Citation Indices: Measuring the 'Impact' of Published Work.

Sir,

The increasing importance of scientific publications for academic growth[1] has put the spotlight on citation indices to evaluate the impact of published material. The total number of citations for a particular paper refers to the number of other scientific articles citing it in their bibliography. This is taken as a measure of the ‘importance’ of the index paper in the scientific community, as work of greater significance is likely to be cited more often. Citation indices are to researchers as impact factors are to journals. For an individual researcher's profile, h index is probably the most often used measure for scientific contribution. The h index refers to a number figure where ‘h’ number of articles of an author has ‘h’ or more citations.[2] The ‘h’ index when initially used for Nobel prize winning physicists showed a peak around 35. The current h indices of representative Nobel laureates (year 2014) in the field of physics (Isamu Akasaki), medicine or physiology (John O’Keefe), chemistry (Eric Betzig) and economics (Jean Tirole) are 58, 48, 32 and 47 respectively.[3] Two other common ways of ascertaining the scientific impact of an individual researcher includes total number of citations (sum of all citations from all published material), and i10 index (the number of articles having at least 10 citations). Other related measures are available which minimize the constraints of the ‘h’ index.[4]

Certain issues deserve consideration with regard to the number crunching involved in computing these citation indices. First, different computation services determine the number of citations of published material in different ways, or rather, using different databases. Citation indices obtained from Google Scholar might be numerically greater than those obtained from other indexing services like Scopus and Researchgate. This makes citation index data difficult to compare if derived from different sources. Second, the number of citations in a particular field would depend upon the total volume of scientific publications in that area. An important work in sociology may be less frequently cited than an important work in medicine. Third, for similar impact articles, those provided through open access may be cited more often than ones requiring paid subscription.[5] Fourth, the provision of search-customized keywords may favour the “discovery” of some articles over others during literature review, and hence they may be cited more often. Finally, senior researchers are likely to be cited more often not only because they have lead time in publication history, but also because of their reputation and camaraderie with other researchers in the field.

To conclude, citation indices are a step forward in determining the impact of research in an objective and computable manner. They assist in the holistic evaluation of research impact of an individual, journal, institution or even an entire country. However, their inherent limitations should also be kept in mind when gauging the research contribution.