116
Testing the Viability of the READ
Scale (Reference Effort Assessment
Data)©: Qualitative Statistics for
Academic Reference Services
Bella Karr Gerlich and G. Lynn Berard
Bella Karr Gerlich is University Librarian in the Rebecca Crown Library at Dominican University; e-mail:
bkarrgerlich@dom.edu. G. Lynn Berard is Principal Librarian, Engineering and Science, and Fellow of the
Special Libraries Association in the University Libraries at Carnegie Mellon University; e-mail: lberard@
andrew.cmu.edu. This paper was presented at the RUSA RSS Research & Statistics Committee’s 14th
Annual New Reference Research Forum at the 2008 ALA Annual Conference in Anaheim, California, on
Sunday, June 29, 2008. Select portions (such as the READ Scale©, concept, and figures) of this paper have
appeared in other publications in shorter, focused, introductory articles. This paper comprises the complete
2007 national study results. The READ Scale (Reference Effort Assessment Data) ©Bella Karr Gerlich
The READ Scale (Reference Effort Assessment Data) is a six-point
scale tool for recording qualitative statistics by placing an emphasis on
recording effort, knowledge, skills, and teaching used by staff during a
reference transaction. Institutional research grants enabled the authors
to conduct a national study of the READ Scale at 14 diverse academic
libraries in spring of 2007 and test its viability as a tool for recording refer-
ence statistics. The study data were collected from 170 individuals and
24 service points with over 22,000 transactions analyzed. There was a
52 percent return rate of an online survey of participants, with more than
80 percent of respondents indicating they would recommend or adopt the
Scale for recording reference transactions. The authors suggest that the
READ Scale has the potential to transform how reference statistics are
gathered, interpreted, and valued. This paper presents the findings of a
nationwide study testing the Scale in spring 2007 and suggests practical
approaches for using READ Scale data.
eference transactions are on
the decline, as documented
by librarians and their institu-
tions, yet reference activities
taking place beyond traditional service
desks are on the rise. Librarians are
reporting that they are as busy as they
have ever been. According to an Asso-
ciation of Research Libraries (ARL) 2002
study conducted to reveal best practices
in reference work, the findings exposed
a general lack of confidence in current
data collection techniques as “failing to
capture and accurately reflect reference
activities overall.”1
What factors account for this change
in reference work? Technology has
transformed our ability as information
providers to serve our user communi-
ties, structure our facilities, and conduct
Testing the Viability of the READ Scale 117
our work. The introduction of online
information resources has heightened the
need for instruction in the classroom, as
well as instruction via e-mail, over chat
services, and at point of use. Reference
librarians are being sought out for their
knowledge management expertise and
subject specialization at the reference desk
as well as increasingly in their offices and
hallways. 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.
Figure 1
reAD Scale—reference effort Assessment Data Scale©
1: Answers that require the least amount of effort and no specialized knowledge skills
or expertise. Typically, answers can be given with no consultation of resources. Length of
time needed to answer these questions would be less than 5 minutes. Examples: direc-
tional inquiries, library or service hours, service point locations, rudimentary machine
assistance (locating or using copiers, how to print a document or supplying paper).
2: Answers given that require more effort than the first category but require only
minimal specific knowledge skills or expertise. Answers may need nominal resource con-
sultation. Examples: call number inquiries, item location, minor machine and computer
equipment assistance, general library or policy information (how to save to a disk or
e-mail records, launching programs or rebooting).
3: Answers in this category require some effort and time. Consultation of ready reference
resource materials is needed; minimal instruction of the user may be required. Reference
knowledge and skills come into play. Examples: answers that require specific reference
resources (encyclopedias or databases); basic instruction on searching the online catalog;
direction to relevant subject databases; introduction to Web searching for a certain item; how
to scan and save images; more complex technical problems (assistance with remote use).
4: In this category, answers or research requests require the consultation of multiple
resources. Subject specialists may need to be consulted and more thorough instruction and
assistance occurs. Reference knowledge and skills needed. Efforts can be more supportive
in nature for the user or, if searching for a finite answer, difficult to find. Exchanges can be
more instruction based as staffs teach users more in-depth research skills. Examples: in-
structing users how to use complex search techniques for the online catalog, databases, and
the Web; how to cross-reference resources and track-related supporting materials; services
outside of reference become utilized (ILL, Tech services, etc.), collegial consultation; as-
sisting users in focusing or broadening searches (helping to redefine or clarify a topic).
5: More substantial effort and time spent assisting with research and finding informa-
tion. On the high end of the scale, subject specialists need to be consulted. Consultation
appointments with individuals might be scheduled. Efforts are cooperative in nature,
between the user and librarian and/or working with colleagues. Multiple resources used.
Research, reference knowledge and skills needed. Dialogue between the user and librarian
may take on a “back and forth question” dimension. Examples: false leads, interdisciplin-
ary consultations/research; question evolution; expanding searches/resources beyond
those locally available; graduate research; difficult outreach problems (access issues that
need to be investigated).
6: The most effort and time expended. Inquiries or requests for information can’t be
answered on the spot. At this level, staff may be providing in-depth research and services
for specific needs of the clients. This category covers “special library” type research
services. Primary (original documents) and secondary resource materials may be used.
Examples: creating bibliographies and bibliographic education; in-depth faculty and Ph.D.
student research; relaying specific answers and supplying supporting materials for publi-
cation, exhibits etc; working with outside vendors; collaboration and ongoing research.
118 College & Research Libraries March 2010
Gerlich developed the READ Scale at
Carnegie Mellon University as a proposed
quantitative measurement method de-
signed to capture all occurrences of refer-
ence activity.2 The READ Scale (Reference
Effort Assessment Data) is a six-point
scale used for recording vital supplemen-
tal qualitative statistics gathered when
reference librarians assist users with their
inquiries or research-related activities by
placing an emphasis on recording the
skills, knowledge, techniques, and tools
used by the librarian during a reference
transaction (figure 1).
Institutional grants received in 2006
enabled the authors to expand the study
beyond one institution to fifteen academic
libraries in the spring of 2007 with the
goal of testing the viability of the READ
Scale as an adaptable tool for gathering
qualitative statistical reference data on a
national level.
Study Objective
Our objective was to test the viability of
the READ Scale as an additional tool for
gathering reference statistics. The READ
Scale was launched at Carnegie Mellon
University as a trial in the spring of 2003,
followed by an academic year study in
2003–2004. The READ Scale emphasizes
effort, skills used by staff at the time the
reference transaction occurs. This method
is especially appealing in a profession
where the current industry standard for
recording statistical data is a hash mark
that records and recognizes quantity as
opposed to quality.
Literature Review
A review of literature and studies on ref-
erence librarians, reference services, and
reference statistics was used to inform and
support the design of the READ Scale, as
well as the contribution of qualitative study
to librarianship. There are two distinct areas
of study in reference assessment that di-
rectly influence our work: the measurement
and evaluation of reference service and the
means of recording reference transactions
(both traditional and automated practices).
