Less is More

A Case Study of Consolidating Two Database A–Z Lists for Better Staff and User Experiences

Todd Quinn

Todd Quinn (tq@unm.edu) is the Business & Economics Librarian at the University of New Mexico in Albuquerque.

Manuscript submitted January 13, 2020; returned to author for revision April 20, 2020; revised manuscript submitted April 20, 2020; manuscript returned to author for minor revision May 27. 2020; additional revisions requested by editor June 8, 2020; revised manuscript submitted by author June 9, 2020; accepted for publication June 15, 2020.

This case study describes the consolidation and migration of the University of New Mexico’s University Libraries’ two database A–Z lists. A subject librarian led the nine-month project that included most subject librarians, the electronic resources team, the Director of Collections, and the Web & Discovery Librarian. The project also provided the University Libraries with the opportunity to review the resources on the lists, update descriptions, and create new workflows for adding and managing a single list.

The University of New Mexico’s University (UNM) Libraries’ (UL) first robust database A–Z list was developed by a library IT staff member who also worked shifts at the reference desk. This experience of working with users and other public services librarians provided him with insights as he developed the A–Z database tool to meet users’ and staff needs. Many libraries have developed homegrown A–Z tools using different methods.1 The UL’s list was created around 2005, and enabled the electronic resources (e-resources) team to track and maintain subscription databases and push the resources to the public via the library’s website. The tool allowed the e-resources team to add internal notes, descriptions, and create persistent URLs (PURLs) for each resource. The PURLs allowed the e-resources team to maintain consistent URLs to the public even as database vendors changed the resources’ URLs. These PURLs also saved librarians time since they did not have to update library pages and guides when the vendors changed these URLs.

The official list’s content, which lacked consistency, was determined by the Director of Collections, the e-resources teams, and subject librarians. Brisbin, Parlette-Stewart, and Oldham explain, “one of the challenges that arose over time was a lack of coordination. This led to inconsistent tagging, lack of naming conventions, outdated descriptions, dead links, and a lack of a shared understanding around the intentions of the list. It was clear that the database list was suffering from a lack of care and attention,” which was the UL’s exact situation.2 Unfortunately, as noted, there are few papers in the professional literature on managing A–Z lists.3 These two lists coexisted for a few reasons: each was managed by two different UL employees, the LibGuides A–Z tool made it easier for librarians to add databases to guides, and there was a lack of time by the author to fully investigate the LibGuides A–Z tool. Additionally, the individual who had served as the UL’s web librarian departed to accept a position at another institution. She worked with the author on LibGuides tools, including the A–Z list, and her departure created a gap in knowledge and skills. The parallel management of the two A–Z lists arose when the UL subscribed to LibGuides V2. The LibGuides A–Z list was initially populated via its import process and grew as subject librarians added free resources to the list to enable them to easily add these databases to their guides. This secondary A–Z list was public, but diverged from the official A–Z list, managed by e-resources, with these additions.

The UL used Springshare’s LibGuides for many years, mostly to provide subject and course guides for students and faculty. In 2014, Springshare released LibGuides V2, and the author led the UL’s migration from V1 to V2. He got consensus from colleagues to delete all the UL’s guides in V1 and to start fresh for V2 since many guides were outdated or obsolete. This allowed the UL to incorporate best practices for all guides in V2. As part of the migration, the author started the secondary A–Z list in LibGuides in collaboration with the creator of the homegrown A–Z list. The LibGuides A–Z list allowed the librarians to easily add/reuse databases with consistent descriptions to course and subject guides. Librarians could also change a database description from the general description to something more specific for a particular guide or audience. For example, EBSCO’s Business Source Complete contains articles/journals useful for many disciplines, so most guides contained the general description of the databases, but a librarian may change the general description to focus on its management content and yet another librarian may change the general description to focus on its legal content. Even though librarians are using the same database in their guides, the descriptions differ because the audiences for the guides are different. This flexibility is useful for subject librarians, but unimportant for e-resources management.

After the subject librarian, who managed the LibGuides, investigated LibGuides’ A–Z tool, he discovered it had several features that were available in the UL’s homegrown tool, plus other advantageous features for both internal and public use. Internally, the LibGuides A–Z list would allow the e-resources team to track resources, add notes, export the list, add the proxy URL, and to highlight trials. On the public facing side, the LibGuides tool allows patrons to browse the list alphabetically, sort the list via the three drop-down menus (subject, type, and vendor), see “best bets” by subject, and use the search box. The homegrown application had few of these features, and was no longer being developed, only maintained. Switching to the LibGuides A–Z tool saved the Library IT department time for other projects since they would no longer need to maintain the application or server for this purpose.

