Title: E-measures Committee, Final report, September 16, 2004
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00089879/00001
 Material Information
Title: E-measures Committee, Final report, September 16, 2004
Physical Description: Archival
Publication Date: September 16, 2004
Copyright Date: 2004
 Notes
General Note: Committee members: Shellet Arlen, Joe Aufmuth, Rich Bennett, Michele Crumpe, Carol Drum, and Erich Kesse
 Record Information
Bibliographic ID: UF00089879
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.

Full Text




















E-Measures Committee,
Final Report, September 16, 2004

Members:
Shelley Arlen, Joe Aufmuth, Rich Bennett,
Michele Crump, Carol Drum, Erich Kesse










Background


As a member of the Association of Research Libraries (ARL), the George A. Smather's
Libraries at the University of Florida moved to participate in ARL's project, ARL New
Measures Initiative The E-Metrics Project, and implement the associated report,
Recommended Statistics and Measures for Library Networked Services ARL E-Metrics
Phase II Report, Association of Research Libraries, www.arl.org. Consequently, on
December 1, 2003 the E-measures Committee was formed and charged with collection of
the new ARL e-metrics project as a "dry run" for the present academic year. The
committee was also charged with creating a set of repeatable procedures, and suggested
policies, for collecting the statistics on an ongoing basis. Additionally the committee was
requested to consider whether the Library can combine traditional statistics with the e-
measures and while the Library is free to report any and all statistics, the committee was
requested to determine which UF Fact book statistics are reliable, which ones have value
and which ones can used as substitutes for currently reported statistics.








Committee Assignments


Seven action items were generated from the committee's initial January 14, 2004
meeting. Those items were:

1) Produce a list of people who collect and keep statistics, and which statistics they
collect.
2) Produce a list of all databases owned by UF or owned by the state and used by UF and
which statistics are available for database services
3) List affiliated library statistics for Health Science and Law
4) Review ARL Statistics -- non electronic, which are 1) General, and 2) Preservation
related
5) Produce a matrix of each database and potential overlap of statistics
6) Summary of Statistics collected
-- UF internal statistics
-- ARL statistics
-- UF library's internal statistics
-- Stats for links from SFX -- times used and database target
-- Details about counters used by vendors
-- Library web statistics counter
-- Email sources and counts of questions
7) Recommendation on stats not collected

In the ensuing committee meetings the members researched available database and
service statistics, which met ARL's e-measures criteria. The committee attempted to
create a list of all databases and vendors and a list of all stats gathered from each. As for
statewide databases, the Florida Center for Library Automation (FCLA) extracted UF's
usage statistics. After investigating implementation of the ARL e-measures project
through an active attempt to collect statistics outlined in the ARL report, the committee
discovered many issues, which need to be resolved with staff assignments, available e-
metrics tools, and the ARL e-metrics criteria. Statistics collected are presented in Table
1. A summary of e-measures data collection is presented in Table. 2. A summary of
findings and issues to be resolved is presented at the end of the report.








Print and insert Table 1 here








Print and insert Table 2. here. Table 2 is 2 pages








Review of ARL E-Metrics Statistics and Committee Comments


R1 Number of electronic full-text journals

ARL Definition:
"Number of electronic full-text journal subscriptions that the library provides to users
either through an individual institutional licensing contract with the provider of journals
or through other arrangements (e.g., regional or state consortium) for which the library
pays a reduced or no fee for access.

The full-text journals should provide both search and browse capabilities by title
and issue. This is different from journal article databases, such as Expanded Academic
ASAP in INFOTRAC, that do not provide browsing capability.

This includes electronic full-text journals offered by established scholarly journal
publishing houses (e.g., Elsevier's ScienceDirect and Academic Press's IDEAL),
scholarly societies (e.g., American Chemical Society journals and American Institute of
Physics Online), and services which aggregate content from smaller publishers or from
those publishers that prefer to use an external delivery platform (Highwire, OCLC ECO,
and EbscoOnline).

This should exclude general-purpose periodicals such as magazines and
newspapers."

ARL Issues:
None

ARL Procedures:
"Create a master list of full-text electronic journals from all the sources available. Use a
spreadsheet or database program to organize and maintain the list. Remove titles that do
not meet the above-mentioned definition but keep duplicate titles resulting from
multiple subscriptions (e.g., main library and medical or law libraries). Record the
counts and be sure they are updated regularly.

Possible sources for the information include library catalog records (those records that
point to web addresses), library web pages that list the database titles, the internal
electronic resource management database, and vendor records websitess and contract
documentation)."

ARL Special Considerations:
"It is time-consuming to establish procedures to collect this statistic for the first time.
Include journal titles that come with print subscriptions or print plus online subscriptions.
Do not include free government publications and free electronic journals to which
the library provides links."








Committee Comments:
Number of electronic full text journals is probably one the most time consuming to
collect. Most electronic journals at the Library are provided through publishing houses.
The number of titles available and withdrawn varies through out the year and by
subscription level. Additionally, journals may be available through more than one
subscription service. Another aspect that makes the statistic difficult to collect is the
criteria "the full-text journals should provide both search and browse capabilities by title
and issue." Over 45,000 journals are electronically available. Each journal would have
to be evaluated against the ARL criteria.