Measurement and Evaluation of
Reference Service
Beyond efficacy; the exemplar librarian as a
new approach to reference evaluation by Quinn
(1994)3 takes an interesting approach as
it suggests using qualitative methods of
evaluating reference librarians by first
asking ”what makes a reference librarian
great?” Quinn asserts his study implies that
good reference behavior is learned and that
cultural preparation is a must. The study
also found that not one single factor made a
librarian great: it is a combination of skills.
Quinn’s article focuses on behavioral as-
pects of reference librarianship. This study
will add to those findings by determining
if participants’ using the READ Scale find
that ranking and recording their efforts
results in positive feelings as their effort,
skills, and knowledge are being recognized
during the reference transaction.
Quality Reference Service: A Preliminary
Case Study, Stalker and Murfin (1996)4 stud-
ied the results of the WOREP (Wisconsin-
Ohio Reference Evaluation Program)
survey at Brandeis University, which
demonstrated the highest level score to
date of a general reference department
using the WOREP, to determine to what
extent the high quality of professional
service was demonstrated, due to use of
the WOREP model. This article found that
allowing for sufficient time for the consult-
ing role of reference librarians led to the
high success rate when using the WOREP
Model at Brandeis; other factors included
contents and configuration of the reference
area, and strong support for services by
administration. The READ Scale likewise
acknowledges the interactive nature of the
reference transaction, the time element and
records the service component.
Perspectives on Quality of Reference Ser-
vice in an Academic Library: A Qualitative
Study was a study done by Mendelsohn
(1997)5 to explore the concept of quality
as it applies to reference service. Four par-
ticipants in humanities and social sciences
areas were interviewed and perceptions
of quality discussed. This paper supports
earlier works that emphasize willingness
Testing the Viability of the READ Scale 119
to help, knowledge and skills, morale and
time as vital components in the quality of
the reference transaction from the librar-
ian point of view.
Work in Motion / Assessment at Rest:
An Attitudinal Study of Academic Reference
Librarians; A Case Study at Mid-Size Uni-
versity MSU A, written by Gerlich (2006)6
is a study that focuses solely on reference
librarians and their attitudes about their
work: what they value, how they perceive
themselves, how they perceive others view
them. This study supports the notion that
reference, or the transaction interaction,
is the primary function of the reference
librarians’ position and the highest valued
task by both the reference librarians and
administrators. The study also reveals a
lack of assessment or reward for this work
outside of the anecdotal, with librarians
and administrators in agreement that cur-
rent statistical data gathered for reference
work is not adequate for recording effort,
knowledge, and skill.
Testing Classification Systems for Refer-
ence Questions, Henry and Neville (2008)7
follows the University of South Florida,
St. Petersburg study using Warner’s clas-
sification system at the Nelson Poynter
Memorial Library in comparison to Katz’s
traditional reference categories described
in Introduction to Reference Work (direc-
tional, ready reference, specific search
questions and research). The results of this
study support the idea that the adoption
of new measures for reference statistics
seems warranted to be more exacting, rel-
evant, and reflective of reference services.
The conclusions also reached similar find-
ings of the READ Scale that by recording
actual effort means reexamining staffing
of the reference desk as a service point.
The Recording of Reference
Transactions
Usage-Based Staffing of the Reference Desk:
A Statistical Approach, Dennison (1999)8
discussed the importance of staffing
decisions for reference desks, and how
measuring usage of service can inform
those decisions. At Winona State Univer-
sity Library (WSU), Dennison reports on
using direct measurement applications
to reference statistics. WSU employed
categories to record reference statistics
and determine peak times for staffing
the reference desk based on the category
assigned to each transaction.
A New Classification for Reference Statistics
by Warner (2001)9 describes a test of an al-
ternative reference data-gathering model.
The impetus for creating the classification
model in Warner’s case was borne out of
need for training and triage at a new single
point-of-service desk at Eastern Carolina
University. Warner’s study changed from
a daily collection of data for the first three
months to being randomly selected once a
month. Warner’s research and subsequent
implementation of a classification system
in this case lays a foundation for this study
by introducing alternative methods for
gathering statistics.
SPEC Kit 268, Reference Service Statistics
& Assessment, Novotny (2002)10 paints a
picture of changing reference services
and stagnant assessment measures of the
same in research libraries by surveying
and documenting how ARL libraries were
collecting and using reference service
transactions data. This survey described
in its executive summary the confusion
and angst surrounding modern refer-
ence work as libraries scramble to collect
data. There is no mention of improving
reference quality, developing employees,
or recognition of work effort—the study
did not distinguish between a successful
or unsatisfactory transaction. While it
recognizes the use of electronic tools to
gather data, there is a failure to recognize
the librarian’s use of electronic tools to dis-
tribute information in any sense outside
the narrow confines of the “transaction”
definition. This study was most useful for
this work in that it painted a picture that
the system of reference assessment in use
by ARL libraries appears to be in flux.
Reference Use Statistics: Statistical Sam-
pling Method Works (University if Tennessee
at Chattanooga) by Murgai (2006)11 supports
one of the findings of the Novotny study
120 College & Research Libraries March 2010
that librarians felt busier than ever helping
patrons, despite a decline in the number
of patrons served. Murgai suggested that
most reference librarians would like refer-
ence statistics to reflect all aspects of refer-
ence but would also like statistic recording
to be simple, while acknowledging that
reference service is anything but simple.
The University of Tennessee at Chatta-
nooga (UTC) reviewed other academic
libraries’ sampling methodologies and
employed sampling for a year to compare
to daily data gathering. The results of the
statistical analysis showed that the num-
bers gathered for a set period of time are
very close to data gathered over a longer
period of time, supporting the results of
the 3-week period of data capture selected
for the READ Scale study. The limits of
the UTC study also support the need for
a tool like the READ Scale, noting that
the classifications for reference statistics
used in the UTC study did not capture
the types of questions, resources, off-desk
questions—measures that are used in the
READ Scale—were needed to get a com-
plete picture of reference services.
Methodology
Timeline
The preparation of this study occurred
in the summer and fall of 2006, with
participation commitments in place by
late November 2006. The Institutional Re-
view Board (IRB) approval and pre-study
exercises took place between December 1
and February 4.
