Astrophysicists on Twitter: An in-depth analysis of tweeting and scientific publication behavior
This paper analyzes the tweeting behavior of 37 astrophysicists on Twitter and compares their tweeting behavior with their publication behavior and citation impact to show whether they tweet research-related topics or not. Astrophysicists on Twitter are selected to compare their tweets with their publications from Web of Science. Different user groups are identified based on tweeting and publication frequency. A moderate negative correlation (p=-0.390*) is found between the number of publications and tweets per day, while retweet and citation rates do not correlate. The similarity between tweets and abstracts is very low (cos=0.081). User groups show different tweeting behavior such as retweeting and including hashtags, usernames and URLs. The study is limited in terms of the small set of astrophysicists. Results are not necessarily representative of the entire astrophysicist community on Twitter and they most certainly do not apply to scientists in general. Future research should apply the methods to a larger set of researchers and other scientific disciplines. To a certain extent, this study helps to understand how researchers use Twitter. The results hint at the fact that impact on Twitter can neither be equated with nor replace traditional research impact metrics. However, tweets and other so-called altmetrics might be able to reflect other impact of scientists such as public outreach and science communication. To the best of our knowledge, this is the first in-depth study comparing researchers' tweeting activity and behavior with scientific publication output in terms of quantity, content and impact.