bibliometric indicators of quality?
TRANSCRIPT
INDICATORS OF
QUALITY??
OUTLINE
INDICATORS OF QUALITY??
• Methods of quantitative evaluation• University level – Macro bibliometric analysis• Individual level – Micro bibliometric analysis
Lorna WildgaardRoyal School of Library and Information [email protected], F2016
STUDY OF QUANTITATIVE ASPECTS OF:
informetrics
bibliometrics
scientometrics
webometrics
cybermetrics
© Björneborn
altmetrics
Information in any form by any social group
Production, dissemination & use of recorded information
Science as a discipline or economic activity
Politico-economical aspects.
Construction & use of information resources & tech. on the internet
Web documents, text & multi-media, blogs, tags, wikis..
Dissemination, use & online activity of scholarly content
SO METRICS ARE:
”The application of mathematical and statistical methods to books and other media of communication.”Pritchard, A (1969), Journal of Documentation, 25(4):348-9
e=18
h=8
m=1
=3 CPP=18
t=4
PERCENT DISCIPLINARY COLLABORATION
C=
Nm = Number of multi-authored papers in the discipline, eg. 22929Ns=Number of single authored papers in the discipline, eg. 9325
C= = 0.71 * 100 = 71%
S c (i) = γ (Y (now) − Y (i) + 1)−∗ δ |C(i)|∗
Weighted citation score of an article
Actual Year minus publication year
Citation count
Coefficient set at 4, i.e. for an article published during the current year, its citations account four times
WHAT IS A PUBLICATION?
A scholarly publication is a an expression of the state of the researcher at a particular time (Price 1970)….the product of scientific research (Cawkell 1976; Lazarev 1996)….what consitutes a publication is different in the scientific disciplines (Castellini 2014; Hicks 2012; Tinkler 2011; Frandsen & Nicolaisen 2011)….different publication types indicate different methodologies (Grant & Booth 2009;Dzombak 2013)….Publications should be conceptualized as a discrete set of objects so they can be counted in aggregate (Skupin 2009)….output in a citation index (Antonakis & Lalive (2008).
What constitutes a ”publication” should be clearly defined to ensure representative operationalization in the indicator and the extraction of meaningful relationships (Wouters 1999; Skupin 2009; Colledge 2012)
WHAT IS A CITATION?Citations reflect som sort of cognitive influence (Moed 2005)….the strength of the influence is not known (Martyn 1964)….income or reward (Ravetz 1971)….meaning of citation is governed by paradigmatic & social norms (MacRoberts & MacRoberts 1996; Cronin 2000)….acknowledgement of intellecual debt to other works (Merton 1971)…. Markers or symbols of influence in the aggregate (Small 1978;1987)….citation impact (Martin & Irvine 1983)….rhetorical devices (Luukkonen 1997; Latour 1987)….authority (Gilbert) quality sensor mechanisms (Lindsey 1978)….A manipulable measure of something or other (Singleton 1976)….influence on average (Nicolaisen 2004)…citations are arbitrary (Leimu & Korcheva 2005)… markers (footprints) of scientific communication (Garfield 1979)….income (Albarrán 2011)……
The concept of citation is open for interpretation, of which there are numerous, and it is impossible to exclusively link citations to quality
WHAT IS A RESEARCHER?
Dias 1001/05/2023
WARM UP EXERCISEhttps://todaysmeet.com/Warm_Up
Handout: Researcher evaluation1. Suggest indicators that could measure
attributes, skills, interests and achievements.
