The thematic orientation of publications mentioned on social media. Large-scale disciplinary comparison of social media metrics with citations
| Date | 18 May 2015 |
| Published date | 18 May 2015 |
| DOI | https://doi.org/10.1108/AJIM-12-2014-0173 |
| Pages | 260-288 |
| Author | Rodrigo Costas,Zohreh Zahedi,Paul Wouters |
| Subject Matter | Library & information science,Information behaviour & retrieval |
The thematic orientation of
publications mentioned on social
media
Large-scale disciplinary comparison of social
media metrics with citations
Rodrigo Costas, Zohreh Zahedi and Paul Wouters
Center for Science and Technology Studies (CWTS), Leiden University,
Leiden, The Netherlands
Abstract
Purpose –The purpose of this paper is to analyze the disciplinary orientation of scientific
publications that were mentioned on different social media platforms, focussing on their differences
and similarities with citation counts.
Design/methodology/approach –Social media metrics and readership counts, associated with
500,216 publications and their citation data from the Web of Science database, were collected from
Altmetric.com and Mendeley. Results are presented through descriptive statistical analyses together
with science maps generated with VOSviewer.
Findings –The results confirm Mendeley as the most prevalent social media source with similar
characteristics to citations in their distribution across fields and their density in average values per
publication. The humanities, natural sciences, and engineering disciplines have a much lower presence
of social media metrics. Twitter has a stronger focus on general medicine and social sciences.
Other sources (blog, Facebook, Google+, and news media mentions) are more prominent in regards to
multidisciplinary journals.
Originality/value –This paper reinforces the relevance of Mendeley as a social media source for
analytical purposes from a disciplinary perspective, being particularly relevant for the social sciences
(together with Twitter). Key implications for the use of social media metrics on the evaluation of
research performance (e.g. the concentration of some social media metrics, such as blogs, news items,
etc., around multidisciplinary journals) are identified.
Keywords Citation analysis, Bibliometrics, Altmetrics, Science indicators, Science mapping,
Social media metrics
Paper type Research paper
Introduction
Web-based applications are starting to have an impact in scholars’daily practices
(Wouters and Costas, 2012), involving a broad set of activities, from managing their
literature using Mendeley, CiteULike or Zotero (Li et al., 2011), to writing and reading
blogs (Shema et al., 2012), sharing publications in Facebook or Google+(G+) (Zhu and
Procter, 2012), tweeting about scientific papers (Haustein et al., 2014c), or commenting
on and rating books in Goodreads (Zuccala et al., 2014, 2015).
Aslib Journal of Information
Management
Vol. 67 No. 3, 2015
pp. 260-288
©Emerald Group Publishing Limited
2050-3806
DOI 10.1108/AJIM-12-2014-0173
Received 15 December 2014
Revised 18 March 2015
Accepted 18 March 2015
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/2050-3806.htm
The authorswish to acknowledgethe technical support byHenri de Winter and Erik vanWijk from
CWTS in collecting all the altmetric data, LudoWaltman and Nees Jan van Eck fortheir help in the
understanding of the VOSviewer maps, and Euan Adie from Altmetric.com for his help in the
collection and understanding of the data from the different altmetric sources. The authors also
acknowledge the comments by the two anonymous referees and the editors ofthe journal.
260
AJIM
67,3
One important characteristic of these web-based practices is that they often leave
“traces”in the form of saved publications in online reference managers, tweets, blogs,
Facebook wall posts, etc. The collection and study of these traces is the main target of
the so-called “altmetrics”(Priem et al., 2010), which have opened the door to new ways
of studying scientific communication and its different forms of perception by diverse
audiences. However, “altmetrics”is not considered as a proper term for a series of
metrics that are very diverse and complex (Rousseau and Ye, 2013). Instead “social
media metrics”has been proposed (Haustein et al., 2015a) as these metrics come from
sources that are embedded in the social web (Bar-Ilan et al., 2012), although the
discussion on the proper term (or terms) for the different “traces”and events captured
by these sources is still open (Haustein et al., 2015b).
Research in altmetrics has mostly focussed on aspects about the description of the
different sources and metrics (Galligan and Dyas-Correia, 2013; Khodiyar et al., 2014;
Piwowar, 2013), the coverage of publications by the different sources (Peters et al., 2014;
Robinson-García et al., 2014; Zahedi et al., 2014) and the adoption/use of social tools
by different communities (Haustein et al., 2014b; Mas-Bleda et al., 2014; Thelwall and
Maflahi, in press); correlations (particularly with citations) (Costas et al., in press;
Haustein et al., 2014c, 2015a; de Winter, 2015) or data problems and quality
(Chamberlain, 2013; Zahedi et al., 2014).
In several of these previous works some disciplinary analyses have been performed
(Hammarfelt, 2014; Haustein et al., 2014c; Mohammadi and Thelwall, 2014; Zahedi and
van Eck, 2014), pointing to differences in social media metrics across fields of science.
However, a broader and detailed disciplinary analysis in a global map of science is
still missing, which is essential to better understand the presence and role of social
media metrics across disciplines. This paper intends to fill this gap.
Objective and research questions
The main objective is to analyze the disciplinary orientation of scientific publications
that received mentions from different social media sources, and particularly to establish
their main differences/similarities with citations. The following research questions are
targeted:
RQ1. What is the presence and density of social media metrics across scientific
disciplines? (with “density”here we mean strictly the average of metrics per
paper (see similar terminology in Haustein et al., 2015a), not to be confused
with the “density view”from the VOSviewer tool explained below).
RQ2. How is the distribution of social media metrics across fields?
RQ3. What are the scientific disciplines that have a higher propensity to present
some social media activity vs. citation impact?
Methods
For this study we have considered the same set of publications analyzed in a previous
study (Costas et al., in press). This set is composed of 500,216 Web of Science (WoS)
publications (articles and reviews) from July until December 2011 with a Digital Object
Identifier (DOI). The DOI is used as the linking element across the different data
sources. Citation data have been collected up to week 39 (August) 2014, considering a
citation window of more than 2.5 years. Mendeley readerships have been collected
(using the Mendeley REST API) up to mid October 2014 and Altmetric.com data has
261
Thematic
orientation of
publications
To continue reading
Request your trial