, researchers at NCL aimed
to design and deploy a voice- and videoconferencing-over IP application that enhances
the SIP-based voice communication standard. This DS research artifact was to be
deployed across 202 universities.
Problem Identiﬁcation and Motivation
There were three particular technical problems that emerged in discussions within
the IETF and the Internet2 consortium. First, while SIP standards were emerging,
there were no actual SIP-based software artifacts that would provide telephony and
video functionalities and features. Second, because universities and companies use a
variety of vendor products, technologies, and standards, there was a need to develop
middleware that provided a uniform way for storing and ﬁnding information related
to video and voice users, as well as devices and technologies in enterprise directories.
This problem was particularly relevant to Internet2 because universities were imple-
menting diverse technology solutions. Third, there was a need to solve the security
problem: some applications including SIP cannot traverse ﬁrewalls and fail to work
when private IP addresses are used behind network address translators (NATs).
Objectives of the Solution
Several requirements drove the research effort. First, researchers needed to follow
SIP technical standards closely. Second, the performance of the artifact could not be
allowed to overwhelm the capabilities of a typical desktop computer of the time. Also,
there were functions and features necessary to meet the requirements of the end users,
including point-to-point calls, instant messaging, and videoconferencing. Furthermore,
the middleware software for storing user and device information had to be compat-
ible with existing directory services within participating campuses. Finally, a security
solution was required that would be implemented within the application in such a way
that no external measures were required within ﬁrewalls and routers.
Design and Development
The design and development process followed that of an IS development research proj-
ect. It started with a requirements-gathering process, in which a diverse set of potential
end users participated, resulting in requirements documentation, which was later used
for designing a detailed technical architecture through Internet2 member meetings and
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 65
Figure 4. DSRM Process for the CGUsipClient v1.1.x Project
66 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
mailing list discussions. In particular, the middleware that was developed was standard-
ized through the International Telecommunications Union’s standardization section
(ITU-T), which required participation from several European and other international
participants. The software was developed, based on computer science and networking
literature, to provide a proof-of-concept and a fully working client application.
The implemented artifact includes the SIP application and its directory middleware.
Implementation details serve as a demonstration of the approach. CGUsipClient v1.1.x
is a Java-based application implemented on a commercial SIP stack. It uses Java
Media Framework (JMF) application programming interfaces (API) for voice and
video operations. It provides point-to-point telephony, video calls, directory service
lookup, click-to-call, and secured authentication. It uses technologies to solve the se-
curity problems mentioned above and utilizes a lightweight directory access protocol
(LDAP)–based solution for providing directory information. It uses an H.350 direc-
tory to offer “white page,” “click-to-call,” and “single sign-on” facilities. White page
displays user information. Click-to-call enables a user to call another user by clicking
on the other user’s SIP uniform resource identiﬁer (URI). Single sign-on provides an
authenticating facility with an SIP-based proxy or a registrar based on the credentials
fetched from the LDAP structure instead of explicitly providing the user name and
the password for registration.
Once the software was developed, researchers started a thorough testing process. First,
the artifact was extensively tested for debugging purposes within a closed group. Next,
the application was shared with the entire Internet2 community via a Web portal,
where users were able to download the software after providing information about
themselves. The information provided was automatically linked to the middleware
directory. More than 250 institutions downloaded the software artifact. The researchers
found that 30 percent of those who downloaded the software used it for one or more
hours daily. In addition, the CGUsipClient was tested for performance, usability, and
usefulness. The researchers measured the call setup time, CPU usage, end-to-end delay;
all results were satisfactory. The test of the H.350 middleware standard implementation
showed that the directory service performed well and lookup time was satisfactory.
Finally, the security mechanism that was developed to open and close pinholes in the
ﬁrewall/NAT for active SIP sessions was successfully tested, indicating only minimal
and acceptable delays. The Internet2 working group was pleased with the efforts,
judging the design process successful.
