From Name:
From Email:
To Name:
To Email:

Optional Message:

New Metrics and Open Data Could Improve Reproducibility

Marcie Granahan,
NFAIS Executive Director
A recent study at the Georgia Institute Technology found that metrics which use tweets as a measurement of engagement with scientific literature, such as those used by Altmetric, may be flawed [see full article here]. Of the 4,350 dental research papers in the study, most tweets were not used as a source of academic discussion but instead as a marketing dissemination tool, with the overwhelming majority of tweets coming from a single account.

A promising new measure is Authorea’s R-factor, whose metrics are based on whether later work corroborates the original findings [see full article here]. Instead of how many citations a paper gets, the R-factor focuses on the reproducibility of the work. As the demand for reproducibility increases, open data becomes more available, and AI reduces manual calculation, the R-factor could hold real value.

The Trump Administration sees open data as a source for economic growth [see full article here], and the Office of Management and Budget (OMB) hosted a roundtable this summer to address the use of open government data and its potential as a major economic resource. In addition to commercial uses, open data also provides a source for secondary research. It will be interesting to see whether government grants will begin to support this as a priority.

Technology could also play a role as blockchain evolves from an exchange of value (Bitcoin) to an exchange of information [see full article here]. New innovations, such as Kim Dotcom, allow for the enforcement of copyright while users access content from any platform. Along with digital research object identifiers, could blockchain offer a solution to managing provenance and copyright ownership?

With or without the R-factor, for the time being, researchers should beware. Sloppy, unreproducible research can have an impact on the overall reputation of academic scientists, particularly in the case of retracted articles. A recent study found that scientists who have published papers that are retracted subsequently suffer a 10 percent drop in citations of their remaining work [see full article here]. more

Powered by MultiBriefs
7701 Las Colinas Blvd., Ste. 800, Irving, TX 75063