One year of MarXiv: Our first annual review

MarXiv (rhymes with "archive") celebrated its first birthday in early November 2018. Over the course of the last year, we've hit a number of important milestones.

A look back at our first year

  • 313 papers have been shared;
  • We requested 420 papers from their respective authors via 839 email messages;
  • 62% of authors opened at least one of (up to three) email requests to share their paper in MarXiv;
  • Our email requests to authors were successful for 26% of the papers we asked for;
  • Only 12 contacted authors have explicitly stated that they will not share their paper in MarXiv or elsewhere;
  • 51% of the papers in MarXiv are pay-walled outside of the archive by their respective publishers;
  • Papers in MarXiv were downloaded 14,136 times;
  • Free downloads of these "Green Open Access" papers (which are available for free in MarXiv, but pay-walled outside of MarXiv by their respective publishers) have saved ocean conservationists $320,379.19 in per-article download fees.

We have a lot to celebrate — but also a lot of room to improve! Sadly, these numbers show that our email requests (with the subject line, "Paper request") are ignored by 38% of authors we contact. And that's even after asking a second time in about three weeks, and again once more roughly three weeks after that. These findings align with anecdotal evidence that authors respond to paper requests roughly 50% of the time. It seems that for every author who is happy to give away free copies of their papers over email, they have a colleague who ignores these requests.

All authors who have stated that they will not share a freely-available version of their papers have either 1) stated that the publisher's formatted, pay-walled version is the only version that should ever be read, or 2) hold on to inaccurate information about self-archiving in MarXiv (i.e. that doing so would be illegal, even though that's 100% false). Much like talking to climate-deniers about climate-change, some academics can harbor personal opinions about scholarly publishing which they apply in their decision-making over actual facts. That said, we do not know what logic was driving the decision-making of the roughly 70% of authors who simply ignored our repeated requests.

Moving forward to 2019

We have a number of strategies we'll be employing for the second year of MarXiv:

  • More MarXiv Ambassadors
    • Six MarXiv Ambassadors shared their time and energy advocating for MarXiv during our first year. Next year, we're opening the pool to any number of interested candidates. Hopefully with more volunteers aiding our efforts, we'll see more unsolicited sharing of research in MarXiv.
  • Growing awareness of MarXiv
    • For 2019, we'll be experimenting with various ad campaigns to get the word out about self-archiving preprints and postprints in-general, and especially sharing relevant papers in MarXiv. We hope that by doing so, our email requests to authors won't be the first time they've heard about preprints/postprints or self-archiving.
  • Improved A|B testing of requests
    • We're in the process of rebuilding our request database from scratch to make analyses of which request templates resonate with authors easier, and more accurate. This will allow us to implement A|B testing to hone-in on particular phrases or terminology that dispel misconceptions about self-archiving.
  • MarXiv Bookmarklet
    • Request more, get more! At least that's what we hope for. Install the MarXiv Bookmarklet in your browser to tell us which papers you want to read in MarXiv. Crowdsourcing the papers we request should help us find many more relevant papers, thus helping us to grow the MarXiv archive.

Other ideas?

Do you have suggestions or feedback to offer? We'd love to hear from you! Email Nick Wehner at nick@octogroup.org. Thank you!