Demographic and Citation Trends in Astrophysical Journal papers and Preprints

Abstract

We have used data from ADS, AAS, and astro-ph, to study the publishing, preprint posting, and citation patterns for papers published in the ApJ in 1999 and 2002. This allowed us to track statistical trends in author demographics, preprint posting habits, and citation rates for ApJ papers as a whole and across various subgroups and types of ApJ papers. The most interesting results are the frequencies of use of the astro-ph server across various subdisciplines of astronomy, and the impact that such posting has on the citation history of the subsequent ApJ papers. By 2002 72% of ApJ papers were posted as astro-ph preprints, but this fraction varies from 22-95% among the subfields studied. A majority of these preprints (61%) were posted after the papers were accepted at ApJ, and 88% were posted or updated after acceptance. On average, ApJ papers posted on astro-ph are cited more than twice as often as those that are not posted on astro-ph. This difference can account for a number of other, secondary citation trends, including some of the differences in citation rates between journals and different subdisciplines. Preprints clearly have supplanted the journals as the primary means for initially becoming aware of papers, at least for a large fraction of the ApJ author community. Publication in a widely-recognized peer-reviewed journal remains as the primary determinant of the impact of a paper, however. For example, conference proceedings papers posted on astro-ph are also cited twice as frequently as those that are not posted, but overall such papers are still cited 20 times less often than the average ApJ paper. These results provide insights into how astronomical research is currently disseminated by authors and ingested by readers.