Initial efforts to collect statistics on the number of electronic full text journals regardless
of their search and browse capability have been made. However, vendors do not
consistently supply electronic resource statistics as requested by ARL. Consortium
purchases through the Florida Center for Library Automation (FCLA) have been distilled
and reported by participating Universities (http://www.fcla.edu/system/intro.html).

A matrix of journal vendor vs. journal title could be created and a code such as s
(searchable), b (browsable) and SB (searchable and browsable) could populate the table.
Using MS Excel, a count of journals fitting the criteria could be created. We would also
discover duplicate items provided by multiple vendors.








R2 Number of electronic reference sources


ARL Definition:
"This includes citation indexes and abstracts; full-text reference sources (e.g.
encyclopedias, almanacs, biographical and statistical sources, and other quick fact-
finding sources); full-text journal and periodical article collection services (e.g.,
EBSCOhost, ProQuest, Academic Universe, and INFOTRAC OneFile); dissertation and
conference proceedings databases; and general-purpose magazines and newspapers.
Licensed electronic resources also include those databases that institutions mount
locally."

ARL Issues:
None

ARL Procedures:
"Possible sources for the information include library catalog records (those records that
point to web addresses), library web pages that list the database titles, the internal
electronic resource management database, and vendor records websitess and contract
documentation).

Create a master list of electronic databases from all the sources available. Use a
spreadsheet or database program to organize and maintain the list. Remove titles that do
not meet the above-mentioned definition but keep duplicate titles resulting from
multiple subscriptions (e.g., main library and medical or law libraries). Record the
counts and be sure they are updated regularly."

ARL Special Considerations:
"The number of databases provided by a vendor is reported and not just the total
[number of] vendors. If 1 vendor supplies 3 databases, the number reported is 3. This
count should not include freely available databases to which the library provides links
or library-created finding aids."

Committee Comments:
The Number of electronic reference sources is another time-consuming statistic to report.
While the restrictions on search and browse capabilities of the databases have not been
applied, duplicate titles from multiple subscriptions are allowed. Since the definition also
states "Licensed electronic resources also include those databases that institutions mount
locally," every database purchased on CD and DVD should be included. Many databases
are available in Government Documents, but those free through the depository program
should not be counted.

The FCLA Technical Information page (http://www.fcla.edu/svstem/intro.html) contains
links to some statistics. The ETD usage statistics on the FCLA page have not been
updated since February 2002. Many aspects of the statistic are not currently collected.








R3 Number of electronic books


ARL Definition:
"Number of electronic full-text monographs that the library offers to its users either
through an individual licensing contract with the content providers or through other
arrangements (e.g., regional or state consortium) where the library pays a reduced or no
fee for access.

This includes electronic books purchased through vendors, such as netLibrary and
Books24x7, and electronic books that come as part of aggregate services. It excludes
internally digitized electronic books, electronic theses and dissertations, digitally created
archival collections (e.g., Early English Books Online), and other special collections.
This also excludes publicly available electronic books to which the library provides web
links. It does not include machine-readable books distributed on CD-ROM, or
accompanied by print books."

ARL Issues:
None

ARL Procedures:
"For each electronic book collection, get the electronic title counts from either the
providers or catalog records. Unlike electronic full-text journals and reference databases,
it is not necessary to list the titles for each electronic book collection. Count any
duplicate titles resulting from multiple subscriptions (e.g., main library and medical or
law libraries). Record the counts and be sure they are updated regularly."

ARL Special Considerations:
"Do not include book collections that are a part of aggregate services and function more
as a reference collection (e.g., MD Consult reference books, ProQuest's Early English
Books Online, and books@OVID). They should be reported in the electronic reference
databases.

Do not include freely available electronic books such as titles available from the National
Academy Press."

Committee Comments:
Perhaps easiest and most straight forward statistic to collect. The e-book statistic is
currently collected through ARROW reports. Collection of the statistic through ALEPH
will have to be developed.








U1 Number of electronic reference transactions


ARL Definition:
"Number of electronic reference transactions conducted via email, a library's website, or
other network communications mechanisms designed to support electronic reference. An
electronic reference transaction must include a question either received
electronically (e.g., via e-mail, WWW form, etc.) or responded to electronically.
Those transactions that are both received and responded to electronically are counted as
one transaction. This count excludes phone and fax traffic unless either the question or
answer transaction occurs via the described manner. It includes the counts accrued from
participation in any local and national projects, such as DigiRef and the Library of
Congress's CDRS (Collaborative Digital Reference Service)." Note: This service is now
called QuestionPoint.

A reference transaction is an information contact, which involves the knowledge, use,
recommendations, interpretation, or instruction in the use of one or more information
sources by a member of the library staff."

ARL Issues:
None

ARL Procedures:
"Count the number of requests, not the number of questions." Establish an administrative
procedure to report electronic reference transaction counts to a designated staff person,
no matter who receives the questions or answers the reference requests. Select a typical
week (or month) to run a sample study. Be sure to vary the specific week (or month)
chosen over the course of a year or from year to year to account for seasonal
fluctuations." Note: If 1 request contained 3 questions the number reported would be
1.

