Big data-driven theory building: Philosophies, guiding principles, and common traps

Abstract

While data availability and access used to be a major challenge for information systems research, the growth and ease of access to large datasets and data analysis tools has increased interest to use such resources for publishing. Such publications, however, seem to offer weak theoretical contributions. While big data-driven studies increasingly gain popularity, they rarely introspect why a phenomenon is better explained by a theory and limit the analysis to data descriptive by mining and visualizing large volumes of big data. We address this pressing need and provide directions to move towards theory building with Big Data. We differentiate based on inductive and deductive approaches and provide guidelines how may undertake steps for theory building. In doing so, we further provide directions surrounding common pitfalls that should be avoided in this journey of Big-Data driven theory building.