The Design Process in Information System Research

Table of Contents

Figures

Tables

1 Introduction
1.1 Motivation and Research Questions
1.2 Structure

2 Theoretical Background of Design Science
2.1 Introducing the Design Concept
2.2 Design Science as an Area of Research
2.2.1 The Separation of Design Science and Natural Science
2.2.2 Theorizing in Design Science
2.3 Establishing a Philosophical Basis for Design Science

3 Developing the Design Science Discipline
3.1 Three Different Design Science Frameworks
3.1.1 Building an Information System Design Theory for Vigilant EIS
3.1.2 Design and Natural Science Research on Information Technology
3.1.3 Design Science in Information System Research

4 Identifying the Core Subject Matter of Design Science in IS research
4.1 The IT Artifact
4.2 The IT Work System

5 Conclusion and Outlook

References

Figures

Figure 1: Knowledge Building and Knowledge Using Process (Owen 1997)

Figure 2: The Outputs of Design Research (Purao 2002)

Figure 3: The Research Paradigm Concept (own Diagram following Gregg et al. (2001))

Figure 4: Design Science Research Framework (March and Smith 1995)

Figure 5: Relation between Research Outputs (own Diagram following March and Smith (1995))

Figure 6: Organizational Design and Information System Design Activities (Hevner et al 2004) .31

Figure 7: Information System Research Framework (Hevner et al. 2004)

Figure 8: Information System Research Cycle (Becker et al. 2007)

Figure 9: IT Artifact and Its Immediate Nomological Net (Benbasat and Zmud 2003)

Figure 10: The Work System Framework (Alter 2003)

Figure 11: Possible Focus Areas in IS Research (adopted from Benbasat and Zmud 2003)

Tables

Table 1: Major Research Paradigms and the Characteristics of their Major Beliefs (Gregg et al. 2001)

Table 2: Components of Information System Design Theory (Walls et al. 1992)

Table 3: The Use of Essential Elements in the Framework by Walls et al. (1992)

Table 4: The Use of Essential Elements in the Framework by March and Smith (1995)

Table 5: Design Science Research Guidelines (own Table following Hevner et al. (2004)

Table 6: The Use of Essential Elements in the Framework by Hevner et al. (2004)

1 Introduction

1.1 Motivation and Research Questions

The Information Systems (IS) research discipline is undergoing a serious identity crisis, seeking its sphere of activity to be relevant in practice and rigorous in scientific considerations. One reason for this is the strengthening of the Design Science approach. This new discipline developed as a synergy from aspects of engineering, architecture, and industrial design and is employed in the design of IT artifacts and software systems. Design Science is becoming a powerful trend in IS research (Vahidov 2006). It gives the IS discipline a new and more detailed focal point as pertains to the application of software and IT artifact development which is growing in importance in IS research over the time (Weber 2003; Orlikowski and Iacono 2001; Cross 2001). IS practitioners ask for new and innovative design approaches, dealing with the evolving organizational and inter-organizational tasks. The way these tasks are executed, in close cooperation with the practical business world, seems to be insufficiently considered. The debate in IS research is carried out between traditional scientists and the knowledge-producing researchers/practitioners and “it could be argued that research aimed at developing IT systems, at improving IT practice, has been more successful and important than traditional scientific attempts to understand it” (March and Smith 1995, p. 252).

IS researchers are mainly focused on the behavioral impact of new IT solutions within a business unit. These concepts are needed to describe the relationship between the humans and the technology. However, this way of conducting research is descriptive and evaluative. Instead of telling “what is” or “what will be”, Design Science is giving guidance as to “how to do” things (Walls et al. 1992). The importance of this new approach is given through the rapid development of business needs and the increased necessity to solve business problems through the implementation of IT solutions. The knowledge base for designing new solutions has not yet been fully developed. IT consultants borrow knowledge from reference disciplines and apply this knowledge to present problems. This way of providing solutions is not compatible with Design Science as an area of research. A relevant design approach needs to give new answers to phenomena thus far unsolved.

However, the IS discipline has not yet established a solid groundwork for Design Science within its discipline. The Design Scientist needs to develop his/her own routines and competencies in form of standards to establish a framework, which can serve boundaries of and describe the ties within their discipline.

A philosophical underpinning is required to align the researchers of Design Science to the same route. Design Science needs to be effective and efficient (Hevner et al. 2004) to deal with the tasks that lie before it. Instead of applying available knowledge, Design Science has to establish itself as a pure research discipline (Weber 1987).

