Open Access to Refereed Research Publications and Open Access to Research Data: A Crucial Strategic Distinction

Stevan Harnad
Canada Research Chair in Cognitive Sciences
Université du Québec à Montreal
&
School of Electronics and Computer Science
University of Southampton

(1) On the Open Access Impact Advantage for Refereed Research Reports
It has now been repeatedly demonstrated that refereed research articles that are made Open Access (OA) are used and cited significantly more in every scientific and scholarly field tested than those that are not made OA. It has now also been shown that this OA advantage is just as great for mandated OA as it is for self-selected OA. This means that the OA Advantage is not (as some have suggested) simply an artifact of selectively making higher-impact research open access: OA is the cause of the increased research impact. This finding greatly increases the importance and urgency of mandating OA for the sake of increasing and accelerating research uptake and progress.
Gargouri, Y., Hajjem, C., Lariviere, V., Gingras, Y., Brody, T., Carr, L. and Harnad, S. (2010) Self-Selected or Mandated, Open Access Increases Citation Impact for Higher Quality Research. PLOS ONE 5(10): e13636.
(2) On the Importance and Potential of Open Access Data-Archiving
Although there is not yet enough OA data to be able to demonstrate that the same kind of impact benefits will be generated by OA to research data as those that have been demonstrated for OA to research articles, it is highly probable that that will prove to be the outcome. Moreover, the impact benefits of making research articles OA, and the rich new means of measuring research usage and impact that OA is generating will also serve as incentives to encourage researchers to provide OA to both their articles and their data.
Brody, T., Carr, L., Gingras, Y., Hajjem, C., Harnad, S. and Swan, A. (2007) Incentivizing the Open Access Research Web: Publication-Archiving, Data-Archiving and Scientometrics. CTWatch Quarterly, 3 (3).
(3) On the Crucial Differences Between Research Archiving and Data-Archiving -- And Why Immediate Data-Archiving Cannot be Mandated
There is, however, a crucial difference between providing OA to research articles and providing OA to data: Scientists and scholars are not primarily data-gatherers. They gather data in order to data-mine, analyze, interpret and build further findings, theories and applications on it. Hence (except in the rare cases where the data speak for themselves), researchers cannot be expected (or mandated) to make their data OA immediately upon having collected or generated it, for all other researchers to data-mine and analyze. Researchers must be given sufficient time to data-mine their data, having invested the time and effort into collecting or generating it. And the length of the fair embargo interval on Open Access to data will vary depending on the nature of the data and the time, effort and ingenuity required to collect or generate it. This is fundamentally different from the case of refereed research reports, for which there is no justification whatsoever for embargoing Open Access once the paper has been peer-reviewed and accepted for publication.

Hence providing OA to refereed research reports can and should be mandated by researchers' institutions and funders, immediately upon acceptance for publication. Such immediate OA mandates cannot, however, be simplistically extended to research data (nor to unrefereed preprints of research reports) without generating the risk of needless and counterproductive conflicts of interest with the researchers that gathered the data. OA data-archiving, as soon as possible, should be strongly encouraged; in some cases embargo length limits can be set. But it cannot and should not be mandated (except in very special cases where the data-gathering itself is the research that is being funded.)