“Good research involves publishing complete replication files, making every step of research as explicit and reproducible as is practical.” This is the conclusion from a new paper by political scientist Allan Dafoe (Yale University).
Dafoe examines the availability of replication data in political science journals, and concludes that “for the majority of published statistical analyses, […] readers have to trust that the scholars correctly implemented the many stages of analysis since replication files are not available. The reality is that it is not uncommon for the key results of scientific research to be non-reproducible or to arise from errors.”
Requiring scientists to provide complete replication files, from primary data to final output, would help solve this problem. Dafoe outlines concrete recommendations for authors, journals, universities, and funders on how to make this happen.
A blog post about the paper was originally published on the Institution for Social and Policy Studies’ (ISPS) Lux et Data blog, and later cross posted on Nicole Janz’s Political Science Replication blog.