Article
Interactions:
A Study of Office Reference Statistics
Naomi Lederer
Liberal Arts Librarian
Morgan Library
Colorado State University
Fort Collins, Colorado, United States
Email: naomi.lederer@colostate.edu
Louise Mort Feldmann
Business and Economics Librarian
Morgan Library
Colorado State University
Fort Collins, Colorado, United States
Email: louise.feldmann@colostate.edu
Received: 22 Nov. 2011 Accepted: 19 May 2012
2012 Lederer
and Feldmann. This is an Open Access article
distributed under the terms of the Creative Commons-Attribution-Noncommercial-Share Alike License 2.5 Canada (http://creativecommons.org/licenses/by-nc-sa/2.5/ca/), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is
properly attributed, not used for commercial purposes, and, if transformed, the
resulting work is redistributed under the same or similar license to this one.
Abstract
Objective – The purpose of this study was to
analyze the data from a reference statistics-gathering mechanism at Colorado
State University (CSU) Libraries. It aimed primarily to better understand
patron behaviours, particularly in an academic library with no reference desk.
Methods – The researchers examined data from
2007 to 2010 of College Liaison Librarians’ consultations with patrons. Data
were analyzed by various criteria, including patron type, contact method, and
time spent with the patron. The information was examined in the aggregate,
meaning all librarians combined, and then specifically from the Liberal Arts
and Business subject areas.
Results – The researchers found that the
number of librarian reference consultations is substantial. Referrals to
librarians from CSU’s Morgan Library’s one public service desk have declined
over time. The researchers also found that graduate students are the primary
patrons and email is the preferred contact method overall.
Conclusion – The researchers found that
interactions with patrons in librarians’ offices – either in person or
virtually – remain substantial even without a traditional reference desk. The
data suggest that librarians’ efforts at marketing
themselves to departments, colleges, and patrons have been successful. This
study will be of value to reference, subject specialist, and public service
librarians, and library administrators as they consider ways to quantify their
work, not only for administrative purposes, but in order to follow trends and
provide services and staffing accordingly.
Introduction
Reference
services have traditionally been measured in some way in order to collect
evidence, most commonly by a simple tick mark to indicate a transaction. In
late 2006, Colorado State University (CSU) Libraries moved from a traditional
reference desk model to a referral system. Staff and students working at a
library information desk started to refer patrons to librarians for in-depth
assistance, and the librarians wanted to collect data about their in-office reference
consultations in order to capture information about this new service.
CSU
is a land-grant institution located in Fort Collins, Colorado, United States,
with an FTE of approximately 25,000 students. The Libraries consist of a main
library, Morgan Library, and a Veterinary Teaching branch. The CSU Libraries
College Liaison Librarians unit consists of 10 librarians and 2 staff members.
Since 2007, these librarians have used a reference database developed in-house
to record office research consultations. This database provides a place to
input various data and to generate reports for librarians, the College Liaison
unit, and the Libraries administration. Administrators can use the database to
see specific liaison workloads and which subjects have the most inquiries, and
can then use this information for rebalancing of assignments (e.g., subjects
reconfigured or other responsibilities reassigned to compensate for a heavier
load) and justification of budgets for additional librarians and other relevant
resources.
At
CSU, College Liaison Librarians do not staff a public service desk, but provide
reference assistance in their offices via drop-in and appointments.
Additionally, some librarians offer reference services in departments or
colleges for two to four hours each week. CSU Libraries has a help desk at
which staff and students may refer in-depth questions to librarians. The
researchers were curious about how CSU university library patrons are seeking
information. Claims that reference statistics are declining may refer only to
data from the traditional reference desk. Are patrons still seeking librarians
for assistance? Are trends at a national level, such as a decline in reference
desk statistics, occurring locally? The data from the office statistics database
provided an opportunity to identify patterns and to explore how patrons are
seeking reference services, and in 2011 the database statistics were analyzed
to answer these questions. The subject areas of the questions were also of
interest because they might reflect success in outreach or areas that might be
candidates for additional promotion of services. In this study, the researchers
identified overall trends and looked specifically at the subject areas of
Liberal Arts and Business.
Literature
Review
The
broad topic of library statistics often encompasses collection holdings,
staffing, and circulation data. In line with the focus of this article, only
literature relating to library reference statistics was examined. Only one
article was found that discusses the collection of reference statistics
resulting from transactions originating from multiple sources (reference desk,
email, phone, instant messaging, etc.); the majority of articles focus on
public service desk statistics, and those which have relevant ideas are
discussed below. Few articles consider how statistics are gathered, but rather
focus on the results of the statistics gathering. Furthermore, no close
analyses of any particular librarians’ office interactions were found.
