MEASURING THE IMPACT OF SCIENTIFIC PUBLICATIONS. THE CASE OF THE BIOMEDICAL SCIENCES

The bibliometric indicators currently used to assess scientific production have a serious flaw: a notable bias is produced when different subfields are compared. In this paper we demonstrate the existence of this bias using the impact factor (IF) indicator. The impact factor is related to the quality of a published article, but only when each specific subfield is taken separately: only 15.6% of the subfields we studied were found to have homogeneous means. The bias involved can be very misleading when bibliometric estimators are used as a basis for assigning research funds. To improve this situation, we propose a new estimator, the RPU, based on a normalization of the impact factor that minimizes bias and permits comparison among subfields. The RPU of a journal is calculated with the formula: RPU=10(1-exp (-IF/x)), where IF is the impact factor of the journal and x the mean IF for the subfield in which the journal belongs. The RPU retains the advantages of the impact factor: simplicity of calculation, immediacy and objectivity, and increases homogeneous subfields from 15.6% to 93.7%.