Case Studies

'A set of case studies, drawn from real-life examples, documenting the role that metadata plays in the patron-driven acquisition of ebooks.'

If you want to comment or contribute to this wiki use the DISCUSSION tab or feel free to contact
Ken Chad ken@kenchadconsulting.com Tel: +44 (0)7788 727 845 Twitter: @kenchad

Case studies

The following (2013) presentation about PDA at NTNU University Trodheim, Norway was NOT part of the project but has been added as it may prove useful

Bangor

Glasgow

Hertfordshire

Newcastle University

Huddersfield University

Kings College London

Liverpool University

(Evidenced Based Purchasing)

Sussex University

York University

How PDA works on EBSCO



Case study Template


Please feel free to create you own case study so that your experience can be shared with the community. Use the templateabove and send your case study to:-
Ken Chad
Email: Ken@ kenchadconsulting.com
Tel: +44 (0) 7788727845
Twitter @kenchad

Case study headings:

What was/is the problem which PDA was implemented to address (i.e. what does PDA solve?)
How does the PDA scheme work? (include statistical data if available)
What benefit does it have (or might potentially have) over other procurement solutions
  • The benefits to the end user
  • The benefits to the library/institution

The role of metadata in the discovery, selection and acquisition of e-books
  • What is the role of metadata in the discovery, selection & acquisition of ebook
  • How is metadata uses for profiling collection available for PDA
  • Is there a role for evaluative metadata?

What do you think the impact is/will be of 'discovery' services? (Summon, Encore, Primo etc)

What have been the outcomes/lessons
  • Have you achieved a sustainable model for PDA?
  • What factors are important in terms of sustainability?

Case study overview

Analysis of the current case studies

Discovery

Library catalogue

In all cases the case study libraries felt confident that main discovery tool for ebooks is the library catalogue. MARC records are received from the ebook platform vendors and/or publishers and uploaded into the library catalogue.

Reading list tools and VLEs

These are also used for discovery

Library discovery services

Of the four universities only one (Huddersfield) has already implemented ‘discovery service’. (Summon from Serials Solutions) but ebooks are not yet included in its scope. Primo (from ExLibris) is currently being implemented at University of Newcastle. Case study libraries saw opportunity for library discovery services in improving ease of use for the user. Discovery services are increasingly harvest ebook records so they might well supplant the OPAC as the primary discovery service for ebooks. Were this to be the case librarian saw opportunity to reduce the current inefficiencies and resources devoted to loading (and unloading) ebook MARC records for the library catalogue. The ‘profiling’ of ebooks made available for PDA is seen as being managed through the link resolver though the precise way in which this might work is not yet clear.
Keeping any discovery system up to date was identified as an issue as new ebooks are added to collections on a daily basis but most discovery systems update their records monthly. Furthermore one library reported an unexpected problem of discovery for a small but growing number of external readers such as local researchers, who can discover ebooks but authentication requirements prevent access to full-text. This may become more of a problem if local library collaboration continues.

Google Scholar

Case studies were not able to be clear about the extent students and others used Google Scholar to discover ebooks. E-book Platform providers have reported that around 25-30% of accesses to their ebook platforms come from Google/Google Scholar. In the case studies libraies estimates for ebooks discovery for ebooks were low and perceptions were that it was sued primarilty for electronic journal articles. However there was clear awareness that this might change

PDA

All the libraries selected as case studies were using PDA systems. Libraries had all made significant changes to their PDA services following initial trials and following trials held in other libraries. A range of models were used based around rental, purchase and mediated purchase options.

