Search   |   Back Issues   |   Author Index   |   Title Index   |   Contents



D-Lib Magazine
June 2004

Volume 10 Number 6

ISSN 1082-9873

Comparing the Impact of Open Access (OA) vs. Non-OA Articles in the Same Journals


Stevan Harnad
Chaire de Recherche du Canada
Centre de Neuroscience de la Cognition
Université du Québec à Montréal
Montréal, Québec, Canada H3C 3P8

Tim Brody
Intelligents, Agents and Multimedia
Electronics and Computer Science
Southampton University
Highfield, Southampton, SO17 1BJ, United Kingdom

Red Line

(This Opinion piece presents the opinions of the author. It does not necessarily reflect the views of D-Lib Magazine, its publisher, the Corporation for National Research Initiatives, or its sponsor.)

The way to test the impact advantage of Open Access (OA) is not to compare the citation impact factors of OA and non-OA journals but to compare the citation counts of individual OA and non-OA articles appearing in the same (non-OA) journals. Such ongoing comparisons are revealing dramatic citation advantages for OA.

A recent Institute for Scientific Information (ISI) study has reported that traditional journals and Open Access (OA) journals have similar citation impact factors (Pringle 2004). The ISI's press release announced:

"Of the 8,700 selected journals currently covered in Web of Science, 191 are OA journals... [A study on] whether OA journals perform differently from other journals in their respective fields [found] no discernible difference in terms of citation impact or frequency with which the journal is cited" <>.

It is certainly welcome news that there are no impact differences between the 191 OA journals and the 8509 non-OA journals indexed by ISI, equating for comparable journals as closely as possible: <>. This proves that the skeptics who thought OA journals would be of lower quality or impact and would not be indexed by ISI were wrong (at least for those 191 OA journals): OA journals are indexed by ISI, and they do have comparable citation impacts. But obviously this methodology has a certain element of circularity!

To get a realistic estimate of the effect of OA on impact, it is not enough to compare only the 2% of ISI journals that are OA journals with the 98% that are not, to find that they are equal in impact (for this may well be comparing apples with oranges, even if you equate for subject matter).

What further needs to be compared is:

(1) the citation impact of the much higher percentage (perhaps as high as 20-40% according to Swan & Brown's (2004) sample) of articles from the 98% non-OA journals that have been made OA by their authors (by self-archiving them)


(2) the citation impact of articles from those very same journals and issues that have not been made OA by their authors.

As the founder of citation analysis as well as the ISI, Gene Garfield (1998) has often stressed, the article (and author) citation counts need to be analyzed, not just the average citation counts of the journals in which the article appears.

What this kind of analysis is beginning to reveal in the OA era is that there is indeed a "discernible difference" in terms of the frequency with which the article is cited: there is a dramatic advantage in favor of the articles that their authors have made OA (Lawrence 2001; Kurtz 2004; Brody et al. 2004). Results are only available for computer science, astronomy, and physics so far, but all other disciplines are currently being analyzed. The earlier Lawrence (2001) study on the impact-enhancing effects of OA in computer science needed to be replicated in other fields to check whether it was merely an artifact of the fact that computer science is conference-based rather than journal-based, and whether the advantage really reflected OA vs. non-OA rather than just online access vs. paper access. Fortunately, thanks to the ISI database licensed to the Observatoire des Sciences et des Technologies (OST) and a special contract generously provided by ISI to conduct the study, our research team at the Université du Québec à Montréal, Southampton University and Universität Oldenburg is in the process of testing the OA advantage across all disciplines in a 10-year ISI sample of 14 million articles. The physics analyses up to 2001 are already completed (Brody et al. 2004), and they reveal even larger effects than those reported by Lawrence, with OA/non-OA citation ratios of 2.5 - 5.8.

Chart: Open Access vs. Non-Open Access Citation Impact Rations All Physics Fields

Figure 1. The OA Advantage in Physics.

Figure 1 shows the total number of journal articles in physics indexed by the Institute of Scientific Information from 1992-2001 (gray), the percentage of these that have been made open-access (OA) by author self-archiving (green), and the ratio of the citation counts for the OA articles to the citation counts for the non-OA articles (red). (Analyses are by Brody et al. (2004) and are based on the ®ISI CD-ROM* citation database metadata and references from 14,000,000 articles in 7,000 leading journals from 1991 to 2001 licensed to the Observatoire des sciences et des technologies <> and from the 260,000 articles self-archived in <>.)


All indications are that the 2002 and 2003 data will raise these ratios even further, as the biggest effects occur within the first 3 years of publication in scientific disciplines (and both OA and the awareness and visibility of OA articles are also increasing year by year).

