Carson C Woo

Success of information systems (IS) projects is deeply dependent on the quality of the requirements informing their design. Specifically, if the requirements are not precisely determined, IS cannot meet the business needs of organizations. One of the most enduring challenges in managing requirements is ongoing change in requirements. Changes emerge due to various reasons. The main goal of this dissertation is to shed more light on change drivers and change processes in IS requirements. To achieve this goal, we conduct three separate but interdependent studies. The first study elicits and synthesizes various change drivers from the extant literature. By employing a design science research methodology, we propose the Socio-Technical Change Framework and Socio-Technical Requirements Change Method. These two artifacts are drawing upon models in socio-technical systems studies. The proposed framework elaborates on how various social and technical change drivers jointly develop the changes in requirements. The proposed method also provides IS analysts with a new solution to anticipate potential future changes in requirements. The second study investigates the change drivers in contemporary IS projects. We approach this study from a qualitative research methodology. The results from the interviews and surveys reveal 14 categories of change drivers. These change drivers are similar to the change drivers in the first study. However, we explore that the contemporary context includes interesting unique characteristics, which moderate the change processes in IS requirements. The third study dives more deeply into some of the findings from the second study. Specifically, we examine how two of the important change drivers — changes in the environment and changes in user expectations — and one prominent characteristic of contemporary projects — interdependency between requirements — jointly influence the final changes in requirements. To address this query, we employ an agent-based simulation model. We explore interesting insights into the impact of the interdependency between requirements and learning patterns of the population on requirements change. In sum, the findings of all three studies contribute to the extant body of knowledge in IS requirements and open interesting avenues for future research. They also have practical implications for IS analysts, project managers, business owners, and third-party vendors.

View record

This dissertation defines modelling products in information system (IS) analysis in terms of IS analysis models. Motivated by better understanding these models, it identifies and solves some relevant issues, by method of studying relevant concepts and their connections, in the literature. This method is qualitative and requires our subjective interpretations. However, the utility of the research results have been extensively demonstrated. This dissertation consists of three studies. The first study focuses on concepts applicable to all types of IS analysis models. Previous views of modelling adopted in IS analysis suffer from being either generic for not tailored to the specific context of IS analysis, or limited in not applicable to the entire scope of IS analysis. To remediate, we propose a new view of modelling in IS analysis that is neither generic nor limited. We call it the generalized view. It can be used to distinguish IS analysis models from other models, to compare different types of IS analysis models, and to explain the representation and interpretation of IS analysis models. We apply it to compare two IS analysis models of an example project. The second study focuses on the connection among different types of IS analysis models. They were studied in different areas in isolation with inconsistent terminologies, hindering inter-area collaboration. To remediate, we propose another new view of modelling in IS analysis, to connect different types of IS analysis models in a comprehensive picture. We call it the consolidated view. It can be used to help stakeholders to determine whether two IS analysis modelling methods are substitutes or complements, to compare modelling methods that are substitutes, and to coordinate the use of multiple modelling methods that are complements. We apply it to evaluate and compare Gaia and Tropos, two agent-oriented modelling methods. The third study suggests how to apply the two views together to analyze modelling methods. It also discusses the differences and relationships between them in such applications. We apply them to analyze the GRAPPLE modelling method. Through the analysis, we test the generalizability of the two views, and identify and solve issues of GRAPPLE.

View record

The correct and complete identification of requirements for information systems is important to the development and implementation of systems into organisations. These requirements are gathered using conceptual models, which are important tools used by system analysts to understand organisations. However, as more users participate in system analysis, researchers have found that traditional conceptual modeling techniques do not completely represent the contextual information in a user's scenario.In this work we design and develop a structured conceptual modeling method, called the organisational actor method. We define a conceptual modeling method as a set of concepts, rules, and method for representing phenomena in the world. The organisational actor method represents the context of an environment which can be represented by understanding how an actor thinks about the environment and how the environment affects, and is affected by, the actor.A design-science approach was used to develop the organisational actor method. The approach used the theory of affordances, the systems approach, ontology, and software agent literature as theoretical foundations for the development of the organisational actor method. This study's purpose was to develop a method that could represent the internal view of an actor, the external view of the actor, and connect the two views.We also conducted four studies determining the usefulness and usability of the organisational actor concepts and the organisational actor method. We found that our concepts: perceptions/inputs, beliefs, learning, goals, reasoning, intentions, capabilities, actions, output are usable because non-technical people implicitly use the organisational actor concepts when describing domains. Also, our concepts are useful because domain experts judge them to be better representations of the domain than another commonly used conceptual modeling language called i*, and our method is usable because domain experts better represent the domain when using a method than when they don't. Finally, our method is useful because it can be used to gather information from employees about how they perform, and make decisions about, their work.In future research we will use our method to design cooperative information systems and to gain an understanding of processes that involve interdependent but autonomous individuals.

View record

The alignment of business strategies with Information Technology (IT) is optimal when harmony exists between organizational and system goals. Empirical evidence reveals that effective strategic alignment leads to superior financial performance for organizations. This observation has spurred extensive research into business-IT alignment, and the issue of alignment remains a top concern for CIOs. In this thesis, we argue that the parochial view taken by past research into business-IT alignment is a probable cause for continuing system failures. Subscribing to a multi-disciplinary perspective, we present a method for detecting misalignments between business strategies and IT. Our investigation divides into three essays.In Essay 1, we develop a goal-based framework that incorporates goal concepts from multiple disciplines to investigate business-IT alignment. When applied to a case study, the framework revealed several insights to assist systems analysts in understanding the links between goals at the operational level and goals at the strategic level of an organization. One of the novelties of the framework is the explicit distinction between goals assigned to users versus goals interpreted by users. In Essay 2, we explore whether there exist salient factors that influence users when they describe goals to systems analysts. In a laboratory study, we found that motivation and experience of users and the complexity of describing tasks do influence the extent to which users describe their assigned goals to systems analysts. In this research, we also discovered that the complexity of describing goals is highest at the middle management level as opposed to the executive level.In Essay 3, we propose a systematic mapping methodology to complement the framework proposed in Essay 1. The methodology when applied to a case study highlights several insights for the effective alignment of operational level goals with strategic level goals. One of the contributions of the methodology is its capability to explain, in the context of business-IT alignment why some operational level goals do not show direct contributions to strategic level goals.Collectively, the findings of the three essays enrich our understanding of the use of goal concepts to detect business-IT misalignments.

View record