Participation and publication practices in global team science
Jarno Hoekman
School of Innovation Sciences
Eindhoven University of Technology
j.hoekman@tue.nl
Background
Collaboration and globalization are defining characteristics of contemporary scientific
knowledge production. As science is increasingly organized in multi-national teams (Jones et
al. 2008), much attention is paid to the benefits that team science brings. These benefits are
said to be considerable (Sonnenwald 2007) and include the sharing of knowledge and
research infrastructure, the production of scientific knowledge with more diverse intellectual
inputs, and the opportunity to solve issues of global relevance such as global warming and
epidemic diseases. Yet, much of the alleged benefits of team science hinge on the assumption
of equal negotiation power and participation of team members. This is a questionable
assumption, especially when scholars from peripheral locations increasingly become involved
in team science. To shed a light on this issue I study under which conditions participation of
researchers from low-income countries in global team science result in the provision of credit
as measured by scientific authorships.
I take clinical research as an example in this study. Large clinical trials require the
involvement of many individuals and sub-teams, with tasks ranging from conceiving the
study and designing the protocol to recruiting human subjects for data collection.
Standardization and harmonization of research practices - as promoted within an evidence
based medicine paradigm (Timmermans and Berg 2003) - has made the outcomes of research
conducted at geographically dispersed sites portable and acceptable to national regulatory
bodies worldwide. As a consequence, the geography of clinical medicine has rapidly started
to change and clinical investigators from low-income countries are increasingly incorporated
in multi-national scientific teams to test the safety and efficacy of new therapies (Petryna
2009).
However, despite the inclusion of these investigators, worries have been voiced over
their exact role in clinical research. Critics argue that investigators in non-traditional research
locations are only hired to execute protocols using their patients as ‘experimental subjects’
without having significant roles in defining research questions, designing the protocols or
analyzing the data (Petryna 2009). In addition, there are concerns that embedded relationships
between sponsors and clinical investigators fail to materialize, as the decision where to recruit
human subjects is made on a project-by-project base (Azoulay 2003). With such a division of
labor in place, learning opportunities are expected to be limited for investigators in
developing countries and ‘global’ team science continues to primarily serve scientists,
patients, and markets in developed countries.
Data and Methodology
To shed an empirical light on these issues, I study the conditions under which the
participation of sub-teams from low-income countries in clinical trials, result in the provision
of credit for their work as measured by authorships on the publications resulting from these
trials. In doing so, I use detailed protocol information on the funding sources, research
designs, diseases targeted and cities where human subjects are recruited for approximately
32,000 completed clinical trials. The data is extracted from www.clinicaltrials.gov, a publicly
accessible register of clinical trials managed by the US National Library of Medicine.
Registration of ongoing and completed clinical trials in this database is now mandated by
national laws in various countries and is also required before clinical trial outcomes are
considered for publication in peer-reviewed journals. For approximately 10% of this sample I
establish a 1:1 match between a study protocol and the publication that reports on the
outcome of the study protocol. Using this information I compute participation-to-publication
rates (i.e. PtP rates) for countries and cities which serve as the main explanatory variable in
this analysis.
Results
Descriptive analysis shows that PtP rates are significantly lower for sub-teams located in
lower income countries (Figure 1). Lower PtP rates are already apparent when considering
authorship as a whole but become even more marked when focusing exclusively on firstauthorships and last-authorships as an indicator of leadership in scientific projects. At the
city-level, there also tends to be a rural-urban divide in PtP rates, which is particularly
pronounced in low-income countries.
In a preliminary multi-level analysis to explain the determinants of PtP rates, I show that PtP
rates are significantly lower when the research is industry-funded and when the research
targets diseases that are mainly of significance to major markets in high-income countries. I
also study whether more frequent participation in clinical trials - which I interpret as the
opportunity to learn - renders higher PtP rates. Although we find evidence that learning
effects are present, most sub-teams in developing countries and especially those in rural areas
do not profit from this effect because their participation tends to be rather short-lived, before
the sponsor moves on to recruit human subjects at other locations.
Interpretation
The preliminary results confirm a strong division of labor in clinical team science, where
human subjects increasingly become recruited in developing countries but knowledge claims
about the data obtained from these human subjects remain disproportionally produced in
high-income countries. Moreover, due to the mobile character of clinical trial activities,
embedded relationships fail to materialize and learning effects in low-income countries are
generally low. The results signal limitations in the publication system to provide credit where
credit is due. Moreover, as the observed outcomes are particularly pronounced in industryfunded research, they also casts major doubts on the ability to safeguard prevailing norms of
scientific conduct (e.g. access to data, possibility to publish) in industry-funded projects as
opposed to publicly-funded projects.
References:
Azoulay, P. (2003) Agents of embeddedness, NBER Working Paper 10142. Available at
http://www.nber.org/papers/w10142.
Jones, B., Wuchty, S., Uzzi, B. (2008) Multi-university research teams: shifting impact,
geography and stratification in science. Science, 322, 1259-1262
Petryna, A. (2009) When experiments travel: clinical trials and the global search for human
subjects. Princeton University Press: Princeton.
Sonnenwald, D. (2007) Scientific collaboration. Annual Review of Information Science and
Technology. 41, 643-681.
Timmermans, S., Berg, M. (2003) The gold standard: the challenge of evidence-based
medicine and standardization in health care. Philadelphia: Temple University Press.