2. Consider how to indicate the quality of these attributes, skills, interests & achievements.
EVALUATION
OUTPUT
DATA
MODEL
MOTIVATION TO PUBLISH
MOTIVATION TO CITE
INTERPRETATION OF ALI
LEGITIMATE LINK TO REAL WORLD
PERFORMANCE OF RESEARCHER IN SYSTEM REDEFINES WHAT SUCCESS IS
EXOGENOUS VARIABLES
WHAT IS MISSING
INDEXING POLICY
THE PROBLEM
COFFEE BREAK: Meet back here in 10 minutes
UNIVERSITY LEVEL
RESEARCH PERFORMANCE EVALUATION SYSTEMS
-Register the amount & type of a scholar’s publications. REF (UK): Panel-based evaluation every 6th year. Quality=peer review.
BFO (Flanders): Citation-basedQuality=reception & use
ERA (Aus)=Publication-based (udifferentiated). The number of publications dictates the amount of funds given to the university & the assessment of a researcher’s productivity.Quality= not assessed
16
Norwegian indicator 2004 (differentiated)Tiered system of publication types and sources.
Aim: increase publications in international journals and establish a registration system in which institutes can validate & verify publications as well as see other institutions publications
Quality=not addressed
BACKGROUND FOR BFI
BFI (DK) 2009-2012 (differentiated)Tiered system of publication types and sources
Aim: Distribute funding based on the quality of research results(UFM 2015a). Universities are judged based on published research.
Quality= explicit assumption: the journal's status is assumed to say something about the individual publication’s quality (UFM 2015a).
BFI NorgeQuality indicator Publications indicator25% of a sub-pool of basic funds, amount changes each year
2% of all university funds each year
Domain neutral Normalised. Benchmarked against total publication points and domain share
Unclear, complicated Clear, simplePublication performance Stimulate publication in international
resourcesAccountability model, aimed at distributing funds, production measured in Danish Crowns and legitimacy
Registration model
Fractionalised count ”Whole” countingInternational/eksternal collaboration rewarded with more points
Points normalised to the domain w.r.t. speed of publication, collaboration traditions and publication type
Book publishers on level 1 & 2 Book publishers on Level 1
18
TEACHING40%
FACULTY WORK23%
ADMIN.18%
RESEARCH & PUBLICATION
12%
NETWORK7%
WHAT IS THE BFI MEASURING?
HOW RE-SEARCHER
S SPEND THEIR TIME:
WEEKLY AVERAGE
Ziker (2014) https://thebluereview.org/faculty-time-allocation/
BFI
19
GROUP AREA FIELD COVERAGE
1-19 HUM
Language, Culture, History, Linguistics & Semiotics, Literature, Dance, Music, Theater, Art History And Art, Cultural Studies, Media & Communication, Film, Rhetoric, Theology, Philosophy, Archeology, Pedagogy, Education, Anthropology, IT & Humanities
20-27 SOC. SCI
Law, Sociology, Social Work, Political Science, Geography, Social Economics, Psychology, Library & Information Science
28-46 SCI/TEK
Geo-physics, Physics, Astronomy, Math, Chemistry, Biology, Environmental Science, Agriculture, Molecular Science, Pharmacology, Civil Engineering, Transport, Electronics, Energy, Chemical Engineering, Nanoscience, Medical Technology, System Development And Entrepreneurship, Food Science
47-62 SUND
Medicine, Surgery, Anesthesiology, Neurology, Ear-nose-throat, Eye Disease, Gynecology, Dermatology, Pediatrics, Pharmacology, Dentistry, Nursing, Public Health, Veterinary, Pathology, General Medicine
63-68 OTHERGender Studies, Architecture, Town Planning, Sport, Digital Communication, Information Systems & IT Management, Science Studies And Research Analysis
20
BFI STRUCTURE: PUBLICATION TYPESPublication types N. 1 N.2Monography in ISBN title (inc. reports, ex. Text books) 5 8Monography in book series 5 8Contribution to anthology, ISBN title (inc. introductory & final chapters; book chapters; ex., abstracts, lecture manuscripts, PPTs, encyclopedia, preface/afterword, commentaries, editorial work)
0.5 2
Contribution to anthology, bookseries 1 3Contribution to a conference, series 1 3Contribution to a conference 0.5 2Article (inc. letter, review ex., abstracts, editorial letter, commentary, book review) 1 3
Doctorate 5Patent 1
Publication types registered in PURE, PURE4 42 basic publishing types:http://bibliotek.science.ku.dk/forskdok/docs/metadatamodel_P4_LIFE_GR_270312.pdf
21
CALCULATION OF BFI POINTSPoints are shared between universities dependent on how many authors belonging to the university are listed on the publication:
A book on level 2 gives 8 points. There are 3 authors from AU, 2 from CPU & 1 from a foreign university.
