Time to call time on the library catalogue?

Twitter commentators on Guilhem Chalancon’s presentation to this year’s UKSG conference observed an interesting point: not once in the half hour that he spoke about the tools that he uses to find information for his research did he mention library catalogues.

Chalancon is a PhD student at the MRC Laboratory of Molecular Biology (LMB) and the University of Cambridge, UK. In his talk he presented a detailed account of his approach to finding and organising reference resources more effectively:  ‘The key information is rare relative to all the information we absorb. It would be good to maximise the focus on the really useful information. There is lots of innovation to help. From an end-user standpoint we need to adopt new habits, avoid distractions and build workflows to help manage this information.’

Chalancon’s typical workflow has three key steps. The first is to identify relevant information, and spot and save it. The next step is to digest the information, which includes annotating and tagging, assigning to a project or theme. The final stage is to use the information, which includes a second read of the material and citing it. He noted that there is a difference between tagging and active reading.

To help with his information management he uses a range of tools at different stages of the process. These include search tools such as Google  Scholar and reference management tools such as Mendeley and ReadCube. He observed that there are some key recurring principles of good tools, including minimalist and responsive design and cloud synchronisation.

The absence of the library catalogue in Chalancon’s research process was noteworthy but not unique. Simone Kortekaas, project manager, innovation and development at the library of the University of Utrecht, the Netherlands, noted in her presentation how, while traffic to the library’s journal holdings had grown, the proportion of access to these holdings via the library catalogue had dropped dramatically. Instead, patrons were increasingly coming to library resources via discovery services, reference-sharing tools, databases that the library subscribes to and, perhaps most notably, Google Scholar.

This trend was echoed from the other side of the information-provision equation: a delegate from Springer said referrals to the publisher’s resources via library catalogues have been steadily dropping over recent years and now stand at around 15 per cent of the total. He also noted the huge traffic to the publisher’s resources from Google Scholar and from services such as Mendeley.

Bold moves
Despite these trends, however, some may still see the University of Utrecht’s response as bold: in 2012 the library decided to stop running the library catalogue that it had built in house 10 years earlier – and not replace it with a new system, either in-house or from a vendor.
Kortekaas explained: ‘Like every library, we always offered a catalogue but the use of it changed in recent years with things like discovery tools and GoogleScholar.’ She sees the move as a response to user behaviour and preferences. In reply to a question from the floor about why the library ‘chose Google’, she noted that ‘we didn’t choose, our users chose’.

Stopping the library catalogue was not a decision taken lightly. The library surveyed patrons and made communication efforts to help ensure users were informed of the change.
‘Before phasing the catalogue out we had to prepare users because the statistics showed that it was still used by many and we knew there would be disappointed users,’ she explained. Steps to help the transition included removing information about the catalogue from the library homepage beforehand and providing support and information about alternative routes to the content.

‘Six months on from stopping the catalogue we can report that we didn’t face major difficulties or complaints,’ she added.

But this is not the end of the story. ‘Behind the scenes we continue to improve discovery.’ This includes adding holding information to discovery tools, sharing SFX knowledge bases with Google Scholar and Scopus, opening the repository for harvesting, and supporting easy authentication for off campus access. She said that the library would also cooperate with other libraries and groups of libraries as much as possible. In the first phase the library kept its OPAC. However, Kortekaas said that the next step would be to phase that out too: ‘In the world we live in today it seems silly to ask users to start their search in a local catalogue. There is a lot of work for libraries to do instead of worrying about discovery, for example supporting researchers in data management. With discovery, libraries need to accept that others can do it better.’

Raising the game
Not all libraries share Kortekaas’ view. For Birte Christensen-Dalsgaard, deputy director general and head of information technology services at the Royal Library in Denmark, the increase in access via Google is a call for libraries to raise their game. She told the UKSG meeting that she was struck by Chalancon’s talk about how he does research. ‘Where is the library? It’s all about Google,’ she observed.

‘It is easy to see why many prefer Google,’ she continued, giving an example of a search for a famous person in the library catalogue and the same search in Google. The Google results had a wider range of resources and images and were more easy to use. ‘We should do at least as well,’ she observed of libraries.

