CHAPTER 3: RESEARCH DESIGN
Research is defined as the creation of new knowledge by making use of an appropriate process in order to satisfy the needs of the users of the research (Oates, 2006). The previous chapter discussed the problem of the implicit common ground in the understanding of EA by practitioners and researchers. The EA practitioner and researcher represent the two groups of users of EA research. The diversity in the knowledge needs of these groups, namely the practical needs of the practitioner and the academic discourse of the researcher, makes the creation of new knowledge by way of an appropriate process, a complex undertaking. Design science research (DSR) provides an approach to address this complexity, due to its focus on creating new knowledge as a result of building artefacts (Vaishnavi & Kuechler, 2013). The benefit to the practitioner resides in the artefact itself, while the EA researcher benefits from the new knowledge that is discovered during the design process. Because of these benefits, DSR was selected as the methodology to create artefacts as solutions for the research question posed in Chapter
1. The aspect of what constitutes new knowledge is described in the context of a research paradigm where the philosophy of the research effort is understood and categorised. The research paradigm for this thesis falls under the category of a subtle realist position. Guided by the research paradigm, a DSR methodology is employed to design, implement and demonstrate the EAt and SIM conceptual artefacts.
The chapter is structured in two parts, namely a discussion of the components of the research strategy (section 3.2) and the details of the research plan (section 3.3). The subtle realist paradigm that guides the research strategy is discussed in the first part, whereas the DSR methodology is discussed in the second part.
THE COMPONENTS OF A RESEARCH STRATEGY
A research design describes the researcher’s plan for addressing a research problem. Creswell (2003) proposes a research design framework that contains the key components of a research design, namely the elements of inquiry, approaches to research and the research design processes. This framework, illustrated in Figure 3.1, shows the placement of the key elements of inquiry, namely knowledge claims, strategies of inquiry and research methods, as part of the researcher’s conceptualisation process. The elements of inquiry determined the researcher’s approach to research, in terms of the decisions that will influence the execution of the research plan.
According to Crotty (1998), the components of a research design follow a logical sequence (Figure 3.2). An epistemology informs a theoretical perspective that supports a research methodology that, in turn, contains a method or number of methods.
Each component is further explained by Crotty (1998) as follows:
• Epistemology is the theory of knowledge, and serves as the philosophical context from which the validity of knowledge claims is justified.
• The theoretical perspective represents the philosophical stance of the researcher, and informs the methodology that will be followed while doing the research.
• The methodology is the strategy of the research, and impacts on the appropriate choice of research methods.
• Methods are the specific techniques or procedures used to gather and analyse the data for the research project.
The quality of the research results is dependent on the unity between these four components (Figure 3.1). The theoretical perspective should be based on an appropriate epistemology for the researcher’s knowledge claims to be valid. The chosen research methods should, in turn, align with the research methodology as well as the theoretical perspective, in order to guarantee useful scientific results. The relationship between the elements of inquiry in Creswell’s framework and the research plan, is illustrated by mapping the components of a research design as described by Crotty (1998) (Figure 3.3), to the elements of inquiry described by Creswell (2003). The result of the mapping forms the basis of the research design, and is shown in Figure 3.3:
In keeping with the definition of research, as put forth by Oates (2006), the new knowledge produced by the research is grounded on a philosophical foundation that includes an epistemological and theoretical perspective, while the appropriate research process is designated by the specific research process followed by the researcher (Figure 3.4). The sections that follow discuss the philosophical foundation of the research design used to answer the research question.