AU = 3/6 pointsCPU=2/6 points8/6 = 1.3 points to each authorCollaboration with authors from other universities, Danish or foreign, is rewarded, The total number of points are multiplied by 1.25.AU total points= 1.3 * 3 =3.9 * 1.25 = 4.87 pointsCPU total points = 1.3 * 2 =2.6 * 1.25 = 3.25 points
BFI is domain neutral (funds are shared within domain and not across domains)
22
DISTRIBUTION KEY (fordelingsnøglen)
200945% STÅ INCOME 20%
EKSTERNAL10% PhD
25% BFI
2015
200850% STÅ INCOME 40% EKSTERNAL 10%
PhD
2000
Tax catalogue: http://ufm.dk/uddannelse-og-institutioner/videregaende-uddannelse/universiteter/okonomi/uddannelsesbevillinger
23
WHAT HAPPENS AFTER 2015?BFI must only distribute new funds and not alter the existing taxometer for distribution of basic university funds (DU 2009)
• New funds = globalization pool• The globalization pool is expected to be emptied in 2015• Suggested 25% BFI distributions parametre is supplemented by
taking funds from external grants• Basis funding is grouped together with other funding and
therefore effects distribution and share. fx:HUM greatest operational costs, greatest number of students, lowest STÅ tax and lowest amount of funds invested in researcher education, researcher support, administration and traditionally the lowest number of publications.http://www.altinget.dk/forskning/artikel/basismidler-fylder-mindre-paa-universiteterne
24
COLLABORATION PARAMETER
Publications
Level
Total Points%
Nationalpoints
Collaborative publications
% Collaborative publications
1 2
NAT/TEK 2921 1596 1325 2973.99 24 2406 82
SUND 3437 2379 1058 2851.26 42 3056 88
HUM 963 625 338 1502.72 34 185 19
SAM 739 448 291 1141.47 21 225 30
I ALT 8060 5048 3012 8469.44 30% 5872 72%
*Universities share of the total number of national points, given in percent
https://bfi.fi.dk/Publication/NationalAnalysis
25
PUBLICATION POINTS (UFM 2015)
% diff. i p.p 2013-2014 % diff. antal p. 2013-2014Aalborg U +35.3 +37Aarhus U +4.1 +2.6CBS +20.2 +11.6DTU -0.3 +3.7ITU +19.5 +12.5KU +0.3 -3.7RUC +7.7 -4.9SDU +16.1 -16.4
DOMAIN % diff. 2013-2014
NAT/TEK -0.5
SUND +16.9
HUM +15.3
SAMF +12.5
26
EVALUATION OF BFI
Good focus on strategic publishing. BFI works as an incitement to publish in certain resources.
Yet, the development of longterm, dialogue based communication forms with public society could be ignored in a quest to fulfill the requirement of selling science through an unidirectional intemediatory (red.: journals).
(Schneider, 2015)
27
VISIBILITY OF HUM IN CITATION INDICES
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 20150
50
100
150
200
250
300
350
HUM i WOSHUM i Scopus
Publication year
Num
ber o
f pub
licati
ons
BFI
P=963 BFI
P=813 BFI
28
EVALUATION OF BFI
UFM(2015) 1. BFI lacks transparancy2. Need for better communication between stakeholders3. Citation indicators could be advantages in some
domains. This will enable analysis of international competition.
