Appropriate use of bibliometric indicators for the assessment of journals, research proposals, and individuals

Bibliometric indicators provide numerical scales that are intended to quantitatively determine the value of scientific research and the scholarly publication in which that research is published. Since scientific performance cannot, of course, be directly "measured," citations acquired by each published paper are assumed as a proxy for quality, without prejudging the reasons for the citations. The application of bibliometrics to quantify the significance of individual journals dates back several decades,1 and the field has now reached a sufficiently high level of maturity to recognize that the scientific impact of journals as evaluated by bibliometrics is a complex, multidimensional construct and therefore more than one indicator is needed for such evaluation. Nearly all commonly used bibliometric indices, can be classified fundamentally as measuring either popularity or prestige, two concepts for which citation behaviors are valued in different and complementary ways. These indices also offer differing consideration of self-citations and have various levels of susceptibility to potential manipulation. As such, use of a single bibliometric index to rank, evaluate, and value journals is inappropriate. Rather, the use of multiple metrics with complementary features provides a more comprehensive view of journals and their relative placements in their fields. The IEEE, in its desire to fulfill its primary mission of fostering technological excellence for the benefit of humanity, recognizes the above concerns about the inappropriate application of bibliometrics to the evaluation of both scientists and research proposals. Therefore, the IEEE endorses specific tenets in conducting proper assessment in the areas of Engineering, Computer Science, and Information Technology as laid out in this document. The IEEE also recognizes the increasing importance of bibliometric indicators as independent measures of quality or impact of any scientific publication and therefore explicit- y and firmly condemns any practice aimed at influencing the number of citations to a specific journal with the sole purpose of artificially influencing the corresponding indices.