Most libraries have a database A–Z list to help patrons find specific resources, but there are unfortunately few papers in the professional literature on managing A–Z lists. As Hoeppner explains, there are “no articles primarily offering practical advice to librarians about database list management” and electronic resource management articles only gave cursory mention to “A–Z database list functionality.”4 Through a literature review, the author found that five papers were found that specifically discuss database A–Z lists.

Hoeppner, at the University of Central Florida Libraries, conducted a survey of libraries on Database List Systems (her term).5 She discovered that over 72 percent (47/65) of respondents used LibGuides as a database management system, and half used LibGuides as the public facing database A–Z list tool. She found that other respondents used eleven tools other than LibGuides for managing and displaying their A–Z list to users. The second part of her paper provides many useful tips for creating and managing a functional database list.

Brisbin, Parlette-Stewart, and Oldham discussed the approach taken by the University of Guelph’s McLaughlin Library to migrate their database A–Z list from a homegrown ColdFusion system to LibGuides.6 They discuss the many challenges faced in creating the list, including an in-depth discussion on how their list is managed.

Tobias discussed Michigan State University Libraries’ approach to making their database A–Z list more user-centric.7 They conducted many usability testing sessions and physical card sorting with users to receive feedback so that they could create a more user-friendly interface. Furthermore, they researched best practices to develop their new A–Z list using LibGuides as a management tool and public facing site. They developed a process for reviewing and editing database descriptions and determining which titles to include in their list.

A brief paper by Arnold covers the process of migrating from a homegrown system to LibGuides at the University of North Carolina Chapel Hill Libraries.8 They conducted interviews and usability testing with their undergraduate students, researchers, and faculty to understand their usage of the UNC-Chapel Hill Libraries site. The few findings provided do not specifically discuss the use of the A–Z list by their patrons.

Ramshaw, Lecat, and Hodge at the American University of Sharjah provide a coding solution by connecting OCLC’s WorldShare Management Services (WMS) to LibGuides A–Z list for their patrons.9 Initially, they populated their A–Z public facing list using an XML file, then migrated to the WMS library services platform, which necessitated a change in their process. Their A–Z list project was guided by four goals: be automated, be clean, be consistent, and maintain continuity. Since they met these goals, they suggest that other WMS libraries with LibGuides use their solution for internal workflow and populating an A–Z list for their patrons.

It seems that these libraries, among others, want similar things: a tool to manage the technical aspects of database lists with a good workflow, and a public facing A–Z database list that is understandable and useful to patrons.

The UL at the author’s institution had two lists (official and secondary) from 2014 to 2018, which is not uncommon.10 The official list, created in 2005, was on a library server and used a homegrown application to manage the e-resources and push content to a library web page. The secondary list was created in 2014 when the UL migrated from Springshare’s LibGuides V1 to V2. These two lists never included identical content because the lists were maintained by two different people in separate departments and for different purposes: one to manage e-resources and the other to support subject librarians’ work. A subject librarian who maintained the LibGuides list saw the need for a change and led a project in late 2017 to create a single consolidated A–Z database list for both purposes, and to complete the project by August 2018.

Project

The consolidation project began in fall 2017, with the project lead, the author, gaining consensus from subject librarians, the e-resources team, the Web & Discovery librarian, and the Director of Collections. The project lead submitted an official project request to the UL’s administration for review. This project form incorporates project management concepts, which is similar to the University of Guelph McLaughlin Library’s A–Z project.11 The document named all the employees who were involved, their roles, time commitment, timeline, outcomes, budget, and assumptions. After the administration approved the project, work began in November 2017 and concluded in August 2018. The project outline was communicated to the entire UL via email and in various meetings, which allowed people to provide email and in-person feedback and ask questions. The project’s ultimate goal was to create one database A–Z list that provided a better user experience (internally and publicly). Secondary goals included: reviewing and updating database titles and descriptions, updating the subject and type categories, removing outdated free resources, and assigning a champion to each resource who would apply keywords, subjects, and types as appropriate.