Committee Comments:
Electronic reference statistics for e-mail, web forms and RefeXpress (See "Ask A
Librarian" http://www.uflib.ufl.edu/ask.html) are currently collected and reported. The
Library collects reference statistics during 2 randomly selected monthly sampling days
using a paper form, which asks staff to record e-mail reference statistics, as well as actual
and phone reference statistics. Daily e-mail and web-form statistics are also collected
from the libraries general e-mail reference and home page. RefeXpress collects and
reports reference statistics by transaction type using database software.

The committee recommends continued collection of electronic reference statistics. The
Library should consider collecting electronic statistics on a random weekly basis instead
of only 2 days per month.








U2 Number of logins (sessions) to electronic databases


ARL Definition:
"Number of user initiated sessions in licensed electronic resources. A session or login is
one cycle of user activities that typically starts when a user connects to a database and
ends with explicit termination of activities (by leaving the database through logout or
exit) or implicit termination (time out due to user inactivity). Licensed electronic
resources also include those databases that institutions mount locally."

ARL Issues:
"When analyzing the login counts, it might be important to explain any increases or
decreases in the figures. Specify, for example, whether the increase comes from (1) the
addition of new databases, (2) databases which did not report the statistic in the past but
have now begun reporting, (3) increased demand, and/or (4) an increase in the number of
simultaneous users.

Problems with the comparability of login counts from different vendors is a serious threat
to the utility of the combined count. Content providers use different time-out thresholds
(ranging from 7 to 30 minutes on average). Also, because of the IP-based authentication,
several sessions conducted at the same public workstation can be counted as a single
login. Alternatively, libraries can collect attempted logins to various licensed databases
by making users go through a central gateway (which counts all attempted logins). This
will ensure that one login attempt to a database is the same as a login to other databases.
However, what this data collection method misses is user logins that go directly to
content provider sites. It is unclear how many user logins fall into this category, but the
phenomenon certainly results in a substantial undercount of user logins.

While the gross login figure is useful, it is useful only for trend plotting and gross
justification of electronic resources. Within the library, the usage measures of licensed
electronic resources have many users and uses. Circulation of usage statistics on the
database title level (or in an extreme case on the journal title level) and discussion of any
noticeable changes (or lack thereof) need to occur at various levels among the concerned
parties, including collection development personnel, web masterss, technical services
staff, and so on."

ARL Procedures:
"We [ARL] recommend that each library develop an in-house spreadsheet or database to
capture monthly usage statistics of licensed databases. At least on the database title level,
usage statistics should be collected from vendors, entered into the in-house databank, and
maintained for reporting and analysis."

"Process monthly usage statistics from vendors and copy or import the number of
attempted sessions in each database (in each journal collection for full-text journals) to an
in-house spreadsheet or database file. Calculate the total sessions for a given month by
adding the number of sessions from each database or journal collection."








ARL Special Considerations:
Not all vendors report this statistic. Therefore, it will be necessary to qualify the statistic
with a sentence such as this: "We have 150,000 logins recorded from 120 databases out
of 200 subscribing. We cannot report this statistic for the remaining 80 databases
because the vendor does not supply login (session) information to customers."

Committee Comments:
Initial efforts to collect statistics on the number of sessions to electronic databases have
been made. However, many vendors do not consistently supply electronic resource
statistics as requested by ARL. Collecting the number of sessions to electronic databases
will require contacting vendors, which makes the statistic time consuming. Explaining
increased use coming from (1) the addition of new databases, (2) databases which did not
report the statistic in the past but have now begun reporting, (3) increased demand, and/or
(4) an increase in the number of simultaneous users may be more difficult.








U3 Number of queries (searches) in electronic databases


ARL Definition:
"Number of user initiated queries (searches) in licensed electronic resources. A search is
intended to represent a unique intellectual inquiry. Typically, a search is recorded each
time a search request is sent/submitted to the server."

ARL Issues:
Different assumptions about and mechanisms for collecting search counts by different
vendors are potential threats to the combined count.

ARL Procedures:
Same as U2. "Until there is a standardized report generation system that captures
different statistics from different content providers, we recommend that each library
develop an in-house spreadsheet or database to capture monthly usage statistics of
licensed databases. Usage statistics need to be collected from vendors, entered into the
in-house databank, and maintained for reporting and analysis. Process monthly usage
statistics from vendors and copy or import the number of attempted searches in each
database to an in-house spreadsheet or database file. Calculate the total number of
searches for a given month by adding the number of searches from each database or
journal collection."

ARL Special Considerations:
Same as U2. "Because some vendors do not report this statistic, it will be necessary to
qualify the statistic with a sentence such as this: We have 150,000 searches recorded
from 120 databases out of 200 subscribing. The other 80 do not provide this statistic."

Committee Comments:
Unlike electronic reference transactions where a single request with three questions is
counted as one transaction, user initiated queries are counted individually even if on the
same topic. While searches are recorded individually, there is no outcome measure of
search success. These counts are vendor controlled.








U4 Items requested in electronic databases


ARL Definition:
"Number of items requested in all of the library's licensed electronic resources. These
resources may include journal articles, e-books, reference materials, and non-textual
resources that are provided to the library's users through licensing and contractual
agreements. The user requests may include viewing, downloading, mailing, and
printing to the extent the activity can be recorded and controlled by the server
rather than browser. The items reported depend on the type of content. Examples
include citations, abstracts, tables of contents, and full-text articles (ASCII, HTML, PDF,
or PS)."

