A Proposal for Topic‐based Impact Factors and their Application to Occupational Health Literature

The Journal Impact Factor (JIF) is a bibliometric index calculated by dividing the number of current year citations by the number of source items published in a journal during the previous (usually two) years. The JIF is therefore a measure of the frequency with which the average article in a journal has been cited in a particular year. To the extent that the number of citations is accepted as a measure of influence, the JIF can be considered a measure of the influence of a given journal. However, the number of citations of individual papers within a particular journal is known to follow a skewed distribution, so there is a poor correlation between individual citation count and the JIF. This means that the JIF is a poor surrogate for individual citation counts. Nevertheless, the JIF is often considered to be a proxy measure of the individual research papers published in a particular journal. This is partly because JIF values are more accessible than individual citation counts, although both measures are derived from the same database. A further problem is that the JIF is widely accepted, wrongly in our view, as an indication of journal quality, which has an impact on researchers, research sponsors and a more general audience. This can lead to the bad practice of judging researchers and institutions on the basis of the JIFs of their publications. Many researchers have pointed out the problems inherent in the JIF, such as selection of source journals, period for counting citations, and problems related to its misuse and/or abuse. In response to the many criticisms raised against JIFs, several alternative proposals have been advanced but have done little to change the situation. Any alternative method should not only rectify the inherent problems but should also be interpretable, feasible, simple to calculate, and be applicable to all scientific disciplines. These goals can be achieved by simply replacing, as the unit of analysis, the journal by a group of papers dealing with a similar theme. This produces a novel citation index that can be used as a reference standard for comparing individual citation counts with the average performance of papers dealing with similar topics. The basic concept of this method has already been proposed and even evoked a favorable response from the founder of the JIF. The objectives of this paper are to apply this new concept to the occupational health literature and to discuss its implications therein.