Preliminary results of this project were reported in refereed conferences (e.g., [47, 48])
and detailed results appeared in IEEE Journal on Selected Areas in Communications
 and Journal of Internet Technology . In addition, the middleware work received
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 67
recognition in Internet2 and National Science Foundation (NSF) press releases. Trade
magazines, such as Network World, and corporations, such as Packetizer, maintain the
H.350 middleware standard information and details.
This research enhanced and further developed the existing SIP voice-over IP standard
into an SIP-based video-over IP standard. The enhanced standard was successfully
evaluated and made available to the Internet2 working group, which may result in
the commercial use of this new standard. Further, the existence of a video-over IP
standard may serve as a foundation for future research aimed at enhancements of this
Case 4: Developing a Method at Digia to Generate Ideas for
New Applications That Customers Value
Digia Ltd. was a Helsinki-based research and development ﬁrm specializing in inno-
vative software applications for wireless communication that focused on the creation
of personal communication technologies and applications for wireless information
devices. This case reports the efforts of Peffers and colleagues [34, 36] at Digia, also
illustrated in Figure 5, to develop a better IS planning method.
A Client-Initiated Project
In fall 2000, Digia Chairman Pekka Sivonen approached one of the authors with a
request to help deﬁne a portfolio of potential applications for Digia to develop to meet
the need for ﬁnancial services delivered by the next-generation wireless devices .
The researcher accepted this invitation because it ﬁt with his current research objective
and that of colleagues: to develop a method to support the generation of ideas for IS
projects that would provide the greatest impact on achieving a ﬁrm’s strategic goals.
Because few applications for providing ﬁnancial services using mobile devices were
in operation at the time, this looked like a good opportunity to use a new conceptual
method for IS planning that the authors earlier trialed in a business case environment.
Because the client’s objective was a portfolio of applications and the research objec-
tive was the development of a requirements engineering methodology for determin-
ing this portfolio, this meant that the initiative for the project came from a proposed
demonstration of the new methodology.
Problem Identiﬁcation and Motivation
Literature had shown that in most ﬁrms, there was no shortage of ideas for new IS
projects, but most tended to be suboptimal . The problem was to design a method
for managers to make use of the ideas of many people within and around the organiza-
tion, while keeping the focus on what is important and valuable for the ﬁrm. Bottom-up
planning generates so many ideas that it may be impossible to sort out the few that
68 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
Figure 5. DSRM Process for the Digia Study
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 69
have the potential to have a high impact on the ﬁrm, because most are self-serving,
narrowly focused, and of little potential impact. Top-down planning has the beneﬁt of
strategic perspective and better alignment with the interests of owners, but its weak-
ness is an inability to take advantage of knowledge from around the organization and
beyond the organization about ideas that may be important to the ﬁrm. It generally
ignores all ideas except those that originate in the executive suites.
Objectives of the Solution
The researchers’ objective was to demonstrate a new IS planning method in an in-
dustry setting. This allowed the researchers to study how well, in a noncontrolled test
environment, the method would meet the proposed objectives: (1) allow them to make
use of the ideas of many from in and around the organization; (2) include experts
outside the ﬁrm and potential users, but keep the focus on ideas with high strategic
value to the ﬁrm; and (3) transform the resulting data into forms that can be used for
IS planning and development.
Design and Development
The Digia project researchers made use of a pilot study, conducted at Rutgers University
, as the basic template for the new method. They used personal construct theory
(PCT)  and critical success factors  as theoretical bases for the method devel-
opment. For the data collection, they borrowed “laddering,” a PCT-based technique
developed for use in marketing research for structured interviewing, to collect rich
data on subject reasoning and preferences. For the analysis, they adapted hierarchical
value maps, which had been used in marketing to display aggregated laddering data
graphically. They incorporated an ideation workshop, where business and technical
expertise was brought to bear on the task of developing feasible ideas for new business
applications from the graphical presented preferences and reasoning of the subjects.
The result of the design effort was the CSC method for using the ideas of many people
in and around the organization to develop portfolios of feasible application ideas that
are highly valuable to the organization. In the Digia case, these concepts were applied
to a real industry setting, which, in turn, allowed the researchers to extend the CSC
method with concepts relevant to the case organization.