Libraries were given the option of con-
ducting the study for the duration of their
spring semester, and or for the predeter-
TABLe 1
reAD Scale Participating institutions
enrollment Less than
5,000–5 institutions
enrollment greater than
5,000–4 institutions
enrollment greater than
10,000–5 institutions
Clarke College
Clarke College Library
Dubuque, IA
Carnegie Mellon University
(1 Institution, 6 Service
Points) Pittsburgh, PA
Georgia Institute of
Technology
Georgia Tech Library
Atlanta, GA
Eastern Virginia Medical
School
Edward E. Brickell Medical
Sciences Library
Norfolk, VA
Georgia College & State
University
Library & Instructional
Technology Center
(1 Institution, 2 Service
Points) Milledgeville, GA
New York University
Business & Documents
Center – Bobst Library
New York, NY
Lawrence University
Seeley G. Mudd Library
Appleton, WI
Robert Morris University
(1 Institution, 2 Service
Points)
Moon Township, PA
West Virginia University
(1 Institution, 3 Libraries)
Morgantown, WV
Lewis & Clark College
Aubrey R. Watzek Library
Portland, OR
Washburn University
Mabee Library
Topeka, KS
University of California,
San Diego
Science & Engineering
Library
La Jolla, CA
Our Lady of the Lake
University San Antonio
(OLLUSA)
Sueltenfuss Library
San Antonio, TX
University of Nebraska
Love Library
(Chat Service only)
Lincoln, NE
Testing the Viability of the READ Scale 121
mined three-week duration: February 4–
February 24, 2007. These two options were
selected to accommodate those institutions
that normally only sample reference statis-
tics as well as those that collect data daily
for an entire semester. All institutions had
to commit to the February data collection
period. These three weeks were selected to
limit the chance for spring breaks to occur
within the study time frame.
Study Participants
The research team decided on the follow-
ing parameters for seeking participants in
the study. The universities must:
• Number between 9 and 15 academic
libraries
• Be diverse geographically
• Contain diverse enrollment figures,
grouped as follows: ≤5,000, >5,000 and
≤10,000, and ≥10,000
• Include both public and private
institutions
The number range 9–15 was deter-
mined with a minimum acceptance rate
of 9 participating, with at least three for
each enrollment figure represented. One
institution that initially agreed to partici-
pate had to withdraw for reasons unique
to that university, leaving the number of
participants at 14 institutions, with 170
individual participants total. See table 1
for participating libraries. Each institution
was asked to identify an onsite coordina-
tor at each location who would commit to
disseminating information and managing
the activities, timelines and follow-up as-
sociated with conducting the study.
Pre-Study Calibration of Sample Scale
Questions
To familiarize and prepare participating
librarians with the READ Scale and its
proper use, a list of pre-study test ques-
tions was developed and sent to onsite
coordinators. Each site received the same
set of questions; however, the coordinators
were instructed to select some questions
from the list but were given the flexibility
to substitute others for those localized to
the institution. The addition of a sample
question(s) that occurs frequently at the
home institution reference desk provided
a common ground for a discussion of
how to apply the scale when rating the
effort level of the transaction. The least
number of questions distributed was six,
and we asked that a range of project effort
be represented (1–6 levels on the READ
Scale) to acquaint participants with the full
range of scale levels. All participants were
asked to answer and rank their effort for
each of the sample questions. It was agreed
that onsite coordinators would evaluate
responses and respond to participants’
questions regarding all aspects of apply-
ing the scale. Participants were also asked
to record time during this exercise so that
the researchers could average the length of
time per transaction, per scale rank overall.
Table 2 represents those questions from the
researchers’ test list along with the average
time it took to complete the transaction.
Across the board, the pre-study rat-
ing effort for transactions at the 1, 2, or
6 level were typically unanimous, while
the 3, 4, and 5 ratings revealed some dif-
ferences between individuals’ perceived
rankings. Differing of individual rankings
for the same type of reference transaction
was thought to be due in part to subject
specialization and how individuals tend
to “grade” (hard or easy). Coordinators
met with their participants and summed
up how the transactions were resolved,
the recommended rating to assign, the
time it took to answer the question, and
the reason for the rating. This enabled
individuals to adjust their personal grad-
ing habits for traditional inquiries. It
was important to recognize that where
subject specialization is the norm, effort
associated with customer service should
be recognized. This is why the number of
elements (the definition for each number
on the scale) and time associated with the
scale rankings are important to note. Staff
helping someone out of their area of ex-
pertise should feel comfortable assigning a
higher scale point than the librarian with a
specialization in the subject area. As noted
later in this paper, the criteria of time and
122 College & Research Libraries March 2010
how it is applied using the READ Scale is
an area considered for further research.
Additionally, reference librarians were
asked to conduct the study in their offices
during “off-desk” times. The term “off-desk”
is used to note reference transactions
handled by a reference professional that
occurs away from an established, regularly
TABLe 2
Common Test Questions
Common Question, Academic Libraries Most Common ranks
& Average Time
I need a translation for an Italian aria. 3, 7 min
Where is the bathroom? 1, 1 min
I have a laptop—where I print out library records? 2, 1 min
I am researching postwar suburban housing development in the
(City) region—can you show me what you have in that relates to
this topic, or where I should look?
5–6, 90 min
I need the issue number for this citation: Le Goff, Jacques Or-
dres mendiants et urbanisation dans la France Annales: *cono-
mies, soci*t*s, civilisations, vol. 25, (1970)
4, 15 min
I am trying to place a hold on a book in process by using the
online catalog request form. Kept receiving error message re-
questing item info—please help!
2, 5 min
I'm trying to find out about the philosophy of St. Benedict. Do
you have any suggestions on which of his books or writing I can
download?
4, 15 min
I need to find some contemporary criticisms for the play Fences
by August Wilson—both the writing of the play and a produc-
tion.
4, 12 min
I am looking for some help getting started on a research project
—gender roles and the selection of college majors in the south—
where do I start? How do I conduct a study?
5, 23 min
Do you have a book with pictures of kitchen utensils used in
colonial times?
4, 28 min
PsycInfo says that we have this journal, but it isn’t in the library
—please help!
2–3, 5 min
Common Questions, Academic Medical Libraries
Curriculum models for teaching medical students about medical
ethics: (1) What should be the learning objectives; 2). What the
curriculum content should entail.
5–6, 90 min
Need recent (up to 10 years) clinically relevant articles on the
patient care of thrombolytic therapy and antiplatelet therapy and
anticoagulation in the treatment of peripheral vascular disease
3–4, 15 min
I need a list of drugs that affect lymph flow or lymph vessel
contractions.
3–4, 15 min
I’m looking for medical licensure lookup, medical school etc and
if there are any malpractice proceedings against Dr. _____—can
you help?
3, 10 min
Testing the Viability of the READ Scale 123
scheduled reference desk. Anecdotal evi-
dence suggests that this is where the major-
ity of higher-level scale effort in assisting
patrons is being conducted, especially for
those clients served by a liaison librarian
with subject-specific responsibilities. These
data were gathered and compiled to help
determine at which service point users
sought assistance; it was theorized that
transactions at the 4, 5, and 6 levels would
be recorded by individuals while working
from their offices rather than at a tradi-
tional service point. The recording of “off-
desk” statistics is a nontraditional activity
and one not often employed by reference
librarians or reported institutionally. It is
the case then that this valuable effort has
not been seriously studied or credited to
the work effort of reference professionals.
The READ Scale data recording meth-
od is such that it allows institutions to
use their local paper or online form that
captures day, hour, and approach type
for both directional and reference ques-
tions, on and off desk. Participants in the
study were asked specifically to use their
existing forms to test the adaptability or
translation of the READ Scale in using a
number from the scale in place of a hash
mark when recording a reference transac-
tion. On the researchers’ end, there was
little difficulty in recording data onto the
statistics spreadsheets, and the benefit for
participating institutions was the ease of
adoption of the scale into existing local
recording instruments.