The IS discipline is an amalgamation of different and complex research areas that cannot be easily divided into single parts without losing their identities and the practical use of this discipline. However, scholars of the IS discipline are still in search for the core subject matter of IS research to be able to overcome the current identity crisis (e.g. Orlikowski and Iacono 2001; Alter 2003; Weber 2003).

The Design Scientists have developed various concepts and frameworks to describe the relevance of their discipline, which are described in the following chapters to give an insight in one of the research areas of IS research.

To develop this insight this paper investigates the following research questions:

1. Does the Design Science approach need to be linked to laws and areas of research of the Natural Sciences to be able to produce relevant outcomes?

This question is posed because a pure research discipline needs to develop clear boarders of the discipline. On the one hand this separation can be useful to focus on the core concept of Design Science, but on the other a separation can lead to irrelevant outcomes that lack realistic reference to the organizational contexts.

2. What should be regarded as the core subject matter of the Design Science discipline and why?

The core subject matter defines of each scientific discipline. The design approach risks focusing too much on a single element within the IS research cycle. Without incorporating various objects and possible questions, an IT artifact is unable to solve a class of problems. The core subject matter is needed as fixture with which to adjust the discipline. However, Design Science has not yet defined its core subject matter. Different approaches need to be analyzed.

3. Do current approaches to developing important Design Science frameworks focus on a core subject matter?

Finally this paper analyzes which core subject matter has been applied in the academic papers most often cited in Design Science history and how the discipline could possibly develop in the future.

1.2 Structure

This paper gives insight into the development of Design Science – one special part of IS research that has gained a lot of relevance within the last decades (e.g. Hevner et al. 2004; Walls et al. 1992, March and Smith 1995).

Chapter 2 highlights the relationship between Design Science and the Natural Sciences by regarding the historical development of design from one of the earliest Design Science definitions presented in ‘The Sciences of the Artificial’ (Simon 1996) through to the use of design in the IS world today. The chapter attempts to answer the question whether the Natural Sciences and Design Science can be regarded separately or only in combination within IS research? In closing Chapter 2 briefly examines the different philosophical approaches of Design Science and incorporates the question if Design Science is a paradigm itself.

Chapter 3 addresses the different Design Science concepts in academic IS literature. The considered frameworks construct different Design Science concepts that are interdependent in their combined view of design in IS research. The focus of this chapter lies on the use of design to create efficient and effective IT artifacts (Hevner 2004). The question is posed as to whether suggested frameworks see Design Science in IS as a separate and pure area of research. The role of the laws and approaches of the Natural Sciences in Design Science are also explored. The conclusion of this chapter presents different views of Design Science are have been offered by other scholars to give a more complete picture of the existing academic literature.

Chapter 4 examines the essential role of the core subject matter of the Design Science discipline in its defining and overcoming its identity crisis. The question is posed whether the IT artifact should be regarded as the basic research area in Design Science. The mostly cited academic articles in this area give an understanding of the possible use of the IT artifact within their frameworks. It is shown how the frameworks introduced in chapter 3 treat the IT artifact in their research projects. To give a brief contrast the idea of the Work System as core subject matter of Design Science is demonstrated in this chapter accompanied with the criticism of this approach.

Chapter 5 gives a conclusion of the paper and an outlook of a possible progress of Design Science in IS research.

2 Theoretical Background of Design Science

This chapter gives a brief introduction of the theoretical background of Design Science. The basic concept of design that emerged over time is described in section 2.1. This chapter gives also an insight into the common use of design and Design Science in the IS research discipline. Section 2.2 questions whether Design Science and Natural Science need to be connected or separated in order to pursue their aims of producing new knowledge and/or describing knowledge. The background for finding an appropriate answer of this question is deciding whether, and if so, under which conditions design can be regarded as a unique science discipline. The chapter compares the requirements of a science discipline with the characteristics of Design Science. This chapter focuses on the role of theory and theorizing in Design Science. Section 2.3 contrasts the different approaches of philosophical paradigms that can describe Design Science and poses the questions as to whether Design Science is a paradigm itself.

2.1 Introducing the Design Concept

Design is the process of changing present situations into preferred ones. Design creates something new that does not yet exist in nature and for this reason design is not comparable to the Natural Sciences that describe the objects and phenomena of the natural world (Simon 1969; Vaishna and Kuechler 2004). The study of designing things has long been a fundamental part of art, engineering, architecture and other disciplines in the business and industrial sectors. Design plays a key role in these areas by distinguishing the practical and professional side from the sciences of this particular discipline (Simon 1969). Only the artificial processes and products in the form of newly arranged artifacts can achieve the desired outcome in the practical world.

