In October 2001, the Information Services Management Committee (ISMC) initiated a study of the feasibility of introducing virtual reference service. The study was broadly intended to ensure that we are aware of developments associated with virtual reference elsewhere and that our offering or not offering such a service is the result of an informed decision. More specifically, through the study we intended to learn more about what virtual reference really is, why we might consider offering it, what impact it would have on current Library services, who the stakeholders are, what staff and resources would be required to offer the service, how much the necessary hardware, software, furnishings or equipment would cost, and what risks might be associated with the service. We planned to summarize our findings in a report that would also include an explanation of issues associated with what we learned, a recommendation about whether to proceed with a pilot project, and recommendations related to implementation of such a project (if, in fact, we recommended going forward) along with a list of next steps. This is the report.
Virtual reference is one of many terms used to describe a way of providing traditional reference service on-line. Other terms used to describe the service include digital, on-line, electronic, real-time, live, and chat. For the sake of consistency ISMC decided to refer to the service as virtual reference, with the understanding that we might choose a different term if and when we start to deliver and promote such a service.
Regardless of the term used, a key feature of virtual reference is on-line chat. Using chat software, staff communicate with patrons in real time, interactively. With the chat feature, virtual reference can be a reasonably effective way of helping someone who has a fairly specific or simple question and at the end of the session the librarian can send the user a transcript for later reference. In addition to this basic feature, some virtual reference services also enable staff to take over a patron's browser and "push" web pages to the patron, and "co-browse" (patron and staff member view the same web pages simultaneously). With the push/co-browse features, virtual reference also becomes a powerful tool for helping with more complex questions, i.e. ones that require the development of search strategies and that lend themselves to instruction about some facets of information literacy.
Virtual reference service may be further enriched through the use of voice over internet protocol - a feature that would enable the librarian and patron to talk with each other while also co-browsing. As we understand it, development of this technology is still in the embryonic stage.
To answer this question we first asked and explored answers to other questions. Would introducing virtual reference be consistent with general University and Library directions and trends? What assumptions can we make about our users that might suggest that virtual reference is a service that they would value? What advantages would virtual reference offer to our users?
Introducing virtual reference would be consistent with broad University and Library directions and trends, in particular those associated with using technology. Through initiatives such as LT3 and the development of on-line courses, the University encourages and supports the increasing use of technology for learning and teaching; and the Library's statement of directions for 2002-2005 identifies a committment to "effective, focused and innovative use of information technology".
A specific example of virtual reference being consistent with directions and
trends relates to the University's plans for increased student enrolment.
More students will lead to more competition for classroom space, and we understand
that the University may respond to that competition by requiring students to
take some of their courses as distance education courses. There will also be
more competition for study space and workstations in the Library. Because virtual
reference would be available to users wherever they are, the service has the
potential to help offset some of the difficulties that may be associated with
increased enrolment.
Another specific example derives from something that we are seeing in reference
service as we currently provide it. The number of questions asked at the Information
Services desks has been continually declining for several years while the number
asked through our email reference service has been growing. Most reference questions
do still come through one of the Information Services desks, and the number
that come via email is still relatively small. However, the fact that the email
service is used, and is growing in use, suggests that an even more effective
on-line service would be used, especially if it was effectively promoted.
Based on what library staff have learned about our users by working with them in various ways, we made a number of assumptions about our users that suggest that they would be interested in virtual reference. These assumptions are: users want services available to them at time of need; they want easy access to services; they want services delivered quickly; they want services available regardless of where they are; they are increasingly computer literate and welcome services delivered on-line.
Virtual reference could be advantageous to our users in a variety of ways. In at least the foreseeable future, virtual reference would be offered in addition to in-person, email, and telephone reference and as such would extend a user's choice of ways to access reference service. Those who want immediate and/or in-depth service from their home or office, using their computer, could choose this option during service hours. Others might want to use the service even when they are in the Library -- for a variety of reasons such as not wanting to risk the loss of a workstation in order to go to an Information Desk for assistance, being more comfortable with written communication than with oral, or wanting more privacy than is possible at an Information Services desk.
