Broadening the Conceptualization of Theory in the Information Systems Discipline

Abstract

The traditional and most prevalent view of theory in the information systems (IS) discipline is conceptualized in what has been termed ?variance theory," or more properly ?variance meta-theory," where a meta-theory is a means by which to conceptualize the types of constructs and relationships used to develop a specific theory (instance). Variance meta-theory conceptualizes the constructs and relationships in a theory strictly as properties of entities, interrelated in a static (statistical) correlational manner. Theories formulated using variance meta-theory generally seek to explain or predict immutable inferred causal relationships. This view of theory is limiting and constraining in an applied field like IS where phenomena of interest are complex, constantly changing, and reliant on proactive human actions, as well as underlying forces. As new technologies are developed, users adapt and learn, take actions, and respond to results. We argue that alternative meta-theories more readily address this wider range of IS research questions. We examine three such meta-theories ? network, process, and co-evolution, comparing and contrasting their underlying conceptualizations, entities, relationships, and methodologies. We argue further that theory (instances) formulated in the language of each of these meta-theories can be used as part of a broad scientific process of proposing, testing, and reevaluating theory to develop increasingly nuanced understandings of IS phenomena, ultimately resulting in the accumulation of knowledge that is relevant to both theory and practice.