1
Research emphasis and collaboration in Africa
Anastassios Pouris
1
and Yuh-Shan Ho
2,*
1.
Institute for Technological Innovation, University of Pretoria, Pretoria, South Africa
2.
Trend Research Centre, Asia University, Taichung 41354, Taiwan
*Corresponding author: e-mail: ysho@asia.edu.tw
Abstract
Scientific co-authorship of African researchers has become a fashionable topic in the recent
scientometric literature. Researchers are investigating the effects, modes, dynamics and
motives of collaboration in a continental research system which is in an embryonic stage
and in different stages of development from country to country. In this article we attempt to
provide some additional evidence by examining both patterns of collaboration at country
and continental levels and the scientific disciplines emphasised. Our findings indicate that
the continent’s research emphasises medical and natural resources disciplines to the
detriment of disciplines supporting knowledge based economies and societies. Furthermore,
we identify that the collaborative patterns in Africa are substantial higher than in the rest of
the world. A number of questions related to research collaboration and its effects are
raised.
Keywords
Africa; Collaboration; Research; Co-authorship; Scientometrics
Introduction
Research collaboration is a sociological phenomenon that is receiving the attention of
researchers and governments internationally (Yeung et al. 2005). Researchers are
investigating the effects, modes, dynamics and motives of collaboration, while governments
utilise research collaboration as a policy instrument for technology transfer from
universities and research councils to industry, for knowledge transfer from abroad, as a
means to improve diplomatic relations with other countries by creating goodwill, and to
gain political capital (Wagner et al. 2002a). Researchers collaborate with each other for
various reasons. This can be to improve their visibility and recognition (Narin et al. 1991), to
utilise expensive equipment that is not under their control (Meadows and O’Connor 1971;
Schubert and Sooryamoorthy 2010), or to acquire expertise and new ideas (Beaver and
Rosen 1978) needed for research.
In the policy domain, scientific collaboration has become an important component of
science, technology and innovation policy internationally, with substantial resources being
allocated by governments for this objective. Wagner et al. (2002b) estimated that the USA
was spending US$3.3 billion in the mid-1990s on international research collaboration.
2
Similarly, other developed countries were spending substantial amounts as a percentage of
their gross domestic product (Wagner et al. 2001). Russell (1995) and Wagner et al. (2001)
have suggested that international collaboration is replacing other models as the preferred
method of building scientific capacity in developing countries.
While investigations identify the benefits to be derived from collaboration (at least in the
currency of science i.e. citations), this collaboration is not without debate related to the
risks and benefits of such activities. Arguments expressed include the concern that the
spending on international collaboration is not always to the benefit of the paying country
and that critical technologies and key knowledge for competitiveness are given away to
competitors. Additional concerns have been voiced that collaborative agreements are
subordinate to the interests of science and technology to strategic or political ends.
Similarly, in the academic domain, researchers have argued that collaboration may be an
endogenous self-perpetuating outcome of science, with substantial costs and no
commensurate benefits (Jones et al. 2008).
An issue that has received attention and is of importance in the context of Africa is the
dependency of the size of collaboration on the size of the scientific community. Narin et al.
(1991) found that international co-authorship is higher for scientifically small countries.
They argued that scientists in scientifically small countries have far more scientists outside
their country with whom to cooperate and far fewer inside their country than scientists in
much larger scientific countries do. The argument appeared to be that the collaborative
effort is initiated by researchers in small countries who cannot find collaborators.
However, Melin (1999) concluded that “the results indicate that the situation is much more
complex than that large country researchers collaborate less internationally than small
countries as their scientists more easily can find their partners within the national borders
than in smaller countries.” Similarly, Boshoff (2009) identified that north–south
collaboration takes place in a particular format with the south collaborator basically
assisting in fieldwork and data collection. In other words the developed countries’
researchers seek collaboration in order to access data and conditions available in the
developing countries.
Historically studies on research collaboration were focused on or used data from
industrialised countries. More recently, a number of such studies include developing
countries in general (Arunachalam and Viswanathan 2008) and African countries (Boshoff
2009; Sooryamoorthy 2009) in particular. Sooryamoorthy (2009) investigated the
collaboration patterns of South African researchers and Boshoff (2010) identified the
collaborative patterns in the Southern African development community (SADC) countries.
Onyancha and Maluleka (2011) found out that knowledge production through collaborative
research among sub-Saharan African countries is minimal. Schubert and Sooryamoorthy
(2010) showed that “a theory of scientific collaboration building on the notion of marginality
and centre-periphery can explain many facets of South African-German collaboration, where
South Africa is a semi-peripheral region, a centre for the periphery, and a periphery for the
centre”.
3
In the context of African collaboration it should be emphasised that scientometric studies in
general and collaboration studies in particular are in an embryonic stage on the African
continent. Even South Africa, which is the major producer of research publications on the
continent, produces few publications in the field of scientometrics (Pouris 2012).
In this article, the authors use co-authorship analysis to identify the state of research
collaboration on the African continent. The questions they attempt to answer are as follows:
Which scientific disciplines are emphasised in Africa?
How did research collaboration evolve in Africa during the period 2007–2011?
Who are the main research partners of African countries?
Are the patterns of collaboration (extended and disciplinary) in Africa similar to
those in the rest of the world?
How do the various African countries perform in terms of collaboration?
Which are the main African institutions that are actively engaged in collaboration?
