As more journal articles go online, only more recent articles tend to be cited, according to a study published today in Science. In addition, only a small group of journals and articles are being cited, the study found.

James Evans, a sociologist at the University of Chicago, surveyed a database of 34 million articles, their citations over the past 50 years, and their online availability. Evans' results suggest that online literature searches yield more recent articles, and cause researchers to cite a less broad and diverse range of articles.

Not everyone agrees. Carol Tenopir, a professor of information sciences at the University of Tennessee, told The Chronicle of Higher Education that her group has found exactly the opposite result when it comes to how many papers researchers are reading. According to her research, which studies when researchers read articles as opposed to cite them, the number of older articles being read has actually increased in the past 10 years, along with the number of different journals read. She attributed the difference in the two studies to the fact that older journals are being put online only recently, and that researchers may read papers for research purposes but not necessarily cite them right away.

"We are looking at things differently, but it is surprising how different [the results are]" Tenopir told The Scientist in an Email. "Maybe the...inherent lag for citations explains it a bit."