The researchers used the opportunity at Digia to demonstrate CSC’s feasibility and
efﬁcacy [34, 36]. They started by recruiting and interviewing 32 participants, ap-
proximately evenly divided between experts and potential end users. They conducted
individual structured interviews, using stimuli collected from the subjects ahead of
time. The interview method was intended to encourage participants to focus on the
value of ideas.
70 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
The laddering interviews provided rich data about applications the participants
wanted and why. Using qualitative clustering, data were used to create ﬁve graphical
maps, containing 114 preference and reasoning constructs. The next step was to conduct
an ideation workshop with six business and engineering experts and managers from the
ﬁrm to convert the participant preferences to feasible business project ideas at a “back-
of-the-envelope” level. In the workshop, conducted in isolation in a single ﬁve-hour
stretch, the participants developed three business ideas, with application descriptions,
business models, and interaction tables. These were further developed by analysts in
postworkshop work to be integrated into the ﬁrm’s strategic planning effort.
The CSC method met the project’s objectives. It enabled the researchers to use rich
data collected by a widely representative sample of experts and potential lead users
from outside the ﬁrm, to keep the focus on ideas of potential strategic importance to
the ﬁrm, and to analyze the data in such a way so as to make it useful for IS planning
in the ﬁrm. Digia representatives were enthusiastic about the results of the workshop
. This feedback and the successful implementation of the method in practice en-
abled the project researchers to present initial “proof-of-concept”-level validation of
the new method [34, 36]. The ﬁrm intended to use the resulting applications to plan
its continued product development efforts.
The case study was reported in the Journal of Management Information Systems 
and Information & Management . The structure of these papers closely follows
the nominal sequence of activities presented in the DSRM. In addition, the ﬁndings
were presented in several practitioner-oriented outlets, including a book chapter ,
technical reports (e.g., ), and trade magazine articles (e.g., ).
The artifact developed as a result of this research is a method for IS planning that
can be used to make use of the knowledge of many people from in and around the
organization, maintain focus on potential systems and applications of strategic value to
the ﬁrm, and produce outputs of use to designers and managers in the IS development
process. Consequently, it may be valuable for use in IS planning practice. In research,
the method may be extended to enable planning efforts that focus on the development
of requirements at the feature level, particularly to develop requirements engineering
methods for such contexts as the development of cross-cultural feature sets and for
special populations, such as for disabled persons.
Evaluation of the DSRM Process
WE EVALUATE THE DSRM PROCESS IN TERMS of the three objectives for the DSRM de-
scribed above. First, it should be consistent with prior DS research theory and practice,
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 71
as it has been represented in the IS literature, and with design and DS research, as
it has been conveyed in representative literature in reference disciplines. Second, it
should provide a nominal process for conducting DS research in IS. Third, it should
provide a mental model for the characteristics of research outputs. We will address
each objective below.
First, the DSRM process is consistent with concepts in prior literature about DS
in IS. Because we used a consensus method to design the DSRM, this consistency is
an inherent outcome of the process. For example, Nunamaker et al.’s  ﬁve-step
methodology can be mapped roughly to the DSRM process. Likewise, Walls et al.’s
[55, 56] “components of an information system design theory,” Takeda et al.’s 
“design cycle” solution for intelligent computer-aided design systems, Rossi and Sein’s
 steps, Archer’s  process for industrial design, Eekels and Roozenburg’s 
process for engineering design, and Hevner et al.’s  guidelines for the required
elements of design research are all consistent with the DSRM.
Second, the DSRM provides a nominal process for conducting DS research. In
addition, in the demonstration of four cases, we showed how each of the four DS
research projects described in the cases followed a process consistent with the DSRM.