ARL Issues:
"Different vendors apply different assumptions and mechanisms in collecting items
requested counts. This lack of standardization makes it difficult to calculate an aggregate
count. We do not have good measurement of in-house materials usage, particularly
journal usage. However, electronic journals and databases allow libraries to find out how
often materials are requested. Having in-house usage figures is important for
understanding the dynamics of usage between print and electronic journals, so that we
can ascertain any correlation between them."

ARL Procedures:
Same as U2. "Until there is a standardized report generation system that captures
different statistics from different content providers, we recommend that each library
develop an in-house spreadsheet or database to capture monthly usage statistics of
licensed databases. Usage statistics should be collected from vendors, entered into the in-
house databank, and maintained for reporting and analysis. Process monthly usage
statistics from vendors and copy or import the number of items selected for viewing,
downloading, and mailing in each database. Count the number and type of items users
selected: abstracts, citations, and full-texts. Calculate the total number of items for a
given month by adding the number of items requested from each database or journal
collection."

ARL Special Considerations:
Same as U2. "Because some vendors do not report this statistic, it will be necessary to
qualify the statistic with a sentence such as this: More than 150,000 items were requested
from 120 databases out of 200 subscribing. The other 80 do not provide this statistic.
Among the requested items, 100,000 were some form of full-text records."

Committee Comments:
Some vendors consider the number of items requested and method of acquisition as
confidential information. U4 could actually be considered a certain measure of outcome
success, but it does not address how useful the requested information was to the searcher.
The committee is not sure how many vendors report: viewing, downloading, mailing,
and printing. Currently, the Library does not log in-house usage figures on items








requested from Library web pages. Some statistics exist for Digital Library Center items,
but the counter has been inconsistent.








U5 Virtual visits to library's website and catalog


ARL Definition:
"This is defined as user visits to the library's website or catalog from outside the physical
library premises regardless of the number of pages or elements viewed. If a user looks at
16 pages and 54 graphic images while at a website, that user registers one visit on the
web server. All visits to the website should be counted regardless of repetition by one
user. A visit is usually determined by a user's IP address, which can be misleading due
to Internet Service Providers (ISPs) and Firewalls or Proxy Servers. Thus, this measure is
actually an estimate of the visits."

ARL Issues:
"This measurement requires a relatively high degree of technical skills either on staff or
available from the library's website host."

ARL Procedures:
"* Identify all sources of virtual visits to the library. This may involve activities that
take place on more than one web server. Some of the web servers may be owned by the
library and some may be owned or maintained by another department in the university,
an Internet Service Provider (ISP), or other library vendors (e.g., library OPAC provider).
* Exclude internal use within the premises of the library from the counts for this
measure when possible. Two common approaches are using IP addresses or some form
of authentication tagged to each transaction. In terms of external visits to the library,
three common sources are: external access to the library's web page, remote logins
(sessions) to non-web-based library databases, and remotely accessible library OPAC.
* Develop strategies for collecting the data from each of these sources of virtual
visits. Different software may be needed to measure each electronic source of virtual
visits. In some cases, the library may calculate the virtual visits using one or more log
analysis software packages. In other cases, the external owner of the web server or
service (the ISP) must provide the data. Discussions may need to be held with these
service providers to obtain the needed data. In still other cases, custom programs may
have to be developed.
* In the case of library web pages housed on the library server, identify, configure,
and install appropriate log analysis software. Determine log analysis software definition
that corresponds to the virtual visit definition.
Note: Different log analysis software packages may count virtual visits in different ways,
so the count obtained will by necessity be an estimate. Arrange with the server technical
staff for regular (monthly) reporting of internal visits at the various user access Internet
workstations, external library user virtual visits, and total virtual visits (internal visits plus
external visits). Run the log analysis software.
* In the case of library web pages housed on an ISP's server, identify the log
analysis software the ISP uses. Determine the definition of "visit" used by the log
analysis software that corresponds to the virtual visit definition with the assistance of the
ISP. Arrange with the ISP for regular (monthly) reporting of internal library visits at the
various user access Internet workstations, external library user virtual visits, and total
virtual visits (internal visits plus external visits).








* Where virtual visit counts include the aggregate of internal and external visits,
indicate this in your report."

ARL Special Considerations:
"Count all visits to the website regardless of repetition by one user as long as each
visit meets the criteria for this statistic.

After one user connects to the Internet, several users could conduct multiple different
searches in the electronic service. In some cases, e.g., Internet-accessible OPAC use
inside the library, several users, one after the other, might make use of the same
established connection. In most systems, a connection is cut off after a specified period
of non-use, thus solving part of the problem. The best existing method of collecting
virtual visits is to use log analysis software. The log analysis software producers may
define virtual visits differently. For example, does a visit end after a time-out period of
30 minutes, 15 minutes, or some other time? The recommended time-out period is 30
minutes, but a local library may have to accept the available log analysis software's
definition even if it varies from the above.

Some libraries will find it difficult to report every virtual visit. For example, libraries
may have difficulty counting the use of library OPACS because their vendors do not
provide this information. Make a record of those sources of virtual visits not counted.
Do not estimate virtual visits for which data are not available."