Data Collection
As all of the institutions had different
methodologies in place for recording sta-
tistics, researchers developed a common
table to compile data by Scale number and
approach type (table 3).
Some institutions had numerous cat-
egories that identified inquiry types, such
as “equipment” or “database search.”
These were placed into the “Walk-Up
Reference” category for the study. READ
Scale definitions do not distinguish the
kind of question, but they reflect the effort
expanded, knowledge required, and even
the teachable moment that occurs during
the transaction.
The time of day that the transaction
occurred was not reported cumulatively
by the researchers, as reference desk
hours and personnel schedules varied by
institution and could not be normalized.
They were recorded for each individual
institution as reported and made available
to the respective organization so that as-
sessments could be made locally.
The approach type for transactions was
recorded to establish frequencies for how
transactions occurred. As suggested by the
ARL study, some academic institutions are
experiencing a decline in reference trans-
actions. Recording approach frequency
here would help determine the most
popular method for seeking reference
help and where that transaction occurred.
At the conclusion of the three-week
data collection period, an online question-
TABLe 3
Sample Form with Categories and Approach Type
Date: Walk-
Up Dir
Walk-
Up Ref
Phone
Dir
Phone
Ref
E-
mail
SCALE 1 2 3 4 5 6
Walk-Up Directional
Walk-Up Reference
Phone Directional
Phone Reference
Chat
E-mail
124 College & Research Libraries March 2010
naire was sent to all study participants.
The survey was designed to assess the
participants’ experience when applying
the scale, to gain their feedback on the
value of the scale in demonstrating effort
when recording reference transactions via
this method, and to inquire how the scale
might be changed to improve the data col-
lection instrument. While the researchers’
individual institutional experiences with
the scale were very positive, one desired
outcome for conducting a national study
was to determine the viability of the 1–6
point Scale.
Results
Three-Week Study
Fourteen institutions participated in the
READ Scale Study during the spring
semester of 2007. There were a total of
24 service points and 170 individual
participants. All institutions submitted
statistics using the READ Scale for the
same three-week time period, February
4–February 24, 2007. Seven institutions
elected to continue using the Scale for the
duration of their respective semester after
the initial study period. Table 4 illustrates
the cumulative number of transactions,
READ Scale category assignment, ques-
tion and approach type for all service
points, and institutions for a total of 8,439
transactions during the three-week study
period. All institutions were encouraged
to use the READ Scale for recording
off-desk statistics as well, if appropriate.
Seventeen out of a possible 170 individu-
als reported off-desk statistics for a total
of 1,531 off-desk transactions recorded in
the three-week period (table 5). Combined
transactions for service points and off-
desk totaled 9,970.
The study illustrated that the majority
of inquiries continue to be by physical
approach (figure 2). “Off-desk” the per-
centage of e-mail is considerably higher
(figure 3) and almost equal in percentage
to in-person interactions.
Comparisons between service points
illustrate that the highest majority of
TABLe 5
Cumulative Data, Off-Desk, All institutions, 2/4–2/24/07
READ SCALE 1 2 3 4 5 6
Walk-Up Directional 23 4 0 3 1 0
Walk-Up Reference 196 197 157 74 41 18
Phone Directional 44 6 2 0 0 0
Phone Reference 85 109 41 20 5 2
E-mail 193 142 93 49 21 5
Totals 541 458 293 146 68 25 Total 1,531
TABLe 4
Cumulative Data, All Service Points, All institutions, 2/4–2/24/07
READ SCALE 1 2 3 4 5 6
Walk-Up Directional 2,260 337 23 2 0 0
Walk-Up Reference 1,693 1,750 1,067 397 89 34
Phone Directional 148 38 4 5 0 0
Phone Reference 111 113 85 17 7 5
E-mail 47 44 44 19 0 2
Chat 13 19 44 22 0 0
Totals 4,272 2,301 1,267 462 96 41 Total 8,439
Testing the Viability of the READ Scale 125
transactions that occur at the reference
desks are in the READ Scale number-one
category (figure 4), followed by number
category two revealing that most inquiries
at the public service point require the least
amount of effort, knowledge, and skills of
library personnel.
Off-desk comparisons show a different
but consistent pattern (figure 5): that the
percentage of questions answered off-
desk for most of the institutions require
a much higher level of effort, knowledge,
and skills from reference personnel than at
the public service point. Only three of the
seventeen off-desk comparators in figure
5 have more level READ Scale categories
representing the bulk of their transactions
off-desk, with only two of those recorded
in the higher than 40
percent range, overall.
The majority of the
off-desk ratings for
the remaining group
were at category two,
three, and four respec-
tively, suggesting that
users actively seek out
the expertise of par-
ticular reference staff.
These data further
support the research-
ers’ theory that most
of the higher-level
effort, knowledge,
and skill required of
reference personnel
will take place away
from the public ser-
vice point. The need
to increase efforts to
record off-desk ref-
erence statistics was
also expressed by
the many of the re-
spondents in the ARL
Study.12
Semester-Long Study
Seven of the institu-
tions elected to con-
tinue to use the READ
Scale for the duration of their respective
semesters. Figures that follow represent
fourteen service points and ninety-four
individual participants. There were a to-
tal of 15,194 transactions recorded (table
6). Data was collected through May 11
and includes three-week study figures
reported previously. Approach type for
this group was also recorded (figure 6).
All institutions were encouraged
to continue to use the READ Scale for
recording off-desk statistics as well, if
appropriate. Seven institutions, eight ser-
vice points, and a possible 66 individuals
reported off-desk statistics, for a total of
1,156 transactions recorded for the dura-
tion of their respective semesters (table 7).
Data includes three-week study figures
Figure 2
Approach Type, All Service Points,
Three-Week Study Period
Figure 3
Approach Type, Off-Desk, Three-Week Study Period
126 College & Research Libraries March 2010
Figure 4
Comparative illustration of the Percentage of each reAD Scale Category,
per Service Point
Figure 5
Comparative illustration of the Percentage of each reAD Scale Category,
Off-Desk
Testing the Viability of the READ Scale 127
reported previously. Approach type for
transactions that occurred off-desk was
also recorded (figure 7).
As with the three-week study period,
the semester-long group’s data compila-
tion showed that the preferred approach
type was “in person” overall. However,
when separated out, the use of e-mail as an
approach type came very close compara-
tively to that of in-person approach type
when a transaction took place off-desk.
Comparative illustrations coincide
with the three-week dataset; the major-
ity of the transactions that occur are at
category 1 of the READ Scale at service
points for all institutions (figure 8).
Off-desk comparisons again show a
different but consistent pattern (figure
9). The percentage of questions an-
swered off-desk for the semester-long
group participants required a much
higher level of effort, knowledge, and
skills from reference
personnel than at the
public service point.
