The Limits of Reproducibility: Strategies for Transparent Qualitative Research
The social sciences are currently going through a reproducibility revolution (and/or crisis), changing expectations for the availability of code and data and the expectations that third parties can reproduce published findings. But what does “reproducibility” mean for qualitative research? And is it an achievable or even worthwhile goal for qualitative researchers?
This talk presents a two-part argument for prioritizing underlying values such as transparency and understandability over reproducibility. Contrary to current discourse, reproducibility in and of itself is not of particular importance for scientific validity. Instead, it is a) a (quasi) automatic byproduct of transparent quantitative research and b) a useful facilitator for the statistical evaluability of qualitative research. Moreover, while most quantitative social science research focuses on explaining social phenomena in cause/effect terms, qualitative research often focuses on their understanding and interpretation instead.
This does not mean, however, that qualitative researchers should disregard recent calls for transparent research. Credible qualitative research should take heed of the underlying challenges and possibilities of the reproducibility revolution. The talk ends with practical suggestions for enhancing the transparency and understandability of qualitative social science research.