Committee Comments:
Collecting virtual visit statistics should be easier to collect than what the ARL procedures
outline. Unfortunately, the current electronic counter software used to measure Library
Web page access does not consistently function. Virtual visit statistics for this report
have not been segregated by IP address. The library should investigate purchasing new
logging software to collect web site and catalogue statistics. The ARL criteria do not
seem to require differentiating between visits to the Library's home page and direct visits
to other associated Library pages.








C1 Cost of electronic full-text journals


ARL Definition:
"Expenditures for electronic full-text journal subscriptions that the library provides to its
users. Include both initial purchase cost, membership fees (such as JSTOR) as well as
annual access and service fees paid directly or through consortia arrangements."

ARL Issues:
None

ARL Procedures:
"Refer to the definition of R1. Current library accounting systems do not support coding
of materials expenditures by the categories used in the manual. Therefore, it may become
necessary to create an in-house spreadsheet or database file to keep track of cost
information according to the types of resources (e-journals, reference databases, and e-
books). Gather reports and invoices related to electronic databases and resources for the
reporting period. These documents are typically handled by the library's accounting
office. Even if a licensing contract or consortium arrangement period is different from
the reporting period, use the annual licensing fee to calculate the statistics. If a fee is paid
to a consortium or other joint arrangement, include the amount. In the case where a fee is
paid for an aggregate service and the service contains different categories of resources
(full-text journals and reference sources) as a bundle, use an estimate based on expected
or historical use, or list prices. Note any major commitments (such as JSTOR one-time
costs) that do not occur year to year and that significantly influence the reported amount."

ARL Special Considerations:
"Some electronic full-text journals come either as a free service with a print subscription
or as part of a print-plus-online-access subscription (the library pays extra for electronic
access). In the first case, the problem is whether or not to post any amount for the cost
of electronic access. In the latter case, the question is how much of the cost can be
attributed to electronic access."

"You can report the annual amount paid without prorating on the basis that over the years
the figures will even out."

Committee Comments:
The Cl statistic is currently collected. Collection of the statistic through ALEPH will
have to be developed. When considering cost of full-text journals ARL's Special
Considerations identifies print/electronic bundled journal costs as a problem which need
to be resolved.








C2 Cost of electronic reference sources


ARL Definition:
Expenditures for electronic reference sources and aggregate services that the library
provides to users either through individual licensing contracts with content providers or
through consortia or other arrangements where the library pays some fees. These fees
include both annual access fees and other service costs paid to the vendor directly or
through consortial arrangements.

ARL Issues:
None

ARL Procedures:
"Refer to the definition of R2. Current library accounting systems do not support coding
of materials expenditures by the categories used in the manual. Therefore, it may become
necessary to create an in-house spreadsheet or database file to keep track of cost
information according to the types of resources (e-journals, reference databases, and e-
books). Gather reports and invoices related to electronic databases and resources for the
reporting period. These documents are typically handled by the library's accounting
office. Even if a licensing contract or consortium arrangement period is different from
the reporting period, use the annual licensing fee to calculate the statistics. If a fee is paid
to a consortium or other joint arrangement, include the amount. In the case where a fee is
paid for an aggregate service and the service contains different categories of resources
(full-text journals and reference sources) as a bundle, use an estimate based on expected
or historical use, or list prices. Note any major commitments that do not occur year to
year and that significantly influence the reported amount."

ARL Special Considerations:
"You can report the annual amount paid without prorating on the basis that over the years
the figures will even out."

Committee Comments:
The C2 statistic is currently collected. Collection of the statistic through ALEPH will
have to be developed.








C3 Cost of electronic books


ARL Definition:
"Expenditures for electronic full-text monographs that the library offers to its users.
Include both initial purchase costs and membership fees as well as annual access and
service fees paid directly or through consortia arrangements."

ARL Issues:
None

ARL Procedures:
"Refer to the definition of R3. Current library accounting systems do not support coding
of materials expenditures by the categories used in the manual. Therefore, it may become
necessary to create an in-house spreadsheet or database file to keep track of cost
information according to the types of resources (e-journals, reference databases, and e-
books). Gather reports and invoices related to electronic databases and resources for the
reporting period. These documents are typically handled by the library's accounting
office. Even if a licensing contract or consortium arrangement period is different from
the reporting period, use the annual licensing fee to calculate the statistics. If a fee is paid
to a consortium or other joint arrangement, include the amount. In the case where a fee is
paid for an aggregate service and the service contains different categories of resources
(full-text journals and reference sources) as a bundle, use an estimate based on expected
or historical use, or list prices. Note any major commitments that do not occur year to
year and that significantly influence the reported amount."

ARL Special Considerations:
"Traditionally books are purchased on a one-time payment in exchange for permanent
ownership by the library. However, with regard to electronic books, it appears that some
arrangements allow libraries to subscribe to an e-book collection at a predetermined fee
and for a predetermined interval of time. We are concerned with the format of the
material, not the subscription or payment arrangement. These materials should be
counted as books, not serial publications."

Committee Comments:
The C3 statistic is currently collected. Collection of the statistic through ALEPH will
have to be developed.








