This paper examines issues of theory development, study design, and data analysis for the study of mediated interaction. Building from previous traditions and approaches in communication, ethnology, and social psychology, a framework is suggested for the study of electronic mediated interaction. From the theoretical discussion of mediated interaction and interactivity, a methodology is proposed for the rigorous testing of new theories of mediated interaction. Behavioral observational and psychometric measurement techniques are discussed in relation to their benefits and drawbacks for coding mediated interaction. Exemplar approaches are discussed for their applicability to the study of mediated interaction. The paper concludes with an example of the suggested methodology with results from the application and implications for future research.