The project consisted of four phases: (1) create documentation and guidelines; (2) subject librarian review and update of the 500 databases in the lists; (3) create a workflow for the e-resources team; and (4) provide a clean searchable/browsable A–Z list to the public, which were similar to the goals of the University of Guelph McLaughlin Library.12

Phase 1: Documentation and Guidelines

To make the project successful, workflows, timelines, and documentation were needed. First, the project lead consolidated the two lists into one master spreadsheet in a workbook. The master spreadsheet listed all titles in both lists, eliminating most duplicates. In addition to a locked master spreadsheet in the workbook, the databases were divided among spreadsheets for general databases, and one for each subject cluster: Arts and Humanities, Sciences, and Social Sciences. The twenty-two columns for each spreadsheet were color-coded to signify who should edit the various columns: gold for subject librarians, pastel blue for the e-resources team, green for the web librarians, and no color meant that they were not to be touched. The columns included title, description, subjects, PURLs, Reviewers, among others (see table 1).

Second, documentation on the process for reviewing each database was drafted by the project lead and edited by subject librarians. The documentation provided explained how to review the resources in the A–Z list (see table 2) and provided guidelines/format for editing and rewriting current database descriptions (see table 3). Brislin, Parlette-Stewart, and Oldham, along with Tobias, created similar guidelines for their projects.13

Finally, all project participants were provided read/write access to a folder via Office 365 that included the workbook and the documentation files. Arnold describes a more complicated review for the University of North Carolina-Chapel Hill Libraries.14 The group chose Office 365 since all participants had access to it via their UNM account, and it allowed them to see the project’s progress. Original copies of all files were also stored on the project lead’s desktop.

Phase 2: Review and Updating A–Z Content

In 2016, the UL’s subject librarians were grouped into three subject/outreach/liaison clusters: Social Sciences, Sciences, and Arts and Humanities. Each group has a lead who convenes meetings and sets agendas. Of course, not all disciplines easily fit into one of these clusters, and many colleges, departments, and programs are interdisciplinary in nature. Two examples that illustrate this point are the Water Resources program, which is in the Economics Department, and the program covers sciences and social sciences; and the Urban Planning program in the School of Architecture + Planning, includes all three discipline areas. Although the UL has these three clusters, individuals and the groups must work across disciplines. For this project, the sciences cluster eventually included librarians from the Health Sciences and Informatics Library Center, which is a separate entity at UNM and not part of the UL.

With these clusters in place, the project lead divided the five hundred databases among the three clusters, plus a fourth group for “general” databases, led by the Director of Collections. Many databases could fall into multiple clusters, but each database had to be placed somewhere in the workbook. Each group was assigned between 120 and 180 databases to review and edit. The project lead met with each cluster to discuss the project and documentation, and the workflow. After these meetings, each cluster devised their own internal workflow. Librarians in all the clusters worked independently, and the sciences and social sciences clusters also held working lunches. The Arts and Humanities cluster was more challenging because liaison duties are secondary to most librarians in the cluster, thus most of their work was solo. Additionally, to encourage progress, the project lead held a few editing/writing meetings that were open to all participants.

The “general” group had the most databases (approximately 180) to review. General databases are those that are applicable to multiple disciplines, and included titles such as Academic Search Complete, JSTOR, Web of Science, and WorldCat. This was led by the Director of Collections and included volunteers from all clusters.

Each librarian reviewed the assigned databases, wrote a concise description, provided a coverage date range, and reviewed/assigned subjects and types. This basic description format worked for most titles. Following a librarian’s review/rewrite, it was reviewed by a second librarian or their cluster. Columns were provided in the spreadsheet for each title’s review status. This review/rewrite process took seven months.

Issues arose during the review. Some database vendors provide a parent URL for all indexes, and secondary URLs for individual titles or small groupings. This is a nice option but can lead to confusion for patrons. The librarians needed to decide whether to collapse the titles or to keep them separated. For example, the three main Web of Science indexes were listed as individual databases, along with a listing for Web of Science Core (the three main indexes and some other indexes). After discussion, it was decided that only the Web of Science Core would be listed since most users knew the title Web of Science, not necessarily the individual index titles. Other Web of Science titles (e.g., Biosis, Zoological Record) continued to be listed separately since they have distinct audiences and are not part of the core. Another example is ProQuest’s Global Newsstream, which has over two thousand sources and includes prominent news sources specifically sought by name, such as New York Times (current), Wall Street Journal, and the Los Angeles Times. A decision was made to list these individual titles separately, along with a Global Newsstream listing, to help users find specific content or search the entire database.

