Article-Level Metrics for 
 Impact Assessment ! Martin Fenner, http://orcid.org/0000-0003-1419-2405 Technical Lead Article-Level Metrics http://fivethirtyeight.com/interactives/world-cup/

2 Discovery http://commons.wikimedia.org/wiki/File:Space_Shuttle_Discovery.png! http://en.wikipedia.org/wiki/ List_of_video_telecommunication_services_and_product_brands! http://en.wikipedia.org/wiki/Big_data! http://simple.wikipedia.org/wiki/Test! Assessment Scholarly Metrics Use Cases Online conversation Business Intelligence

3 Impact Assessment for Individual Researchers Journal Impact Factor Article citation counts Alternative metrics for articles Peer Review No Evaluation Metrics for other research outputs Altmetrics Article-Level Metrics

4 Don't reduce individual research performance to a single number ! Don't use impact factors as measure of quality for individual researchers! Don't apply (hidden) bibliometric filters for selection, 
 e.g. minimum IF for inclusion in publication lists! Don't apply arbitrary weights to co-authorship. 
 Algorithms bases on author position might be problematic! Don't rank scientists according to one indicator. 
 Ranking should not be merely based on bibliometrics! ! Wolfgang Glänzel and Paul Wouters, July 2013 http://www.slideshare.net/paulwouters1/issi2013-wg-pw

5 Do not use journal-based metrics, such as Journal Impact Factors, as a surrogate measure of the quality of individual research articles, to assess an individual scientist's contributions, or in hiring, promotion, or funding decisions. ! ! San Francisco Declaration on Research Assessment, May 2013 http://am.ascb.org/dora/

6 Yet how do we actually know what excellence is and where it is worthwhile to foster a scientific elite? In reality, no one actually knows, least of all the politicians who enthusiastically launch such excellence initiatives. ! ! This is where the idea of artificially staged competition comes in. It is assumed that these competitions will automatically make the best rise to the top—without the need to care about neither content nor purpose of research. ! ! Mathias Binswanger, January 2014! ! Binswanger, M. (2014). Excellence by Nonsense: The Competition for Publications in Modern Science. In Opening Science (pp. 49–72). Springer-Verlag. 
 doi:10.1007/978-3-319-00026-8_3

7 PLOS tracks 30 different metrics for every article Viewed PLOS Journals (HTML, PDF, XML) PubMed Central (HTML, PDF) Figshare (HTML, Downloads, Likes) Cited CrossRef Scopus Web of Science PubMed Central PMC Europe PMC Europe Database Links Recom- mended F1000Prime Discussed Twitter Facebook Wikipedia Reddit PLOS Comments ResearchBlogging ScienceSeeker Nature Blogs Wordpress.com OpenEdition Saved Mendeley CiteULike http://articlemetrics.github.io The data and software are openly available, and are used by several other publishers and CrossRef Labs.

3 Social Media Metrics • quick (often hours instead of years) • demonstrate reuse beyond the scholarly community • reflect how we communicate in 2014 • can often be easily collected through open APIs ! but … ! • poorly predict future citations, scholarly impact unclear • have often not been standardized • may come and go over time • are sometimes easy to game ! ➜ limited value for impact assessment of individual researchers

3 Other Research Outputs • Research datasets • Scientific software • Posters and presentations at conferences, in particular in some disciplines, e.g medicine or computer science • Electronic theses and dissertations (ETDs) • Performances in film, theater, and music • Blogs • Lectures, online classes, and other teaching

3 NISO Alternative Assessment Metrics (Altmetrics) Project Three-year (2013-16) project funded by the Sloan Foundation, first year to collect input from community ! After talking to more than 400 experts and stakeholders, the project published a white paper with potential action items for further work this week: 
 http://www.niso.org/topics/tl/altmetrics_initiative/ National Information Standards Organization