The impact factor: its place in Garfield's thought, in science evaluation, and in library collection management

This paper is a response to that of Vanclay, who proposes, that since the impact factor (IF) is so seriously flawed, Thomson Reuters should either correct the measure or-preferably-no longer publish it and restrict itself to journal certification. It is argued here that Vanclay's analysis is itself seriously flawed, because he appears totally ignorant of the thought structure of Eugene Garfield, IF's creator. As a result, Vanclay appears unaware of the importance of total cites and the close connection of IF with review journals, where the paradigms of science are defined. This paper's author agrees that IF is a defective measure, analyzing its defects from the perspective of the frequency theory of probability, on which modern inferential statistics is based. However, he asserts that abandoning it would be counterproductive because of its demonstrated ability-even with its defects-to identify small important journals like review journals, giving it an important role in science evaluation and library collection management.