When Are Interviews Good Enough? a Reflection on Threats to Retrospective Methods Posed By Automaticity and Performativity

Research on respondent performance in the research setting and on automaticity in human cognition directly challenge the validity of interviews. Interviews, however, efficiently produce what looks like rich, relevant data and they are particularly appealing when studying complex work in difficult to research environments. This paper looks for evidence of the practical relevance of performativity and automaticity in transcripts of interviews with civilian and military staff just returned from their rotations in a reconstruction mission in a (post?)conflict environment. Re-analysis of these interviews suggests that that the validity of retrospective methods may degrade rapidly as complexity increases.