One major task in the development of the design field was the construction of a joint view of design and science to establish a theoretical foundation for design practices. The idea of design using parts of the traditional science disciplines has continuously emerged during the early decades of the 20thcentury. This resulted in the use of design based on scientific knowledge in industrial sectors, i.e. the so called “scientific design”. It was not yet a synonym for Design Science but already an application of modern design practice (Cross 2001).

One highlight of this new way considering design was the book “The Sciences of the Artificial” by Herbert Simon published in 1969 (Simon 1969). Simon was one of the first to give the “Science of Design” a basis and definition, which will be considered at a later stage in this paper.

Some authors distinguish between two ways of looking at design in a scientific context by describing “Science of Design” as the study and the understanding of design and “Design Science” as a systematic approach to design that leads to a scientific design activity itself (Cross 2001). This might be important when arguing that the act of designing can never be a scientific activity itself in contrast to the study of design that can be subject of scientific investigation (Grant 1979).

In this paper the expressions “Design Science” and “Science of Design” will be used synonymously due to the fact that the important academic articles use Design Science in a scientific way without varying in their view of design as a scientific discipline (e.g. Hevner et al. 2004; Nunamaker et al. 1991; Walls et al. 1992). The idea of Design Science as a scientific activity will be further developed in this paper.

The IS discipline is strongly interrelated to design and Design Science due to its historical origins with a key focus on information and communication technology in organizational settings (Kuechler et al. 2007). The IS discipline is confronted with special requirements of the professional and practical business areas and must align these requirements in connection with the aspects of Behavioral Science that characterize the internal organizational structures.

Technology is practical and useful, therefore, IS practitioners and IS users are involved in design to improve the effectiveness and efficiency of their organization (March and Smith 1996; Walls et al. 1992; Hevner et al. 2004). However, the IS discipline uses various methodologies in a complex setting. Design Science displays the prescriptive component that can improve, for example, IT performance as a knowledge-using activity (March and Smith 1995). In the Computer Science discipline every machine is recognized as an artifact that has been designed to describe an experiment. Experiments can produce new knowledge by demonstrating the possible implementation of designed artifacts into the real world. This all is realized by using Design Science (Newell and Simon 1976).

This shows that Design Science is primarily a problem-solving domain leading to an innovative way of implementing and managing IS efficiently (Hevner et al. 2004). Benbasat and Zmud (1999) show that this problem-solving criterion is one of the core tasks of IS. They argue that an information system is irrelevant unless IS professionals can use its knowledge and to create practical value through implementation. The challenges for Design Science in information systems grows, as on the one hand IS professionals ask for further adaption of business and organizational knowledge into IT artifacts and on the other hand the needs and the expectations of newly designed knowledge for their business areas grows continuously.

2.2 Design Science as an Area of Research

The newly emerged Design Science discipline was developed on the basis of scientific knowledge that was used to develop innovative design products in an industrial context (Cross 2001). The available scientific knowledge was taken from the well-established Natural Sciences to substantiate the usefulness of the designed elements and to fill in for the missing theoretical foundation of designing elements. From this starting position from the “Scientific Design” the question is asked whether Design Science needs to be fully separated from the Natural Sciences in order to establish a strong and independent scientific discipline and whether an influential position of the Natural Sciences and their laws is essential to produce relevant new knowledge.

To give an answer to these important questions it is necessary to discuss the characteristics and the differences of Design Science and the Natural Sciences.

2.2.1 The Separation of Design Science and Natural Science

This section is addressing the first research question mentioned in the introduction. The characteristics of Design Science and Natural Sciences developed over time. This paper tries to find an answer and highlights already existing answers of whether both approaches need to be regarded separately or not.

It is a common opinion of IS scholars that the Natural Sciences are regarded as the descriptive way of looking at the world and its phenomena (e.g Hevner at al 2004; March and Smith 1995; Walls et al. 1992; Vahidov 2006). Assigned to the IT discipline the Natural Sciences form the understanding of the nature of IT. Although IT phenomena, mainly in the Computer Science discipline, are man-made and therefore not natural, the Natural Sciences can address these artificial phenomena by analyzing their “nature” of being (March and Smith 1995). This understanding consists of several approaches to produce scientific knowledge. Natural Science research methods are normally composed of a developmental part and a justifying part (Hevner et al. 2004). Through the developmental part, the scientists attempt to characterize naturally occurring phenomena with the help of special sets of concepts and ways of describing the natural reality (March and Smith 1995). This side of the Natural Sciences concludes with the formulation of scientific theories and the proposition of new natural laws. These theories and laws form an understanding of the explored phenomena, which is not yet fully described in the terms demanded by the Natural Sciences. The missing part is the explaining or justification part of analyzing phenomena (March and Smith 1995; Hevner et al. 2004). The testing of theory plays a key role in the Natural Sciences (Venable 2006) as only the tested, justified, and validated theory or law can be included into the knowledge base of the accordant discipline.