The Philosophical Foundation
A philosophical foundation consists of an epistemological as well as a theoretical perspective (Figure 3.4), and is understood as a research paradigm. Guba (1990) defines a paradigm as a basic set of beliefs that guides action. A paradigm includes the researcher’s epistemological, ontological and methodological position. Epistemology is described as the study of human knowledge (Mouton & Marais, 1988), and consists of sentences that claim knowledge about reality. Reality, in turn, as far as it speaks of a world and the things in it, is described by way of an ontology (Scott & Marshall, 2009). Hevner et al. (2004) present design science research (DSR) as an appropriate research paradigm for information systems research, because of the use of artefacts in solving problems. Iivari (2007) proposes Karl Popper’s three-world model as a basis to construct an ontology for DSR research. In Iivari’s ontology, information technology (IT) is viewed as the core artefact of information systems research. In Popper’s three worlds model (1978), World 1 consists of physical bodies, and is called the physical world (another way to view this world is to see it as the world of everyday existence); World 2 is called the mental (or psychological) world, and contains subjective human experiences; and, finally, World 3 contains the products of the human mind. The DSR-oriented ontological perspective proposed by Iivari is summarised in Table 3.1:
March and Smith (1995) created a general typology of IT design artefacts that includes constructs, models, methods and instantiations. Better theories are listed as a fifth artefact in the list of design research artefacts by Vaishnavi and Kuechler (2013). Table 3.2 shows a description of each artefact type:
Iivari (2007) also proposes an epistemology for DSR that includes three types of knowledge, namely conceptual, descriptive and prescriptive. Table 3.3 shows a summary of this epistemology, with each type of knowledge’s associated research goal:
Niehaves (2007) argues in favour of a pluralistic approach to using positivist and interpretivist epistemologies for design science. The position taken in this research thesis is based on Niehaves’s standpoint of a paradigm that a) accepts a reality outside the realms of human cognition (positivist ontology), and b) regards knowledge as determined by the subject (interpretivist epistemology) – thus accepting that objective knowledge of the world is not possible. This paradigmatic position is descried by Ritchie and Lewis (2003) as a subtle realist position.
As a theoretical perspective, the science of the artificial (Simon, 1996) frames the understanding of artificial things (artefacts) as a ‘knowing by making’, and is designated as foundational to the field of design science research (Baskerville et al., 2011). Design in this context is defined as the act of creating something new that does not exist in nature (Vaishnavi & Kuechler, 2013). DSR’s defining feature is learning through building or learning though artefact construction, by making use of design as a research method (Vaishnavi & Kuechler, 2013). The knowledge that is the result of this learning process, relates to the process of making the artefact, as well as understanding the effectiveness of the artefact in adhering to the purpose it is designed to achieve. Hevner and Chatterjee (2010) capture this relationship between knowledge, design and artefact as the fundamental principle of design – namely knowledge and understanding of a design problem and its solution are acquired in the building and application of an artefact.
The interpretivist epistemology, as illustrated in Figure 3.5, informs a science of the artificial theoretical perspective that is situated in Design Science Research. The knowledge claim of this research thesis is therefore an understanding (interpretation) of what can be learned by making and using the artefact (Science of the Artificial), where the artefact is understood to be the solution to the research question (section 1.4.2) posed in this thesis.
The Research Process
The purpose of the research process is to actively engage in the execution of the research, and includes a methodology and methods (Figure 3.5). Methodology is defined as the science of method, where method is understood as the procedure that is followed to gain knowledge (Wyssusek et al., 2003). For the purpose of this chapter, the specific meaning of methodology is based on the definition by Mingers (2001), and is taken to mean the actual research method or methods used in a certain piece of research. The aspects of a research project that a methodology should address and describe, are the identification of appropriate data sources, the collection and analysis of data, and the justification of conclusions based on this analysis. The specifics of each research method in turn address aspects such as the collection of data by way of interviews, or the interpretation of results from a data-gathering exercise. A methodology may therefore contain and combine multiple methods or parts of methods. The most important role of the methodology is that it should make logical sense, and stay true to the philosophical principles underlying each method and, ultimately, the study as a whole. In accordance with the artificial nature of EA, the research process is based on the design science research (DSR) model described by Vaishnavi and Kuechler (2013). The DSR model consists of five steps – namely awareness of problem, suggestion, development, evaluation and conclusion. The meaning of each step is briefly discussed as follows:
• Awareness of problem:
In this step, an interesting problem is identified for potential solution, by developing an artefact. The awareness of the problem is documented in a proposal for a new research effort.
During this step, a tentative design for a solution is suggested. The suggestion step is essentially creative in nature, due to the task of the researcher to envision a possible solution to the problem stated in the research proposal.
During the development, the tentative design of the suggestion step is enhanced and implemented. A range of theories can be used to inform the design of the solution.
