Tools for citation analysis
FAQ: How can I find out which works are most cited?
Even though the decision has been made not to use citation counts as the primary measure of research excellence in REF2014, there are several good reasons why you might be interested in the number of times your work, or that of others, has been cited. For example:
- to measure and track the uptake of specific research outputs by the scholarly community
- to enhance bids for funding
- to demonstrate personal, group and institutional research performance
- to support career advancement
In this series of posts I will introduce some of the tools that you can use for citation analysis and explain briefly how you can access them.
But first, a word of warning:
Citation indicators are easy to benchmark, objective and globally comparable.
However, they all suffer from a number of limitations; including:
- Incomplete and uneven coverage – each citation index draws its information from a fixed pool of data (i.e. citations). The resulting metrics are entirely determined by the composition of this pool.
So, for example, to achieve a high citation count in Web of Science both your work and the works that cite it must be in publications that are indexed by Web of Science. It therefore follows that if your discipline or publication type is not well covered by Web of Science then your Web of Science citation count will not be accurate.
- Negative citation – an incorrect or controversial paper may attract as many citations (or even more) as a highly respected paper.
- Non-unique and inconsistent author naming – inconsistencies are rife in the use and citation of full names, shortened names, one or multiple initials, changes of name (e.g. on marriage) and so forth. Add to this the likelihood that there will be other authors sharing the same name and the opportunities for miscalculation are enormous.
- Differences in disciplinary citation practices – if your research field is very small or the normal method of disseminating research is not via citable journal articles and conference papers then citation indicators will not provide a true reflection of scholarly impact. For this reason you should not attempt to compare citations in different disciplinary areas.
- Citation ‘circles’ – the practice among some researchers and research groups of habitually citing their own work and that of their colleagues in order to increase citation counts weakens the validity of the measure.
Putting these drawbacks to one side, if you still wish to perform citation analysis, these are some of the tools that will help. Each is addressed in a separate post:
Image credit: diylibrarian on flickr