The performance of BFI in the model to distribute funds was not assessed
29
CONSIDERATIONSBFI results in an increase of research publications in level 1 and 2 sources, and increased collaboration (Ingwersen & Larsen, 2014)
BFI results in an increased attention on the researcheremploymentpromotion funding applications
Under contract to publish in BFI resources and produce publication points(AAU Bestyrelse, 2014; FOU., 2013; Strandskov, 2011)
LOOKING TO THE FUTURE…
Does BFI prioritize what is measurable over what should be measured but is difficult to quanitify? (Aagaard 2012)
• Leiden manifesto: http://www.leidenmanifesto.org/
• Snowball Metrics: http://www.snowballmetrics.com/
• Responsible Metricshttps://responsiblemetrics.org/
COFFEE BREAK: Meet back here in 10 minutes
INDIVIDUAL LEVEL SCIENTOMETRICS
TYPES OF INDICATORS
Hybridmetrics
PUBLICATION METRICS count production, collaboration, &
contribution
CITATION METRICS count citations, self-citations, adjust for
collaboration
Table of indicators: http://tinyurl.com/nj4mvca
Ranking, growth, proxy for excellence, effect, field comparisons & quality
INTERNAL CHARACTERISTICS OF ALI
Wildgaard et al (2014) A review of the characteristics of 108 author-level indicators. Scientometrics, 101(1):125-158
Age: years since first publication (WoS) = 2016-2009 = 7Papers: all publications (WoS)=13Citations: total citations (WoS)= 190
1st subject area=Anesthesiology2nd subject area=Neuroscience3rd subject area=Surgery
INDEX OF QUALITY AND PRODUCTIVITYhttp://tinyurl.com/zlxzyz9
Antonakis, J. and Lalive, R. (2008), Quantifying Scholarly Impact: IQp Versus the Hirsch h. J. Am. Soc. Inf. Sci., 59: 956–969. doi: 10.1002/asi.20802http://www.hec.unil.ch/jantonakis/IQp%20calculator%20version%202008.xls
INDEX OF QUALITY AND PRODUCTIVITY
190
Cited 1.44 times more than average paper in the specialty
Number of papers performing above average for the specialty
JIF of 3 subject categories
ANOTHER ASPECT OF QUALITY?WEB/ALTMETRICS
WEB 2.0 = PARTICIPATIVE WEB
”… supports, extends, or derives added value from human social behaviour …”
TWO ELEMENTS IN NETWORKS
NodeNodes represent things that relate somehow to one another
TieTie represents relations between nodes
© James Cook https://www.youtube.com/channel/UCerAw4EfTOnYYxLLPZAzMxQ
TWO KINDS OF NETWORKS
DIRECTED NETWORKSNetworks in which the tie has a driection (digraphs), eg. Linking to a webpage, downloading a document, citation networks
UNDIRECTED NETWORKSNetworks in which the tie has no direction eg. friends on Facebook,
© James Cook
INDUCED HOMOPHILYTendency for ties to form
similar others because similar others present the
social environments (group, community,
society)
No wonder blues are mostly tied to blues…there are hardly any reds out there!
© James Cook
CHOICE HOMOPHILY
A tendency to choose to form ties with similar others even when different others are available in the social environment (group, community, society)
© James Cook
PRINCIPLES OF NETWORKS• Ties should be straight• Ties should not be far from each
other• Ties should not cross or touch• Ties should be easy to follow from
node to node• Nodes that connect should be close• Similar nodes should be close to
one another• Central nodes should be in the
center
© James Cook
http:
//no
dusla
bs.c
om/r
esea
rch/
know
ledg
e-gr
aphs
-typ
e-do
cum
ent/
LINKED IN: CONNECTIVITY
Quick-Diverse-Broad
A record of attention: how many people have been exposed to and engaged with a scholarly output in the news, blogs, and on Twitter; article pageviews and downloads; GitHub repository watchers.