In addition, the cases demonstrate each of the four research entry points described
in the DSRM, including a problem-centered initiation, an objective-centered initia-
tion, a design- and development-centered initiation, and a client-/context-centered
initiation. In each case, the process worked well, and it was effective for its intended
Third, the DSRM provides a mental model for the presentation of research out-
comes from DS research. The explication of the CATCH data warehouse, reported in
Berndt et al. , incorporated all of the elements of the DSRM process, although it
did not use the DSRM terminology. Rothenberger and Hershauer  followed the
general outlines of the process in the structure of the paper, including a statement of
the problem in the introduction, an explicit “purpose” section to outline the objectives
of a solution, a design section called “creating the measure,” demonstration sections
called “application of the measure to a speciﬁc problem,” and “example case data.”
Peffers et al. used a structure based on Hevner et al.  and consistent with the
DSRM to report the Digia case . Chatterjee et al.  incorporated elements of the
DSRM in presenting the research. The paper identiﬁed the problem and deﬁned the
potential objectives or “beneﬁts” of a solution in the introduction; it also incorporated
the other elements of the DSRM in “Design, Implementation, and Performance of
CGUsipClient” [9, p. 1924].
WE DEFINED DESIGN SCIENCE EARLIER IN THIS PAPER. Recently, however, researchers 
have raised questions about similarities between DS and action research. Both Cole
et al.  and Järvinen  concluded that the similarities between these research
approaches are substantial. Cole et al.  argued that the approaches share im-
portant assumptions regarding ontology, epistemology, and axiology. Järvinen 
pointed to many similarities, although they employ different terminology, and went
72 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
so far as to suggest that we cannot clearly differentiate between them. Perhaps the
clearest distinction between them is found in their conceptual origins. DS research
comes from a history of design as a component of engineering and computer science
research, while action research originates from the concept of the researcher as an
“active participant” in solving practical problems in the course of studying them in
organizational contexts. In DS research, design and the proof of its usefulness is the
central component, whereas in action research, the focus of interest is the organizational
context and the active search for problem solutions therein. Resolution of this point
will have to remain outside the scope of this paper, but it presents an interesting and
perhaps fruitful area for further thought. It would appear that the DSRM could be used
as a structure to present action research. Likewise, the search for a designed artifact
could be presented as action research. Clearly the side-by-side existence of the two
methodologies presents the researcher with choices for the structure of the research
process and the presentation of the resulting solution. This discussion also raises an
interesting question about whether the DSRM could be used in an action research study,
whether researchers could use it to design new innovations based on technical, social,
or informational resources or their combinations , and whether action research
and DS research could be conceptually and methodologically integrated.
The DSRM is intended as a methodology for research; however, one might wonder
whether it might also be used as a methodology for design in practice. There would
appear to be no reason it could not be so used; however, there are elements of the
DSRM that are intended to support essential DS research characteristics that might
not always apply well to design in practice.
A design artifact, such as a curved wooden staircase, a kitchen appliance, or a surgical
knife, is not necessarily required to embody new knowledge that would be conveyed to
an audience through a scientiﬁc publication outlet. Consequently, there is no inherent
requirement that a designer employ any rigorous process to create it. There may, on
the other hand, be organizational, evidentiary, regulatory, or other reasons why some
level of process rigor may be required. The designer of the curved staircase might
be free to work from a simple sketch with a few measurements, while the designer
of the surgical knife might be required to proceed through a careful process of data
collection, consultation, documentation, and testing. Thus, for design in practice, the
DSRM may contain unnecessary elements for some contexts, while being much too
general to support design in others.
An important step in the evaluation of the DSRM was its application to four cases
of previously published DS research. The four studies were chosen in part because
they represent examples of DS research with four different entry points as speciﬁed
in the DSRM. None of the articles in which these studies were reported used the
language of the DSRM to explain its research approach. Instead, they each used ad
hoc arguments to support the validity of the research. We found this to be common in
prior DS research in the IS ﬁeld, because hitherto no generally accepted framework
for conducting and presenting DS research existed, at least not until Hevner et al.’s
 guidelines provided characteristics of good DS research outcomes. Like Hevner
et al. , we have used secondary data, in the form of four cases, to demonstrate
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 73
the application of the DSRM. Results of the analysis of the four cases show that they
are all instances of DS research that can be well framed in terms of the DSRM. Thus,
we used the case discussions as a vehicle not only to evaluate the DSRM but also to
transfer established DS research into a formal research framework and to illustrate
its applicability. We expect that the case studies will provide useful templates for
researchers who want to apply DSRM to their efforts. The development and evalua-
tion of the DSRM was heavily inﬂuenced by design research, thus DSRM concepts
have guided us in the conduct and presentation of this work and this is reﬂected in the
structure of this paper. Clearly, the next step would be to directly adopt the proposed
methodology in new DS research. This is something that we are currently working
on in our ongoing research.