This article goes on to outline the approach the researchers followed and the data sources
used. It follows a results and a discussion section and the article ends with conclusions.
Data sources and methodology
Since Price and Beaver (1966) used co-authorship as an indicator of research collaboration,
it has become an established method, and a multitude of articles have investigated this
phenomenon. The approach has gained popularity, even though it is not without criticism
(Katz and Martin 1997; Laudel 2002). In this article we use co-authorship analysis in order to
identify the collaborative patterns of African researchers.
Data used in this study was retrieved from the Thomson Reuters Web of Science. Again it
should be mentioned that bibliometrics in general and the use of particular databases in
particular may have their own shortcomings (Roland 2007; Leydesdorff 2008). For this
investigation it may be relevant that African countries may publish their research in local
journals and languages which are not covered by the Web of Science. However, we should
emphasize that in South Africa the government and the university authorities take actions
and provide incentives so the researchers publish in the web of Science indexed journals.
The online version of the Science Citation Index Expanded (SCI-Expanded) was accessed on
18 March 2013. In this study, all journal articles in the SCI-Expanded version that were
published by authors on the African continent were selected and analysed in order to
identify publishing institutions and countries, and to classify articles as collaborative and
single-authored publications. The database was searched using the keywords “Algeria”,
“Angola”, “Benin”, “Botswana”, “Burkina Faso”, “Burundi”, “Cameroon”, “Cape Verde”,
“Cent Afr Republ”, “Chad”, “Comoros”, “Congo”, “Cote Ivoire”, “Dem Rep Congo”,
“Djibouti”, “Egypt”, “Equat Guinea”, “Eritrea”, “Ethiopia”, “Gabon”, “Gambia”, “Ghana”,
“Guinea”, “Guinea Bissau”, “Kenya”, “Lesotho”, “Liberia”, “Libya”, “Madagascar”, “Malawi”,
“Mali”, “Mauritania”, “Mauritius”, “Morocco”, “Mozambique”, “Namibia”, “Niger”,
“Nigeria”, “Rwanda”, “Sao Tome and Prin”, “Senegal”, “Seychelles”, “Sierra Leone”,
“Somalia”, “South Africa”, “South Sudan”, “Sudan”, “Swaziland”, “Tanzania”, “Togo”,
4
“Tunisia”, “Uganda”, “Western Sahara”, “Zambia”, “Zimbabwe” and “Zaire” in the address
field.
The researchers limited the publication year to between 2007 and 2011, and articles were
the only document type considered. Document information such as names of authors, title,
year of publication, source journal publishing the articles, contact address, research areas in
the Web of Science subject category were downloaded using Microsoft excel. Additional
coding was performed manually in order to identify the institutional address of the
collaborators.
Affiliations originating from England, Scotland, Northern Ireland and Wales were reclassified
as being from the UK (United Kingdom). “Dem Rep Congo” and “Zaire” were reclassified as
being from the Democratic Republic of the Congo. Collaboration type was determined by
the affiliation of the authors, where the term “internationally collaborative publication”
(ICP) was assigned to those articles that were co-authored by researchers from at least two
countries. The term “inter-institutionally collaborative publication” was assigned to those
articles that were co-authored by researchers from at least two institutions (Li and Ho
2008). The term “institutional independent article” was assigned to articles where the
researchers’ affiliation was from the same institution. Similarly, the term “African
collaborative publication” (ACP) was assigned to articles if authors’ affiliations were from
different countries on the African continent. The term “outside African continent
collaborative publication” (OCP) was assigned if articles were co-authored by authors from
Africa and authors from countries outside the African continent. The identified articles were
further allocated to the Web of Science subject categories. The journal citation reports (JCR)
of 2011 indexes 8,336 journals, classified across 176 web of science categories.
Results and discussion
Language of publication
A total of 112,576 articles were identified. In order to confirm that these articles were
published by authors on the African continent, the researchers further examined the
affiliations of authors, and excluded articles that were not published by authors in countries
on the African continent, which had been accidentally included in the original set. A total of
111,877 articles published by authors in African countries between 2007 and 2011 were
therefore analysed. These articles were published in 17 languages, with the majority of
them (97 %) being published in English. The non-English language articles were published in
French (3,396 articles), German (51), Spanish (39), Portuguese (16), Italian (6), Korean (5),
Chinese (4), Russian (3), Arabic (2), Croatian (2), Dutch (2), Japanese (2), Turkish (2),
Hungarian (1), Polish (1), and Welsh (1). The importance of the French language was not
surprising, since a number of countries on the African continent were French colonies
(Chuang et al. 2011).
Output in research areas
Figure 1 shows the distribution of research articles in the various countries. The African
publications were allocated to various research areas as categorized in the web of science
5
categories. Table 1 shows the research areas emphasised in the continent, while Table 2
shows the areas that are underemphasised. These tables also show the number of world
publications in the particular fields, the number of African publications, the African share
and the activity indices. The activity index characterises the relative research effort a
country/region devotes to a given field. It is defined as the country’s share in the world’s
publication output in the given field, divided by the share of the country/region in the
world’s publication output in all science fields. An index above one means that the region
overemphasises the particular field above the world average. An index below one indicates
an effort below the world average. An index of one indicates that the region’s effort in the
particular field corresponds precisely to the world average.
Fig. 1. Distribution of articles published
