The Lack of Stability of the Impact Factor of the Mathematical Journals

Introduction Although the 2-year Thomson-Reuters Impact Factor (IF) has become a usual tool for measuring the scientific productivity of all fields of the natural sciences (see Aleixandre-Benavent, Valderrama Zurián, & González Alcaide, 2007), its behavior in the particular case of the journals of pure mathematics (the area MATHEMATICS in the thematic directory of Thomson-Reuters) is far from being stable when its values in consecutive years are considered. If we consider the changes of the values of the IF of a given journal in the last decade, it can be easily seen that the variation of the values is surprisingly high if we compare with other disciplines. Mathematical journals seem to have the worst behavior regarding the time stability both of the IF and the position in the IF list. A series analysis of a set of journals uniformly distributed in the IF list shows that the variations of the values of the IFs are very big when compared with other scientific disciplines, e.g., APPLIED PHYSICS and MICROBIOLOGY. The reader can see a representation of this behavior for three mathematical journals together with three journals of physics that have been chosen as representatives of these groups in the following graph (Fig. 1).