The rationale for PDA

Ebooks are seen as an effective solution for a range of academic library problems including:-
- 24/7 access to library collections to all students.
- Accessibility across all campuses and other locations where students were working.
- Meeting the immediate books need for academic essays and assignments.
- Not enough copies of books when they are needed, ‘ebooks provide more resources to go round’ especially for large undergraduate student cohorts.
- Space problems, not enough space to extend print collections. Ebooks in general, but PDA in particular mean demand is satisfied without filling shelves.
Ebook collections are answering a number of concerns raised by students in National Student Surveys.
PDA services for ebooks provide a wider choice of books to students without having to make upfront purchasing commitments. PDA services ‘make sure we buy the right materials’ was noted by one library when relegating materials to their archive revealed a higher than anticipated numbers of books in collection were not used. The ‘try before you buy’ purchase model ensures the items purchased ‘earn their keep’ with fewer book purchases made on non-borrowed items.
Libraries also reported the benefits of increased and improved information on the use of their e-collections. One library noted that students, using PDA systems, had purchased some of the highest downloaded items in the library collections and a higher percentage of PDA ebooks are used in comparison to non PDA ebooks with above average usage and smaller tail (University of Huddersfield).
Data provided by PDA allows libraries to collect more detailed information on collection requirements, identifying gaps in their collections and allowing them to respond quickly to user information needs.
In some instances Ebook and PDA services were replacing other traditional library services. One library had removed short-loan services as e-book collections were fulfilling most of the service needs. PDA services are better than traditional Inter-Library Loan (ILL) services as books are instantly discoverable from the OPAC and delivered quickly to the user. It was noted that ‘users don’t need to know about the device they just get the material.
As academic librarians are preparing for significant reductions in library budgets e-book and PDA systems were also identified as offering cost saving solutions. Devolving procurement frees staff hours spent on stock purchasing though it was noted that more time is spent on invoice issues until EDI invoicing is introduced. In addition there is a saving of up to 6 weeks waiting time to acquire book compared to print (information provided by University of Huddersfield). Whilst e-book purchases are initially more expensive there are cost savings on processing time and staff costs. The University of Huddersfield calculated that their average e-book cost is £50 (before discounts) and their average print book cost are (£33)* these figures based only on local figures.

Funding PDA

All libraries indicated that the funds they allocated to PDA were quickly used with librarians commenting that they could easily be a black hole for the book fund. Finding a sustainable funding method for PDA remains a challenge. Libraries found PDA funds were quickly, requiring them to remove the collection records from the OPAC. It was identified that a purchasing model spread across the year was desirable. The proportion of spend on rental created a dilemma, outright purchases provides a better user experience and was preferred by librarians. After an initial trial period all libraries used chose to reduce the catalogue of ebooks available to provide a more sustainable purchasing methodology.

Sustaining PDA

Mediated purchasing of collection profiling are the two most successful methods currently used to apply sustainability to purchasing methods. Mediated purchase methodologies were employed to provide control over purchase, in some cases purchases were only allowed through mediation. Suppliers are developing new systems to improve the information available to librarians to allow mediated purchases such as the use a text notes fields where user can add comments like ‘needed this for essay’ allowing users to justify their selections for purchase. University of Newcastle believe mediation is the most important factor is PDA sustainability, which provides them with an effective and sophisticated way to control demand. Profiling is the method used by the libraries in the case studies which reduces the records held in the collections accessible by users to make PDA selections. One case study library (Huddersfield) reported that e-book purchasing was now the preferred purchasing model for a number of academic departments. PDA collections were now more heavily biased in these areas. Libraries are developing looking to mixed economy of PDA and mediated recommendations. Currently top-slicing of funding or identification of unused budgets have been used to fund PDA developments but this may move to subject teams in future (Univ Hudds).

The role of metadata

Metadata is currently loaded into the OPAC as the main method for PDA discovery. Most of the libraries in the case studies used basic metadata only and they remain uncertain over role of evaluative metadata. Basic metadata would include author, title, date, edition, subject, keyword and place of publication. These are considered by most libraries to be sufficient for reading list items which form the staple of most e-book libraries. Some also included Table of Content (TOC) and abstracts which makes keyword search more effective. For some libraries the basic records in OPAC was the only option as they do not employ cataloguers and rely purely on what suppliers provide.

