This is a guest blog post by Tamar Loach. She is a Research Metrics Analyst at Digital Science.
It’s fairly easy to see from the Altmetric database that medical articles receive a disproportionate amount of online attention. In fact, 60% of tracked tweets from the last week pointed to articles from journals publishing Medical and Health Science research. Interestingly, 63% of these were directed to articles from journals tagged as relating to Clinical Medicine or Public Health specifically. This is a significant clustering, and food for thought.
Is the intensity of medical research attention simply a consequence of the relatively large publication output of the medical community? Or is it a sign that this community is particularly active on Twitter? Is the emphasis on clinical or practical — over basic — medicine an indicator of practitioner and public engagement? Drilling down into the attention data collected by Altmetric can potentially help us answer questions like these, which can in turn shape how we use and value altmetrics.
The interest in analysing altmetrics data began with the aim to capitalise on the rapid feedback that altmetrics can provide compared to that of the much slower citation lifecycle. More and more research has begun to emerge in the past year; a nice example, using correlation and principal component analysis to compare altmetrics and bibliometrics, can be found in this article by researchers at Leiden University. This is a solid and interesting starting point (and will need to be repeated as the data changes in time, with community behaviour changing in response).
Although making comparisons with citation-based metrics is interesting, there’s a lot to be said for investigating altmetrics in their own right, so that we have the background for a more informed interpretation of attention, and are able to value altmetrics as complementary but distinct insight into research output and scholarly communication. We can look at the demographics of tweeters — and variations across fields — for insights into any associated online communities. We can compare highly tweeted papers with those that receive more attention from bloggers or mainstream media outlets. There is a wealth of knowledge to be uncovered by making new connections between different parts of the data.
Any analysis that leads to a better understanding of what altmetrics data indicate is valuable. As part of the Digital Science Metrics team, I’m working on various projects, including one that looks to link various other datasets (citations included) to Altmetric scores, so that we can better understand the numbers and relationships. Others interested in tackling similar questions should take a look at the Altmetric API, as it provides a valuable window into the data behind the donuts and is a great place to start analysing.