CARDIOVASCULAR JOURNAL OF AFRICA • Vol 23, No 8, September 2012
424
AFRICA
quartiles prior to (1981–1985), with after (1996–2000) the
incorporation of publications in the government formula.
Although the share was stable within each quartile prior to the
change in the funding formula, increases in the share occurred
within each quartile after the change. However, of concern
were the dramatically greater increases in the shares within the
lower two quartiles (~100 and ~50%, respectively) compared
to the increases in the shares in the higher two quartiles (only
~20% each).
13
Hence, although the change in the funding
policy resulted in noticeable improvements in the total number
of publication outputs (an increase of 25%) at a time when
academic staff numbers were stable,
3
this increase was at the
expense of quality.
3,13
A study comparing the research outputs by two Australian
universities lends further support to concerns of increased
quantity at the expense of quality when rewards were based
solely on the number of research outputs.
3
In this study, the
research outputs in terms of quantity and quality of two
Australian universities, which were using different systems to
enhance research outputs, were compared. In the university that
provided financial incentives based only on quantity, although
research outputs increased, there was a simultaneous decline
in publications in high-impact factor journals (a measure of
research quality).
3
In the university that used a strategy in which
the brightest young researchers were recruited and employed,
the total number of publications increased as well as the number
of publications in high-impact factor journals.
3
These studies
clearly demonstrate the negative impact of rewards based solely
on the counting of the number of publications.
In comparison to the Australian approach, in Spain a system
in which only high-quality publications were rewarded was
instituted in 1989.
6
To quote the Spanish parliamentary record,
a bonus is awarded to individual researchers only for ‘…those
articles of scientific worth in journals of recognized prestige in
the field. …. In those disciplines for which international systems
of quality of publications exist, reliance on these systems should
be obligatory…’.
6
Spanish research productivity doubled after the system of
publication bonuses was passed into law.
6
Indeed, twice as many
publications were produced between 1991 and 1998 compared to
between 1982 and 1990, and the number of Spanish publications
in ISI databases was increased.
6
These dramatic increments were
not attributed to factors such as increased financial support,
international collaboration or an increase in the number of staff.
25
The consequences of this law were therefore that research
productivity increased in both quantity and quality. Hence, the
Spanish system of rewarding only high-quality publications
appears to overcome the pitfalls of the Australian system,
whereby rewarding research outputs irrespective of quality
results in improved quantity at the expense of quality.
Further support of the use of a system based on rewarding
researchers for publication in journals with a high impact factor
was provided by Lomnicki
10
in his discussion on the positive
impact of such systems on research productivity in Germany
and France. Lomnicki
10
goes so far as to state that ‘abandonment
of objective methods of science evaluation derived from the
SCI’, would be most dangerous as ‘it would remove a tool for
rewarding researchers who attempt to do good science and for
eliminating those who do not’.
A lesson to be learnt from these previous experiences, is
that in order to avoid possible ‘salami slicing’ of research (data
that should constitute one publication is divided into many
smaller publications),
14,21,26
choosing in-house journals above
international journals,
21,27
and preferences for choosing journals
which are perceived to be easier to publish in,
21,26,27
the rewarding
of research outputs as a means to encourage and enhance
research outputs should be based on an assessment that includes
the quality of publications.
The question therefore arises as to why high-quality
publications are so important? The reason that some journals
have higher impact is that they generate greater citations. In other
words work published in higher impact journals is more highly
cited than work published in lower impact journals. Hence,
publications in high-quality journals are of greater value and
have a superior impact on the scientific/research community.
However, the favouring of high-impact factor journals for the
submission of manuscripts runs the risk of the journal choice
becoming more important than the scientific message of the
manuscript.
28
Indeed, it has been suggested that authors wishing
to have their work published in the journal
Nature
claim novelty
in their work, which is not entirely true, in order to enter into
the review process.
12
During the course of the review such
‘
false novelty’ is then identified and the manuscript is changed
accordingly prior to publication.
12
However, the result is that
work is published in
Nature
that is no more novel than that
published in high-quality specialised journals.
12
Nevertheless, there are other important benefits of sending
manuscripts to high-impact factor journals.
11
Firstly, authors are
more likely to receive meaningful feedback, as the top journals in
each field are most likely to consult the top reviewers in the field.
Secondly, the review process is generally more efficient from a
time perspective, in that reviewers for top journals are generally
given a maximum of 10 to 14 days to review a manuscript.
Collaboration
International collaboration between higher educational
institutions is indeed recommended in that it enhances the
citation and hence quality of research outputs.
3,29,30
However,
collaboration creates a dilemma with regard to which individuals
or higher educational institutions should be credited with
research outputs generated by collaboration. If each country
or institution involved is allocated a share of the collaborative
publication (fractional counting) then the publication count will
decrease.
Indeed, in a study of research outputs from Australian higher
educational institutions, it was shown that collaboration resulted
in a 17.8% reduction in the publication numbers for the period
studied.
3
Moreover, as the average citations per publication
were greater in those publications involving collaborators (5.53
vs 4.22), the total number of citations assigned to Australia
decreased by 24.5% for the period studied.
3
In the South African context, only those researchers affiliated
to a South African higher educational institution are credited by
the Department of Higher Education and Training. To translate
this into research output units, below are some examples of the
research output units generated by accredited journal publications
with and without national or international collaborators:
•
four authors all affiliated to one South African higher educa-
tional institution: 4
×
0.25
=
1
unit