The ad hoc justiﬁcation of prior DS research suggests the difﬁculty that authors
faced, for lack of reference to a commonly accepted DS methodology, in supporting the
validity of DS research in IS. Without a framework that is shared by authors, reviewers,
and editors, DS research runs the danger of being mistaken for poor-quality empirical
research or for practice case study. The DSRM completes a DS research paradigm
with a methodology that is consistent with the DS research processes employed in the
IS discipline, in this way establishing a common framework for future researchers to
validate DS research, without making ad hoc arguments for its validity.
IN THIS PAPER, WE SOUGHT TO DEVELOP A METHODOLOGY for DS research in IS. We wanted
this methodology to be well grounded in existing literature about DS in IS and related
disciplines. In addition, we wanted a methodology that would provide guidance for
researchers who work on DS research and provide a mental model for the presenta-
tion of its outcomes.
Interestingly, other research paradigms have been adapted for use in our discipline
without such a formal deﬁnition. This is hardly surprising because IS research, if one
counts from the tenure of our senior journals, is only about one-third of a century
old. As a result, it was handed behavioral and natural science traditions from much
older research disciplines in the business academe and adopted them without much
adaptation. Consequently, this paper represents a unique effort to formally deﬁne a
research methodology for use in IS.
We should emphasize that this paper represents one general methodological guide-
line for effective DS research. Researchers should by no means draw any inference
that the DSRM is the only appropriate methodology with which to conduct such
research. We can imagine that the efforts of others could result in at least ﬁve other
types of DSRM:
1. A methodology to support curiously motivated DS research, although such
research is not common in business disciplines, might look quite different than
the DSRM. Some research in the social and natural sciences is driven primarily
by curiosity and may therefore lack explicit outcome objectives.
74 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
2. A methodology to support research within a speciﬁc stream in IS might in-
corporate elements speciﬁc to the context of that research. For example, a
methodology to support the design of methods for requirements analysis might
provide guidelines for speciﬁc expected elements of requirements analysis,
including organizational context, data gathering, modeling, and the form of the
requirements speciﬁcation. We observed a number of context-speciﬁc design
research methodologies in engineering.
3. Whereas the motivation for the research is to solve problems in a speciﬁc
organizational context, action research, as suggested in preceding paragraphs,
may be an alternative or complementary paradigm through which to design IS
4. With respect to speciﬁc activities in the research process, future researchers
may enhance the DSRM, for example, by developing subsidiary processes.
5. Finally, circumstances, such as context-speciﬁc constraints, may motivate re-
searchers to develop and implement ad hoc processes that, while inconsistent
with this DSRM, may, nonetheless, be well justiﬁed and produce valid results.
While these ﬁve examples come readily to mind, it seems likely that there are other
ways that DS research could be well done. We present these alternatives here, with-
out recommendation and without knowledge of their prior use, in speculation about
what valid alternatives to the DSRM might be subsequently developed and used. In
doing so, we are suggesting that the DSRM should not be used as a rigid orthodoxy
to criticize work that does not follow it explicitly.
The case studies we provided with this paper demonstrate its use within the scope
of four research problems. Further use will tell us whether there are problem domains
where it requires extension or where it does not work well. Another interesting problem
is that of the research entry point. We demonstrated that there are multiple possible
entry points for DS research. Of course, this issue is not unique to DS research. We do
not recall reading a theory-testing paper where the authors say that they decided on the
research questions after they collected the data or even after they did the analysis, but
we have all observed that this happens with no ill effects. The “scientiﬁc method” is
an espoused theory, approximated but not always matched by theory in use. We think
that a research methodology should account, as far as it is practical, for the research
process in use.