C4 Library expenditures for bibliographic utilities, networks, and consortia

ARL Definition:
"Expenditures paid by the library for services provided by national, regional, and local
bibliographic utilities, networks, and consortia such as OCLC, RLG, excluding fees paid
for user database access and subscriptions, which should be reported in Cl through C3."

ARL Issues:
None

ARL Procedures:
"Gather reports and invoices with bibliographic utilities, networks, and consortia of
which the library is a member for the whole or part of the reporting period. These
documents are typically handled by the library's accounting office. Identify only those
expenditures paid to the bibliographic utilities, networks, and consortia for membership,
maintenance, and other infrastructure. Do not include expenditures that are directly
attributable to access of electronic resources. Those expenditures should be included in
Cl through C3. For instance, if your library paid a total of $100,000 to OCLC for its
various services and your best guess of electronic database access portion of the services
is 80%, then you should report $80,000 for C2 and the remaining $20,000 for C4. Even
if a membership or consortium period is different from the reporting period, use the
amount of the membership or consortium agreement."

ARL Special Considerations:
"It may be difficult to separate pure membership fees from value-added services of
membership (e.g., original catalog credits from OCLC that may be used to offset costs of
databases, purchase of catalog records, etc.). Report the annual amount paid without
prorating with the rationale that over the years the figures will even out."

Committee Comments:
The C4 statistic is currently collected.








C5 External expenditures for bibliographic utilities, networks, and consortia

ARL Definition:
"Expenditures paid by external agencies, such as state government agencies, on the
library's behalf for access to computer files, electronic serials, or search services through
a centrally funded system or consortial arrangements. Examples include state- (or
province-) supported networks such as VIVA (Virginia), CNSLP (Canadian National Site
Licensing Project), and the University of California's California Digital Library
Expenditure."

ARL Issues:
None

ARL Procedures:
"Gather reports and invoices with bibliographic utilities, networks, and consortia that are
related to electronic databases and resources for the reporting period. These documents
are typically handled by the library's accounting office. However, they can be
maintained outside the organization and, in some instances, may only be provided to
libraries upon demand. Find out how much of the central funding is attributable to your
library. For example, if your library contributes a total of $60,000 over a period of three
years to a state consortium that has a matching contribution of $120,000 for the same
period, the amount to report as C5 for a given year during the three-year period will be
$40,000 ($120,000 x 1/3). The library's contribution ($60,000) has to be divided
annually and posted in Cl through C3. If the specific dollar amount is not known, but the
total student FTE for the consortium and the amount spent for the academic members are
known, divide the overall amount spent by your institution's share of the total student
FTE. Alternatively, if the consortium is comprised of different types of institutions
(academic, public, or corporate), but the library has information about the portion of its
own use among the consortium participants, multiply the total amount by the percentage
of known (or estimated) usage rate. As a last resort, consult with a staff member
overseeing the consortium or the central funding system to get an estimate of the portion
of the central funding that is attributable to the library. Please make a note of this."

ARL Special Considerations:
None

Committee Comments:
The C5 statistic is currently collected.








D1 Size of library digital collection


ARL Definition:
"Library digital collection refers to digital materials (texts, images, and audio-visuals)
created in or converted from different formats (e.g., paper, microfilm, tapes, etc.) by the
library and made available to users electronically. This includes electronic theses and
dissertations (ETDs), special collections materials, maps, sound recordings, films, and
other digital materials that are not purchased or acquired from outside through individual
or consortial licensing agreements. Includes the number of titles and size (in gigabytes)
by sub-categories (ETD, visual materials, texts, multimedia), and as an aggregate at the
end of the reporting period. Also includes the number of items (titles) added during the
reporting period.

The types of formats in Appendix C, Figure C.4, refer to original formats rather than the
digitized outputs. Examples of visual materials include photos, maps, and postcards.
Examples of text include books, journal articles and pamphlets. Examples of multimedia
include audio, video, and other interactive materials. However, this statistic does not
include any back up copies or mirror sites because items should be counted only once."

ARL Issues:
"Because of the wide variations of the types and features of digital collections
constructed at ARL institutions, this statistic may be more useful locally than for
comparison across ARL member libraries. Benchmarking may, however, be possible
from the data collected to produce some qualitative and quantitative indicators as to the
extent of digital library collection activities and different emphases across the ARL
membership."

ARL Procedures:
"For cases in which multiple digital formats (derivatives) were produced from an item,
count it only once based on the type of item that was digitized. For example, if a 100-
page book was digitized in 100 TIFF files, each containing a page, a 100-page PDF file,
and 10 PDF files (one PDF file for each of 10 chapters), count it as a single text with 100
pages. If a derivative item was used as the source, do not count the outputs. But in the
total size (in gigabytes) include all versions of derivatives. Designate a staff member to
coordinate the collection of this statistic. The person should be well aware of library
digital collection activities. Identify library staff in charge of various digital library
projects and initiatives. It is necessary to conduct an inventory of digital material stock.
If this inventory information is already available, enter it into the worksheet. After
obtaining the inventory information, ask staff members to keep track of additional output
regularly. At the end of the reporting period, collect the worksheets and calculate the
total production during the reporting period."

ARL Special Considerations:
None.








Committee Comments:
Calculating and reporting the size of the Digital Library's collection in bytes should be
reconsidered.








