TANZANIT – towards a comprehensive quality meta-model for information systems : case studies of information system quality modelling in East Africa

Abstract

The main goal of the study is to overcome the problems of traditional quality models by developing a higher level conceptual framework, called the quality meta-model, to assist in modelling quality of information systems as end products of the development process. The aim is also refine the definition of information system quality and quality requirements. As a by-product the study presents principles of using the proposed meta-model as a governing element in system development process. The main research question concerns the validity of the conceptual framework. The approach is a combination of conceptual analysis, theory-creating and theory-evaluating research. In addition to an initial literature review and references to relevant writings, the research activity comprises three quality modelling case studies in East-Africa. Constructs found through a selective literature review are first generalized and combined into a conceptual framework. Thereafter this initial meta-model is used in case studies as a template for system specific quality models. Based on findings and further literature review the conceptual framework is eventually developed into its final form. The study shows that a comprehensive metamodel, needed to fully account for information system quality, is a hybrid model that in addition to the core quality elements includes parts from the development process, system and context models. The quality of an information system is in the first place determined by the existence and intensity of certain desired relationships, or state of affairs, between the system and its context. These requirements have priority over others that can be derived from the former. A system specific quality model is the driving force of development. The three case studies support the general validity of the proposed quality meta-model through the usefulness and comprehensibility of the quality models created on basis of it. It proved also to be general and flexible enough to be applied in three different contexts. Comparison to traditional quality models and newer alternatives, in turn, proves the meta-model’s comprehensiveness.