One discipline of the Natural Sciences that should be regarded in this context is the discipline of Behavioral Science. Behavioral Science should not be seen in contrast to the broader view of the Natural Sciences that include e.g. physics, biology, and the social sciences; in fact it is one of the specialized disciplines that is closely related to IS research. Simon (1969) argues that many disciplines tend to focus on Behavioral Science and not on Design Science, this is also true for IS research (Carlsson 2006). IS mainly deals with the interaction between people, technology, and organizations and attempts to describe their interdependencies (Hevner et al. 2004).

In contrast to the descriptive and explanatory character of Natural Science, the Design Science approach plays a prescriptive part of the scientific interest in IT (March and Smith 1995). Instead of trying to understand the existing phenomena with the use of scientific instruments, Design Science gives an answer to heretofore unsolved organizational problems. The design approach in the sense of IS research being aimed at broadening the possibilities for implementing of the use of IT in organizational contexts. It is therefore a problem-solving discipline (Simon 1969; Hevner et al. 2004). Technical knowledge grows rapidly, and the need for IT solutions for new application areas grows continuously (Markus et al. 2002). Design Science gives an answer through developing knowledge that had been non-existent up to this point. This knowledge is not only represented in theories or the way defining it, it is also a process of developing innovative artifacts that comply with the ideas, practices, technical capabilities, and products that are needed for the special requirements of IS in form of analysis, design, and implementation (Hevner and March 2003). Design Science is “a body of intellectually tough, analytic, partly formalizable, partly empirical, teachable doctrine about the design process” (Simon 1969 p.113). In other words, the research in Design Science is the design process and the way of building a new artifact is similar to building a new theory in the Natural Sciences. One major task in the design process is the description of the desired organizational capabilities that need to be provided by the artifact. Without the identification of the organizational problem, the design process is not able to address it effectively (March and Storey 2008).

Beside this design process, Design Science always produces a physical artifact. This product is essential for the evaluation process that is necessary to give a proof of feasibility of the (theoretical) constructed artifact in its organizational context (Hevner et al. 2004). The implementation evaluates the effectiveness and efficiency of the artifact in its desired environment, as we need to recognize that the business organization, which forms the environment in most of the cases, is an artifact itself designed and constructed to achieve specific purposes of the business world (Srinivasan et al. 2005). Each organizational framework differs from each other with implications for the design process. It is not possible to set up a general law or theory that describes the operational sequences within a business unit. The hypotheses and desired specifications can be proven only by constructing the artifact (Walls et al. 1992). The designed artifact needs to be implemented to give a concrete assessment of the artifact’s suitability to its intended purpose (Hevner and March 2003).

This necessity supports the idea of Design Science being an area of research with the characteristics “build” and “evaluate” (March and Smith 1995). Research is normally associated with the idea of generating knowledge that can be transferred to general settings (Gregg et al. 2001). The design process needs to demonstrate that no other already existing IT artifact is able to present a solution to the problem at hand. The new IT artifact should not be an adaption of existing knowledge into the design process; instead it has to be the development and creation of new knowledge that will be included into the knowledge base to demonstrate a research process in IS research. Scientific research is defined as the “use of scientific knowledge directed toward the production of useful material, devices, systems, or methods, including design and development of prototypes and processes” (Blake 1978 cited in Gregg et al. 2001).

The IT artifact can be subject of further evaluation to give a better understanding of the value the IT artifact has added to the knowledge base and to the organization. This way of evaluating should not be confused with the Natural Sciences activity of explaining how and why an artifact works (March and Storey 2008; March and Smith 1995).

