How can we create reliable and replicable political science data? A recent article in the American Political Science Review focuses on text analysis and suggests ways to make these data sound and reproducible.
I recently received an email where a graduate student is trying to ask original authors for data. There is some evidence that authors withhold data due to time constraints (Tenopir et al. 2011), but they may also decline to share data because they fear a damaged reputation when a replication of their work fails (Lupia and Elman 2014). From my experience in the Cambridge Replication Workshop, authors are more willing to share their data when the replicator is perceived as trying to be helpful rather than cross-checking results. Here are my tips.
An article in the American Journal of Political Science was corrected after the coding of a political attitude variable was accidentally the wrong way around. Pre-publication cross-checks by the authors and the journal, as well as publication of the original data and variable transformations can avoid such problems.
Following an article in the New England Journal of Medicine, which portrayed scientists who re-use data as parasites, we now hear more on this from Nature. Apparently, data transparency is a menace to the public. The Nature comment “Don’t let transparency damage science” claims that the research community must protect authors from harassment by replicators. The piece further infects the discussion about openness with more absurd ideas that don’t reflect reality, and it leads the discussion backwards, not forward.
Journal editors can enforce replication policies. Authors can decide to work transparently. Most initiatives for open science and reproducibility agree that editors and authors are are the key actors to enforce the gold standard of research integrity. However, peer-reviewers can use their leverage as well: just say you will only review an article once the author provides the data.