Research Roundup: A look at Studies into Altmetrics 


by Amy Rees, Product and Client Support Specialist at Altmetric

Thursday 27 September

On the second afternoon of the conference, 5:AM was treated to a series of presentations from researchers using altmetrics to explore how and why research is communicated online. Each presented their work and gave a great insight into how you can use the data:

Making Headlines: An Analysis of US Government-Funded Cancer Research Mentioned in Online Media - Lauren Maggio (Uniformed Services University)

Lauren Maggio, from the Uniformed Services University, presented her group’s research about which government-funded cancer research is mentioned online. Interestingly, the group came together after Lauren’s conversation with Juan Pablo at 4:AM! They then expanded the team to include other researchers. The study they did involved searching for papers about Cancer in Pubmed, and then reviewing the type of news coverage those studies were receiving. Breast, Lung, and Prostate cancer were the most discussed, with Breast Cancer far and away more discussed than any other topic. The group is going to scale this project up to work towards more research in this area of study.

Compatibility of Altmetrics Indicators for Science Politics and Management - Andreas Meier (Forschungszentrum Jülich)

The second presenter, Andreas Meier, discussed a project undertaken to review the feasibility of using altmetrics for science policy and science management. The German government requested a report reviewing the possibility of altmetrics including in their PFI. The study used personal interviews with experts such as Dr. Stephanie Haustein to explore this, asking the participants their opinions on altmetrics and their role in evaluating the outcomes of research. The group determined that the concept of societal impact has not yet been developed to the extent that they would wish to see, and that collecting data from German-language sources still has room for improvement.

The Effect of Social Media Exposure of Scholarly Articles on Facebook and Twitter:Perspectives of Social Media Engagement and Click Metrics - Zhichao Fang (Centre for Science and Technology Studies (CWTS))

This research sought to evaluate how users on Twitter and Facebook are sharing and discussing academic outputs. Journals have started to use Facebook and Twitter to share their publications, and the research wanted to build on this idea. The journal Cell was used as a case study as it already has quite a substantial online following and engagement. To conduct their analyses, the researchers tracked posts of articles that were made simultaneously by Cell on their Facebook and Twitter pages. They found that exposure of scholarly outputs on Facebook got more attention than those on Twitter, but this must been considered in context: Cell has more more followers of their Facebook page than of their Twitter account. Once they had controlled for this variation in followers, they found that Facebook and Twitter had similar level of shares and retweets.

Examining the Nature and Relationship of Tweet Sentiments with Citation Impact and Altmetric Indicators - Aravind Sesagiri Raamkumar (Nanyang Technological University)

Aravind, from Nanyang Technological University, and his co-researchers conducted a study to look at a now-familiar source of altmetrics: Twitter. In contrast to many other previous studies, they opted to go beyond likes and retweets, instead choosing to evaluate the sentiment of Tweets that contained links to published research. Their group manually classified tweets that discussed academic research, specifically Computer Science, using the TextBlob library and setting their own efficacy thresholds. The results? Neutral sentiment is found in most of the tweets, with 97%+ being classified as neutral or no sentiment. Where sentiment could be identified, words that were determined to indicate positive engagement include ‘interesting’ and ‘awesome’. Those with negative sentiment included words of a slightly less cheerful nature! This initial work will likely be built on with further studies, as sentiment analysis remains a very popular topic in the altmetrics field.