Twitter matters as science seeks to make a social impact

Scientists are increasingly turning to social media to publicise their research. While some see Twitter and Facebook as a bit of fun, others find online tools are becoming integral to their work. In an…

Your water-cooler moment 2.0. 'Pong

Scientists are increasingly turning to social media to publicise their research. While some see Twitter and Facebook as a bit of fun, others find online tools are becoming integral to their work.

In an attempt to investigate whether social media can help increase the scientific impact of scholarly articles, a group of researchers from Canada has looked at how often biomedical papers are tweeted about. It turns out that Twitter doesn’t predict scientific success very well at all.

The study’s authors, led by Stefanie Haustein of the University of Montréal, tracked more than 1.4 million biomedical science articles published between 2010 and 2012. They found that less than 10% of these articles were tweeted even once, suggesting that only a small percentage of scholarly articles are reaching the social media airwaves.

They also found only a very weak correlation between the number of times an article was tweeted and the number of times it was cited by other scientists – a more traditional measure of scientific impact.

Social value

This is actually good news because social media was never meant to replace traditional statistics like journal impact factors or article citations. Instead, a new wave of alternative metrics, shortened as altmetrics, are being used to trace the threads that link tweeted articles between scientists and their appearance in traditional media outlets. Scientists can now follow their work as it goes viral.

Funders are also starting to understand the value of this new type of impact. The US National Science Foundation, for example, is moving towards valuing all research “products”, including altmetrics, when it reviews grant applications. Where a researcher was once asked to list their research publications, they are now called upon to detail other results, such as online data sets, software or online communication through blogging or social media. This can bring in different flavours of outreach when we think about a paper’s impact.

Altmetrics can describe an article’s relevance to policymakers or its outreach to media and the general public. One of the important findings of Haustein and colleagues’ study is that it confirms altmetrics and more traditional bibliometrics, such as citations, as two separate but complementary parts of an article’s social and scientific impact.

Not surprisingly, tweeting about papers is getting more popular. In their recent study, Haustein and colleagues found that around 20.4% of the papers they looked at that were published in 2012 were tweeted about. That’s around 10 times more than the number tweeted in 2010.

Scientists should keep tweeting about papers. Twitter allows your research to reach a “global faculty lounge” that can be seven times larger on average than a scientist’s typical academic department. It allows your papers to reach, through tweets and retweets, not only other scientists, but also journalists, decision makers and the general public.

In fact, tweeting about your research can play a role in the entire life cycle of a scientific publication, from the birth of an idea to peer review and dissemination of the final article. Being a good communicator on Twitter not only increases your scientific networks and productivity, but it can also support the social impact of your research.

What makes a scholarly article go viral?

Haustein and her colleagues found that the top 15 most frequently tweeted articles had curious or funny titles. Penile Fracture Seems More Likely During Sex Under Stressful Situations and Penetration of the Oral Mucosa by Parasite-Like Sperm Bags of Squid both struck a chord with Tweeters, for example. Readers were also keen on articles that might have potential applications in human health, such as those looking at acne in teenage athletes and required amounts of physical activity for life expectancy.

The top two most tweeted articles referred to the Chernobyl and Fukushima nuclear incidents, which appears to highlight the timely and relevant nature of social media for natural catastrophes.

Not all scientific articles will be tweeted equally and these findings show that if you want attention on social media for your work, you should think about the PR aspect of your research to increase its eventual pickup and popularity.

Altmetrics evidently do a poor job at predicting the scientific success of scholarly articles. But that does not mean they are not a valuable tool in other ways. It is becoming increasingly apparent that research can contribute to both scientific and social outcomes. Altmetrics and bibliometrics are two important complements for how we measure and value our science and both are needed to help us shout about it.