The developed solution is evaluated against a set of criteria, to test its ability to solve the problem. Any deviation in expected performance of the solution is noted and explained. The results of the evaluation step constitute the lessons learned, and can lead to more cycles of the DSR model.
The final step in the process is used to consolidate and communicate the results of the DSR cycle. The knowledge acquired during the execution of the DSR process informs the creation of design theories that can be used in further DSR cycles. Figure 3.6 illustrates the complete DSR model, and indicates the knowledge flows as well as cognitive processes.
The core of the DSR model consists of five process steps which produce outputs and consume resources during the execution of a design cycle. The process steps of a design cycle proceed from Step 1 through the rest of the steps to Step 5, in a sequential manner. Each of the process steps in a design cycle produces outputs and consumes resources. The development (Step 3) and evaluation (Step 4) produce knowledge according to a circumscription process. The circumscription process is described by Vaishnavi and Kuechler (2013) as a formal logical method that assumes the validity of knowledge fragments as part of specific situations. Furthermore, the applicability of knowledge can only be determined by the analysis of contradictions (Vaishnavi & Kuechler, 2013). The usefulness of circumscription to the DSR researcher, is that learning takes place when something does not work according to theory (see section 2.5.3 for an example of the circumscription process). Circumscription allows the researcher to learn by making – a process that is in keeping with the research process described in Figure 3.5.
Resources consumed by each activity in the design cycle can be knowledge, data, a theory or an artefact. Activity 1, for example, consumes knowledge about the state of a problem, as well as the importance of a solution to a research problem. As illustrated in Figure 3.7, the design activities mostly consume knowledge. When considering the link between data and knowledge, it is possible to establish the data collection and verification methods, as well as the data sources, for each resource in the design cycle (Figure 3.8). The knowledge about the state of the problem was produced by a literature review (section 2.3), while the selection of a proposed solution was achieved by way of an argument (section 2.5.2). The data sources used to prepare the literature review in section 2.3 included the EA literature associated with the research problem. The evaluation results of existing artefacts, such as, for example, the SIM (Table 4.11) were used as a data source to complete the EAt artefact. Figure 3.7 and Figure 3.8 are combined to produce a research map that will chart the process of the execution of the research plan (Figure 3.9):
THE RESEARCH PLAN
In this section, the research design framework (section 3.2) is used to create a detailed research plan in a sequence of three steps. During the first step, the research goals are specified in order to facilitate the development (Step 3 in Figure 3.7) of the proposed solution (section 2.5.2). The second step discusses the execution of the five process steps in the DSR model, by emphasising the outputs of each activity, as well as the resources needed to successfully complete an activity. The third and final step consists of a discussion about assumptions, data sources and data analysis techniques needed to produce the resources for each activity. The section concludes with a detailed map (Figure 3.15 and Figure 3.16) of the research plan, to aid the navigation of the research process.
Step 1: Research Goals and Objectives
Three research goals were stated in Chapter 1 (section 1.4.1), and are repeated here for the sake of clarity in the research design discussion:
1. Describe the theoretical background of EA research in terms of EA definition efforts and the difficulties with EAF selection.
2. Determine an understanding of EA by interpreting the key works of three prominent EAFs.
3. Construct an enterprise architectonic (EAt) to structure the understanding of EA, in terms of fundamental concepts and their relationships.
The first research goal constitutes the theoretical background to the research problem, and culminates in creating awareness of the research problem (section 2.5.1). The second and third goals address the proposed solutions to the identified research problem (section 2.5.2). According to the activities in the research design cycle (Figure 3.7), the first goal is associated with the first and second activities. The discussion in Chapter 1 (section 1.8.1 and section 1.8.2) and Chapter 2 (section 2.5.2) proposed two artefacts (SIM and EAt) to address the research problem. Each of the artefacts is associated with a separate and unique research goal, as follows:
1. SIM: Determine an understanding of EA by interpreting the key works of three prominent EAFs.
2. EAt: Structure the understanding of EA in terms of fundamental concepts and their relationships.
The SIM artefact is a method type artefact as described by March and Smith (1995), and the EAt is a conceptual artefact as described by Bereiter (2002). Both the SIM and the EAt take into account that EA exists as a concept in the everyday world of human activity. In other words, EA is not understood as a physical object. In ontological terms, the existence of EA is expressed as a concept created by the human mind towards a purpose that is situated in, and determined by, the needs of the business world. The business world (also referred to as the enterprise) is understood as the external reality, described by Popper (1978) as World 1 (Table 3.4).