Novotny
(2002) shows how some libraries collect reference statistics on paper,
including example sheets with categories that in some cases are used away from
services desks. Examples include separate telephone and email reference
question sheets, weekly summaries, and a question sheet with options for
multiple types of contact with the patron available for each question. The
summary of reference statistics covers public desks, not office numbers. Measures for Electronic Resources
(E-Metrics) (2002) discusses digitally based reference (and other)
transactions. Possible statistics are provided for networked and electronic
services and resources, but the emphasis is on electronic resources, not on the
work that librarians might be doing somewhere other than at a reference or public
service desk. Electronic reference is just one aspect of the paper, and in any
case it has changed substantially since 2002. In providing guidelines for
gathering digital reference statistics, McClure, Lankes,
Gross, and Choltco-Delvin (2002) point out that
“libraries have seriously underrepresented their services in terms of use of
digital services being provided . . . by not counting and assessing these uses
and users. As more users rely on digital library
services—including digital library services—this undercount will continue to
increase” (p. 8). The measures in these guidelines focus on digital
reference, rather than on any kind of off-desk assistance. Nevertheless, this
type of statistical gathering could be a useful starting point for a library developing
a statistics-gathering database. The majority of articles on digital reference
services do not focus on the methods patrons use to contact librarians
directly. Instead they mention digital reference in passing or provide a
careful analysis of where and how the services are available at specific
locations (Lederer, 2001; Pomerantz,
Nicholson, Belanger, & Lankes, 2004; White,
2001), or focus on nonaffiliated users of the service
(Kibbee, 2006). Articles on digital reference outside
of North America and collaborative reference efforts are not closely enough
related to the current topic to be included.
Some
researchers have classified types of questions asked at reference or email
reference services. Henry and Neville (2008) discuss how the Katz classification,
as detailed in Introduction to Reference
Work, and the Warner classification, as detailed in his article “A New
Classification for Reference Statistics,” measured experiences at a small
academic library. Henry and Neville include references to Association of
Research Libraries (ARL) statistics and comparisons of newer types of access
such as chat, email, and instant message services used in public as well as
academic libraries. Meserve, Belanger, Bowlby, and Rosenblum (2009)
applied the Warner classification at their institution, evaluating it
favourably and using it to support their tiered reference arrangement. Greiner
(2009) responds to the article by Meserve et al. by
questioning some of the conclusions and notes the decline in questions overall,
attributing it partly to incorrectly interpreted questions, but also citing
relationship building by librarians as a necessary component of good reference
service. Meserve (2009) replies with overall agreement, while emphasizing that at
his library the paraprofessionals are well versed in their role, and reference
services were in decline before paraprofessionals were put on a service desk.
The
evaluation of reference service is a frequent topic in the literature. Logan
(2009) provides a good overview in which he starts from the beginnings of
reference services in 19th-century America, and points out that although the
tools have certainly changed, the functions of reference have not. Reference
was not often discussed in publications until the 1970s, with an emphasis on
assessment and evaluation of reference services in the 1990s, and more recently
on “‘learning outcomes’ and ‘information literacy’” (p. 230). Logan recommends
the “establish[ment of]
flexible criteria for good service,” which include components related to
“behavioral characteristics . . . basic knowledge of resources and collections,
subject knowledge, and reference skills” (p. 231). Welch (2007) describes and
discusses the National Information Standards Organization’s (NISO) Z39.7-2004 standard,
and highlights the importance of counting email, web page, and other reference
transactions. Library services have progressed beyond traditional desk
transactions, and a method for tracking all reference transactions is
necessary. Library administrators need to be convinced of the relevance and
importance of these new methods for providing research services. As Welch
(2007) writes, “including electronic reference transactions and visits to
reference-generated web pages in statistical reports are ways to demonstrate .
. . our continuing usefulness to our patrons” (p. 103).
The
amount of effort expended for different types of questions is explored by Gerlich and Berard (2007, 2010).
They outline 6 levels of effort and provide charts of questions by type for the
2003-2004 academic year (2007); and further broaden
the collection of data to 15 libraries in 2010. One of the main points is that
collection of statistics only from a traditional reference desk does not
capture all reference transactions that are taking place – many transactions
are via email and other methods. Gerlich and Berard (2010) argue that “reference transactions are on the
decline as documented by librarians and their institutions, yet reference
activities taking place beyond traditional service desks are on the rise” (p.
