Final versions carry little weight with librarians

Share this on social media:

In the future, librarians may be happy to point their researchers towards the author versions of papers in repositories instead of to the publishers’ versions. This is one of the main findings of a recent study carried out by Scholarly Information Strategies on behalf of the Publishing Research Consortium (PRC), a group that was set up more than a year ago to commission studies in areas of publishing that are topics of political debate.

This international study was intended to predict the future impact that widespread free availability of open-access content could have on journals. It also looked at the possible effects of aggregating content with an embargo period. ‘We were trying to get a survey of how librarians will react in the future, not recording what is going on now,’ explained Bob Campbell, chairman of the PRC’s steering group and president of Blackwell Publishing.

‘It was a survey of librarians rather than end-users,’ added Chris Beckett, director of Scholarly Information Strategies. ‘We were trying to test not just their main choices but how these choices interact.’ Although people might say that one particular factor such as price or availability is the most important one in deciding to stick with a journal subscription, he believes that in reality a combination of factors is important. Beckett and colleagues identified six key factors for these studies from interviews with librarians and then presented those involved in the survey with product concepts. ‘We did not use words like open access or subscription,’ he added.

Beckett sees the librarians’ attitude to the publishers’ role as a particularly significant finding. ‘Librarians are content with anything that has been peer-reviewed. After that they see very little value in the further work of a publisher, especially in the copy editing,’ he pointed out.

Embargo periods are important

But the length of time before the publishers allow authors to make their versions available from their own sites is an important issue for librarians. ‘The attractiveness of content delivered in different ways depends on the embargo period. If an article is significantly embargoed, i.e. for 12 to 24 months, then this has a significant impact,’ explained Beckett.

The picture is, of course, more complicated for publishers than a simple one-size-fits-all, however. ‘Publishers know that patterns of download depend considerably on the subject area,’ pointed out Blackwell’s Campbell. ‘In molecular biology, for example, most of the downloads come in the first few months after publication. In contrast, downloads of maths papers are more steady over the first few years.’

Nonetheless, Campbell sees a message to publishers from this study being that an embargo period of six months will not provide significant protection for publishers’ revenue. ‘We might see changes in policy about self-archiving of authors’ versions following this research,’ he predicted. ‘Blackwell publishes on behalf of societies so we don’t have a single policy across all our titles. Most of our customers prefer an embargo period of 12 months or more.’

Librarians are not the readers

The attitudes of librarians to the extra work publishers put into papers beyond peer review may be surprising. However, as Beckett pointed out, while librarians have to make the decisions about what publications to pay for, it is not their job to actually read the papers.

‘It is a strange market where librarians buy on the behalf of the users,’ agreed Blackwell’s Campbell. ‘It would be interesting to find out what users want. If it can be shown that users do care which version they use then this might make a difference for librarians.’

To look into this further, Blackwell has carried out a separate study that is due to be published soon. ‘We have studied what value copy editing adds to articles and found that many little things are done at the proofing stage,’ commented Campbell. ‘Many of these changes relate to references that are cited wrongly and which would cause problems for the readers if they were not corrected. Another thing that can be picked up is data that differs between the text and a table. Such mistakes could have serious repercussions if they were, for example, dosage levels of a certain drug.’

The importance of this extra value to readers was noted in a 2002 study for the Association of Learned and Professional Society Publishers (ALPSP). Alma Swan & Sheridan Brown of Key Perspectives, who carried out the study for ALPSP, found that ‘authors do value not just peer review, but also the various functions by which traditional publication adds value to content.’ And, in terms of electronic content, they valued long-term preservation of and access to journal content. Nonetheless even five years ago when this study was carried out, most respondents wanted electronic journals to be free in the future and few saw publishers as adding much value to electronic journals, with the exception of providing citation linking.

This linking aspect of electronic articles is still a strong component of the final published versions. Only the publishers’ versions are assigned digital object identifiers (DOIs), and this helps with linking. Even with this, however, the situation is more complicated than it might at first seem. As Blackwell’s Campbell pointed out: ‘I am a strong believer in CrossRef and DOIs but, of course, these links do not work if the articles are printed out.’

Watch and see

In terms of the recent PRC results, it is still early days yet. As Chris Beckett explained, ‘A&I databases and search engines are quite agnostic in terms of what they index but self-archived copies have to become as available as the published version for the threat to become more apparent.’

And publishers can take heart from another recent study, carried out by Mark Ware Consulting for the Association of Learned and Professional Society Publishers (ALPSP). That study concluded that the three most important factors in the decision to cancel a journal subscription were the faculty no longer requiring a journal, usage and price. Availability of the content via open-access archives and availability via aggregators were ranked equal fourth, but they came some way behind the first three factors, and availability via delayed open access were seen as relatively unimportant.