Enhanced metadata

TOC data in catalogue can be problematic with some case study libraries reporting it as ‘too messy’ in the catalogue display to be of use to users. Whilst other libraries felt TOC was very useful in particular for searches in the humanities where titles were often problematic for effective searching, an example Piper Alpha oil rig disaster found in chapter heading metadata.
TOC data encouraged wider reading on subjects. Libraries reported that better and more evaluative metadata meant e-book citations were ranked higher in catalogue search results. There had been concerns that print would be ignored and libraries initially debated removing this metadata but it was generally agreed that it was helping relevant material to be retrieved.
Libraries did not report that they bought synthetic solutions or metadata solution services. One library reported the use of book-covers metadata on their catalogue. The book cover option was very popular with students and heavily used.
It was felt that discovery services, such as Summon, would be more useful to exploit additional metadata options particularly full-text options but would remain restricted to what suppliers provide.

Key metadata for discovery

One library was particularly concerned about duplication in collections. Metadata could be particularly useful for removing duplicate entries (de-duping) on OPACs (University of Hertfordshire). It was felt generally that metadata had a significant role in balancing the competing demands for discover and budgetary constraints on PDA. However librarians were unsure how their patrons made use of metadata. Some librarians suspected that patrons made use of only the most basic record/display to make valued judgements of texts.
One library relied totally on metadata for patrons to discover the PDA collections, no additional publicity or methodology was used. They still reported that the funding for PDAs was quickly used. Another library suggested that mediation services would become more important as improvements were made in metadata for discovery purposes.
Librarians reported that MARC records are currently not sufficient as the main metadata method for discovery of items for wider reading. They reported more reliance on table of contents to get into chapters.

Metadata for profiling PDA collections

A more surprising role for metadata quickly emerged from the research. Metadata was used by many libraries for demotivating the amount of PDA or reducing discovery success of ebooks. Metadata was used by most libraries in the research for profiling collections and reducing the availability of e-book records made available to users. Dewey decimal classification number ranges were used by a number of libraries for selective purposes, whilst other libraries used date or edition fields to prune the collections. Place of publication was also identified as an exclusion field for some subjects such as law. One library indicated they would like vendors to develop more sophisticated methods for profiling their PDA offerings.
Profiling is time consuming and can make demands on system processing effort and high degree of manual intervention.

Metadata to support recommendations

Other libraries prefer to use a ‘recommend’ systems which track the borrowing history of patrons and using this data to make recommendations on the OPAC. The catalogue record displays the use data to patrons for example ‘people who borrowed this also used xxxxxxxx’. The library feels this data is of more use to readers than other forms of metadata. Other libraries also reported they would like to make better use of patron opinions to develop sustainable budget solutions for the future. Further research on user behaviour in a PDA environment is recommended. Some libraries and suppliers are developing systems to gather user profile information, matching metadata from users and ebook use could be powerful for future discovery services.

Managing metadata: Obtaining, editing and quality:

Metadata is mainly obtained from suppliers via as MARC records made freely available when libraries sign up for PDA. One library reported that they do not subscribe to collections where free MARC records are not available.
Editing metadata was essential to some libraries, whilst other libraries reported that they would not edit records with the only exception to add data about reading list status or to add notes to retain for CLA purposes. This was problematic for libraries, particularly where libraries were reliant on supplier metadata. It was generally considered that metadata provided was fairly constant and not ‘too bad’ it was reported that the quality of metadata had improved significantly over the previous year.. Libraries reported that E-journal metadata remains consistently better than e-book metadata.If discovery services and their associated link resolvers become the primary means of discovery the may be issues about the quality of metadata need to enable knowledge bases/resolvers to make unambiguous links to the full text

Improving metadata

Specific suggestions for improvements to metadata were:
  • An unambiguous 020 field to allow matching and de-duping. De-duping records is real problem as libraries cannot successfully match ISBNs for this purpose. They are currently using the 002 field in Marc to’ clean up the junk’. Some suppliers even have their own ISBNs. Deduping is particularly vital where libraries are using large collection/package of ebooks in addition to collections to prevent users selecting the PDA item not the purchased version.
  • 776 field used for standard ISBN to assist in de-duping. (Herts)
  • Metadata which included references would be a great improvement. (Hudds)
  • Inclusion of DOI would help in references and in discovery (Hudds)