Context: The globalisation of activities associated with software development and use has introduced many challenges in practice and for research. While the predominant approach to research in software engineering has followed a positivist science model, this approach may be sub-optimal when addressing problems with a dominant social or cultural dimension, such as those frequently encountered when studying work practices in a globally distributed team setting. The investigation of such a team reported in this paper provides one example of an alternative approach to research in a global context, through a longitudinal interpretive field study seeking to understand how global virtual teams mediated the use of technology. Objective: Our focus in this paper is on the conduct of research in the context of global software activities, particularly as applied to the actions and interactions of global virtual teams. Method: We describe how we undertook a substantial field study of global virtual teams, and highlight how the adopted structuration theory enabled us to deliver effectively against our goals. Results: We believe that the approach taken suited a research context in which situated practices were occurring over time in a highly complex domain, ensuring that our results were both strongly grounded and relevant to practice. It has resulted in the generation of substantive theory and techniques that have been adapted and applied on a pilot basis in further field settings. Conclusion: We conclude that globally distributed teamwork presents a complex context which demands new research approaches, beyond the limited set customarily applied by software engineering researchers. We advocate experimenting with different research methodologies and methods so that we have a more rounded repertoire to address the most important and relevant issues in global software development research.(Abridged)