Metrics and document versions: how should altmetrics providers handle metrics for preprints, author manuscripts, and the published version?

The 2015 Altmetrics Workshop
Amsterdam, 9 October 2015

William Gunn


Altmetrics, like open access, is largely a child of digitization. As more and more scholarly communication has moved online, there are more and more opportunities to instrument the process to understand what is happening, what impact that has, and even in some cases to predict what is going to happen. However, one fundamental principle, baked in to the core of the protocols that run the internet itself, is redundancy. Many, many resources exist that are duplicates of one another and there’s not necessarily a known link between the two nor an indication of the authoritativeness of a given copy. For the literature, we have solved this partially through the use of DOI registries such as Crossref, but as scholarly output grows beyond the literature as a category and DOIs are assigned to various objects, things start to get fuzzy again. If a preprint is uploaded to Figshare and gets a DOI, then that document is uploaded to Mendeley and accumulates readers, how does one treat the various metrics that accumulate on the various versions. Now that scholarly publishers have given the official green light to sharing in the STM Association Scholarly Sharing Principles statement (, this sort of issue will grow. This talk will share Mendeley’s perspective on the issue and outline a few things we’re doing to help altmetrics scholars. We will also share data on a proposed new comprehensive survey of Mendeley users & we will describe the dataset which we will make openly available.

Supplementary materials