Acknowledgments: The ﬁrst two authors made substantially similar contributions to this paper.
First authorship was determined by rotation among papers.
1. Adams, L., and Courtney, J. Achieving relevance in IS research via the DAGS framework.
In R.H. Sprague Jr. (ed.), Proceedings of the Thirty-Seventh Annual Hawaii International Confer-
ence on System Sciences. Los Alamitos, CA: IEEE Computer Society Press, 2004 (available at
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 75
2. Archer, L.B. Systematic method for designers. In N. Cross (ed.), Developments in Design
Methodology. London: John Wiley, 1984, pp. 57–82.
3. Aumann, H.H.; Chahine, M.T.; Gautier, C.; Goldberg, M.D.; Kalnay, E.; McMillin, L.M.;
Revercomb, H.; Rosenkranz, P.W.; Smith, W.L.; Staelin, D.H.; Strow, L.L.; and Susskind, J.
AIRS/AMSU/HSB on the Aqua mission: Design, science objectives, data products, and process-
ing systems. IEEE Transactions on Geoscience and Remote Sensing, 41, 2 (2003), 253–264.
4. Berndt, D.J.; Hevner, A.R.; and Studnicki, J. Data warehouse dissemination strategies
for community health assessments. Upgrade, 2, 1 (2001), 48–54.
5. Berndt, D.J.; Hevner, A.R.; and Studnicki, J. The CATCH data warehouse: Support for
community health care decision-making. Decision Support Systems, 35, 3 (2002), 367–384.
6. Berndt, D.J.; Hevner, A.R.; and Studnicki, J. Bioterrorism surveillance with real-time data
warehousing. In H.E.A. Chen (ed.), IEEE International Conference on Intelligence and Security
Informatics. Los Alamitos, CA: IEEE Computer Society Press, 2003, pp. 322–335.
7. Berndt, D.J.; Fisher, J.W.; Hevner, A.R.; and Studnicki, J. Healthcare data warehousing
and quality assurance. IEEE Computer, 34, 12 (December 2001), 56–63.
8. Berndt, D.J.; Bhat, S.; Fisher, J.W.; Hevner, A.R.; and Studnicki, J. Data analytics for bio-
terrorism surveillance. In H.E.A. Chen (ed.), IEEE International Conference on Intelligence and
Security Informatics. Los Alamitos, CA: IEEE Computer Society Press, 2004, pp. 17–27.
9. Chatterjee, S.; Tulu, B.; Abhichandani, T.; and Li, H. SIP-based enterprise converged
network for voice/video-over IP: Implementation and evaluation of components. IEEE Journal
on Selected Areas in Communications, 23, 10 (2005), 1921–1933.
10. Cole, R.; Purao, S.; Rossi, M.; and Sein, M.K. Being proactive: Where action research
meets design research. In D. Avison, D. Galletta, and J.I. DeGross (eds.), Twenty-Sixth Inter-
national Conference on Information Systems. Atlanta: Association for Information Systems,
2005, pp. 325–336.
11. Cooper, R.G. Stage-Gate Systems—A new tool for managing new products. Business
Horizons, 33, 3 (1990), 44–54.
12. Cooper, R.G. Winning with new products: Doing it right. IVEY Business Journal, 64, 6
(July–August 2000), 54–60.
13. DMReview. Glossary. SourceMedia, Brookﬁeld, WI, 2007 (available at www.dmreview
14. Eekels, J., and Roozenburg, N.F.M. A methodological comparison of the structures of
scientiﬁc research and engineering design: Their similarities and differences. Design Studies,
12, 4 (1991), 197–203.
15. Evbuonwan, N.F.O.; Sivaloganathan, S.; and Jebb, A. A survey of design philosophies,
models, methods and systems. Proceedings Institute of Mechanical Engineers, 210 (1996),
16. Fulcher, A.J., and Hills, P. Towards a strategic framework for design research. Journal
of Engineering Design, 7, 1 (1996), 183–193.