The ideal reference transaction takes place in person, with the requestor and a staff member working at a computer together to locate and access resources. Since virtual reference more closely simulates in-person reference than does either email or telephone reference, it may be more effective than telephone or email service for the user who does not want to, or cannot, come into the Library.
Another advantage that virtual reference software offers is the ability to help a group of students on-line, all at the same time. A librarian could, for example, work with a professor to arrange for a scheduled, on-line, interactive information literacy session for students in a particular class. Alternatively, a librarian could let students in a particular class know that he or she would be available to them on-line at pre-scheduled times to provide assistance with their research. Virtual reference software may enable us to reach more students than we now do, in part simply because students may prefer being able to work from their computer.
The primary stakeholders are those for whom the service is intended and those who would provide the service. As with reference service currently offered, our primary clientele would be University of Waterloo students, faculty members, and staff members and within that group we would expect undergraduates to be the largest user group. Those who would provide the service are some or all of those whose responsibilities now include providing reference service.
Other stakeholders would include others to whom the service might be available and those who would provide support of one type or another. A list of identified stakeholders and the nature of their stake is appended to this report.
As noted above, for the foreseeable future virtual reference would be offered in addition to current reference services. At this point, it is not clear what impact virtual reference might have on other services, but the following might happen if the service is offered and becomes popular: questions coming via email may not continue to increase as people become used to getting immediate assistance through virtual reference; in-person reference service may decline even more dramatically, especially if students already working in the Library prefer to seek reference assistance from a public workstation; more distance education students may take advantage of reference service because virtual reference makes it easier for them; we may want to purchase electronic versions of more reference materials so that on-line users have easy access to them; we may see a shift in the way information literacy sessions are delivered if virtual reference software is used for this purpose. All of this might, in turn, lead to an overall increase in the use of Library services and resources - as it becomes easier for students to access and use these services and resources they may be better able to see their value and more inclined to take advantage of them.
To begin to understand staffing and related implications of virtual reference, a sub-group of ISMC conducted a telephone survey of three Canadian universities currently offering this service (Ryerson, Windsor and Alberta). Key findings from the survey include the following points:
The sub-group also conducted a literature search to obtain similar information about the experiences of U.S. institutions offering a virtual reference service. The survey responses were compared to findings in the published literature and found to be consistent. In addition, we have had opportunities to talk with staff from Guelph and have learned that their experience with virtual reference is consistent with what we learned through the survey and literature.
As noted above, those already offering reference service have concluded that it is best to offer the service from staff offices, rather than from a public desk, and this is what we would recommend if we introduce the service. We also know that the software that we would recommend is fully compatible with the workstations that our staff have. Since we already have the furniture, equipment and hardware necessary to offer virtual reference using the software that we will recommend, we anticipate no additional expenditures for these.
To investigate virtual reference software options, another ISMC sub-group developed a list of criteria to use as the basis for evaluating various packages. Features of software used for virtual reference are frequently described using language that we were not previously familiar with. To help us understand the terminology and to use it consistently and appropriately, the sub-group also developed a list of definitions for the terms in its criteria. The sub-group then compared the criteria to various software packages and prepared a table outlining their findings. The comparison table and the list of definitions are appended to this report.
The group had not concluded its software investigation when its members were invited to join a meeting of librarians from Guelph, Windsor, York and Ryerson. The libraries at each of these universities are already offering virtual reference and the librarians were meeting to share ideas about their experiences as background for recommendations that they planned to present at the April meeting of the OCUL (Ontario Council of University Libraries) directors. This meeting was fortuitous for us because what the sub-group learned enabled us to fast track our thinking about the software that we would recommend.
The key reasons for introducing virtual reference service are noted above, in the "Why might we introduce virtual reference?" section. Introducing it as soon as possible will enable our staff to gain experience with the service and to be well positioned to keep pace with, and influence, broader developments that may, in turn, influence how we offer the service. There is, for instance, growing interest within OCUL about the possibility of some form of collaboration in offering the service. Local experience is needed before we can decide whether participating in an OCUL collaboration, and perhaps other ones, makes sense.
