بهبود کیفیت از مدل پروسه مرجع: گسترش کارکرد کیفیت

Little academic work exists on managing reference model development and measuring reference model quality, yet there is a clear need for higher quality reference models. We address this gap by developing a quality management and measurement instrument. The foundation for the instrument is the well-known Quality Function Deployment (QFD) approach. The QFD-based approach incorporates prior research on reference model requirements and development approaches. Initial evaluation of the instrument is carried out with a case study of a logistic reference process. The case study reveals that the instrument is a valuable tool for the management and estimation of reference model quality.

Today's hyper-competitive and increasingly regulated markets see organizations place significant focus, and thus resources, on managing and improving their business processes [42]. Such improvements and innovations are considered to be an important factor in creating organizational wealth [49]. Indeed, recent Gartner studies show that CIOs now consider Business Process Management (BPM) to be the top priority in the coming years [41], [42] and [43]. The high prioritization of process management in the recent years is also due to today's regulatory climate, which is forcing organizations to document processes and ensure their compliance. Many recent regulations (e.g. Anti Money Laundering Act [4]), however, are principle-based, as opposed to being prescriptive in nature, and require significant interpretation on account of the regulatee [36]. Anecdotal evidence from the Australian finance sector suggests that organizations are seeking reference models (RM) to help ease their compliance management pain and reduce the significant spending brought on by compliance requirements. RM are blueprints of recommended practice and, thus, are sources of reusable and efficient business processes on which organizations can model their own [58]. Their main purpose is to streamline the design of enterprise models and enable organizations to apply ‘best practice’ knowledge. The use of high quality RM can result in cost and risk reductions, as well as an improvement of the organization's business processes [58]. It is estimated that the use of RM in projects can reduce the project duration and required financial resources by 30% [60]. Clearly, while there is much potential for savings with the use of RM, using a low quality RM can be damaging to the performance of the organization and to the quality of its decision making. Business processes, and therefore also RM, contain decision making components, such as policies or business rules for example [54], hence a high quality specification of the RM is important to ensure compliance with various requirements. In other words, an organization should ensure that the considered RM is complete, accurate, and easily configurable (i.e. flexible) for their purpose. To date, however, little work has been carried out that might provide guidance for the selection of high quality RM, let alone guidance for the development process that leads to high quality RM [45]. Only a few studies have focused on the quality of RM, despite reference modeling being an established field in Information Systems research. This situation is despite the fact that prior research has explicitly identified the need to close this gap [70]. For example, according to Fettke and Loos [20], the selection of models is increasingly complicated while being ‘a crucial task for the project’. Frank [24] concludes that “… the evaluation of reference models is a challenging, yet important task”. Accordingly, the organizations that develop RM (e.g. standardization or regulation bodies), and also those that are potential RM users, would value an instrument that aims to increase the quality of RM, through guiding its development, and also provides an easy measure of model quality that can be used in communication between the RM provider and RM user organizations. Indeed, the research presented in this paper was incepted by a request from a German standardization body that required such an instrument despite already having a quality control process in place. The organization was interested in obtaining an RM quality management and measurement instrument that would incorporate a best practice RM development process while also taking into consideration RM user requirements. In response to the clear gap in RM quality research, and in response to the request of the aforementioned standardization organization, we present an interrelationship matrix-based artifact for increasing and measuring the quality of RM. The measurement evaluates the steps that are taken to develop an RM with respect to a set of required model characteristics and also considers the ‘voice of the customer’. We adapt the first phase of the Quality Function Deployment (QFD) approach (also referred to as ‘narrow QFD’) for this purpose and derive an artifact that not only helps organizations develop high quality RM but also measures the achieved quality level. QFD, which originates from Japan, is an approach aimed at satisfying the users through the provision of high quality products that fit the users' requirements. The approach involves collecting user demands and converting them into design targets and major quality assurance points to be used throughout the development phases [3]. We see a QFD-based approach as most suitable here due to QFD's user-centric nature that captures the mapping of user requirements into product design [26]. The paper is structured as follows. Section 2 discusses related work — its main contribution is the extensive literature analysis and synthesis of academic literature related to RM quality and RM development, much of which is published in various German publication outlets and, hence, not easy accessible by international researchers. Research methodology is presented in Section 3. Section 4 describes the proposed instrument and Section 5 presents its application in a case study. Section 6 discusses findings related to the development process and RM characteristics. Last, Section 7 summarizes contributions, limitations and outlines future research.

This study is the first to recognize the usefulness of a QFD-based approach for conceptual models, in particular RM. The study provides three main contributions. First, the proposed QFD-based approach incorporates the voice of the RM users and presents a compressed measure for the quality of such models as well as a means for the management of quality RM development. Thus, it allows for better communication between the users and the developers, an easier means of comparing quality of different RM versions, as well as development of higher quality RM. Second, our research consolidates disparate literature on RM characteristics and identifies five required characteristics of such models, viz. generality, flexibility, usability, completeness, and understandability. Third, our research contributes to the creation of a cumulative tradition in RM development. We consolidate guidelines for RM development from various works into a seven-phase development approach, in which each phase is characterized by different activities that need to be carried out in order for the phase to be successfully completed. In this paper, we also present the initial evaluation of the proposed QFD-based artifact in one in-depth case study. Through the empirical component of this study, we are able to not only show the accuracy of the measurement instrument, but also discuss the user and developer assessments of different aspects of the RM development. Moreover, due to lack of data at this stage, we run a simulation to determine benchmarks for RM quality and compare the chosen RM to these values. We see our work as relevant to the research and practice communities. The contributions are two-fold. First, through the innovative adaptation of the QFD approach, hence, adding to the QFD body of research. Second, through the consolidation of existing English and German literature on RM characteristics and development approaches, and the development of the measurement instrument that incorporates both these factors. The contribution to industry is also twofold. First, for users, the QFD-based approach is an instrument that enables easy comparison of quality of RM versions, and, hence, helps guide the RM selection and revision process. Second, the identification of how to develop RM may assist development organizations in increasing the maturity of their work, consequently, positively influencing the quality of the RM they develop. Finally, the activities defined in the various development phases can guide the developers in the development phases and hence, contribute to the overall RM quality. The limitations of the study are related to data collection and analysis. While every effort was taken to eliminate bias, it is possible that the focus on organizations from one region and from one industry is a source of bias. Also, the data collection carried out by the research team with the developer organization may have been influenced by the researchers' background. One might perceive the subjectivity of the QFD-based approach as another limitation because non-objective measures are used to determine the quality. Future research can proceed in a number of directions. The QFD-based approach is a Design Science artifact that should be applied in further case studies to evaluate its usefulness. Further research can also focus on the application of the instrument for in-depth longitudinal case studies. Also, while the QFD-based approach assesses the quality of an RM developed for German speaking organizations of the consumer goods industry, it would be interesting to see the differences of the user importance and fulfillment assessments in other industries and regions with the same RM.