published in Science and Technology Libraries, 25, no 1/2 (2004): 55-72.


As the number of retrospective digitization projects of journal content increases, there is a need to assess the impact of these projects on the productivity of researchers. Librarians making collection development decisions about acquiring these backfiles need to know how useful they are to researchers. This study provides data on usage of a range of years of the Physical Review, and citation information from Physical Review Letters to other Physical Review articles. The usage of the online archive of Physical Review articles indicates that articles are accessed all the way back to the first issue, with an average number of downloads on the order of ten per article per year. Both usage and citation rates show exponential decay rates, however, with different intrinsic time scales. The citation half-life is consistent with previous studies of the physics literature, while the usage half-life computed here is in conflict with older analyses of print usage of the physics literature, although in line with some recent online usage studies in medicine. An analysis of the citation data indicates a potential order of 10% enhancement in citations to articles available in the online archive, but the statistical error is of the same magnitude, so no firm conclusions can be drawn from that data. A few more years of citation data may be able to resolve the question of impact of the online archive on citation rates.


scientific journals, citation analysis, online usage

Date of this Version

August 2006