D2 Use of library digital collection


ARL Definition:
"Number of times library digital collection titles and physical files were accessed and the
number of searches (queries) conducted (if there is such a capability) during the reporting
period."

ARL Issues:
"This statistic needs to be collected and used locally instead of across ARL member
libraries because of the wide variations of the types and features of digital collections
constructed at ARL institutions."

ARL Procedures:
"Do not report web page hits. Instead, count how many times the digitized items were
accessed. If a search capability is a feature of a library digital collection, the total number
of searches submitted needs to be collected. If continuous collection of use statistics is
not possible or desirable, select a typical week (or month) to run a sample study."

ARL Special Considerations:

Committee Comments:
Use of the Digital Library collection, while reported, is intended for local or internal use.
The statistic asks for the number of times an individual item is accessed. If possible,
Title access also is to be recorded. If the institution has the capability, the statistic also
asks for the number of queries conducted. Collecting the number of file queries provides
no information on the success of the query or on the usefulness of the files retrieved. In
the electronic reference measure 1 request with many questions is treated as one 1
transaction. However, with this statistic "hits" on individual chapters (many questions)
of the same book title (1 request) are recorded separately.

The current electronic counter used to measure Digital Library Web page access does not
consistently function.








D3 Cost of digital collection construction and management


ARL Definition:
"Annual direct costs (personnel, equipment, software, contracted services and similar
items) spent to create digital materials (texts, images, and multimedia) or to convert
existing materials into digital form for the purpose of making them electronically
available to users. Include expenditures related to digitization, OCR, editorial, creation
of markup texts, preparation of metadata for access to digitized materials, data storage,
and copyright clearance. Exclude expenditures for information resources purchased or
acquired from outside the institution through individual or consortial licensing
agreements."

ARL Issues:
"This statistic needs to be collected and used locally instead of across ARL member
libraries because of the wide variations of the types and features of digital collections
constructed at ARL institutions"

ARL Procedures:
"Designate a staff member to coordinate the collection of this statistic. Direct the
designated staff member to contact library staff members who are in charge of digital
collection projects. Ask all library staff members involved in any digital collection
projects as part of their official responsibilities to fill out the worksheet in Appendix C,
Figure C.8, for the reporting period. Ask them to estimate how much of their time was
spent on planning, implementing, and managing digital collection projects. This
information will be entered in the worksheet as FTE (full-time equivalent). A further
breakdown of activities may be necessary if the library wants to have more detailed
information on the distribution of efforts. Note that annual salary should not be asked of
the staff members filling out the worksheet and should not include fringe benefits. When
all the worksheets are collected, the salary information will be obtained from the library
accounting or personnel department. Direct staff cost will then be calculated. The
personnel cost should also include wages paid to non-salaried staff, including student and
other hourly workers. Cost of equipment should be amortized. For example, if a $3,000
scanner was purchased at the beginning of the reporting year and has a depreciation
period of three years, register $1,000 as the equipment cost. Costs of software should be
reflected in full amounts based on the time of the purchase. If a subcontracting period is
different from the reporting period, prorate the amount for the reporting period. If the
payment is based on percent to completion, include only the amount that belongs to the
reporting period."

ARL Special Considerations:
None

Committee Comments:
Probably the most onerous statistic to collect, it involves a detailed business analysis of
the Digital Library Centers activities.








P1 Percentage of electronic reference transactions of total reference

ARL Definition:
"Percentage of annual electronic reference transactions to total reference transactions.
An electronic reference transaction must include a question either received electronically
(e.g., via e-mail, WWW form, etc.) or responded to electronically. Count excludes phone
and fax traffic unless either the question or answer transaction occurs via the described
manner."

ARL Issues:
None

ARL Procedures:
"If continuous collection of this statistic is not possible or desirable, select a typical week
(or month) to run a sample study. It is recommended that you sample a week in a
different month or several months to account for seasonal fluctuations. Extrapolate based
on the sample data. Designate a staff member to coordinate the collection of this
measure. Key tasks include distributing a daily tally sheet, collecting the daily tally
sheet, adding each day's total to a weekly figure, and being available to respond to data
collection problems should they arise. For electronic transactions, use the count obtained
by following the procedures for Ul. See Table 2.7 or section 3.2.2.1 of this report for
additional information on Ul. Total the overall number of transactions. Divide the
number of electronic reference transactions by the total number of transactions. Multiply
by 100. Indicate and describe any additional methods used outside of this definition and
these guidelines."

ARL Special Considerations:
"Count the number of transactions, not the number of questions. That is, if one request is
emailed with three questions, it should be counted as one transaction, not three."

Committee Comments:
Percentage calculation, which also uses number of actual reference transactions.

(Electronic reference transactions/(actual reference transactions + electronic reference transactions)) X 100








P2 Percentage of virtual library visits of all library visits


ARL Definition:
"A virtual library visit is when a user visits the library's website or catalog for any length
of time or for any purpose from outside the physical plant of the library, regardless of the
number of pages or items viewed or requested. The term "virtual visit" excludes in-
library visits where a patron or a staff member uses electronic resources. If a user
looked at 16 pages and 54 graphic images while at a website, that user registers one visit
on the web server. A visit is usually determined by a user's IP address. Due to various
server management issues and differing software, this measure is an estimate of the visits
to the library site."

