The Construction of Quantitative Evidence in Political Science: A Turn to Practice

The novelty of our approach from the standpoint of this discipline is to bring in methods developed in the Science, Technology and Society (STS) literature: including studies of the tacit, in situ practices involved in producing scientific knowledge. To our knowledge this will be the first attempt in political science to empirically study embedded practices of research and the production of scientific facts. Moreover, this work has an interdisciplinary appeal as it would be one of the first ‘laboratory studies,’ after the model pioneered by Bruno Latour and Steve Woolgar (1979), to deal with political science.

The motivations for a practice-sensitive, ethnographic approach are many-fold, but center around widening the definition of transparency beyond that of existing work. Current definitions of transparency focus on written documentation of processes: registration, pre-analysis plans, and the posting of replication data. In this book, we will argue for the need to expand this definition beyond the written to include the unwritten contexts and cultural factors that lead to the production of political science knowledge. There are no rules for following the rules, and scientific work is subject to fundamental regress in experimental replication. Such concerns are immediate and material: we do not have a consensus on the scope of definitions of replicability and transparency, leading to confusion in the field.

For this work we will examine the decisions and experiences of graduate students, junior faculty, senior faculty, journal editors, and research funders. We will draw on observations of their work and in-depth interviews to ask about transparency in the context of three dimensions of research: teaching & training, protocols & methods, and peer-review & publication.

Contact me for our book proposal and Table of Contents.