Conceptualizing collaboration and community in virtual reference and social question and answer services.

Introduction

Virtual reference services are well-established offerings that have existed for more than twenty years including email and live chat (Sloan, 2006). virtual reference services are accessed through library websites and are evolving to enhance users’ experience, increasingly allowing entre through text-messaging, mobile devices, and social networking sites (Cole and Krkoska, 2011). Empirical data are needed to gain understanding of possibilities to enhance the effectiveness of virtual reference services given the economic downturn, which has negatively impacted library funding (Goldberg, 2010; Henderson and Bosch, 2010), and prompted “devastating service reductions” (“State Budgets Hammer Public Libraries Nationwide,” 2009, p.19). Standalone virtual reference services for individual libraries and some collaborative services, although they effectively leverage consortia resources, have been discontinued or endangered due to reduced human or budget resources (Radford and Kern, 2006).

Social question and answer services provide an opportunity for crowd-sourcing in lean economic times that increase the need for collaborative resource sharing and revenue generation. This short paper reports preliminary findings from Cyber Synergy: Seeking Sustainability between Virtual Reference and Social Q&A Sites, a multi-year project funded by IMLS, OCLC, and Rutgers University that explores how to better employ existing and frequently underutilized virtual reference services (see also Naylor, Stoffel, and Van Der Laan, 2008). It also investigates how to harness deep subject expertise of librarians through technical innovation and virtual collaboration, and to connect with potential users (and experts) through social question and answer services to create more open, innovative, and sustainable future pathways. Insights from in-depth interviews with virtual reference services librarians on topics including referrals, collaboration, social question and answer services, and the application of the communities of practice theoretical framework (Wenger, 1998, 2004) will help to understand current professional practice and to inform new service model designs.

Literature Review

Extant research examines the use and effectiveness of virtual reference services, including user perspectives, interpersonal dimensions, and accuracy (Connaway and Radford, 2011; Radford and Mon, 2008; Ross, Nilsen, and Radford, 2009). A rapidly changing information landscape and economic pressures invite innovative ways to address user needs and to provide a more sustainable virtual reference services model. One possibility is to turn to “social reference,” a collaborative endeavor that “may differ significantly from the traditional (and digital) reference encounter by its social and collaborative nature that involves group effort and open participation” (Shachaf, 2010, p. 67). A natural question follows as to how to achieve this goal, especially in light of barriers to collaboration which include geographical dispersion, electronic dependence, structural dynamism, and national diversity (Gibson and Gibbs, 2006).

Connaway, Dickey, and Radford (2011) and Pomerantz (2006) note that it is imperative for collaborative virtual reference services to provide desired answers, not just referrals or instruction. A possible venue for virtual reference services collaboration is social question and answer services which operate under a different model, allowing anyone (as opposed to credentialed librarians) to provide answers. social question and answer services have three components: a mechanism to submit natural language questions, a venue to submit answers, and a community built around this exchange (Shah, Oh, and Oh, 2009). Investigators have studied social question and answer services users’ information-seeking behaviour (Kim, Oh, and Oh, 2007), selection of resources (Harper, Raban, Rafaeli, and Konstan, 2008), social annotations (Gazan, 2008), motivations (Shah, Oh, and Oh, 2008), and comparisons to other question-answering services (Su, Pavlov, Chow, and Baker, 2007).

Theoretical Framework: communities of practice

This research employs communities of practice as a theoretical construct. Previously applied in library and information science (Davies, 2005), communities of practice was originally developed by Lave and Wenger (1991) and situates “groups of people who share a concern, a set of problems, or a passion about a topic, and who deepen their knowledge and expertise in this area by interacting on an ongoing basis” (Wenger, McDermott, and Snyder, 2002, p.4). Wenger (1998, 2004) describes three distinct dimensions of a communities of practice: 1) they are joint enterprises, meaning they are created and maintained by their members; 2) they feature mutual engagement, meaning all members come together to form a social entity; and 3) the members have a shared repertoire of resources and sensibilities that have been communally developed over time.