Since the UL adds free resources to the A–Z list, these needed to be reviewed as well. The same guidelines were followed for this part of the project. The review led to a number of deletions of outdated and “dead” resources. A new workflow was developed (see below) for adding free resources to the A–Z list.

Phase 3: Workflow for the E-Resources Team

The e-resources team includes two staff members who handle much of the technical communication with vendors and maintain the database A–Z list. Once the project was initiated, it was decided that they would maintain both lists. This meant that any changes to the A–Z list were made in both the homegrown system and the LibGuides system. Although this was a duplicate effort for several months, it allowed the two staff members to learn the LibGuides system and to create a workflow with support from the project lead and their supervisor for the LibGuides system.

As part of this workflow, the e-resources team conveyed two concerns to the project lead: (1) the team never knew specifically who to contact when a database needed attention (e.g., renewal, updates), and (2) there were no guidelines for adding free resources. When a database needed attention, they would contact the subject librarian who they assumed was responsible for the topic, which was not always the correct person. This was not their fault since no list of subject specialists existed. This was resolved by assigning each current database title to a specific individual. These assignments, which were recorded as a column in the spreadsheets, were determined by the clusters and the Director of Collections. Any new subscription databases or free resources would also be assigned to a specific individual after the migration. Adding free resources to the A–Z list was another issue. Typically, the e-resources team would add titles suggested by any library employee, which were usually subject librarians. This was the practice for more than ten years, which led to bloat in the list, among other issues. The project lead devised a process (see table 4) for these types of resources. After the migration, all free resource suggestions were reviewed by the most appropriate cluster (similar to a database review) before they were added to the A–Z list by the e-resources team.

Phase 4: Web Design

The web design phase was led by the Web and Discovery Librarian. His responsibilities entailed managing the display of the A–Z list and reviewing/revising the current database subject and type terms. Since the UL had two A–Z lists, each A–Z list had its own subjects and types, which again did not correspond with each other. The review of subjects and types was more complicated. This was difficult since each specialist had a different interpretation of the information. The Web and Discovery Librarian led the discussion, and a decision was made to eliminate some narrow subjects and to limit the number of types.

The LibGuides A–Z tool provides the option to include a description of each database, keywords, and alternative names. This was an opportunity to assign keywords not in the description and to add keywords for common misspellings of database names. A few sub-database titles were removed from the A–Z list and were added in the description or as keywords to the “parent” database. For example, the Web of Science’s three main indices were removed, and these index titles were added to the Web of Science Core description. It was found that these keywords displayed on the LibGuides A–Z public page under the database title and description, which cluttered the display and could be confusing. Following discussion, the Web and Discovery Librarian added programming that suppressed keywords and alternative names from the public view. These keywords were still indexed and searchable, and the suppression provided patrons a cleaner view of the A–Z list.

The e-resources team notified the project team that there was an issue regarding UNM’s four branch campuses’ libraries as it related to the Database A–Z list. These branches are separate from the main campus, are located outside of the Albuquerque metropolitan area, and have access to most, but not all, UL subscription resources. Before the project, the UL provided a web page to list the databases specifically available to the branches, and this page would not work after the project. To resolve the issue, a keyword was created specifically for the branches. This keyword was added to all the databases available to the branch campuses. By adding this keyword, a user could find these databases with the keyword via the LibGuides A–Z search feature. More importantly, the search created a persistent URL listing for these databases, which allowed the branches to display this list on their library sites. Changes are automatically displayed without any additional work required from the branch campuses.

Issues with Springshare

The LibGuides A–Z tool is a Springshare product and has provided the UL with a better tool to use internally and to serve the public. They provide quality support and are responsive to their customers. However, the tool has presented some issues for the UL: (1) it would be ideal to provide a toggle switch that lets libraries decide to show or hide keywords and alternative titles from the public review; (2) most public users will never use the vendor’s drop-down menu, and an option to hide this menu would be useful; and (3) when one selects from any of the drop-down menus, it not only updates the A–Z list for this selection, but it also updates all the drop-down menus. For example, when one selects Business from the subject menu, and then wants to select Art from the subject menu, the user sees only Art databases that include Business as a subject. This may confuse patrons who want to see all the Art databases. A patron must select the “clear filters” button to see the full list again for each subject, type, and vendor in each menu. Potentially, Springshare could provide an option to update the display based on a patron menu selection, but not update the menus. This would allow libraries to choose which option they prefer.