How soon is soon? Ideally we would begin to offer the service as early as possible in the upcoming spring term. Launching the service when other public service activities are relatively quiet will allow staff time to develop the necessary skills and to identify and resolve unanticipated problems before the usual demands of the fall term must take priority. If we are not able to begin this spring, we will have to delay until either the winter term or next year's spring term.
As the appended software evaluation chart indicates, the subgroup considering software investigated seven options, comparing each one to their established criteria. The group obtained detailed information by reviewing virtual reference literature and vendors' websites, contacting vendors for further information, attending a demo of one package, NetAgent, that was arranged by the OCUL Projects Director and held at UW, and joining the meeting of librarians from Guelph, Windsor, York and Ryerson referred to earlier.
During the meeting with librarians from Guelph, etc., the group learned more about LSSI and concluded that it would best meet our needs. This conclusion was based in large part on an interest in three things: introducing virtual reference with full functionality software (i.e. not just chat but also functions such as co-browsing); ensuring that users would be able to access the service using a variety of web browsers and browser versions, from PCs and Macs, and that they not be required to load any software or register with a third party; supporting opportunities for co-operative ventures.
LSSI supports each of these interests and is already in use at three of the four OCUL libraries currently offering virtual reference service. Using the same software as other OCUL libraries will ensure that we are well positioned to take advantage of any collaboration opportunities that develop within OCUL. Additionally, there will be experienced staff nearby with whom we can consult and share experiences. We have already been invited to participate in further discussions about virtual reference developments with the librarians from Guelph, Windsor, York, and Ryerson.
LSSI is available through a licence that determines how many staff may work with the software, to provide service, at any one time. The licence refers to "seats" and one seat allows for one staff member providing the service at any one time; libraries can arrange licences for multiple seats. The price for one seat per year is US $6,000. There is also a one-time installation and training fee of US $9,000, plus the trainer's travel expenses.
Because of the cost, we are recommending that we start with just one seat. The software can be accessed from any number of computers and by any number of people, but only one computer and one person at a time. Until we have a better understanding of the demand for the service, licensing only one seat seems prudent. This is, however, less than ideal because of our multiple library locations and because staff are currently trained to provide reference service for the types of questions, and using the types of resources, particular to one location. How we will manage provision of the service in our environment with only one seat requires further thought. Some possible solutions have been identified and discussed briefly, and while this is a significant issue we are confident that it is a manageable one.
ISMC invited all staff who provide reference service to a meeting on March 25 to talk about matters related to staffing virtual reference. All the library assistants from ISR and from UMD were at the meeting, as were most of the ISR librarians and two staff members from Cataloguing who work at an Information Services desk.
At the end of the meeting, participants were asked to send individual email messages to ISMC, or any member, with their response to two questions: do you think that the Library should provide virtual reference service, and are you interested in participating in provision of the service? ISR librarians who missed the meeting were subsequently also asked to respond to the two questions.
All the assistants and librarians sent responses indicating support for introducing the service; most of the assistants and all the librarians also expressed interest in providing the service - although a few had reservations about adding new activity on top of already heavy workloads. The two people from Cataloguing did not have an opinion about whether we should introduce the service but indicated a tentative interest in participating if we do.
Knowing that introducing virtual reference is widely supported by those whose work would be most affected, but also knowing that there was a significant concern about workload on the part of some library assistants, ISMC decided that it would be best to introduce the service with librarians only. We know that some of the assistants are very enthusiastic about participating, but they also carry an especially heavy reference load because they are responsible for email reference service, and they frequently work extra shifts at the Information Services desks when there are scheduling problems. In addition, some of the assistants would need more training and experience with reference work in general before they could assume responsibility for the broad range of questions that we anticipate. We very much appreciate the interest and enthusiasm that the assistants have shown and will find opportunities for them to participate as soon as reasonable. Later on we will also give further thought to the possibility of involving those from Cataloguing and ILL who currently provide reference service.
