Access, Readership, Citations: A Randomized Controlled Trial Of Scientific Journal Publishing

摘要

This dissertation explores the relationship of Open Access publishing with subsequent readership and citations. It reports the findings of a randomized controlled trial involving 36 academic journals produced by seven publishers in the sciences, social sciences and humanities. Between January, 2007 and February, 2008, 712 articles were randomly assigned free access status upon publication from the publisher's websites (the treatment), leaving 2,533 control articles that were accessible by subscription (the control). Article usage data was gathered from the publishers' websites and article citations were gathered from ISI's Web of Knowledge. At the time of this writing, all articles have aged at least two years. Articles receiving the Open Access treatment received significantly more readership (as measured by article downloads) and reached a broader audience (as measured by unique visitors), yet were cited no more frequently, nor earlier, than subscription-access control articles. A pronounced increase in article downloads with no commensurate increase in citations to Open Access treatment articles may be explained through social stratification, a process which concentrates scientific authors at elite, resource-rich institutions with excellent access to the scientific literature. For this community, access is essentially a non-issue. The real beneficiaries of Open Access are the communities that consume, but do not contribute to, the scientific literature. The focus on information consumers requires us to advance the theory of the attention economy. The linear transmission model, where information flows from the sender to the receiver is rejected for a two-sided market model, with authors on one side, readers on the other and journals fulfilling the role of the intermediary agent. The primary purpose of the journal-agent is to transmit quality signals to potential readers. I argue that this model is able to explain both author and reader behaviors as well as the persistent role of journals in an information environment that decouples certification from dissemination.