Unlike the three-week
s t u d y, h o we ve r, n o
scale category level 1
exceeded the 40 per-
cent mark, and two
of the off-desk insti-
tutions recorded no
level 1 transactions at
all. The semester-long
off-desk group also
TABLe 6
Cumulative Data, All Service Points, Semester-Long Participants
reAD SCALe 1 2 3 4 5 6
Walk-Up Directional 3,787 899 56 6 0 0
Walk-Up Reference 2,203 2,606 1,784 501 153 29
Phone Directional 377 148 10 4 1 0
Phone Reference 375 358 231 40 22 3
E-mail 465 423 238 85 69 4
Chat 19 76 150 66 6 0
Totals 7,226 4,510 2,469 702 251 36 Total 15,194
Figure 6
Approach Types, Semester-Long Participants
TABLe 7
Cumulative Off-Desk Data, Semester-Long Participants
reAD SCALe 1 2 3 4 5 6
Walk-Up Directional 2 1 1 1 0 0
Walk-Up Reference 89 134 153 87 51 28
Phone Directional 0 0 0 0 0 0
Phone Reference 30 53 48 22 3 2
E-mail 43 95 181 105 24 3
Totals 164 283 383 215 78 33 Total 1,156
128 College & Research Libraries March 2010
reported a higher percentage of level 3
category transactions than 2 with level
4, 5, and 6 following.
As stated earlier, these data sup-
port the researchers’ theory that most
higher-level effort, knowledge, and
skill required of reference personnel
take place away from the public service
point. Furthermore, increases in the
percentages of READ Scale categories
3 and 4 off-desk data
s u g g e s t t h a t a s t h e
semester continues,
there is a likelihood
that the opportunity
for off-desk transac-
tions increases and
the need for a level of
expertise, knowledge,
and skill likewise in-
creases. This coincides
with curriculum ex-
pectations that typi-
cally have fewer dif-
ficult assignments at
the onset of a semester but demonstrate
an increase in complicated assignments
and prolonged research projects as the
term progresses.
Online Survey Results
An anonymous survey was constructed
to solicit feedback on the READ Scale
that included its ease of use, partici-
pant difficulty distinguishing between
Figure 7
Approach Types, Off-Desk,
Semester Long Participants
Figure 8
Comparative illustration of the Percentage of each reAD Scale Category,
per Service Point, Semester-Long Participants
Testing the Viability of the READ Scale 129
Figure 9
Comparative illustration of the Percentage of each reAD Scale Category,
Off-Desk, Semester-Long Participants
TABLe 8
Degree of Difficulty
Question: Please Rank Your Degree Of Difficulty Using The Read Scale
Responses Not
Difficult
Somewhat
Difficult
Moderately
Difficult
Difficult Very
Difficult
Skipped
Question
Number
Responded
52
(51.0%)
38
(37.3%)
10
(9.8%)
2
(2.0%)
0 0 102
TABLe 9
Application ease / Question: Was the reAD Scale easy to Apply?
Responses Very
Easy to
Apply
Easy to
Apply
Moderately
Easy
Somewhat
Easy
Not
Easy
Skipped
Question
Number
Responded
16
(15.7%)
39
(38.2%)
38
(37.3%)
8
(7.8%)
1
(1.00%)
0 102
TABLe 10
Scale Adds Value to Statistic gathering
Question: Please rank the level of perceived “added value” the reAD Scale
placed on statistics gathering for reference transactions.
Responses Extreme
Value
Added
High
Value
Added
Moderate
Value Added
Minimal
Value
Added
No Value
Added
Skipped
Question
Number
Responded
7
(6.9%)
46
(45.5%)
35
(34.7%)
9
(8.9%)
4
(4.0%)
1
(.99%)
101
130 College & Research Libraries March 2010
categories, and participants’ perception
regarding added value to reference work.
All participants (170) were sent an online
survey to complete. The response rate for
the survey was high, with 102 (60%) total
respondents. The questions and their re-
sponses are detailed in this paper.
The majority of participants had no
difficulty using the READ Scale (table
8) and found the READ Scale easy or
moderately easy to apply (table 9). When
asked to rank perceptions of added value
to statistical data gathering, the major-
ity of responses fell in the “high-value
added” category (table 10). The favor-
able response rate, with the majority of
respondents in agreement that the READ
Scale’s added value to reference statistics
is “high” (45%) or “moderate” (35%) ac-
counts for a total of 80 percent of the study
group’s opinions.
Participants were asked about difficul-
ties they may have experienced in decid-
ing between categories. Most implied
difficulty deciding between ranks 3 and 4;
participants were also asked how they felt
about evaluating their own efforts (table
11) with the majority responding that they
were comfortable with the process.
Asked if they would recommend the
READ Scale to another reference librar-
ian, 67 percent of the study participants
answered in the affirmative as is, with an-
other 20 percent who would recommend it
with modifications, bringing the favorable
response rate to more than 80 percent.
A follow-up question inquired if the
study group would likely be in favor of
having the Scale adopted in their library
as is, or with modifications. A total of 50
percent responded affirmatively to “as
is,” with another 30 percent who would
adopt with modifications, bringing the
favorable response rate to 80 percent.
The survey group was also given an
opportunity later in the survey to sug-
gest modifications, of which 24 deposited
comments and two optional questions
asked for specifics about what the study
group liked and disliked about the READ
Scale.
TA
B
L
e
1
1
D
if
fic
ul
ty
B
et
w
ee
n
R
an
ki
ng
s
Q
ue
st
io
n:
D
id
y
ou
h
av
e
di
ffi
cu
lt
y
in
d
ec
id
in
g
be
tw
ee
n
ra
ti
ng
s?
I
f s
o
ch
ec
k
al
l t
ha
t a
pp
ly
.
R
E
A
D
S
ca
le
1-
2
2-
3
3-
4
4-
5
5-
6
N
o
D
if
fic
ul
ty
R
es
po
ns
e
C
ou
nt
To
ta
l
R
es
po
ns
es
Sk
ip
pe
d
Q
ue
st
io
n
R
es
po
ns
es
12
(7
.6
%
)
32
(2
0.
4%
)
46
(2
9.
3%
)
31
(1
9.
7%
)
15
(9
.6
%
)
21
(1
3.
4%
)
15
7
99
3
Se
lf
-e
va
lu
at
io
n
Q
ue
st
io
n:
H
ow
d
id
y
ou
fe
el
a
bo
ut
e
va
lu
at
in
g
yo
ur
o
w
n
ef
fo
rt
s?
R
es
po
ns
es
E
xt
re
m
el
y
C
om
fo
rt
ab
le
V
er
y
C
om
fo
rt
ab
le
M
od
er
at
el
y
C
om
fo
rt
ab
le
M
in
im
al
ly
C
om
fo
rt
ab
le
N
ot
C
om
fo
rt
ab
le
Sk
ip
pe
d
Q
ue
st
io
n
N
um
be
r
R
es
po
nd
ed
12
(1
1.
