Measures of Progress of Science in India

SCIENCE, VOL. 86, NO. 9, 10 MAY 2004 1195
in physics had received low citations, between
0 and 7 per paper.
These facts only seek to highlight the
point that citations to a paper do not depend
upon journal IF, although journal
IF does depend upon citations. Certainly,
journal IF cannot be taken as the only
indicator for measuring the quality of a
research paper. It is more so because acts
of citation only seek to underscore the
theoretical and practical significance of a
paper, or its intrinsic value for future
research4. Citation count probably sounds
a more accurate and objective indicator
of research quality. Statistically also, the
Pearson correlation coefficient between
journal IF and citation count computed
on 902 records was low (0.21).
The pockets of excellence in research
seem to be confined to a select few papers
and not to any larger set of papers. For
example, only 10% papers accounted for
42% citations, 20% for 61% citations,
30% for 73% citations, and 50% for 89%
citations (Figure 1). This observation further
confirms the prevalence of a wide
gap in research evaluation based on journal
IF and citation frequency.
The study finds that there is wide disparity
in the evaluation of research measured
on journal IF and citation count.
As expected, research papers published
in high impact journals failed to receive
proportionately high citation, in six years
since their publication in 1997. This trend
was applicable to the bulk of Indian physics
output considered in this study. Majority
of such papers had their citation frequency
below the average citation count
for the whole sample. Nearly 13% papers
did not receive even a single citation in
six years since their publication.
The disparity is because citations do
not depend upon journal IF, although journal
impact does depend upon citations.
Citations depend mainly on the theoretical
and practical significance of the research
reported in the paper. Citation...