Chapter 3. A Strategy Framework for Cloud Data Governance
3.2. Theoretical Foundation
This section provides an insight into the theoretical foundation of this research. A theoretical basis provides a guide for the researcher in the interpretations of the results of their study (Simon & Goes 2011). Both academic researchers and practitioners consider that the loss of governance of data in cloud computing has several impacts on cloud users’ strategies and on the capacity to meet their mission and goals. Both sides share the consensus view that data governance is not a one-size-fits-all proposition (Weber et al., 2009). However, to the best of the author’s knowledge, there are no empirical studies in the academic literature that address data governance for cloud computing.
This research aims to close this gap by developing a strategy framework to understand how to implement a data governance programme for the cloud services on the basis of a comprehensive analysis of the data governance in both science and practice. In the following section, different approaches and concepts found in the literature will be considered as part of the theoretical foundation of this research, enabling afterwards the construction of the conceptual framework of the study. These approaches and concepts are the following:
3.2.1. Analytic Theory
Analytic theory is useful for understanding the data governance topic, and for understanding the existing data governance frameworks (Otto, 2011). Gregor (2002) showed that "descriptive theories are needed when nothing or very little is known about the phenomenon in question"(Gregor, 2002, p.7). The analytic theory is the most basic type of theory used to analyse a phenomenon (Gregor, 2006). Gregor (2006) postulated that the analytic theory is useful for describing or classifying specific dimensions or characteristics of individuals, groups, situations or events by summarising the commonalities found in discrete observations. With the popularity of frameworks, classification schema, and taxonomies in IS, the variants of the analytic theory are referred to as classification schema, frameworks or taxonomies (Gregor, 2002). In this study, the analytic theory has been chosen as a concept with which to make a strategy framework for implementing data governance for cloud services. Since this study will be based on the deductive approach, the analytic theory will be suitable for conducting the research. In this context, the deductive approach and analytic theory are used as they allow the researcher to acquire a more complete view and different perspectives of the research problem being studied. The research approach comprises three steps:
Step 1: Analysis of the scientific and practice-oriented literature related to data governance in general and for cloud computing in particular – this was explored in the literature review (see Chapter Two).
Step 2: All data governance frameworks in the sources of literature are analysed and coded according to the dimensions of the cloud data governance.
Step 3: The analysis of both scientific and practice-oriented literature combined with a comparison of existing data governance frameworks. This will allow an insight into the dimensions, concepts and relationships within this area. Once further analysed and generalised, these features will then be developed into a novel strategy framework for cloud data governance. Table 3.1 gives gives a summary of using analytic theory for cloud data governance.
Table 3.1 Use of analytic theory for data governance
Theory Overview
This theory is one of the types of theory used in information systems. Gregor (2006) describes this theory by saying that it "provides a description of the phenomena of interest, analysis of relationships among those constructs, the degree of generalizability in constructs and relationships and the boundaries within which relationships, and observations hold".
Theory Scope
The scope of this theory is the methodologies and procedures that have been proposed in the scholarly literature – these are systematic reviews, taxonomy and process.
Theory Component Task Means of
Representation Analysis of the
literature review
To understand the state of data governance in both science and practice, and to identify any gaps in research.
Words, diagrams, tables.
Taxonomy To identify and classify specific dimensions or characteristics of data governance for non cloud (traditional IT) and cloud computing.
Words, diagrams.
Prescriptive Statements Identify the important processes for designing, deploying and sustaining cloud data governance programme.
Words, diagrams, tables.
Framework Design a strategy framework for an effective cloud data governance programme.
3.2.2. Critical Success Factors (CSFs) Concept
In combination with the steps of analytic theory (as mentioned in section 3.2.1), CSFs are also used in the development of the strategy framework for cloud data governance. They allow a different aspect for the framework development that will not be addressed within the existing frameworks. Therefore, the CSFs for implementing data governance for the cloud will be considered in this chapter. The CSF concept is a good theoretical basis to support this objective. The concept of CSFs has been established over the last 30 years by a number of researchers, particularly by Rockart in 1979 (Forster & Rockart, 1989). Currently, this approach is increasingly used to support IS strategic planning by consultants and IS departments (Amberg et al., 2005). Pinto & Prescott, (1988, p.8) argued that “the majority of the studies in the critical success factor research stream have been theoretical and have assumed a static view of the importance of various factors over the life of a project”. The literature also showed that the CSF concept is important for overall organisational objectives, missions and strategies. The CSF concept is appropriate to each unit of business and the overall organisational aim in the fulfilment of the organisation’s objectives (Amberg et al., 2005). Establishing clear CSFs would be a significant element of risk management and of eventual data governance programme success (Ladley, 2012). This requires a repetitive process for CSF identification, validation and analysis of the constraints underlying each CSF, and a determination of the measures needed for each identified CSF (Amberg et al., 2005). Thus, a successful data governance programme requires a number of CSFs.