The LibGuides A–Z tool’s search function is a useful addition to A–Z lists. Patrons may now use the A–Z search box in addition to browsing alphabetically or by subject or type. The search may be faster for patrons who know the database title or those seeking to find a list of databases by keyword. The addition of common misspelled database names as keywords has proven useful. For example, PsycINFO is often misspelled with an added “h” as in PsychINFO. This misspelling was added as a keyword and the A–Z search data has shown that the misspelling is the more common search term than the correct spelling. As part of an annual review, the UL can discover potential other keywords (and common misspellings) by reviewing all the A–Z searches, which are collected in the system.

Although the search function is useful, it has a glaring problem as it relates to subject assignments (selected by librarians). According to the UL’s Web and Discovery Librarian, David Hurley (personal communication, April 22, 2020), the search algorithm assumes a database with only one subject assignment is more relevant than a database with more subjects assigned. In theory this may seem appropriate but does not work in practice. Some databases are the best database for multiple disciplines and are assigned multiple subjects. For example, Web of Science Core is a top research literature database for multiple sciences, but when a patron searches “biology,” Web of Science Core is listed last, thirty-second, in the search results. Springshare is aware of the issue but has not provided a solution.

Springshare resolved one issue during the project. When the e-resources team added a trial database, they used the trial checkbox provided on the form, which included an expiration date. This highlighted to the public which titles were under review. Unfortunately, the LibGuides default for expired trials was to remove the trial icon and leave the title in the A–Z list. After the UL raised concerns, Springshare changed the default to suppress trial databases on the expiration date.

Discussion

Overall, the project was a success. The UL now has one A–Z list with more concise and consistent descriptions (see table 5) for the public with some extra features (drop-down menus and a search box). At the end of the project, the number of resources in the A–Z list dropped from 504 to 420. The drop is attributed to the deletion of outdated and “dead” free resources, and the consolidation of some subscription database titles. Unfortunately, there was no statistical method to track the impact of these changes on usage for a few reasons: (1) the homegrown application did not provide usage statistics; (2) database titles were consolidated and separated; (3) LibGuides consolidates all database usage statistics from the A–Z list, guides, direct friendly URLs, and more; and (4) user satisfaction surveys were not conducted due to lack of time.

There were other positive outcomes. The e-resources team now has one A–Z list (LibGuides A–Z) to maintain, has a better workflow, and there is improved communication between the subject librarians and the e-resources team concerning additions and edits to the list. The library IT department no longer needs to maintain the homegrown application and server, which saves them time for other projects. Plus, during the review, the subject librarians learned more about individual resources. Librarians were able to convey their knowledge about specific resources to their colleagues as titles were discussed. This positive, unforeseen outcome has provided opportunities for librarians to expand their resource toolboxes. The project has not resolved all issues (e.g., subject and type lists), nor fixed all database descriptions or titles. An annual review of the list and guidelines is probably necessary since the list is dynamic and there is always personnel flux.

As Brisbin, Parlette-Stewart, and Oldham state, “When library staff use the term ‘database,’ they are referring to something very specific, yet this nebulous term is often quite meaningless to our users.”15 It would be easy enough to use the term “resources” instead of “databases,” but the profession has chosen “databases,” and it is the term we have taught to our users over the decades. These “databases” include subscription databases, e-reference titles, and free resources (e.g., US Census, Hispanic American Periodicals Index (HAPI)) selected by librarians. When the profession uses the term “databases,” it also means various types of content (e-books, journal articles, news articles, indexes, datasets, maps, finding aids, catalogs, etc.), which adds to the confusion. Initially, many libraries used a homegrown A–Z database tool to feed their subscription databases on a public-facing web page. But the list changed once many quality free/open access resources became available. Additionally, many databases index multiple formats, and provide full text. These additions are quite valuable to patrons, if they can locate the resource, but cause technical and workflow issues for libraries.