Figure 1 shows how the knowledge base is connected with the design process (Owen 1997). The design process is given through the channel on the knowledge using side. Through a structured process (for example one of the Design Science frameworks that are introduced later in this paper) available IT knowledge is used to set up the basic model of each IT artifact. As we noticed before, the design process uses an immense amount of new knowledge that is produced within the design process; however, some basic ideas and structures are taken out of the existing knowledge base. The produced IT artifact can be seen as the “works” depicted in Figure 1. The artifact itself is a piece of knowledge, a proof of concept in the organizational framework (Hevner et al. 2004). The evaluation process of the innovative artifact within the Design Science discipline makes it possible to transfer the new knowledge back into the knowledge base through a structured channel defined by Design Science frameworks.

illustration not visible in this excerpt

Figure 1: Knowledge Building and Knowledge Using Process (Owen 1997)

The channels on both sides can also be influenced by other disciplines. As we noted before, Behavioral Science aspects play a major role in the justification of Design Science in IS research. Without this context, Design Science is not able to produce artifacts that make business organizations more effective and efficient (Hevner and March 2003).

Simon (1969) describes the “artificial sciences” as constrained from two directions at simultaneously: (1) the outer environment and (2) the inner environment. The outer environment is given in form of parameters, in this case through the organizational composition in which the artifact has to operate. The other dimension is the inner environment, the organization of the artifact itself. “The bringing-to-be of an artifact, components and their organization, which interfaces in a desired manner with its outer environment, is the design activity” (Vaishna and Kuechler 2004 p. 2). This basic design definition makes clear that the design process is constrained through natural laws (i.e. technological, organizational, and behavioral) of its outer environment; therefore, it is necessary to understand these laws to create an artifact. However, occasionally the laws constraining the design results are not fully understood, this is the reason why implementations are necessary in order to gain insight into the laws affecting the possible reaction of the artifact in its provided environment (Srinivasan et al. 2005).

Each IT artifact contributes to the IT knowledge base and can be subject of the evaluation of Natural Science disciplines. Design Science informs e.g. the Behavioral Science discipline with its final act of implementing a new artifact in a business organization. The artifact becomes the focus of expanded and continuing research (Nunamaker et al. 1991). Behavioral Science is able to recognize why the artifact works and can challenge the design discipline to build more effective artifacts, this constitutes progress in IS research.

A possible answer to the posed research question is that the design process is able to create bridges between the human processes and the technological capabilities of IS research (Gregg et al. 2001). It is not useful to isolate both science approaches of Design Science and the Natural Sciences as otherwise the described bridge is not able to connect the strength of both disciplines. The outcomes of the joint view are able to solve relevant problems in the real business world without sacrificing the traditional science criteria. However, it is not possible to give a binding answer to this question as it is constrained by the focus point each scholar sets for his or her own research. This focus point is connected with the understanding of the core subject matter of IS research. This question will be analyzed in chapter 4.

To be able to install the obtained knowledge out of the design process into the knowledgebase, a theoretical foundation must be set up to provide a background for further scientific research in the analyzed area. The question needs to be answered whether Design Science must develop a strong theoretical foundation within the discipline itself or if the theoretical foundation laid by the Natural Sciences suffices.

2.2.2 Theorizing in Design Science

This section provides brief insight into the use of theory and theorizing in Design Science discipline. Several authors have addressed the role of theory in Design Science. Although some authors exclude theorizing from Design Science and regulating it rather to the Natural Sciences (e.g. March and Smith 1995, Hevner et al. 2004), this paper recognizes the importance of theory building in the design process. The possibilities of theorizing in Design Science discipline are highlighted to focus on an area that is part of the overall Design Science and IS research discussion. This focus also contributes to the demand of Design Science being an area of research because theory plays a key role in the traditional Natural Sciences. The use of theory helps to distinguish between research and other efforts such as consulting or programming (Gregg et al. 2001). However, this section seeks not to forestall special objectives of the Design Science frameworks covered in Chapter 3. Instead, general insight into the role of theory is the basis for this analysis.

This paper already discussed the physical artifact as the key element of each Design Science process. The prescriptive nature of Design Science pretends to give answers to real world phenomena and problems, posed mainly by the business world. Therefore, the necessary relevance is provided by the proof-of-demonstration through the implementation of IT artifacts that are designed to provide a solution for the stated problem. The artifact remains the most visible output of Design Science, but it is not necessarily the only output of Design Science research processes (Purao 2002).

Nunamaker et al. (1991) has already noted that Design Science (here: Systems Development) is in need for a multimethodological approach where the building of an artifact is only one part of the research cycle. Figure 2 demonstrates one possible conception of the outputs of Design Science described by Purao (2002). Beside the artifact as an output of the design process and as the commonly used design element in an organizational context, two other levels are introduced in this framework. The focus in this section lies on the dotted oval of Figure 2 that displays theory as a potential output of the Design Science process. Knowledge as operational principle is outside the scope of this discussion (not all three forms will be produced in every design process) (Purao 2002).

[...]