She noted that Google enriches its searches using semantics. ‘Libraries could do the same and can help identify quality and what you can trust.’

Such things are important because, while Google may be easier to use, the library has a wealth of specialist knowledge. ‘It is important to teach students to evaluate quality, relevance and value to them,’ she said.

However, she added that people will only use the library to search for external resources if the library adds value – and here she sees a need for the library to do marketing. She also suggested that libraries could put a price tag on the resources they supply. This is not that users have to pay but to make them aware of costs and the library’s role in supplying quality content. This, she said, would be useful in, for example, communicating about open access. ‘Libraries need to add value, support users to judge quality, provide semantics and cloud technology and support different delivery channels,’ she explained.

‘Either we can all go to Google or can look at how we can add value ourselves – and I think it’s worth doing,’ she said. Some might be more comfortable with this approach than with Utrecht’s. As someone observed from the floor, ‘What if everything’s via Google and then Google pulls Google Scholar with one month’s notice? They’ve done it before.’

There is also plenty of interest in accessing library content using dedicated discovery tools. This is an area that researchers at the LISU department of Loughborough University, UK have been looking into, as Valérie Spezi, explained in her talk. Speaking about a recent LISU study that she was involved in into the impact of library discovery technology in the UK, she noted that 77 per cent of UK higher education libraries are already using resource discovery systems (RDS), with a further 11 per cent in the process of implementing one. She added that the market is dominated by Primo (from Ex Libris), EDS (from EBSCO) and Summon (from ProQuest). ‘Libraries want to offer a Google-like experience – and this seems to be working; user feedback is very positive,’ she said.

She noted that LISU’s study showed that the impact of RDS technology is mainly seen in the increased use of e-books, more than journals, although she said that it is hard to isolate RDS from other factors.

When it comes to publishers, she said that visibility and discoverability are the main drivers for publishers to engage with RDS technologies. However, she added that they have no clear evidence of the effect of such tools on usage, an observation that was disputed from the floor by a delegate from Springer who said that the publisher did know how much traffic comes via RDS tools.

Spezi also suggested that RDS tools might bring more benefits for smaller publishers than the bigger ones. She added that there are challenges for publishers, including that RDS tools focus more on their library customers than on publishers and that work may be required on the metadata.

‘Publishers should engage and work closely with libraries and RDS suppliers to optimise discoverability,’ she said.

The impact of discovery systems on journal usage has been the subject of another study, by Michael Levine-Clark of the University of Denver, John McDonald of the University of Southern California and Jason Price of the SCELC Consortium. The researchers carried out a longitudinal study of journal use that focused on the use of Primo, Summon, EBSCO Discovery Service (EDS) and WorldCat Local. The study included 33 libraries, six that use each of the four discovery tools and nine controls that have not yet implemented discovery tools. The libraries were all in English-speaking countries: 28 in the USA, two in Canada, one in Australia, one in New Zealand and one in the UK.

The study compared usage, based on the COUNTER JR1 (full text downloads per title) reports for the year before and the year after implementation of the discovery service (or the two consecutive years in the same period but without implementation in the case of the controls). The survey included data from six publishers and 9,206 journals.

In most cases journal usage increased over the two years of the study, although usage dropped for six of the libraries, including at least one of the controls. The study results showed that journal usage at Summon and Primo institutions increased more than with the other two services in the study and the control group. Increases observed with EDS and WorldCat Local were roughly in line with those observed with the control group, according to McDonald. The study also showed that the impact of discovery services differs across publishers. ‘It is interesting to note that the mean usage change was less than or equal to zero for at least one publisher’s sets of journals for each discovery system.’

The researchers now plan to look at the effects on the results of factors such as aggregator full text availability, size of publisher, journal subject, overall usage trends and configuration options in the discovery services. They also hope to widen the group of libraries studied and look for explanations for the changes in usage observed.

Such studies provide a snapshot of the ways the information discovery is changing as research behaviours and technologies change. Whichever way libraries choose to proceed in helping their users to find relevant information for their research, the knowledge-seeking behaviour of Guilhem Chalancon and many researchers like him will be – and should be – hard to ignore.