Access is not a sufficient condition for citation, but it is a necessary one. OA dramatically increases the number of potential users of any given article by adding those users who would otherwise have been unable to access it because their institution could not afford the access-tolls of the journal in which it appeared; therefore, it stands to reason that OA can only increase both usage and impact. The ratio of "reads" to "cites" will no doubt vary by field. For example, Kurtz (2004) and co-workers report it as 17:1 and even 12:1 in astrophysics. Odlyzko (2002) predicts analogous trends in mathematics. Tim Brody's remarkable correlator/predictor1 gives the size of the download/citation correlation by field, which can be used to predict citations 6-24 months later from downloads today (with an adjustable time-window): <>.

The proportion of articles for which their authors provide OA is likely to increase dramatically now, in part because of the mounting evidence for the impact advantage OA confers. OA will also increase because of the growing number of journals that have already given their official "green light" to author self-archiving, partly because journal impact factors also benefit from increased article impact, and partly because journals are eager to demonstrate that they have no wish to stand in the way of OA and its benefits to research and researchers. (The proportion of "green" journals rose from 55% to 83% from 2003-2004: <> and <>.

RoMEO Directory of Publishers/Journals who have given their
Green Light to Self-Archiving

Proportion of journals already formally giving their green light to
author/institution self-archiving (currently 83%) continues to grow:
Chart: RoMEO Direcotry of Publishers who have given their Green Light to Self-Archiving

Figure 2. Directory of Publisher/Journal Policies on Author/Institution Self-Archiving.


Percentage of green
PUBLISHERS grew from
42% - 58% from 2003-2004
Chart: Percentage of green publishers growth 2003-2004
Percentage of green
JOURNALS grew from
55% - 83% from 2003-2004
Chart: Percentage of green journals growth 2003-2004

Figure 3. Percentage Increase in Green Publishers and Journals: 2003-2004.

A JISC survey (Swan & Brown 2004):

"asked authors to say how they would feel if their employer or funding body required them to deposit copies of their published articles in... repositories. The vast majority... said they would do so willingly."

It is hence clear that the ball is now in the universities' court: The sooner they extend their existing publish-or-perish policies to require also providing OA for all those published articles, the sooner the entire research community will enjoy the benefits of maximizing its research impact by maximizing user access to its research output: <>.


Brody, T., Stamerjohanns, H., Harnad, S. Gingras, Y. Vallieres, F. & Oppenheim, C. (2004) The effect of Open Access on Citation Impact. Presented at: National Policies on Open Access (OA) Provision for University Research Output: an International meeting. Southampton University, Southampton UK. 19 February 2004.

Garfield, E. (1998) The use of journal impact factors and citation analysis in the evaluation of science. Presented at the 41st Annual Meeting of the Council of Biology Editors, Salt Lake City, UT, May 4, 1998 - April 17, 1998

Harnad, S., Brody, T., Vallieres, F., Carr, L., Hitchcock, S., Gingras, Y., Oppenheim, C., Stamerjoanns, H., & Hilf, E.R. (2004) The green and the gold roads to Open Access. Nature (web focus)

Kurtz, M.J. (2004) Restrictive access policies cut readership of electronic research journal articles by a factor of two, Harvard-Smithsonian Centre for Astrophysics, Cambridge, MA

Lawrence, S. (2001) Online or Invisible? Nature 411 (6837): 521

Odlyzko, A.M. (2002) The rapid evolution of scholarly communication." Learned Publishing 15: 7-19

Pringle, J. (2004) Do open access journals have impact? Nature (Web Focus). <>.

Swan, A. & Brown, S.N. (2004a) Authors and open access publishing. Learned Publishing 2004:17(3) 219-224.

Swan, A. & Brown, S.N. (2004b) JISC/OSI Journal Authors Survey Report. <>.


1. Correlator/Predictor: The correlator counts downloads and citations for an article within an adjustable "window of time" as well as adjustable value ranges for downloads and citations, and calculates their correlation. For example, it could calculate the correlation between downloads and citations within a two-year window for articles in the top quartile of downloads and the bottom quartile of citations. The windows and ranges can be made as wide or narrow as one wishes. The correlation can be as high as .4 - .6, which means that one can predict from today's downloads what citations two years from now will look like. A series of narrow windows can also test cyclical causal effects and their time-constants: downloads cause citations, and then citations cause more downloads.

Copyright © 2004 Stevan Harnad and Tim Brody

Top | Contents
Search | Author Index | Title Index | Back Issues
Previous Opinion | Conference Report
Home | E-mail the Editor


D-Lib Magazine Access Terms and Conditions