9%
)
50
(4
9.
5%
)
35
(3
4.
7%
)
4
(4
.0
%
)
0
(0
%
)
1
(.9
9%
)
10
1
Testing the Viability of the READ Scale 131
The likes listed by the participants
where coded into the six most common
reoccurrences: Effort/Value; Approach to
Evaluation; Types/Levels; Time; Staffing
Levels; and Reporting to Administration:
Sample comment, Effort/Value (17 oc-
currences noted):
It gave me a quick visible check of
my recent efforts. This made my
deskwork more rewarding, since I
sometimes feel like I do so many
1s and 2s—but I could see that I
was actually doing a higher level of
reference than I realized. It added
value to the statistics—literally
Sample Comment, Approach to Evalu-
ation (13 occurrences noted):
It qualifies what we were only
quantifying and therefore is a more
realistic indicator of what we do at
the desk.
Sample Comment, Types/Levels (9 oc-
currences noted):
I like that it makes a qualitative
distinction between types of refer-
ence interactions; it gives credit to
more challenging transactions. The
differences between the kinds of
interactions are flattened in a typi-
cal “hash mark” approach to noting
reference interactions.
Sample Comment, Time (5 occurrences
noted):
I thought that it was a good way to
see how the time was being spent on
the question. It gives a better picture
of what you are doing instead of just
a tally mark for each question.
Sample Comment, Staffing Levels (6
occurrences noted):
Using the scale made me think
about the types of questions we
were receiving via the various
formats and how we might need
to change staffing patterns to better
serve our users.
Sample Comment, Reporting to Admin-
istration (5 occurrences noted):
It will give a better contour to sta-
tistics as read by administrators and
funders, and help to make better
staffing decisions.
Dislikes were coded into the following
categories: Difficult to Apply/Subjectivity;
Types/Levels; Approach to Evaluating;
Knowledge of the Staff; and Effort/Value.
Sample Comment, Apply/Subjectivity
(19 occurrences noted):
My assessments were somewhat
subjective. I’d like to have some ses-
sions to compare notes with peers
on how to apply the scale to practice
questions to get some common un-
derstanding of how to use the scale.
Sample Comment, Types/Levels (16
occurrences noted):
The criterion for each level should
have had more concrete benchmarks.
Sample Comment, Approach to Evalu-
ating (9 occurrences noted):
It assumes that a question has an
inherent difficulty factor. There is
no taking into account the experi-
ence or inexperience of the librarian.
Sample Comment, Knowledge of the
Staff (6 occurrences noted):
Being uncertain about how effective
my rating was when dealing with
questions far outside the realm of
my normal subject areas—patent
questions, etc. would be more com-
plex for me but a piece of cake for
our patents librarian. I wasn’t sure
how to “figure in” that factor.
Sample Comment, Effort/Value (4 oc-
currences noted):
I also didn’t feel like it was clear how
to assign a number on the scale when
more time than expertise was in-
volved with a reference interaction.
The comments around difficulty be-
132 College & Research Libraries March 2010
tween determining Scale levels reflects
the outcome of an earlier question, which
asked participants to indicate which, if
any, categories they had trouble deciding
between.
A follow-up question encouraged
the participants to suggest alterations to
the Scale for future modification. These
modifications were put into the follow-
ing categories: Delivery Method/READ
Scale Appearance; Time Element; Skill
Level Element; Clarity of Categories;
Discussion Component; and Comments/
Observations:
Delivery method/READ Scale Appear-
ance (9 occurrences):
Automate it! It would be great to
have on the computer.
Time Element (5 occurrences):
Additionally, the numbers in the
scale (1–6) may have more meaning
and value if time were a factor—
I’ve had “3” interactions that can
last anywhere from 5 minutes to
20 minutes, but they are all simply
marked “3.”
Clarity of Categories/more descriptive/
fewer categories (4 occurrences):
The degrees of gradation of reference
questions were important, but not
very clear. I wish they had been more
concrete… like a checklist for each
category or more defined descrip-
tions for each category. A revision
will reduce the variable/error margin
between scoring librarians.… The
criterion for each level should have
had more concrete benchmarks.
Skill Level Element/experience of refer-
ence staff (4 occurrences):
Though it added some context to
reference statistics, it could stand
a little more context. What may
be a 3 or 4 level for someone with
little or no experience (a graduate
student assistant, for example) may
be a 2 or 3 for someone with a great
deal more experience. The scale
may have more use if each person
at the desk kept [his or her] own
statistics, so that experience could
be factored in.
Discussion Component/requirement (2
occurrences):
Reference staff should talk openly
and often about how to apply ques-
tion scale levels to make sure we are
all on the same page. The descrip-
tions are helpful—but everyone
reads things differently. There are
gray areas. There are things we
all do differently—so I think open
discussion would be helpful.
General Comments/Observations (2
occurrences):
The simple nature of the READ scale
works to do two contrary things:
point out the variability of the work
that we do, while showing how
limited we are in tracing the ways
in which we make knowledge avail-
able to each and every patron on an
individual level. Statistics, by nature,
are too broad and contain not quite
enough depth at the same time.
Finally, the study group was asked if
their approach to reference changed in
any noticeable way during the period they
applied the READ Scale to measure their
reference work.
The number of the overall participant
study group that changed their approach
to reference was low, only 10 out of a
total 98 responses, but these responses
are worth including here, as it provides
a snapshot of the online survey partici-
pants’ range of experience. A small per-
centage of the participant group indicated
difficulty with incorporating the READ
Scale into existing reference procedures,
while a high level of study respondents
experienced more satisfaction, increased
awareness, and an appreciation of the
effort, knowledge, and skills involved
with reference work by applying the
Testing the Viability of the READ Scale 133
scale to aid in measuring their reference
work effort.
I experienced an increased aware-
ness of differing levels of reference
work.
Frankly, it complicates the process.
Trying to delineate between a 1 or a
2, a 3 or a 4, etc., is tedious.
I was more likely to think about the
level of service being provided.
I gave more [conscious] thought to
the processes or steps involved in
order to rate each interaction.
I was more aware of the level of
effort that could be applied to ques-
tions vs. what I actually did.
It made me keep statistics regularly.
M o r e a wa r e o f t i m e s p e n t o n
transaction(s).
I think I worked a bit harder to make
sure that I recorded everything.
I had to think about the level of
effort.
I was more self-conscious of the
level of help I was providing, with
the net result that interactions
improved. My level of empathy
and understanding (dare I say
“patience”) improved along with it.
Using READ Scale Statistics: Practical
Approaches
The READ Scale was developed as a tool
for capturing vital supplemental quali-
tative statistics when reference librar-
ians assist users with their inquiries or
research-related activities by placing an
emphasis on recording the skills, knowl-
edge, techniques, and tools used by the
librarian during a reference transaction.
The researchers propose that there are
a number of practical approaches to us-
ing the statistical data derived from the
READ Scale for both strategic planning
and the assessment of reference services.
