Transforming Libraries
Occasional Papers
Statistics and Measurements
Ordering Information
ARL Home

LineLineLineComplete ListLineLineOnline TextLineLineLineOnline SurveysLineLiaisonsLineParticipating

SPEC Kit 251
Electronic Reference Service
October 1999


Executive Summary

Research libraries are increasingly committed to using computer technology for everything from online catalogs to "friends of the library" retail stores. Reference librarians have found many ways to take advantage of technology in extending service beyond the walls of the library. One approach growing in popularity is reference service through electronic mail and web forms.

Library literature demonstrates the growing interest in exploring the benefits of and barriers to delivering reference service by electronic means. Areas of research and writing include the electronic reference interview, delivery of instructional tutorials via electronic methods, and analyses of individual libraries experiences with electronic reference service. Despite this interest, a comprehensive picture of the implementation of electronic reference services in ARL libraries has not been available.

The goals of this SPEC Kit were to report on the extent to which ARL libraries provide electronic reference services and offer a snapshot of the types of users reached, questions received, policies established, data-gathering techniques utilized, and innovations implemented.

For purposes of this survey, electronic reference was defined as reference service designed for remote users and identified by a specific link from a librarys website. Electronic reference queries that come directly to individuals or via other library web links were not included in the definition.

Extent of Electronic Reference Service
Seventy-eight of the 122 ARL member libraries responded to the electronic reference service survey (64% of the total membership). Seventy-five (96% of respondents) reported offering electronic reference, while three (4%) do not. For the 51 respondents who supplied data for the relevant survey question, an average of 177 reference questions are submitted via an electronic reference service each month. Those 51 respondents reported numbers ranging from 2.3 to 3,200 questions per month, with a median of 35. Closer examination of the data shows significant outliers to be the Library of Congress (2,500 questions per month) and the National Library of Medicine (3,200 questions per month). When the three national libraries are removed from the calculation, the average number of questions submitted monthly via electronic reference service is 67.

While still a recent phenomenon, the growth of electronic reference has been developing over a number of years. Very few of these programs are new: 22 (30%) of the programs are more than five years old, 37 (50%) fall into the three to five year range, nine (12%) are one to two years old, and six (8%) reported that they are less than one year old. Most libraries--54 or 75%--reported a gradual rate of increase in the number of questions since the service began. One (1%) reported a rapid increase, while 15 (21%) reported no increase, and two libraries (3%) reported a decrease in use of the service since its inception.

Most of the libraries--36 or 48--link their electronic reference service at the top level of their libraries' web pages. Twenty-six (35%) reported links on the second level, and 13 (17%) reported links on the third level or lower. Many libraries reported multiple locations for electronic reference links. Email is the most common method for submitting questions, even when users submit them through a web form.

Seventy-four of 75 respondents reported that potential users learn about the service from the librarys home page. Other tactics are employed to advertise the service as well: 61 of the respondents noted library instruction activities; 52 respondents noted that service desk staff refer users to the service; and 37 utilize posters, flyers, and other written promotional materials. Even such creative promotional techniques as inclusion on the answering machine for the library hours and TV screen ads in the university community center were mentioned.

Thirty-eight (51%) of the libraries reported that they will accept any question. However, many libraries qualified this by saying that the library's response might be to suggest search strategies and sources, recommend that the patron meet with a librarian, or suggest that the question is not appropriate for an email response. Respondents mentioned excluding questions with foul language, requests for medical and legal information for personal use, questions that can be answered with resources generally available in local libraries, and broadcast queries that can be handled better by local public libraries. Thirty-one respondents (41%) accept only basic factual questions, but not included in this class are such items as requests for lists of factual information (for example, long lists of addresses or bibliographical information) or shelf checks for availability of materials.

Most respondents reported that the reference unit manages the electronic reference service. Teams within public services departments or cross-unit teams were mentioned as part of alternative management strategies for the electronic reference service. In some branch libraries where a reference unit does not exist, the librarians who provide reference assistance manage the electronic reference service.