116). Data collection techniques all too frequently do not take these
additional assistance points into account, and “counting traffic numbers at the
traditional reference desk is no longer sufficient as a measurement that
reflects the effort, skill, and knowledge associated with this work” (p. 117).
Gerlich and Berard discuss expended effort and difficulty in their
larger Reference Effort Assessment Data (READ) experiment (2007). Murgai (2006) describes one library’s sampling of number
and types of questions, which both notes some disadvantages of sampling, but
also shows that the results of the sampling were within acceptable ranges of
accuracy (though reference questions beyond the desk are not included). Another
statistic that is more difficult to collect is the often multiple and varied
types of resources used by a librarian to answer a single question (Tenopir, 1998). Thomsett-Scott
and Reese (2006) examine whether there is a relationship between changes in
library technology and reference desk statistics. They note the changes in the
number of questions when CD-ROMs and Web-based resources were first introduced,
and report that while reference statistics may be declining, the types of
questions are “more intricate” or “complex” (p. 148):
A review of
the literature suggests that reference questions are taking longer to answer
and are more extensive, yet the actual number of questions is declining. Reference
managers may need to reconsider how reference services are measured. Statistics
may be lower due to issues with the traditional recording method of “one
patron, one tick” (p. 149).
In
other words, in the past a patron might come to the desk asking about books on
a topic, then return to ask about articles, then return to ask for help with
citations (three transactions). In the electronic world, this one patron is
likely to be helped in a single transaction. Additionally, the authors point
out that “traditional statistical recording systems also may not include
reference questions answered beyond the reference desk” (p. 162). They also
examined gate counts for 1997-2004, circulation counts for 1998-2004, and
various reference counts and types from 1989-2004 from their own library. Not
all types of statistics were gathered for all years, as email and chat started
only in 1998. They conclude that “statistics should include online reference
methods and possibly web page statistics as the proliferation of library-based
web pages may . . . be answering many of the questions that face-to-face
reference services answered in the past” (p. 163).
Some
articles describe in-house databases created to collect reference statistics.
Aguilar, Keating, and Swanback (2010) describe the
thinking behind their library’s in-house database as a “need to discover new
ways to gauge the needs of our patrons and employ concrete data to make
decisions” (p. 290). Statistics are gathered at multiple service points –
mostly reference desks, but also offices and remote locations as well – and
used to justify collections purchases and increased staffing of their “Ask a
Librarian” service during a specific time of day. Data from ARL and other
reports are easily gathered from Aguilar, Keating, and Swanback’s
database. A second in-house database is described by Feldmann
(2009), which was created to capture the number of reference questions that
were successfully referred from the new information desk after the reference
desk was disbanded. The database evolved to be a useful tool for gathering
information on librarians’ office transactions. The author cites articles that
discuss referral services and various staffing models for tiered services.
Smith (2006) describes a Web-based system for collecting statistics and
discusses various reasons people have collected reference statistics, as well
as the problems associated with collecting them, such as apathy and the wide
variation in the parts and types of questions. The author describes how the database
was developed and the types of information it collects, including screen shots
and HTML coding of and for the database. The references and further reading are
substantial. Todorinova, Huse,
Lewis, and Torrence (2011) describe one university
library’s choice of a commercial product, Desk Tracker, after using a system of
clickers that did not record the time of transactions. The data collected
included type of patron, form of the transaction (in person, email, or phone),
and type of question, and it was used to assign appropriate staffing levels and
to inform collection development decisions. Some output weaknesses were found
in the software, but the data have been proposed as potentially useful for
decision-making and improving services and operations.
Although
some of the literature examines reference statistics closely, it is in specific
contexts such as health or medical libraries or GIS systems (e.g., Parrish,
2006), and has a more focused audience and set of questions. The literature
still lacks a close examination of reference transactions away from the
reference desk. This study looks closely at not only how the questions were
asked, but how long it took to answer them, their subject areas (broadly and
more specific, depending upon the topic), the status of the questioner, and
whether or not the question was referred from someone else (e.g., via a service
desk in the majority of cases).