Individual institutions can use READ
Scale statistics for staffing; training and
continuing education; renewed personal
and professional interest; outreach; and
reports to administration.
Staffing
Comments from the study:
We’ve always known empirically
that a large percentage of our refer-
ence transactions were quick and
easy. This study provided concrete
evidence of this, with possible staff-
ing implications.
It shows a much clearer picture of
what we are actually doing with ref-
erence. It is possible to see where the
true “busiest times” are in the day.
By using the READ Scale, it is possible
for libraries to alter staffing patterns to
best serve the users and librarians. One
institution involved in the study decided
to “let go” of requiring full-time profes-
sional librarians to staff their reference
desk in the mornings and on Saturdays
after viewing the number of level 1 and 2
questions they received at those days and
times. This empowers student workers
and part-time staff, who took over some
of the duties, and frees the professional
librarians to concentrate on liaison and
collection development duties. Another li-
brary in the study is using the data to pro-
pose reducing faculty librarian scheduled
hours in the evening by ending them at 9
pm instead of 11 p.m., having noted that,
after 9 p.m., transactions not only become
infrequent but are rarely ranked above
category 2 on the READ Scale. Prior to
using the Scale, the evidence for changing
schedules could only be described as anec-
dotal. By the same token, the opposite can
be noted—high traffic times or notations
134 College & Research Libraries March 2010
of higher categories of the READ Scale can
be used to supplement and strengthen the
value of reference desk staffing.
Training/Continuing Education
Comments from the study:
I felt it was very useful because it
challenged me to come up higher in
those areas where I need improve-
ment in certain concentrations like
history, which is not my specialty. I
need to learn so much more.
Not directly related to the READ
Scale itself, but based on the com-
pilation of answers for the sample
questions, we realized that not all
our librarians were approaching
questions in the same way. The rat-
ings could vary from 2 to 6 for the
same question. Based on that, we
have decided to bolster our staff
development and training program
and improve our mentoring of new
librarians.
The READ Scale can be used as a
training tool for librarians at all levels.
The second observation above is a great
example of how using the READ Scale
can assist in the training and mentoring
of reference staff. Another service point
also reported the same experience—they
will also increase training. The research-
ers suggest that this training can be done
throughout the semester or year using
the READ Scale. If, at the beginning of
the training period, scale effort levels re-
corded and the answers provided are not
in line with each other, a training regimen
with outcomes can be developed, and a
similar series of questions can be tested
at a later date to ensure that the staff is
developing the necessary reference skills
and knowledge.
As another study participant observed,
using the READ Scale encourages con-
tinuous learning. The researchers suggest
that reference staff could make the most
of this opportunity by writing down any
questions that elicit an assignment of
a category of 4 or higher on the READ
Scale at the reference desk and then
sharing these questions and how they
were answered with their colleagues,
providing the opportunity to discuss
strategies for assisting users, and learn-
ing from colleagues who have in-depth
subject knowledge in that particular area.
This could also be a great way of recon-
necting with others, for the love of the
job. Gerlich’s case study reveals that the
number-one reason reference librarians
chose their profession was to help people
with research; the second reason was the
aspect of “the detective work.”13
Renewed Personal and Professional Interest
In Gerlich’s case study, reference staff
and administrators acknowledged the
primary function of their profession as
that of providing reference service;14
likewise, they recognized that current
data-gathering methodologies were not
sufficient in recording the importance
of this work or effort.15 The READ Scale
provides a way of revealing and counting
important supplemental data that have
been hidden in the customary tick marks
used to record reference statistics.
Comments from the study:
Using the READ Scale added to my
sense of accomplishment!
The thought required to rank ques-
tions according to the READ scale
made me think a little at the comple-
tion of the reference interaction—
and thus to become more self-aware.
It gives ME a tangible scale on which
to rate my efforts, ultimately spur-
ring me to strive for better service.
By using the READ Scale, reference
staff can rate their effort and receive
acknowledgement for their effort, knowl-
edge, and skills as appropriate. The level
of skill is especially important to note
in a situation where subject or liaison
Testing the Viability of the READ Scale 135
practices are the norm and librarians are
sought out for their expertise and con-
sultation services. In-depth specialized
transactions often happen away from
the traditional service desk, and credit
for expertise is often not recorded or
acknowledged.
Outreach
Using the READ Scale can help develop
outreach activities for librarians. In an in-
stance where a liaison program is strong,
but there is little visible research or library
activity and low or no in-office consulta-
tions, this may be a sign that outreach
efforts should be increased. This would
be especially pertinent in an environment
with research-intensive programs, where
reference staff could expect to assist facul-
ty or upper-class students who would be
expected to have intensive assignments,
to conduct research or need primary re-
search materials. An active campaign or
meeting with the department could elicit
an increase in the types of interactions
that would be assigned level 4 or higher
on the READ Scale.
The same can be said for reference desk
statistics in general. If libraries are only
experiencing inquiries that require efforts
at the 1 or 2 READ Scale categories, then
how are students and faculty getting their
information? Do they know what services
and resources you have? Are there new
ways to market services, facilities, or re-
search assistance? Are there times of the
year when higher READ Scale categories
are showing up in the statistics, and, if
so, can those patterns be predicted and
assignments be noted to facilitate new
research guides, make connections to
teaching faculty, or influence new designs
or products?
Reporting/Statistics
The READ Scale is intended to record
supplemental statistics alongside the
traditional quantitative data gathered
that could be used by administrators to
report the knowledge and skills used in
reference services.
Comments from the study:
I liked that it attempts to record the
intensity of the reference transac-
tion. In my view that was a sorely
missing piece of information when
recording in the traditional fashion.
The READ Scale is an assessment
tool that does a better job of reflect-
ing how reference librarians spend
their time. It gives more value than
tick marks on a page. It’s a tool we
can use with administrators to show
what we really do.
Just as READ Scale statistics can help
determine staffing strategies, the quali-
tative nature of the instrument can help
with the creation of narrative text more
descriptive in nature when developing
reports to stakeholders, especially where
an administrator needs to explain roles
or job functions. This could be particu-
larly meaningful in cases when off-desk
statistics are recorded and reference li-
brarians track communications, research
assistance, and appointments with their
constituents via e-mail. More time and
effort are required for those activities but
are rarely recorded.
The READ Scale could also be useful
in estimating average time spent help-
ing patrons. In the testing phase of the
study, participants were asked to record
the amount of time it took to complete a
transaction. These data enabled the re-
searchers to make rough estimations on
the average length of time per transaction
for each scale category.
Table 12 illustrates the total number
of transactions per category and the esti-
mated number of hours or days needed
to complete a transaction, based on the
pre-study question calibration data where
participants recorded time expended to
answer the test questions. These figures
can only be used as an illustration of
what keeping track of time can be used
for, as these data were gathered from the
test period and therefore do not take into
136 College & Research Libraries March 2010
account “real time”—that is, time spent
talking with a patron, the time involved in
conducting teachable moments, the learn-
ing skill of the recipient, and so forth—
times averaged for these transactions and
efforts did not involve a “live” patron. A
real transaction, with an interview and
resulting conversation, dependent on the
needs or communication skills of the user,
in all likelihood would have taken longer.
