COVPRIMO Final Blog Post

Background
Coventry University is a post-1992 University with around 14,000 FTEs. The student population of undergraduates and taught postgraduates is very diverse studying across a wide range of subject areas and located in five Schools and Faculties (Art and Design, Business Environment and Society, Engineering and Computing, Health and Life Sciences and Lifelong Learning). The emphasis is on teaching, with a strong commitment to e-learning complimenting face to face interaction in the classroom. Research activity is also important.

The Lanchester Library is a busy academic library, often with over 5,000 people passing through the gates each day. Around 350,000 items are stocked and over 500,000 issues / renewals are made per annum. Around 12,000 new items are purchased annually. Over a million full-text journal articles are downloaded each year and there are over 800,000 e-book accesses.

Intended outcome(s)
The main objective of the project was to implement Ex Libris’s Primo discovery tool as a single user interface for Coventry University’s users to make the starting point for all their Library searches. This should help to provide a more consistent user experience than searching the Library Catalogue, MetaLib (our federated search tool) and Equella (our institutional repository) individually. This should help to improve search efficiency and so increase the use of our e-resources.

The challenge
In a context in which the ‘information seeking behaviour’ was perceived to be changing and the expectation of intuitive and effective interfaces was emerging, we were conscious of the need to enhance our resource discovery environment.

It was felt that the Aleph Library Catalogue was outdated and had none of the common web 2.0 functionality that our users expected to see. We had developed it as much as we possibly could, introducing a pretty major overhaul 2 years ago. This was carried out locally by our Library Systems team following some usability testing and was a definite improvement. However there were many other changes that we wanted to introduce that just weren’t possible within the constraints of the Aleph OPAC set-up.

Another concern came from the fact that we were using MetaLib alongside the Aleph OPAC as the main route to access our e-resources. This meant that users had 2 different places to search for Library content, and sometimes lead to confusion over which interface they were searching and which one they should be using. Usability testing of MetaLib was carried out at the same time as the OPAC usability tests and revealed that it wasn’t as intuitive as we would have liked. Added to this we are also aware of the growing numbers of international students studying at Coventry. Our understanding from focus groups is that they tend to come to do postgraduate courses with no previous experience of e-resources and do not find MetaLib intuitive.

Authentication to the OPAC and to MetaLib is carried out using 2 different methods, meaning that users are required to login with different usernames and passwords. OPAC authentication is by barcode from the user’s University card, whereas MetaLib uses LDAP to authenticate with the person’s University username and password. We were also aware from responses to the questions in our ‘user satisfaction surveys’ relating to ease of access to online resources that this was an area which we could improve on.

As a result of this it was decided that a new interface that combined the functionality of both the OPAC and MetaLib, was authenticated using Shibboleth and that was very configurable would be a good investment.

Established practice
The established practice was using the Aleph OPAC to search for books and journals but using MetaLib to search for databases by name or by subject. MetaLib also has a Find e-Journal option. This displays holdings information about e-journals which have been activated in the SFX link resolver. If an e-journal is available and displays in this list a link will also be provided back to the OPAC if the journal is available in print. However, if a journal is only available in print this will not display in the Find e-Journal list within MetaLib. This will only be discoverable from the OPAC, although SFX will provide links to print only journals if an article from one of these titles is found in a database.

The LMS advantage
The introduction of Primo as a replacement for the OPAC and MetaLib should simplify any confusion over which interface users should be searching. It is also designed in a much more intuitive way and works like many other popular websites. During our usability testing favourable comparisons were made to Tesco Online and Amazon. Primo is also much more configurable than the Aleph OPAC and than MetaLib. This enables us to customise it much more easily, something that we have already done based on the usability testing that we carried out as part of this project.

The usability tests that we ran as part of the project were extremely helpful. They enabled us to identify aspects of the out-of-the-box Primo which we wanted to change, before the launch. The testing was also of interest to Ex Libris and we had a conference call with the Primo Product Manager in September. This enabled us to feed back the main issues which we believe our usability tests revealed. This included things which we weren’t currently able to change and Ex Libris are going to consider introducing these as future updates to the Primo product.

We are also introducing Primo Central, which provides pre-harvested and indexed metadata from a range of databases. This enables the user to quickly retrieve journal articles alongside their other content (books, journals, videos, DVDs, databases, etc.). Previously to search for journal articles a federated search would need to be done, which carries out a live search of selected databases, so tends to be slow. Primo Central avoids this problem as the metadata is already harvested. We are just in the process of introducing Primo Central, so hope to carry out more analysis of its functionality over the next few months.

Another benefit is that Primo is being configured to authenticate via Shibboleth. This means that there will be no confusion about what credentials users need to login. It will also have an additional benefit for users who login to Primo and then link to a Shibboleth-authenticated database. These users will not need to login to the database during the same session, as they will have already been authenticated.

