Coding RUSA Behaviors
RUSA behaviors refer to the behaviors prescribed in the revised guidelines. These behaviors were coded through the analysis of 422 chat transcripts that show all patron-staff interactions as well as search activities, including co-browsing. Initially, librarians’ behaviors were coded against all forty-nine items in the revised guidelines, which include all three modes of transactions (general, in-person, and remote). This decision was made because certain items listed under the in-person category also were applicable to the chat reference setting (for example, “Accompanies the patrons in the search”). Later, the initial forty-nine items were collapsed into ten types of behaviors after merging similar items together and removing items that were either unobservable or irrelevant for chat reference. For example, some items describe thought processes rather than actual observable behaviors (for example, “recognizes when to refer patrons to a more appropriate guide, database, library, librarian, or other resource [4.8]“). Some behaviors, such as providing information sources, were observed across multiple items (For example, “Offers pointers, detailed search paths [including complete URLs], and names of resources used to find the answer [4.9]” and “Uses appropriate technology [such as co-browsing, scanning, faxing, etc.] to help guide patrons through library resources, when possible [4.11]“). Finally, some items are applicable for physical setting transactions only (for example, “Be mobile [1.7.1]“). Figure 1 displays the final coding scheme that lists ten RUSA behavioral types along with corresponding item numbers in the revised guidelines.
As shown in figure 1, specific items in the five RUSA behavioral areas were coded into ten behavioral types in the present study. Approachability was observed by two behavioral types: welcoming and the use of patrons’ name. First, welcoming was coded to be present either when an initial word contact was made (such as “Hi, this is the reference librarian”) or when a general welcoming atmosphere was observed from the librarians’ written communications. Other than these examples, the revised guidelines do not include many clear behavioral indicators of approachability. Thus, the second behavioral type, the use of patrons’ name, was adopted from the Guidelines for Chat Sessions within “IFLA Digital Reference Guidelines”;the assumption here is that the behavior could lower the patron’s emotional barriers and thus enhance the librarian’s approachability.9 Currently, there is no consensus about the effectiveness of using a patron’s name during the chat session. Some think it is effective because it could make the reference interview more personable and approachable. Others think that it may intrude upon personal privacy. By observing its use, we will be able to determine whether the use of patrons’ names affects user satisfaction.
Interest was observed by using two items in the revised guidelines: focusing attention on the patron (item 2.2 in the revised guidelines) and maintaining word contact (2.6). The remaining items were excluded from coding because they were mostly applicable to the physical reference setting (2.1, 2.3, 2.4, and 2.5), e-mail (2.7), or pre-interview stage (2.8).
Listening/inquiring was observed by two behavioral types: receptive and cordial listening (3.1 and 3.2); and inquiring by rephrasing, clarifying, or asking questions (3.5-3.8). The rest of the items were excluded from coding because they were not easily observable (3.3, 3.4, and 3.9) or overlapped with other items (3.9).
Searching was observed by two behavioral types: searching for or with patrons (4.3, 4.6, 4.10, and 4.11), and offering pointers or information sources (4.5, 4.9, and 4.11). Some behaviors, such as co-browsing, were coded as both when they involved searching and offering information sources. Four items were excluded from coding because they either pertain to listening/inquiring (4.1 and 4.2–verifying words) or follow-up (4.7), or indicate a cognitive process rather than an observable behavior (4.2 and 4.8).
Finally, follow-up was observed by three behavioral types: proper closing (4.7, 5.1, and 5.9), offering alternatives or making referral (5.4-5.8 and 5.9), and asking to come back for further assistance (5.2).
RUSA behaviors were coded by two independent coders for intercoder reliability in order to ensure the consistency. First, the primary researcher coded the entire 422 transactions. Then, the second coder, a reference librarian who received training for coding the RUSA behaviors, coded every fifth transaction (n = 84). This sample for intercoder reliability comprises 20 percent of the total transactions, which is the recommended percentage for social science research.10 Finally, the percentage agreement between the two coders was calculated for each behavior, which informs the level of agreement between two coders.11 The agreements were Welcoming (91.7 percent), Use of patrons’ name (96.4 percent), Interest (78.5 percent), Listening (85.6 percent), Inquiring (78.6 percent), Searching with or for patrons (76.2 percent), Offering pointers and information sources (78.6 percent), Asking if answered completely (91.7 percent), Offering referrals (85.0 percent), and Asking to come back for further assistance (90.5 percent). All these intercoder percentage agreements were either above or close to the generally acceptable threshold value of 80 percent.12
It should be noted that the use of the RUSA guidelines in the current study is differentiated from the approaches of the previous studies.13 While the earlier studies did not observe librarian behaviors with respect to actual items in the guidelines or measured with a few items only, the present study used the entirety of the items in the revised guidelines as the coding scheme to analyze librarians’ chat reference behaviors.
[...] Behavioral Performance on User Satisfaction in Chat Reference Services (pdf) (source: RUSQ via Stephen’s [...]