The knowledge that results from the demonstration of the SIM and EAt artefacts falls in the categories of conceptual and descriptive, as described by Iivari (2007) and summarised in Table 3.4:
The SIM produces conceptual knowledge, in that the meaning of EA contains concepts and constructs. The EAt, in turn, produces descriptive knowledge, by describing how the understanding of enterprise architecture can be structured.
Step 2: Execution of the Research Plan
The research problem constitutes the research context for the design of both the SIM and the EAt. The EAt artefact is designed to address the problem of the implicit foundations of EA, by transforming an understanding of EA (result of SIM demonstration) into a set of concepts and relationships (section 2.7). The SIM is designed to interpret the key writings of three prominent EAFs (section 2.4) to produce an interpreted understanding of EA. The EAt and SIM artefacts therefore stand in relation to each other in the sense that the EAt’s design and development (Step 3 of DSR model) (Figure 3.7) cannot be completed unless the SIM’s demonstration has first produced an understanding of EA (Figure 3.10).
The SIM’s results are therefore a prerequisite for the EAt’s design and implementation. The execution of the research plan is an application of the DSR model proposed by Vaishnavi and Kuechler (2013) (section 3.2.2), to the task of first designing, developing and evaluating the SIM – after which the EAt can be designed with the SIM’s evaluation results as input.
SIM Design Cycle
The design problem that drives the design of the SIM states that the unresolved discussions of EA terms and definitions show a lack of universal acceptance of a general understanding of EA. Two sub-problems are derived from the SIM’s design problem:
1. The understanding of EA is undecided and assumed.
2. The understanding of EA is localised in the EAF.
In addressing the SIM design problem, two objectives are formulated to guide the design of the SIM:
1. Construct a structured approach to interpret an understanding of EA.
2. Capture the understanding of EA in a set of EA propositions.
The resources needed to complete Step 1 (Figure 3.7) are knowledge of the state of the problem, as well as the importance of a solution. Step 2 (Figure 3.7), in turn, needs knowledge about the state of the problem, as well as existing solutions. During Step 3, the artefact (the SIM) is developed. The resources needed for the third step in the cycle are knowledge of theory that can be used to inform the design, and development of the artefact. In the case of the SIM, the theory of interpretation and understanding is needed. The artefact is evaluated in Step 4, and for this step a comprehensive set of instructions on how to use the SIM, is needed. The SIM design cycle ends after the fourth step, as shown in Figure 3.11:
The description of the design of the SIM is described as follows, in terms of its desired functionality, as well as its architecture:
1. Desired functionality of the SIM contains –
a. Repeatable steps to facilitate ease of use.
b. A recognised interpretation theory foundation to facilitate validity and rigour of interpretation.
c. The ability to interpret EA definitions in a phenomenological way.
d. The means to produce a meaning and understanding of EA definitions.
e. The means to allow for qualitative reflection.
2. The architecture of the SIM embodies –
a. Distinct phases of execution, in terms of method preparation, method application and communication of results.
b. Distinct executable steps with clearly defined inputs and outputs.
c. Points of reflection on results of executable steps.
d. The hermeneutical cycle of interpretation.
e. The influence and impact of the interpreter in the execution of the method.
The net result of the demonstration of the SIM is an understanding of EA, in the form of a claim about the meaning of EA, as well as a set of supportive propositions.