ISMC invited ISR librarians to an April 2 meeting to talk about how much time they could individually commit to the service. All but one librarian (who was on vacation) attended this meeting, and at the end all were asked to let us know by email whether they were interested in participating in the first phase and how many hours they could commit. Most of the librarians want to start during the first phase and they are offering a sufficient number of hours to enable us to start.
Providing virtual reference at night or on weekends would require significant adjustments to other services and/or an increase in staff, or possibly time sharing through collaboration with other libraries. We should start by offering the service during hours that we can manage reasonably well and consider possibilities for expanding after we have some experience. Libraries currently offering virtual reference, including those offering it at night or on weekends, report that their peak hours are weekday afternoons. While a recommendation about precise service hours will be left to the task group referred to below, we should focus on weekday afternoons but include mornings if we can.
If we are to introduce virtual reference, ISMC will establish a task force to develop plans for the pilot or first phase, implement the pilot, monitor the service, and establish evaluation criteria. Draft terms of reference are appended to this report.
As noted above, we have already been invited to participate in discussions with other OCUL libraries and we will take advantage of other opportunities to keep abreast of developments.
Next Steps
Assuming that financial resources are available for an LSSI licence, next steps include:
UW Library clients: students, faculty members, and staff members
Quality, effectiveness and accessibility of the service will
be of considerable interest to them
Other Library clients: those
clients who are not members of the UW academic community
Not our primary client
group but, depending on our service policies, may have some form of access
to virtual reference
Information service providers (initial phase): librarians in ISR
Information service providers (following the initial phase): as the service matures, others who provide reference service and are interested in virtual reference will have opportunities to participate
Stakeholders: Support Roles (precise support roles have yet to be more thoroughly considered; the following is tentative and intended as a starting point for further consideration)
UW Library Systems Dept
Software installation
and maintenance
Software licensing issues
Connectivity and bandwidth issues
Systems Help Desk support
Hardware support
POLARIS issues? Endeavour issues?
Assistants in ISR and UMD
Desk scheduling
Record keeping (statistics)
Webpage creation and maintenance (public and staff areas) including FAQ
E-mail messages from virtual reference clients: routing, responding
Library administrative / management staff in Library Office, ISR,
UMD
Policies and procedures
Monitoring and evaluating virtual reference
Financial, human, physical resources
Decision making
Software vendor (LSSI)
Ongoing technical support
Database providers / vendors
Ongoing issues (e.g.,
access, licensing) re e-journals, journal indexes, e-data, e-texts, etc
Stakeholders: Awareness Only
Those with whom we should communicate in order to provide awareness of VR implementation
Chat - ability for patron and librarian to converse in real-time, by typing messages back and forth
Push pages - ability to push the patron's browser to web sites (e.g. type or paste a URL and cause it to appear on the patron's browser screen)
Send active embedded URLs - ability for the librarian
[and patron if two-way] to send URLs as hyperlinks in the chat window
Queue management - queuing and routing system for distributing
calls appropriately in a networked environment (or alert to patron re call position
in a queue)
Librarian-initiated termination - ability for the librarian
to terminate a session (e.g. in case of an obscene call)
Call alert - a signal alerts the librarian to incoming calls
(pop-up box on screen, computer sound, call to a pager, etc.)
Preset answers for FAQs - librarians can create files of pre-scripted
frequently-used messages and URLs
Email chat transcripts - automatic emailing to the patron of
a transcript of the session, complete with URLs, at the end of a VR session
Reports/statistics - transaction logs are maintained and management
reports are available
Redirection or message when closed - there is a library-controlled
message or action when the service is closed or temporarily unavailable (e.g.
pre-scripted message to patron, automatic default to email)
Co-browsing/Escorting - ability for the librarian to "lead"
the patron around the web as they navigate the web together (where you go, patron
also goes) ("problem" sites may cause the librarian's or patron's
system to freeze and thereby disable chatting)
System Stallers - phrases are generated automatically at specific
intervals to let the patron know that there is still someone on the other end
of the chat ("I'm working on it")
Form sharing - ability to share text search boxes and web forms;
form-synchronization feature
Call transfer - ability to transfer the patron to
another staff member
Pull pages - ability for the librarian to "pull in"
and see what the patron has onscreen (e.g. "just what error message are
you getting?")