Communities of practice have learning as a focus and depend on interactions between members. Lave and Wenger (1991) also specified that communities of practice are voluntary, customizable, individual, and sometimes finite entities that encourage members to help one another solve problems and develop new approaches or tools for their field. This allows community members to safely display their areas of weakness, share their expertise (Faraj and Wasko, 2001; Ardichvili, Page and Wentling, 2002), learn together in the ‘public space’ of the community (McDermott, 1999, p. 34), and handle “problem solving, requests for information, discussion of developments, information seeking and coordination, planning, or negotiation of meaning” (Mills, 2011, p.349).

However, the literature argues that insufficient time is the most significant barrier for active participation in communities of practice (Correia, Paulos, and Mesquita, 2010). Other barriers to participation and sharing include “information hoarding” (Ardichvili et al., 2002, p. 12), low levels of collegiality (Smith, Barty, and Stacey, 2005), shifting group memberships (Gannon-Leary and Fontainha, 2007), lack of opportunity for building trust (Jarvenpaa and Leidner, 1999; Kirkup, 2002; Gibson and Manuel, 2003; Ellis, Oldridge, and Vasconcelos, 2004), and geographical gaps (Gibson and Manuel, 2003; Cramton, 2001). Additionally, the literature challenges Wenger’s theory by suggesting that it promotes heterogeneity in groups (Nincic, 2006; Roberts, 2006), which in turn promotes the ideas of dominant voices and reinforces habits and conditions of groups, which may be restrictive and not be conducive to growth.

Communities of practice is an appropriate theoretical approach to investigate collaboration in virtual reference services and social question and answer services. In accordance with Wenger (1998, 2004), virtual reference services librarians possess a shared domain of interest/engagement in participation in the interests of serving user information needs, operate within a community for sharing information, and have been educated in the tenants of a shared practice through library and information science degree programmes. Wenger also mentions shared “passion” as a feature of communities of practice, which librarians clearly exhibit as a profession (Radford, 2008). Additionally, previous studies embracing the communities of practice approach may explain benefits and constraints of collaboration observed in virtual reference services (i.e., physical distance/geographical gaps and limited time). A qualitative approach, described below, is adopted here to investigate how librarians interact to share information in virtual reference services since communities of practice is suitable for attempting to “recognize the multiple viewpoints characteristics of participation in a community of practice” (Davies 2005, p.106).

Research questions and method

Based on the literature review, the following research questions are addressed:

• RQ1 How can virtual reference services become more collaborative, within and between libraries, and tap more effectively into librarians’ subject expertise?
• RQ2 What can virtual reference services learn from social question and answer services to better serve users and attract potential users?
• RQ3 How can we design systems and services within and between virtual reference services and social question and answer services for better quality and sustainability?
• RQ4 In what ways can the communities of practice (Wenger, 1998, 2004) framework contribute to our understanding of collaboration barriers and opportunities in the virtual reference services environment?

In-depth phone interviews by the research team with 25 virtual reference services librarians investigated attitudes towards current practice and possibilities regarding enhanced collaboration and referrals both internal and external to library systems/consortia. Respondents were recruited through professional listservs, personal contacts, and from OCLC’s QuestionPoint (QP) librarian blog. Verbal responses to demographic questions and a series of closed and open-ended questions were entered into SurveyMonkey by the researchers and made anonymous. Questions addressed these topics: collaboration, referrals, comparison of virtual reference services to social question and answer services, and gathering of critical incidents (Flanagan, 1954). Responses underwent line-by-line qualitative analysis to elicit recurring themes and identify representative quotations.

Results and discussion

Demographics

Table 1, below shows participant demographics. Additionally, among the 25 respondents, virtual reference services were seen as slightly busier than face-to-face (FtF) services (10, 40% vs 8, 32%) and overall reference was reported as busy (11, 44%) or extremely busy (8, 32%). Overall reference volume was reported to be increasing by 10 participants (40%).