EAt Design Cycle
The design problem that drives the design of the EAt states that the foundational meaning of EA (the result of the SIM demonstration) is unspecified in conceptual terms. The EAt design problem is reached by achieving the following two objectives:
1. Identify the foundational EA concepts and their relationships.
2. Structure EA foundational concepts and relationships in an architectonic.
The resources needed to complete Step 1 and 2 are similar to the SIM – namely knowledge of the state of EA terms and definition discussion, as well as knowledge about solutions that make EA’s implicit foundations explicit. In the execution of Step 3, the EAt is developed by making use of Heidegger’s equipment analysis (Heidegger, 2000) to determine the foundational concepts of EA. The SIM evaluation results (i.e. an understanding of EA) serve as a resource for the development of the EAt. In Step 4, the EAt is evaluated according to a set of instructions that capture the knowledge of how to use the EAt. The efficacy of the EAt in solving the research problem is assessed in Step 5, by applying an EAt evaluation instrument to demonstrate the EAt to a group of EA practitioners and academics. The interviewee responses are analysed in terms of the EAt evaluation metrics, and produce observations on the value of the EAt to explain the foundations of EA in an explicit manner. Figure 3.12 illustrates the complete design cycle in terms of outputs and resources.
The description of the design of the EAt is described as follows, in terms of its desired functionality, as well as its architecture:
1. The desired functionality of EAt contains –
a. A set of foundational EA concepts.
b. A description of the relationships between foundational EA concepts.
c. A graphical representation of the EAt, to facilitate its use in explaining the foundational meaning of EA.
2. The architecture of the EAt embodies –
a. Distinct representation of EA foundational concepts.
b. Distinct representation of the relationships between EA foundational concepts.
The evaluation of the EAt produced opinions and observations about the value of the EAt as a tool to explain the fundamental concepts of EA. The EAt evaluation protocol (Appendices C, D and E) was approved by the Unisa School of Computing ethics committee, after which the evaluation was conducted over a period of two weeks to an audience of six EA practitioners and researchers. Due to research scope and time limitations, an exhaustive evaluation of the efficacy of the EAt is outside the scope of this research, and will be listed as an important issue for further research. The results of the EAt evaluation and the research contributions are discussed in detail in Chapter 5 (section 5.5.3).
TABLE OF CONTENTS
STATEMENT OF ORIGINALITY
TABLE OF CONTENTS
LIST OF FIGURES
LIST OF TABLES
CHAPTER 1: INTRODUCTION
1.2 BACKGROUND TO THE RESEARCH PROBLEM
1.3 RESEARCH PROBLEM
1.4 THE RESEARCH QUESTION AND OBJECTIVES
1.5 RESEARCH RATIONALE
1.6 RESEARCH SCOPE AND DELINEATION
1.7 RESEARCH APPROACH
1.8 RESEARCH CONTRIBUTIONS
1.9 ORGANISATION OF THE THESIS
CHAPTER 2: THE PROBLEM WITH UNDERSTANDING ENTERPRISE ARCHITECTURE
2.2 A REVIEW OF ENTERPRISE ARCHITECTURE LITERATURE
2.3 A REVIEW OF THE REPORTED EFFORTS TO DEFINE ENTERPRISE ARCHITECTURE
2.4 OVERVIEW OF ENTERPRISE ARCHITECTURE FRAMEWORKS
2.5 DESIGN SCIENCE RESEARCH MODEL
CHAPTER 3: RESEARCH DESIGN
3.2 THE COMPONENTS OF A RESEARCH STRATEGY
3.3 THE RESEARCH PLAN
CHAPTER 4: A STRUCTURED INTERPRETATION METHOD
4.2 THE PROBLEM AND SOLUTION
4.3 THEORY USED TO DEVELOP THE SIM
4.4 THE DEVELOPMENT OF THE SIM
4.5 THE EVALUATION OF THE SIM
CHAPTER 5: AN ENTERPRISE ARCHITECTONIC
5.2 THE PROBLEM AND SOLUTION
5.3 THEORY USED TO DEVELOP EAt
5.4 THE DEVELOPMENT OF THE EAt
5.5 THE EVALUATION OF THE EAt
5.6 THE IMPLICATIONS OF THE EAt
CHAPTER 6: THE RESEARCH CONTRIBUTIONS
6.2 SUMMARY OF RESEARCH APPROACH
6.3 SCIENTIFIC MERIT OF THE RESEARCH
6.4 DESCRIPTION OF RESEARCH CONTRIBUTIONS
CHAPTER 7: CONCLUSION AND FURTHER WORK
7.2 RESEARCH SUMMARY
7.3 THE RESEARCH RESULTS
7.4 FURTHER WORK
7.5 PERSONAL REFLECTIONS
GET THE COMPLETE PROJECT