November 27, 2013
There has been some debate about delayed Open Access, if it can even be called Open Access, and whether this form of sharing research is more or less effective than immediate OA, especially in the context of the impact factor and citation level. This discussion is important as it focuses on the very essence of OA and the development of standards that should guide the publication of research in this model.
In accordance with the definitions of Open Access, for example the Budapest Open Access Initiative, the delayed publication of research results cannot be called OA. However, many publishers and institutions provide articles under an OA model only after a temporary embargo, which is usually 12 months. The question that arises then is does this type of restriction affect the citation level of articles and can this effect can be described as positive or negative?
According to Stevan Harnad and his post on the Open Access Archivangelism webpage, regardless of whether we are dealing with Delayed Access (free online access after a delay period or an embargo) or other forms of limiting access, it is hard to talk about full Open Access, but it is still better if an article is available after a delay rather than not at all. Stevan Harnad refers to data on citations from arXiv.org, from which it is clear that the sooner the article is available (in this case on arXiv.org), the greater the level of citations. While the sample range is inherently limited, the difference in citations is really big. This data applies to Green Open Access. But what about Gold Open Access?
Mikael Laakso and Bo-Christer Björk talk about the level of citations for articles published in scientific journals in various models in their analysis entitled “Delayed Open Access – an overlooked high-impact category of openly available scientific literature.” The results presented are very interesting.
According to the report, from 111 312 of the articles published in 492 journals, 77.8% were made open access within 12 months from publication, with 85.4% becoming available within 24 months. For the time embargo set most frequently by scientific societies or professional associations, commercial publishers and the university press publishers (see chart).
But what is most surprising in the report, is this, and I quote:
“What this study demonstrates is that delayed OA, which is mainly provided by society publishers and university presses, is important both in volume and in granting access to highly cited scholarly articles. It is interesting to note that the yearly volume of delayed OA articles (111 000) is almost ten times as big as the number of individually freed article in so-called hybrid journals (12 000) (Björk 2012), and that total number of citations to delayed OA article in the Web of Knowledge exceeds the number of citations to articles in full OA articles.”
The authors do not reach a conclusion on why “the total number of citations to delayed OA article in the Web of Knowledge exceeds the number of citations to articles in full OA articles”. It is worth asking whether these results can be generalized and rephrased into a conjecture that the time embargo helps in getting a higher number of citations. For me it would be a too far-reaching thesis, as in this report we have correlation effects rather than a chain of cause and effect. One explanation of the data could be the fact that immediate OA is still not a standard in publishing research results.
What is worth noting, however, is that this way of making research available is still common, and probably will be for a long time. So we should not take umbrage at the reality and instead simply seek a way to ensure that immediate OA eventually becomes the standard.