Qualitative results

Difficult virtual reference services questions within and outside of expertise

Respondents were asked to discuss specific difficult FtF or virtual reference services questions they remembered when the question fell within and outside of their subject expertise. When answering questions within their expertise, participants were also asked why they felt successful. When discussing a difficult question outside of their expertise, the librarians were asked what alternatives they felt they had and why they handled the question in a certain way. Three-quarters of the librarians defined success as filling the user’s information need, as surmised when he/she expressed satisfaction/thanks during the interaction, a “nice feedback comment at the end” (L32)[1], or through emoticons: “There were lots of happy faces, so the user seemed pleased” (L24). Participants felt that successful interactions provided an “opportunity to educate the patron” (L30) in their area of expertise, and some mentioned that they: “knew the subject area” (L24), or were familiar with e-resources. One quarter of participants mentioned referring the question to another librarian. For example, one said: “You’ve exhausted the resources that you are aware of and you don’t want to be a dead-end for the patron” (L23). A typical response when asked about a difficult reference questions within expertise area was:

“Someone requested specific census information and demographics for a specific country, and I knew where to go and the website for the government census. I felt like because I was familiar with the resource beforehand and show them pretty early on in the chat how to navigate and find what they were looking for. Because it took me less time to locate the information, I was able to spend more time in the teaching” (L3).

Difficulties for questions outside participants’ expertise usually involved a lack of content knowledge, or of an e-resource, such as not having a “great facility with…[a] specific school’s website interface” (L28). Legal and medical questions provided the most instances of knowledge outside participants’ expertise, such as: “She needed to know how to perform an artery tap…[so] no matter the searching, I was not going to be confident that my answer was going to be complete” (L15). When faced with a difficult question outside of their expertise, mo st participants decided to refer the user to a specific librarian, such as “a colleague who knows about gov [sic] docs” (L10), or to another library. Two librarians also referred users to other non-librarian experts, such as a student’s instructor. Other difficulties were a lack of lead time, usually because “the paper was due too soon for me to answer” (L32), or because of the user being unfamiliar with e-resources: “It’s kind of challenging when the students don’t even know how to use the database (L13).

Facilitators and barriers to collaboration

A majority of participants collaborated more than once a week with e-mail being the most common mode, followed by FtF, chat, and phone. FtF collaboration was seen as easiest in shared physical settings, as one participant commented:

“We have a combined public service desk that incorporates degreed librarians, grad students, and student assistants, so the very nature of the configuration of the public service desk is designed to promote collaboration. We are constantly passing questions all over that desk and working together to come up with the best and the speediest answer” (L1).

One third of the time librarians collaborated when they were unable to answer the question, but a slightly higher number collaborated because they wanted to give the user a more comprehensive answer. One respondent explained that he/she referred a census question because it was: “about public health and ethnic groups and on a very small level in New York City. It was the kind of question where I could give an okay answer and I knew this colleague could give a great answer” (L8).

Participants were also asked why they generally collaborated, and what facilitated or barred collaboration. In addition to lacking expertise to answer the question, librarians also chose to collaborate when the user needed specific local information “about a library policy” (L32) from their home institution, for instance, or when referring was deemed inconvenient. Two factors involved the user: if the user needed more information or wanted to devote more time.

Facilitators and barriers to collaboration existed on both individual and organizational levels. A perception of other librarians being willing to help was the most common facilitator. The second most frequent facilitator was having knowledge of someone to ask for help. Almost four times as many librarians reported that they believed other librarians in general would help them rather than knowing a particular librarian. This finding is a strong indicator that the communities of practice theoretical framework is a useful lens for viewing librarian virtual reference services practices.

One major barrier to collaboration was in not knowing who to contact or in not feeling that colleagues welcomed collaboration. Although some librarians did perceive that other librarians, either individually or working in a particular place, were unwilling to collaborate, this seemed more the exception than the rule. One librarian said: “There are librarians who are hostile in body language and sometimes verbally if it interferes with their other duties. They have made it very clear that I should not ask and so I do not” (L01).

Virtual reference services and social question and answer services compared

When asked to compare, participants saw virtual reference services as authoritative and objective, more synchronous, and receiving more complex questions. Social question and answer services were viewed as asynchronous, less authoritative, having simpler questions, and providing more opinionated answers. By viewing librarians who provide virtual reference services as a community of practice, it is easier to understand why many view social question and answer services as inferior to virtual reference services.

