Outline

• Background
  • Collaboration amongst researchers
  • What are multiple affiliations
  • The why and wherefore of multiple affiliations
• Research landscape Japan, Germany, UK
• Data/Results
  • Trends in multiple affiliations (WoS)
  • Survey of authors
  • Multiple affiliations and research quality
• Conclusions

Background

• Collaboration in science
  • increase in co-author numbers (Adams et al. 2002; Wuchti et al. 2007)
  • increased research performance due to collaboration (Katz and Hicks 1997; Wuchty et al. 2007)
  • reasons: ease of travel/communication; concentration of resources; funders' support (Ding et al. 2010; Freeman et al. 2014)
• Multiple affiliations
  • attachment to more than one institution (public or private)
• Why and wherefore
  • researchers: tap into new networks/ gain additional resources
  • institutions: assessment / competition (Stephan, 2012 )
    • e.g. part-time positions to attract leading foreign scholars (Xin and Normile, 2006; Bhattacharjee, 2011)
  • institutions/researchers: maintain links

Background

Xin and Normile, Science, 313, 22 Sep 2006
Bhattacharjee, Science, 334, 9 Dec 2011

Background

Combined, part-time researcher positions will allow mobility and direct knowledge transfer and cooperation and may link institutions, disciplines, countries and sectors (industry/academia/public).

ESF - Science Policy Briefing - 2013: New Concepts of Researcher Mobility

NISTEP endeavors to utilize human resources from foreign countries. As part of these endeavors, NISTEP carries out joint research with foreign researchers, accepts foreign students, and accepts international affiliated fellows.

NISTEP - About Pamphlet - 2010

Japan - Germany - UK

• comparable in terms of academic output (BIS 2013; NISTEP 2014) and in innovative and technological capacity (WEF 2014)
• differ in terms of
  • governance of academia (e.g. autonomy, staff structures and career trajectories) (Teichler et al. 2013)
  • importance of the academic research sector compared to business and government research (OECD, 2014)
  • internationalisation (e.g. higher shares of foreign born staff in the UK) (Franzoni et al. 2012)
• in all three changes in
  • accountability (assessments, impact considerations)
  • funding distribution (growing importance of third party funding; performance based distribution of core-funding)
  • employment (growing number of part-time and fixed-term contracts, more flexible pay schemes)

Japan - Germany - UK

Article share (world)

Note: Based on documents listed in Scopus and generated by Elsevier
Source: BIS/13/1297

Japan - Germany - UK

Note: Based on documents listed in Web of Science
Source: 科学技術政策研究所「科学研究のベンチマーキング２０１２」調査資料ー218

Empirical strategy

Data

• Journal publications (articles, proceedings, book chapters) from web of science
  • Five journals randomly drawn from each quartile of the eigenfactor distribution in bioscience, chemistry and engineering
  • Period 2008 to 2014: ~ 30,000 articles, >118,000 author-publication pairs with at least one address in Germany, Japan or UK
  • Semi-manual coding of addresses
    • for Japan use of NISTEP dictionaries
• Survey
  • Sample
    • WoS sample - Articles published 2013-2015; additional journals drawn for engineering and economics/business
    • corresponding authors (~9000)
  • Design/response
    • June-August 2016, LimeSurvey web platform
    • 2,260 responses (response rate: 30.5%; 36.6% Japan; 31.1% Germany; 24.5 UK)
  • Questions on secondary affiliations (past and present)

Data

Data

Recap

• Survey - shows that publications are imperfect measure for MA but confirm some of the observations - authors do not report all affiliations. Selection based on journal, funder, coauthors, etc?
• Why and wherefore
  • researchers: tap into new networks/ gain additional resources
    • we saw evidence for this
  • weakend academic employment market: seek affiliations to increase job prospects
  • institutions: assessment / competition
    • some indication that this may be true
  • institutions/researchers: maintain links
    • not presented today, but 26% of affiliations intitated through prior employment

Multiple affiliations and performance

Conclusions / Limitations / Future Work

• MA and research quality
  • very sensitive to measure, but we can see differential effects for different countries
  • Engagement motivations have weak effect on publication impact
  • Results suggest that networks raises publication quality
    • Why Multiple Affiliations? Is Co-authoship not enough if goal is to overcome knowledge burden?
    • Does prestige of organisaton or what they offer not matter?
    • Initial higher citations - but what about the long run? Do we measure visibility effect instead of quality?
    • Is Institutions' push for MA the real driver?
• Cross-sectional data
  • Causality cannot be known
  • Longitudinal study of MA and publication performance