Adding a measure of the time expended
to handle a transaction was also suggested
by some participants in the modification
section of the online survey. If a library
were to keep track of the time expended
for each transaction within a semester,
then accurate data could be applied. This
is especially useful in terms of real-time
electronic services, such as Chat, where
the back-and-forth communication takes
on a different dynamic than an in-person
communication:
At times, certain aspects of the scale
indicating difficulty level seemed to
conflict, particularly on [C]hat. For
example, there were times when an
answer was relatively easy—I knew
it based on my knowledge—but
because I was working via [C]hat, it
required quite a bit of time to guide
a user through the information ses-
sion when I think less time might
have been required for an in-person
transaction.
Conclusion
Reference staffs appear ready to try new
methods for recording reference statis-
tics that include qualifying their effort,
knowledge, and skills. By continuing to
gather data from institutions that try the
READ Scale for reference services, the re-
searchers can begin to amass a large body
of statistics to normalize the Scale even
more, with an aim to create a dialogue
among professionals.
Future Directions for Research
The authors are invested in continuous
improvement of the READ Scale and wish
to thank our study participants as well
as other libraries that have adopted the
scale for use at their institutions and for
continuing to share their data with us. We
have benefitted from users who suggested
modifications as well as from having the
privilege of being engaged in constructive
and fruitful discussions toward progress in
improving the measurement of reference
work. In our quest to share the READ Scale
and to investigate its viability, several as-
pects of the scale have emerged as elements
worth considering for future research.
The most popular inquiry to arise
when discussing the READ Scale is the
issue of the timing of each category (for
instance: on average, how long does a
scale 3 question take to answer?) The
researchers have considered the element
of “time” as a measurement category and
TABLe 12
Final Total Number of Transactions recorded, using the reAD Scale and
Average Time Spent Total (Based on Common Pre-Study Q & A)
Average Time
(in minutes)
1 5 7 15 90 90
READ SCALE 1 2 3 4 5 6
Service Points 9,497 5,622 3,085 926 303 68 Total 19,501
Off-Desk 658 635 565 295 117 53 Total 2,232
Totals 10,155 6,257 3,650 1,221 420 121 Total 21,824
Hours 169 521 426 305 630 181 Total 2,232
Days (24 hrs) 7 22 18 12 26 8 Total 93
Days (8-hr day) 21 65 53 38 78 22 Total 277
Testing the Viability of the READ Scale 137
encourage adoptive institutions to build a
timing element into their preparation and
calibration training tools for librarians
gearing up for using the READ Scale in
their reference work. We have observed
two dominant schools of thought on the
proposal of using timing as a continuous
measurement. One school favors the tim-
ing of each transaction to later be used
as a performance measurement tool, a
training tool for the calibration of level
of effort when applying the READ Scale
rankings, and for reporting workload
effort to administrators. The opposing
school of thought does not favor the use
of “timing” as a measure of reference
effort as it can vary widely due to the
knowledge, experience, and personality
of the librarian handling the transaction.
The issue of the value of timing reference
transactions bears future investigation.
Survey feedback teased out the ques-
tion of how to take into account an
individual librarian’s level of reference
experience and expertise brought to the
reference transaction and how to score for
varying levels when “rating” a transaction
using the READ Scale. The question of
“level of experience and the rating of ref-
erence transactions” is an area that would
benefit from future research. How does
one build in expertise and knowledge that
is unique to the librarian or staff member,
their familiarity with the resources and
policies of their institution, when using
the READ Scale? More work remains to
address this aspect of the application of
the scale.
Determining the effectiveness of the
READ Scale for recording reference sta-
tistics and applying assessment practices
requires continued, long-term data col-
lection from a variety of institutions. The
researchers welcome any and all inter-
ested libraries to try the READ Scale and
contribute to its ongoing development
as a supplemental tool for qualifying
reference statistics by participating in
the ongoing research collaborative and
sharing experiences with colleagues. For
more information, go to http://www.dom/
edu/library/READ/index.html.
Notes
1. Eric Novotny, Reference Service Statistics & Assessment (ARL Spec Kit #268) (Washington,
D.C.: Association of Research Libraries, 2002).
2. Bella Karr Gerlich and G. Lynn Berard, “Introducing the READ Scale: Qualitative Statistics
for Academic Reference Services,” Georgia Library Quarterly 43 (Winter 2007): 7–13.
3. Brian Quinn, “Beyond Efficacy: The Exemplar Librarian As a New Approach to Reference
Evaluation,” Illinois Libraries 76 (Summer 1994): 163–73.
4. John C. Stalker and Marjorie E. Murfin, “Quality Reference Service: A Preliminary Case
Study,” The Journal of Academic Librarianship 22 (Nov. 1996): 423–29.
5. Jennifer Mendelsohn, “Perspectives on Quality of Reference Service in an Academic Library:
A Qualitative Study,” RQ 36 (Summer 1997): 544.
6. Bella Karr Gerlich, Work in Motion/Assessment at Rest: An Attitudinal Study of Academic
Reference Librarians, A Case Study at Mid-Size University (MSU A) (2006).
7. Deborah B. Henry and Tina M. Neville, “Testing Classification Systems for Reference
Questions,” Reference & User Services Quarterly 47 (Summer 2008): 364–73.
8. Russell F. Dennison, “Usage-Based Staffing of the Reference Desk: A Statistical Approach,”
Reference & User Services Quarterly 39 (Winter 1999): 158–65.
9. Debra G. Warner, “A New Classification for Reference Statistics,” Reference &. User Services
Quarterly 41 (Fall 2001): 51–55.
10. Novotny, Reference Service Statistics & Assessment (#268).
11. Sarla R. Murgai, “Reference Use Statistics: Statistical Sampling Method Works (University
of Tennessee at Chattanooga),” The Southeastern Librarian 54 (Spring 2006): 45–57.
12. Novotny, Reference Service Statistics & Assessment (#268), 10.
13. Gerlich, Work in Motion/Assessment at Rest, 122.
14. Ibid.
15. Ibid.
aCrl 2011 Conference
“A Declaration of Interdependence”
Philadelphia, March 30 - April 2, 2011
Call for PartiCiPation:
Great ideas. Multiplied.
Explore new ideas, engage in active peer to
peer learning and share your latest research,
cutting edge practices or innovative
developments with your colleagues.
Session formats
Deadline May 10, 2010
• Contributed Papers
• Panel Sessions
• Preconferences
• Workshops
Deadline November 1, 2010
• Cyber Zed Shed Presentations
• Poster Sessions
• Roundtable Discussions
• Virtual Conference Webcasts
Submit your proposal online today!
ACRL 2011 registration opens May 2010.
www.acrl.org/acrl/nationalconference
acrl@ala.org
800-243-545-2433 ext. 2522