17. Gemmill, J.; Srinivasan, A.; Lynn, J.; Chatterjee, S.; Tulu, B.; and Abhichandani, T.
Middleware for scalable real-time multimedia communications cyberinfrastructure. Journal
of Internet Technology, 5, 4 (2004), 405–420.
18. Gentner, D., and Stevens, A.S. Mental Models. Hillsdale, NJ: Lawrence Erlbaum,
19. Haran, M.; Karr, A.; Last, M.; Orso, A.; Porter, A.; Sanil, A.; and Fouche, S. Techniques
for classifying executions of deployed software to support software engineering tasks. IEEE
Transactions on Software Engineering, 33, 5 (2007), 287–304.
20. Hevner, A.R.; March, S.T.; and Park, J. Design research in information systems research.
MIS Quarterly, 28, 1 (2004), 75–105.
21. Hickey, A.M., and Davis, A.M. A uniﬁed model of requirements elicitation. Journal of
Management Information Systems, 20, 4 (Spring 2004), 65–84.
22. Hoffman, R.R.; Roesler, A.; and Moon, B.M. What is design in the context of human-
centered computing? IEEE Intelligent Systems, 19, 4 (2004), 89–95.
23. Iivari, J.; Hirschheim, R.; and Klein, H.K. A paradigmatic analysis contrasting informa-
tion systems development approaches and methodologies. Information Systems Research, 9,
2 (1998), 164–193.
76 PEFFERS, TUUNANEN, ROTHENBERGER, AND CHATTERJEE
24. Järvinen, P. Action research is similar to design science. Quality & Quantity, 41, 1 (2007),
25. Johnson-Laird, P., and Byrne, R. A gentle introduction. Mental Models Web-
site, School of Psychology, Trinity College, Dublin, 2000 (available at www.tcd
26. Kelly, G.A. The Psychology of Personal Constructs. New York: W.W. Norton, 1955.
27. Klassi, J. Environmental enhancement of the oceans by increased solar radiation from
space. Oceans, 17 (November 1985), 1290–1295.
28. Krishnamurthy, D.; Rolia, J.A.; and Majumdar, S. A synthetic workload generation tech-
nique for stress testing session-based systems. IEEE Transactions on Software Engineering,
32, 11 (2006), 868–882.
29. Lisetti, C., and LeRouge, C. Affective computing in tele-home health. In R.H. Sprague
Jr. (ed.), Proceedings of the Thirty-Seventh Annual Hawaii International Conference on System
Sciences. Los Alamitos, CA: IEEE Computer Society Press, 2004 (available at http://csdl2
30. Maguire, M. Methods to support human-centered design. International Journal of
Human–Computer Studies, 55, 4 (2001), 587–634.
31. March, S., and Smith, G. Design and natural science research on information technology.
Decision Support Systems, 15, 4 (1995), 251–266.
32. McPhee, K. Design theory and software design. Technical Report, Department of Com-
puting Science, University of Alberta, 1996.
33. Nunamaker, J.F.; Chen, M.; and Purdin, T.D.M. Systems development in information sys-
tems research. Journal of Management Information Systems, 7, 3 (Winter 1990–91), 89–106.
34. Peffers, K., and Tuunanen, T. Planning for IS applications: A practical, information
theoretical method and case study in mobile ﬁnancial services. Information & Management,
42, 3 (2005), 483–501.
35. Peffers, K., and Tuunanen, T. The process of developing new services. In T. Saarinen, M.
Tinnilä, and A. Tseng (eds.), Managing Business in a Multi-Channel World: Success Factors
for E-Business. Hershey, PA: Idea Group, 2005, pp. 281–294.
36. Peffers, K.; Gengler, C.; and Tuunanen, T. Extending critical success factors methodol-
ogy to facilitate broadly participative information systems planning. Journal of Management
Information Systems, 20, 1 (Summer 2003), 51–85.
37. Preston, M., and Mehandjiev, N. A framework for classifying intelligent design theories.
In N. Mehandjiev and P. Brereton (eds.), Proceedings of the 2004 ACM Workshop on Interdis-
ciplinary Software Engineering Research. New York: ACM Press, 2004, pp. 49–54.