A measure of dissemination: where and why a piece of research is being discussed and shared, both among other scholars and in the public sphere.
An indicator of influence and impact: Some altmetrics can signal that research is changing a field of study, the public’s health, or having any other number of tangible effects upon larger society, eg. references in public policy documents; or commentary from experts and practitioners.
An output• Journal Article• Dataset• Poster
An identifier attached to the output• DOI• PubMedID
Mention in a source• Social media• Blogs• Wikipedia• Media
API stands for Application Programming Interface
ALTMETRIC.COM
The colours of the donut each represent a different source of information.
The attention score is a weighted count of all of the attention a research output has received.
http://tinyurl.com/z7zozr5
Demographic breakdown Count %Members of the public 153 18Scientists 31 16Science communicators (bloggers, journalists, editors)
5 3
Practitioners (doctors, other healthcare professionals)
3 2
Indicators are designed to measure particular aspects of the effect of a researcher’s work – over time, to field, as quality, ranking all or selected works, co-authorship etc.
Judgements based on indicators can lead to assumptions about the productivity and impact of a researcher, which can be unsubstantiated, and affect the psychological character of the individual.
ALI have to be methodologically sound
WHAT WE HAVE TO BE AWARE OF
51
ALI GIVE A SNAP-SHOT OF SELF IMAGE AND CORE PERSONALITY TRAITS
COMPARING RESEARCHERS CAN EXPOSE THE INDIVIDUAL
INDIVIDUALS USE ANY DATA TO INCREASE ALI SCORES, TO INCREASE THEIR SUBJECT VALIDITY & SELF-WORTH
ALI BRING OBJECTIVITY TO THE EVALUATION & REDUCE GENDER; CULTURAL AND RACIAL BIAS
DOCUMENTING BEING OUT-PERFORMED IS DETRIMENTAL TO A RESEARCHER’S SELF-DEFINITION
ALI DO NOT ADD CONTEXT BUT CAN ADD REDUNDANT INFORMATION
SUCCESS IS DEFINED AS WELL IN THE SYSTEM
EXTERNAL CHARACTERISTICS OF ALI
1. TransparencyKnow which data is used to compute the
indicatorUnderstand the math and the inferences
2. DemographicsBe aware if demographics affect the indicator
3. MotiveIndicators must fit the objectives of the
evaluation4. Diversity
Choose indicators that fit the discipline/publication
5. OpenessMake the limitations of indicator explicit, use supplementary indicators
APPROPRIATE INDICATORS
“CAN WE ASSESS THE BEAUTY OF
THE MONA LISA BY COUNTING THE
NUMBER OF VISITORS TO THE
LOUVRE?”
GROUP DISCUSSION. WRITE YOUR IDEAS AT https://todaysmeet.com/Indicators_of_Quality
statistics 5,074,920
www.wikimindmap.org
http://en.wikipedia.org/wiki/Wikipedia:Statistics
http://stats.wikimedia.org/EN/TablesWikipediaEN.htm
STATISTICS up to 2009/2010
https://en.wikipedia.org/wiki/Wikipedia:Database_reports
59
REFERENCESAagaard, K & Mejlgaard, N (2012) Dansk forskningspolitik efter årstusinskiftet. Aarhus Universitetsforlag
AAU (2013). Newsletter from the Rectorate, October 2013. Retrieved 23-6-2015 from: http://www.intern.aau.dk/ansatte/ledelsen-AAU/aau-rektoratet/rektoratets-nyhedsbrev/vis-nyhedsbrev//nyt-fra-rektoratet-oktober-2013---newsletter-from-the-rectorate-october-2013.cid97315#bfi
Barbulescu, Roxana. 2011. European Research Council’s Grant for Advanced Investigators in Social and Political Sciences. European University Institute.