Key points for effective practice
Key points to note for other practitioners are as follows:
The timescale we were working to for implementing Primo was rather ambitious. We were trying to introduce it before the end of August, which is what would have been required to allow Subject Librarians to prepare their teaching materials for the new term. This might have been possible if we hadn’t wanted to carry out the usability testing and make significant changes to the Primo front end. However, as our Primo installation was delayed until July, with our first training day taking place on July 22nd, we weren’t able to carry out the testing until the final week of July. We did make excellent progress in customising the front end throughout August, thanks to the hard work of our Ex Libris UK contact who helped us make numerous changes until his deserved vacation in the 2nd half of August. Unfortunately, while he was away there wasn’t much additional support available meaning progress was hindered.

A number of other major concerns meant that the launch of Primo had to be postponed anyway. The main one of these was the setting up of Shibboleth authentication, which required involvement from the Library Systems Team, our IT Services department and one of Ex Libris’s German offices. This work began in August and continued throughout September, and there was no possibility of going live until this had been introduced. The delay was made worse by the fact that IT Services were migrating from Shibboleth 1.3 to 2 in early September

The other really big change which we were waiting for was the introduction of Primo Central. This was being rolled out to Primo customers over the summer, but wasn’t available to us until October. Given that adding this was a major change in the way Primo functioned it seemed sensible to delay the launch until after this had been added and configured to our specifications.

Other problems encountered during the project were the technical challenges of the transfer of data from Aleph and Equella (the institutional repository). It was also difficult to obtain advice on suitable backup routines from Ex Libris.

Another issue was experienced around the installation of Syndetics ICE for indexed content enrichment. We were told by Ex Libris that Primo was compatible with ICE and we were keen to test this functionality. However, when we tried to arrange this we found out that Primo wasn’t currently working with ICE. We were told it wasn’t a major problem, but we’re still waiting for this to be set up.

From a communications point of view it wasn't always easy discussing involved / complex technical issues over the phone with our Ex Libris Project Manager who was based in Germany. It was also not clear what we could expect to be carried out by Ex Libris as part of their basic installation service. We did identify a long list of changes to the interface following our usability testing, and our Ex Libris UK contact was extremely helpful in working with us on these. However, with hindsight more detailed clarification over where responsibilities lay would have been helpful.

Conclusions and recommendations
We are happy with our Primo installation so far and keen to make it available to our users. It’s a shame that we weren’t able to release it before the start of the new academic year, but it was definitely the right decision to delay things until Shibboleth had been configured and Primo Central had been added. We will then be able to launch a beta version of Primo. We plan to display a feedback form on the site to request more input from our users. We also hope to carry out further usability testing on the customised Coventry version of Primo and on Primo Central. We will then be able to make further changes (if necessary) before launching Primo as our single user interface.

The feedback we had from our usability testers was very positive and we believe this will also be the case with our other users. Based on our experience of Primo so far we believe it will be a major improvement over the previous search experience at Coventry, particularly with the introduction of Primo Central.

Recommendations:
• Make sure you have the right mix of people on the project group to ensure that the teams most involved in using the new system can present an equal viewpoint and to ensure project ownership across different sections of staff.
• Under no circumstance underestimate the time & resource such a project requires.
• Involve external suppliers (in this case of our repository, Equella) at an early stage.
• Allow enough time to customise the front end. If you plan to make significant changes then it would be sensible to allow a minimum of 2 months, more if you are planning to carry out usability tests and need to allow time to analyse these.
• If you know you are going to be doing testing over the summer period when there aren’t many students around try to recruit testers who live locally before the end of the summer term.
• Carry out a pilot study before you do your real testing. This can help to reveal any questions that don’t work as intended and will enable you to make changes before doing the test with your volunteers.
• Make sure you involve Academic staff in your testing.
• Try to get access to a usability lab to carry out the tests and record what takes place for accuracy.
• Get feedback and buy-in from Subject Librarians and staff who work on enquiry desks as these will be the people who will be teaching students how to use the system
• Try to ensure you have accurate creation and maintenance of the Name Authority Files in your LMS. We didn’t have this in Aleph and it would have made FRBRisation much more accurate. This is something that we are now having to review so that we get one FRBR group for something instead of 2 or more.

Additional information
Project Blog: http://covprimo.blogspot.com/
Coventry University Primo installation: http://primo.coventry.ac.uk/primo_library/libweb/action/search.do?dscnt=1&fromLogin=true&dstmp=1279105993031&vid=COV_VU1&fromLogin=true
Ex Libris Primo Overview: http://www.exlibrisgroup.com/category/PrimoOverview