A major finding was that although the librarians viewed virtual reference services as different from and superior to social question and answer services , they expressed a willingness to consult non-librarian experts. Librarians most often said that they would feel comfortable collaborating with professors, possibly because of a larger proportion of academic library participants, or because being a professor implies subject expertise, although they are not “card-carrying” members of the librarian community of practice. One librarian stated: “[It] would be great to contact a history professor for a history question or social scientist for political information” (L16). None of the librarians mentioned doctors or lawyers, even though those were two areas where the participants mentioned lacking expertise. Most gave the impression that having confidence in expertise would involve: “some standards… that help [other experts] to provide that information in a way that librarians are used to providing that information” (L32).

When asked, participants identified types of questions appropriate to social question and answer services, including questions that were objective, such as ready reference, fact-based, yes/no questions, or those subjective in nature, based on experience or opinion. Objective questions, with easily verifiable answers, were viewed as appropriate for social question and answer sites because, as one librarian explained: “there is a danger in that it's not clear that it's someone's opinion and not based on a good source” (L32). Experience-based questions were suitable because participants felt it was clear that the user would understand that the answerer was giving an opinion. As one participant explained:

“I think if you're asking for recommendations or looking for suggestions, you know, ‘I'm going on a trip to Cape Cod, where can I have dinner? Where can I go whale watching?’ Those are more personal opinions than trying to come off as an expert. My husband loves Yahoo! Answers. He can stay up all night answering questions. I wonder if he's giving people the right information!” (L25)

This suggests that even if the librarian knows the social question and answer services answerer, that is not, in itself, enough to engender confidence that they will provide a quality answer. A few participants also suggested questions on certain topics, such as “everyday type of questions” (L29), popular culture “stuff like Star Wars” (L24), or medical questions were best for social question and answer services.

As can be seen in the rich data above, these initial findings confirmed that communities of practice is an appropriate theoretical framework to apply to the virtual reference services and social question and answer services comparison and the remaining interview data will be considered through this lens. The communities of practice approach presents a way forward to link virtual reference services to social question and answer services if a system can provide assurances of expert help and a mechanism for connecting questioners to appropriate and qualified answerers.

Conclusion

These results suggest that virtual reference services librarians constitute a community of practice (Wenger, 1998, 2004) in their approach to referrals and collaboration and can inform ways to enhance the sustainability of virtual reference services through enabling easier engagement with other librarians and credentialed subject experts. By summer, 2013, analysis will be concluded for: a) the remaining librarian interview data; b) interview data gathered from fifty virtual reference services and/or social question and answer services users that has elicited their perspective on these issues; and c) a series of three design sessions with industry/library experts to delineate specifications for a prototype system.

Content knowledge, shared professional standards, and technological familiarity are vital in addressing and/or referring difficult reference questions. When faced with a difficult question outside of their expertise, participants usually refer it to another librarian, although occasionally they refer to a different type of expert, if confident in qualifications. Participants refer questions to librarians because they feel that they are willing to collaborate and have shared professional ideals and expertise, as delineated in the communities of practice model. Collaboration is not seen as a last resort, but as a value-added service that librarians are happy to provide. FtF collaboration is most-often chosen, so the geographical gap and time pressures are seen as barriers, in keeping with the communities of practice framework. Knowing who to ask is another major facilitator to collaboration, while unwillingness to collaborate and not knowing who to contact are barriers. Although librarians viewed the answers in social question and answer services as less authoritative, less complex, and less objective, they were not against collaborating with experts, provided they had demonstrated professional expertise or extensive knowledge. This suggests that virtual reference services librarians could be willing to expand their communities of practice to embrace other experts, as long as their areas of expertise could be verified.

The promise of high-quality, accurate, and effective virtual reference services is considered to be endangered by difficult economic times that continue to negatively impact library funding and consortia. Cyber Synergy researchers hope to alleviate such difficulties, by exploring social question and answer services models, which provide crowd-sourced opportunities, and to provide effective ways to connect users’ subject-related questions to specialists. Findings may result in greater collaboration to optimize use of information service resources, harness the passion of librarians (and other experts) for sharing their talents and subject knowledge, and, ultimately, to enhance the sustainability of virtual reference services.

[1] When reporting these results, direct quotations will be indicated by an L(librarian), and the participant number in parentheses. So “L1” indicates a direct quotation from Librarian 1. Participant numbers range from 1-50, as assigned when consent forms were obtained, although only the results from the first 25 interviewed participants are included. So L31, for example, may have been the 25th librarian interviewed.