Responsibility for answering questions is clearly decentralized. Twenty-six of 75 respondents described their environment as one in which responsibility rotates among several people on an established basis. Twenty-five described a situation where questions are received by a single individual and distributed to others. Seventy-one (95%) of the respondents reported that librarians are answering the electronic reference questions; 41 (55%) reported that support staff are participating in answering such questions; six (8%) reported using graduate assistants; two (3%) reported using student assistants; and two (3%) reported using other staff expertise such as computer science specialists or public affairs information specialists. Many libraries assign responsibility for answering electronic reference questions to librarians at the reference desk or on telephone reference duty. Others set up a triage system where basic questions are answered by librarians on the reference desk and others are forwarded to subject specialists. Some services allow users to choose the branch library or department to receive their question, so the staffing arrangement may vary by location.

Response time is a minor issue for the respondents. Twenty-nine (39%) of the libraries stated that they do not specify a guaranteed response time. Twenty-one (28%) libraries specify a response time of one day, and 14 (19%) specify two days.

Data Gathering and Assessment
Despite the current interest in assessment as a management tool, little information about users is being collected by ARL libraries in the area of electronic reference services. Most respondents (75%) collect the number of electronic reference questions, and 69% are including those questions in their general reference statistics. However, only 29 libraries (39%) are collecting user data. The user information that is collected includes:

  • type of patron;
  • type of question;
  • time of day;
  • affiliation of patron;
  • response; and
  • number of questions rejected as being outside of scope or policy.

Sixty-five of the libraries (87%) reported that they had performed no user assessment of their electronic reference service. Of the 10 who report conducting user assessment surveys (13%), both informal and formal assessment techniques have been utilized. Strategies for assessment have included questions on the service as part of a general library user survey and follow-up at the end of individual transactions.

Forty-six of the respondents (62%) reported that they maintain an archive of questions and responses. Email systems (e.g., Eudora, Pine) are used most commonly to archive questions and responses. Web-based database management systems, the Internet Public Library's software, and paper were mentioned as alternative archiving techniques. Of those who maintain an archive, the vast majority (91%) report that no part of that archive is publicly accessible.

This survey shows that electronic reference service has become an established form of reference delivery within ARL libraries. It is yet another example of libraries' efforts to exploit technology and provide services to remote users. Despite expressed fears of being overwhelmed by those users, the workload seems to be manageable thus far, but local policies control who is eligible to receive reference service electronically and what type or level of questions are answered via the service. The age and number of electronic reference services indicate a significant amount of experience in developing, managing, and running a remote user service, and most ARL libraries see the service as part of the reference unit. Nonetheless, there is a great deal of variation in staffing and managing strategies.

Little electronic reference assessment seems to be occurring. Most libraries are counting questions and adding them to their reference statistics. This strategy is simple given the email technology that most ARL libraries are using to deliver the service and create archives. Technological solutions beyond email might encourage and facilitate additional data collection and assessment. One useful area for further study would be to examine the way that patrons are using this service; for example, to what extent do their questions elicit a "no answer response" versus the answer to the question. Online user surveys clearly need to be conducted as well, but few ARL libraries are experimenting with interactive technologies, such as chat rooms, to deliver this kind of service. Hopefully, this situation will change because interactive technologies offer expanded opportunities for reference interviews as well as new possibilities for assessment.

This SPEC Kit was prepared by Lori Goetsch, University of Maryland, with the assistance of Laura Sowers, GCI Information Services, and Cynthia Todd, University of Maryland, as part of the OLMS Collaborative Research and Writing Program.

Order Information
SPEC Kit 251, Electronic Reference Service
Lori Goetsch, Laura Sowers, and Cynthia Todd October 1999 ISSN 0160 3582 67 pp.
$40 ($25 ARL members)

  Back to top

SPEC Kit (ISSN 0160 3582) 1999 by the Association of Research Libraries. ARL grants blanket permission to reproduce this information for educational use as long as complete attribution is given. For commercial use, requests should be sent to the ARL Publications Department, Association of Research Libraries, 21 Dupont Circle, Washington, DC 20036. SPEC Kits are available by subscription and single issue.

The Association of Research Libraries
Maintained by: ARL Web Administrator  
Site Design Consultant: Chris Webster
Last Modified: October 5, 2001