Methods
Data
were gathered from an office statistics database, which is a recording
mechanism used to capture CSU Librarians’ reference transactions, both
in-office and during office hours in a department or college on campus (see
Figure 1). The database was developed in late 2006, when a CSU Libraries
Business Librarian, a staff member, and a member of the library’s technical
services department created it using PHP scripting language and MySQL. It was
originally conceived as a method to track referrals from the newly implemented
information desk (Feldmann, 2009). Starting in 2007,
librarians no longer staffed a reference desk or any other public services
desk, and staff and students working at the information desk (the now sole
public service desk) would refer in-depth questions to librarians. The database
initially provided a method for capturing the number of referrals received by
librarians from the information desk, in addition to providing a place to
record reference transactions. Since librarians at this time placed a renewed
emphasis on departmental outreach, it was also thought that the database would
capture the impact of marketing their research consultation services to
faculty. Over the years, librarians have changed or modified input fields to
reflect needs and improve the database. Reports are easily generated in
Microsoft Excel spreadsheets. The input form contains both required and
voluntary fields. Information collected in the required fields include name,
contact type (email, drop-in, phone, appointment, office hours, or other), help
desk referral, time spent, number of patrons assisted, and status of patron.
Voluntary information includes discipline area, course information, and
comments.
Figure
1
CSU
Libraries office statistics database entry form
For
this study, the researchers extracted numbers from the database in the
aggregate (total from all librarians) for the years 2007-2010. Additionally,
data from the Liberal Arts and Business Librarians were extracted as samples to
examine subject-specific data. The Business Librarian provides assistance for
six departments: Accounting, Finance, Marketing, Management, Computer
Information Systems, and Economics. The primary Liberal Arts Librarian covers
seven departments: English, History, Art, Communication Studies, Journalism and
Technical Communication, Ethnic Studies, and Design and Merchandising, which is
part of the College of Applied Human Sciences. The database allows data to be
pulled directly by various fields, date range, and by librarian. The
researchers extracted data by contact type, number of patrons helped, time
spent, patron status, and whether or not the question was a referral from the
help desk for the years 2007-2010. This information was then examined to
determine trends.
Results and
Discussion
Aggregate
Information
Table
1 shows that both the number of consultations and the
numbers of librarians reporting have decreased between 2007 and 2010. While the
total number of office consultations decreased by year, a corresponding drop in
the number of librarians reporting also occurred, so that the mean (average per
librarian) increased from 127 in 2007 to 154 in 2010. Fewer librarians were
employed in 2010 than in previous years due to attrition.
Table 2 shows that email was by far
the most popular way that patrons received assistance, accounting for 50%
(3,141 questions) of all transactions.
Table
1
2007-2010
Office Consultations by Year
|
Year |
No. |
Librarians Reporting |
|
2007 |
1,517 |
12 |
|
2008 |
1,856 |
12 |
|
2009 |
1,515 |
11 |
|
2010 |
1,395 |
9 |
Table
2
2007-2010
Office Contact Type
|
Contact
Type |
No. |
Percent |
|
Email |
3,141 |
50% |
|
Drop-In |
1,214 |
19% |
|
Phone |
748 |
12% |
|
Appointment |
714 |
11% |
|
Other |
424 |
7% |
|
Office Hours |
40 |
1% |
|
Empty |
2 |
0% |
|
Total |
6,283 |
|
The
contact type of “Office Hours” refers to librarians providing dedicated office
hours to answer questions from drop-in patrons, similar to traditional office
hours that faculty provide. They were recorded only in January and February of
2007 as they were a short-term arrangement where librarians were assigned to be
backups for the then information desk. Referrals from the information desk were
so rare that the concept was abandoned after a short run. “Empty” indicates
that no information was entered. “Other” could mean helping someone in the
library while en route to a meeting or returning to one’s office, service
provided at a non-library location (for instance, the Business Librarian’s
“Librarian to Go” reference in the College of Business), instant messaging
(IM), and so on.
The
status of patrons who directly contacted librarians (Table 3) shows that
graduate students and undergraduate students are the heaviest users with
faculty members in a solid third place.
Table
3
2007-2010
Office Patron Status
|
Patron
Status |
No. |
Percent |
|
Graduate |
2,030 |
32% |
|
Undergraduate |
1,969 |
31% |
|
Faculty |
1,156 |
18% |
|
Community |
557 |
9% |
|
Staff |
348 |
6% |
|
Elsewhere |
80 |
1% |
|
Government |
54 |
0.9% |
|
Empty |
51 |
0.8% |
|
Visiting Faculty |
24 |
0.4% |
|
Administrator |
14 |
0.2% |