Searchable transcript database - transcripts are logged in
a searchable database; librarian can search question in database when chatting
with patron
Built-in Survey/evaluation - ability to automatically send a patron an evaluation form, online or via email, at end of a session; ability for library to create and send a survey
Persistent buttons - "Call for help" button
can be set up on all Library web pages without having to code it into the page
(variation on a server side include?)
Spellchecker - inclusion of a standard spellchecker feature
to automatically check [and correct] spellingVideo - patron can see the librarian
(provides patron visual cues missing in a chat environment)
Voice over IP - voice communication over the internet (librarian
and patron must have the voice receiver required; evolving technology that may
make "chat" obsolete)
Remote control - ability to help a patron by correcting software
settings on the patron's PCs (e.g. to resolve proxy problems)
Traffic monitoring - ability for librarians to see visitors
enter the Library's web site, track what pages they visit and proactively
offer help
No special software needed - requires no plug-in or other software be installed by the user; requires only web browser
Supports multiple browsers - system works the same across various browsers and browser versions (e.g. Netscape & IE, 4.x or higher, Mac & PC)
Proxy compatibility - sign-on procedure allows authentication of patron, compatible with current or future UW proxy system
No need to register with 3rd party - user does not need to register with a third party to be able to use the service (e.g. Netscape Instant Messenger)
Adaptive technology compatible - appropriate for use by visually impaired patrons or compatible with screen-enhancing software
Ease of use (intuitive) - screen presented to the user is intuitive (standard Windows interface, clear instructions, etc.), screen layout & vocabulary are user-friendly
Response time - acceptable demonstrated response time once a session begins, speed pushing pages, etc. (accepting that this will vary by time of day, network load, etc.)
Staff/Support:
Vendor support - vendor understands library terminology and service goals, has a support contact/response mechanism (help desk), client base of other libraries
Handles advanced html features (frames, Java) - screen displays correctly for both patron and librarian when visiting "screen-busting" sites, no problems navigating "problem" sites, etc.
Customization capabilities - library can configure screens (local top page, a UW look & feel, help information, buttons, etc.)
Documentation - vendor provides clear documentation
Training tools provided - vendor provides training and/or training tools
Co-operation opportunities - software could support collaborative reference operations (e.g. TUG, OCUL co-operation)
Ease of use - standard Windows interface, commands; screen layout & vocabulary are user-friendly
# simultaneous operators - number of operators who
can be logged on at the same time; number of workstations on which software
may be installed; number of concurrent chats permitted per operator
S/W on local or remote server - ASP [Application Service Provider]
(software resides with vendor - issues re ownership of knowledge base, FAQs;
data if library changes vendors, etc.) vs installation and maintenance on local
server (issues staffing for set up and ongoing maintenance)
Web or client/server local S/W - web-based or requires installation
of client software on VR staff workstations and server
Identifies patrons by IP address (re tracking & history) - IP addresses needed to create patron histories and compile patron data (privacy issues?)
Cost - costs, options, extra fees (e.g. set-up, customization, training)
Libraries using - examples of libraries using the software
March 4, 2002
The Virtual Reference First Phase Task Group will develop plans for the introduction of virtual reference, will coordinate introduction of the service, will monitor the service through December 2002, and will make recommendations about ways in which the service might be evaluated. The Group will be a sub-group of the Information Services Management Committee (ISMC) and will include some members of ISMC and others - six people in total.
The Group will continue until at least December 2002; in December ISMC will consider what needs to happen next.
The Group will be responsible for considering various matters of policy or practice, for making recommendations about these matters, and for consulting with ISMC, those who will be providing the service, and others as appropriate as part of the decision making process. Specific questions to be considered include:
When the Group begins its work, other policy/practice matters to be considered may become apparent.
The Group will also be responsible for performing various tasks, including:
Other tasks may become apparent once the Group begins its work.
Prepared by ISMC April 9, 2002