38. Reich, Y. The study of design methodology. Journal of Mechanical Design, 117, 2 (1994),
39. Rockart, J.F. Chief executives deﬁne their own data needs. Harvard Business Review,
57, 2 (1979), 81–93.
40. Rossi, M., and Sein, M.K. Design research workshop: A proactive research approach. Paper
presented at the Twenty-Sixth Information Systems Research Seminar in Scandinavia, Informa-
tion Systems Research in Scandinavia Association, Haikko, Finland, August 9–12, 2003.
41. Rothenberger, M.A. Project-level reuse factors: Drivers for variation within software
development environments. Decision Sciences, 34, 1 (2003), 83–106.
42. Rothenberger, M.A., and Hershauer, J.C. A software reuse measure: Monitoring an en-
terprise-level model driven development process. Information & Management, 35, 5 (1999),
43. Simon, H. The Sciences of the Artiﬁcial. Cambridge, MA: MIT Press, 1969.
44. Studnicki, J.; Steverson, B.; Myers, B.; Hevner, A.; and Berndt, D. Comprehensive
assessment for tracking community health (CATCH). Best Practices and Benchmarking in
Healthcare, 2, 5 (1997), 196–207.
45. Swaab, R.I.; Postmes, T.; Neijens, P.; Kiers, M.H.; and Dumay, A.C.M. Multi-party
negotiation support: Visualization and the development of shared mental models. Journal of
Management Information Systems, 19, 1 (Summer 2002), 129–150.
46. Takeda, H.; Veerkamp, P.; Tomiyama, T.; and Yoshikawam, H. Modeling design processes.
AI Magazine, 11, 4 (Winter 1990), 37–48.
A DESIGN SCIENCE RESEARCH METHODOLOGY FOR IS RESEARCH 77
47. Tulu, B.; Chatterjee, S.; and Laxminarayan, S. A taxonomy of telemedicine efforts with
respect to applications, infrastructure, delivery tools, type of setting and purpose. In R.H. Sprague
Jr. (ed.), Proceedings of the Thirty-Eighth Annual Hawaii International Conference on System
Sciences. Los Alamitos, CA: IEEE Computer Society Press, 2005 (available at http://csdl2
48. Tulu, B.; Abhichandani, T.; Chatterjee, S.; and Li, H. Design and development of an
SIP-based video conferencing application. In M.M. Freire, P. Lorenz, and M.M.-O. Lee (eds.),
High Speed Networks and Multimedia Communications. Heidelberg/Berlin: Springer, 2003,
49. Tuunanen, T. Critical success chains method. Technical Report LTT-Tutkimus Oy, Elek-
tronisen Kaupan Instituutti, Helsinki, 2001.
50. Tuunanen, T. Mitä käyttäjät todella haluavat [What do the users really want?]. Tietoviikko
(Helsinki), 2002, 1.
51. Vaishnavi, V., and Kuechler, B. Design research in information systems. Association for
Information Systems, 2005 (available at www.isworld.org/Researchdesign/drisISworld.htm).
52. van Aken, J.E. Management research based on the paradigm of the design sciences: The
quest for ﬁeld-tested and grounded technological rules. Journal of Management Studies, 41,
2 (2004), 219–246.
53. van Aken, J.E. Valid knowledge for the professional design of large and complex design
processes. Design Studies, 26, 4 (2005), 379–404.
54. Vandenbosch, B., and Higgins, C.A. Executive support systems and learning: A model
and empirical test. Journal of Management Information Systems, 12, 2 (Fall 1995), 99–130.
55. Walls, J.; Widmeyer, G.; and El Sawy, O. Building an information system design theory
for vigilant EIS. Information Systems Research, 3, 1 (1992), 36–59.
56. Walls, J.; Widmeyer, G.; and El Sawy, O. Assessing information system design theory
in perspective: How useful was our 1992 initial rendition? Journal of Information Technology
Theory & Application, 6, 2 (2004), 43–58.