Downloads, Software

Can apply to: Research software, scripts, code snippets

Metric definition: File downloads over a period of time.

Metric calculation:  Most downloads are calculated in a straightforward manner (where the number of downloads is simply reported as-is), but others account for–and remove–downloads initiated by bots.

Data sources:  Though almost any web platform that can host files (including researcher websites), the most common and reliable sources of research software download statistics are those reported by software hosting sites (e.g. Bitbucket) or repositories (e.g. Figshare). Some altmetrics tools can report downloads for certain software hosting sites (e.g. PlumX).

Appropriate use cases: Software downloads can be used as an indicator for the reuse of programming code. In some cases, it can be a proxy for the number of users.

Limitations: Software is rarely formally cited, though it is often mentioned in publications. One study has found a weak, but statistically significant, correlation between Scopus citations and download counts for Google Code programs. SourceForge.net does not report downloads as of March 2017.

Inappropriate use cases: Software downloads are not direct measures of usage, quality, or impact.

Available metric sources: Google Code (defunct as of January 2016), Codeplex, Bitbucket, Launchpad, GitHub (API), Figshare, Zenodo

Transparency:  Varies by source

Website: n/a

Timeframe:  Typically immediate, but there may be a reporting delay of a few hours up to a 30 days, depending on the source.