Conclusion

In a larger sense, A–Z database lists will never be able to convey the content of all the library’s subscription and free resources to users. One general description developed by subject librarians is better than vendor marketing language, but rarely conveys all the valuable content of any one resource. Some libraries have provided detailed records for databases to users, but it is doubtful that the majority of users read this content. Also, many of these descriptions and listings appear to be more for the library staff than users. Furthermore, the web or user experience librarian must weigh content depth, layout, and discoverability to produce a valuable list to users. Even a more technical solution, such as Ramshaw, Lecat, and Hodge suggested does not resolve these underlining issues.16 The lack of literature on database A–Z list potential is because the list falls into many areas (e.g,. public services, web, user experience, e-resources) and each library handles it differently.

Thus, a well-designed and maintained A–Z database list, along with a quality public-facing A–Z list/tool, can alleviate part of the issue with the term “database” and the variety of types of database, formats, among other issues. Any A–Z tool will not resolve all the confusion, especially for new students, but an A–Z tool that allows users to browse and search provides users a chance to discover the most appropriate database(s) for their information needs. The A–Z database list is an ever-evolving resource that needs more attention and the hope is that vendors and librarians continue to investigate and communicate on this issue and continue to develop tools (commercial or homegrown) that best serve our users.

The UL’s project started simply to consolidate the current two database A–Z lists into one, and then migrate the new list into the LibGuides A–Z tool. This provided the UL with the opportunity to create new e-resources and subject librarian workflows, update the listed databases, update each database description, and create guidelines for any new additions to the list. The list and workflows will need constant tending by all members. Ultimately, over the nine months all members of the project met the goal of providing a tool that better supports patrons and staff so they may be more efficient and effective in their research and work endeavors.

References

1. Sarah Joy Arnold, “Out with the Old, In with the New: Migrating to LibGuides A–Z Database List,” Journal of Electronic Resources Librarianship 29, no. 2 (2017): 17–120, https://doi.org/10.1080/1941126X.2017.1304769; Kailey Brisbin, Melanie S. Parlette-Stewart, and Randy Oldham, “A–Z List Migration: Employing Collaborative Project Management at the University of Guelph McLaughlin Library,” Collaborative Librarianship 10, no. 4 (2018): 234–50, https://digitalcommons.du.edu/collaborativelibrarianship/vol10/iss4/4; Athena Hoeppner, “Database lists A to Z: A Practitioner’s Tips and Caveats for Managing Database Lists,” Serials Librarian 73, no. 1 (2017): 27–43, https://doi.org/10.1080/0361526X.2017.1320779.
2. Brisbin, Parlette-Stewart, and Randy Oldham, “A–Z List Migration,” 237.
3. Brisbin, Parlette-Stewart, and Randy Oldham, “A–Z List Migration,” 235; Hoeppner, “Database lists A to Z,” 32.
4. Hoeppner, “Database lists A to Z,” 32.
5. Hoeppner, “Database lists A to Z,” 33.
6. Brisbin, Parlette-Stewart, and Randy Oldham, “A–Z List Migration.”
7. Christine Tobias, “A Case of TMI (Too Much Information): Improving the Usability of the Library’s Website through the Implementation of LibAnswers and the A–Z Database List (LibGuides v2),” Journal of Library & Information Services in Distance Learning, 11, no. 1–2 (2017): 175–82, https://doi.org/10.1080/1533290X.2016.1229430.
8. Arnold, “Out with the Old.”
9. Veronica Ramshaw, Véronique Lecat, and Thomas Hodge, “WMS, APIs and LibGuides: Building a Better Database A–Z List,” Code4Lib Journal no. 41 (2018), https://journal.code4lib.org/articles/13688.
10. Brisbin, Parlette-Stewart, and Randy Oldham, “A–Z List Migration,” 236; Hoeppner, “Database lists A to Z,” 32.
11. Brisbin, Parlette-Stewart, and Oldham, “A–Z List Migration,” 237–38.
12. Brisbin, Parlette-Stewart, and Oldham, “A–Z List Migration” 237.
13. Brisbin, Parlette-Stewart, and Oldham, “A–Z List Migration” 238–39; Tobias, “A Case of TMI,” 180.
14. Arnold, “Out with the old,” 118–19.
15. Brisbin, Parlette-Stewart, and Oldham, “A–Z List Migration,” 234.
16. Ramshaw, Lecat, and Hodge, “WMS, APIs and LibGuides.”