Improving human interaction research through ecological grounding

In psychology, we tend to follow the general logic of falsificationism: we separate the ‘context of discovery’ (how we come up with theories) from the ‘context of justification’ (how we test them). However, when studying human interaction, separating these contexts can lead to theories with low ecological validity that do not generalize well to life outside the lab. We propose borrowing research procedures from well-established inductive methodologies in interaction research during the process of discovering new regularities and analyzing natural data without being led by theory. We introduce research procedures including the use of naturalistic study settings, analytic transcription, collections of cases, and data analysis sessions, and illustrate these with examples from a successful cross-disciplinary study. We argue that if these procedures are used systematically and transparently throughout a research cycle, they will lead to more robust and ecologically valid theories about interaction within psychology and, with some adaptation, can enhance the reproducibility of research across many other areas of psychological science.