This is the conclusion of a recent paper by Lutz Bornmann, a sociologist of science working at the Max Planck Society. The paper itself was only published on ArXiv and it was probably not peer reviewed in a formal way, but for me it seems reliable.
Altmetrics are article level metrics – an alternative to a traditional citations count, concerning the impact on a wider group of Internet users (not only scientists). Altmetrics can count tweets, Facebook, Google Plus and blog posts etc. relating to a measured paper (more here).
Lutz Bornmann examined the altmetrics count with a special focus on tweets about papers listed in the Faculty of 1000 database that were published in 2011 or later. Faculty of 1000 (F1000) is a publisher of services for life scientists and clinical researchers. F1000 Premium is a paid recommendation tool for end users. It’s reviewers recommend papers to its subscribers. They may recommend a paper published in any biomedical journal and tag it as “good”, “very good” or “excellent”, along with some additional tags, such as “good for teaching” or “controversial”.
This means that Bornmann’s database includes papers on biology and medicine, which were recommended by F1000 reviewers and were published after 2011. Among the articles those tagged by reviewers as “good for teaching” were significantly more often tweeted than others. What is more, there is strong relationship between quality tags (good/very good/exceptional) and the number of tweets concerning the article. Papers tagged as “new finding” or “interesting hypothesis” were not tweeted more often than others. Tags provided by F1000 reviewers seem to be adequate, which can be verified by their relationship with citation counts (that were measured for a different trial of older articles).
The study confirmed that the majority of papers have 0 tweet counts, even though tweets are the easiest way of gaining the attention of Internet users (even fewer articles were mentioned on blogs or Facebook). Now we know that educational papers are more tweeted and have major overall altmetrics scores. We do not know what to do to get the “good for teaching” tag, which seems to be a good predicator of Twitter success, but I think that all of us can imagine. Tweeter users, like students, enjoy simplicity, summarizing works and of course high quality.