DS 2014: Svagtfald I statens forskningsbudget(Nyt fra Danmarks statistik, 28 jan 2014) http://www.dst.dk/pukora/epub/Nyt/2014/NR039_1.pdf
DU. (2009). Politikpapir om model for fordeling af basismidler. Faggruppe 68. Retrieved 23-6-2015, from: http://faggruppe68.pbworks.com/f/Politikpapir_170409__til_hjemmesiden_.pdf
60
REFERENCES
DU. (2012). Overblik over finanslovforslaget for 2013. Retrieved 23-6-2015, from: http://www.dkuni.dk/Politik/~/media/Files/Politiknotater/notat%20om%20ffl13%20(3).ashx
Emmeche, C. (2014). Den Bibliometriske Forskningsindikator - fordele og ulemper. Faggruppe68. Retrieved 23-6-2015, from: http://faggruppe68.pbworks.com/w/page/6015700/Den%20Bibliometriske%20Forskningsindikator%20-%20fordele%20og%20ulemper
FOU. (2013). Publiceringsstrategi for Rosklide Universitet 2015-2018. RUC Hearing. Retrieved 23-6-2015, from: www.ruc.dk/?eID=minute&m=1948&t=179&docID=30986
Hansen, L. (2009) ‖Hvorfor er en humanistisk artikel mere værd end en sundhedsvidenskabelig?‖ Indlæg på bloggen ‘Forskningsfrihed‘ 12/10-2009.
Ingwersen & Larsen (2014) Influence of a performance indicator on Danish research production and citation impact 2000–12. Scientometrics, Online 19/4 2014; DOI 10.1007/s11192-014-1291-x.
61
REFERENCESMeo, Sultan A., Abeer A. Al Masri, Adnan M. Usmani, Almas N. Memon, and Syed Z. Zaidi. 2013. Impact of GDP, spending on R&D, number of universities and scientific journals on research publications among Asian countries. PloS One, 8(6), e66449. doi:10.1371/journal.pone.0066449
Podlubny, I (2005) A note on comparison of scientific impact expressed by the number of citations in different fields of sciencehttp://arxiv.org/ftp/math/papers/0410/0410574.pdf
Schneider, J (2015) Publications or Citations – Does it matter? Beneficiaries in two different versions of a national bibliometric performance model, an existing publication-based and a suggested citation-based model. ISSI, Bogaziçi University Printhouse, (2015), from: http://www.issi2015.org/files/downloads/all-papers/0477.pdf
Sivertsen, G. (2009). A Bibliometric Funding Model based on a National Research Information System . Norwegian Institute for Studies in Innovation, Research and Education. Retrieved 23-6-0015, from: http://www.issi2009.org/agendas/issiprogram/public/documents/ISSI%202009%20Sivertsen%20Vista-094456.pdf
Strandskov, J. (2011). Referat af Institutledermøde den 6. december 2011, Syddansk Universitet. Retrieved 21-04-2015, from: board.sam.sdu.dk/Download.mvc/2889
UFM (2015a). Basismidler efter kvalitet. Retrieved 22-6-2015, from: http://ufm.dk/uddannelse-og-institutioner/videregaende-uddannelse/universiteter/okonomi/basismidler-efter-kvalitet
62
LINKSThe Ref indicator explained: http://www.ref.ac.uk/
Sherpa/Romeo. Publisher Copyright policies & self-archiving http://www.sherpa.ac.uk/romeo/search.php
Higher Education Funding council for England:http://www.hefce.ac.uk/news/newsarchive/2015/Name,103785,en.html
The BOF indicator explained (Dutch): http://data-onderwijs.vlaanderen.be/edulex/document.aspx?docid=14492
The Norwegian Indicator: https://www.regjeringen.no/no/tema/helse-og-omsorg/sykehus/nasjonalt-system-for-maling-av-forskning/id446980/#Ind
BFI: http://ufm.dk/forskning-og-innovation/statistik-og-analyser/den-bibliometriske-forskningsindikator