ARL Issues:
None

ARL Procedures:
"Obtain the virtual library visits count using the procedures for U5 (virtual library visits).
Obtain physical attendance count from turnstile counts or swipe card records. To the
extent possible, collect comprehensive data from all library branches. Combine the
virtual visit count and the physical attendance count. Divide the number of virtual library
visits by the total library visits. Multiply by 100. For example, a library had 1,000
external virtual visits and 9,000 physical visits for a total visit composite measure of
10,000. 1,000 virtual visits divided by 10,000 total visits equals .10 (or 10%)."

ARL Special Considerations:
None

Committee Comments:
Percentage calculation, which also uses number of actual library visits.

(Virtual library visits/(virtual library visits + actual library visits)) X 100

Our statistics report all virtual visits to the library regardless of location. Virtual visits
should be separated by IP address into Library and Non-library locations. Unfortunately,
there is no similar requirement to segregate electronic reference transactions, electronic
database access, electronic serials searches, or electronic reference searches by location.








P3 Percentage of electronic books to all monographs


ARL Definition:
"Percentage of the number of electronic books available to users (through either an
individual licensing contract or other consortial arrangements) to all the library's
monographs."

ARL Issues:
None

ARL Procedures:
"Identify all types of monographic materials. Use in-house record-keeping sources and
other library sources to determine the number of all monographs, including electronic
books, non-electronic books, and other monographic materials. Identify electronic book
types, including electronic books and electronic full-text aggregate services, using the
sources in step 1 of the procedures for R3. Count individual electronic book titles.
Record the number of individual electronic books from the spreadsheet or record the
number from another source. Exclude electronic reference books, i.e., publicly available
electronic books that are accessed for free. Calculate the total number of all monographs,
including electronic books, non-electronic books, and other monographic materials.
Divide the number of electronic books by the number of all library monographs
(electronic and non-electronic monographs). Multiply by 100. Indicate and describe any
additional methods used outside of this definition and guidelines."

ARL Special Considerations:
None

Committee Comments:
Percentage calculation, which also uses number of actual of monographs.

(Electronic books/(actual monographs + electronic books)) X 100









Summary of Findings and Issues to be Resolved


1) Standardize and centralize, the Library's statistics collection.
2) Document procedures for who collects what statistics and how.
3) Document usefulness of statistics collected
4) Resolve disparities between how reference transactions are determined.
5) Many of the requested ARL e-metric statistics while useful require the Library to
extend significant personnel and software resources to collect.
6) The requested ARL statistics may require an initial 4 month time commitment to
establish baseline statistics.
7) Coordinate collection and implementation of library statistics through a single
individual. A single individual, whose assignment for 0.25 to 0.5 FTE is to
collect statistics, would be best suited to the task.
8) The committee recommends adopting ARL e-measure statistic collection
procedures outlined in the ARL e-metrics report.
9) The committee recommends collaborating with ARL to refine the intent of some
e-measures and simplify some of the e-metrics collection procedures.
10) Eliminate e-metric statistics not used across ARL libraries. While some of the
suggested statistics like D2 and D3 are useful for internal purposes, they are time
consuming and are not used for ARL purposes.
11) The current electronic counter used to measure Web page access does not address
the e-measure statistics requested by ARL. And the current electronic counter
used to measure Web page access does not function consistently.
12) Vendors do not consistently supply electronic resource statistics. The Library
should work with vendors to supply appropriate usage reports
13) The committee recommends the Library adopt the international COUNTER
program (Counting Online Usage of Networked Electronic Resources).
14) ALEPH's electronic usage report potential for e-metric statistics is unknown and
should be explored.
15) Proxy servers for access to electronic databases would allow for access statistics
to be collected.
16) Inclusion of Health Center and Law Library in Library statistics collection.
17) Usage reports supplied by vendors should be coordinated through a single point of
contact.
18) Several categories of the e-metrics statistics are time consuming to collect and
require someone intimately familiar with vendors and databases.
19) Statistics on catalogue links to outside sources were not currently collected
20) The committee concluded no internally useful statistics are presented in the UF
Fact Book.








DRAFT
I. Policy:
Flexible hours / modified work schedules. Each employee has a standard shift that is
documented on their position description. However, there are occasions when an
employees may request flexibility in their normal working hours. Modified work
schedules that last more than 30 days must be documented on the employee's position
description. See excerpt below from UF Human Resources.
"Modified work schedules are those that deviate from the traditional schedule of eight
hours per day, five days per week. These schedules may be approved based on the
needs of the individual workplace but must still adhere to the requirements of the
FLSA as well as UF policy. Any modified work schedule that will be maintained for
more than 30 days should be reflected on the appropriate employee's position
description. All work schedules are established at the discretion of management and
require appropriate vice presidential approval."
However, if the flexible schedule is to last less than 30 days, a memorandum must be
placed in the employee's library personal file documenting the reasons for the shift
change and the duration of that change. Reference should also be made to verify that
all other conditions of UF policy are being met -- that an appropriate supervisor will
be available during the flexible hours, that the time will be accurately reported, and
that others in the immediate workplace who may be affected have been informed.
This policy is applicable to TEAMS, and USPS positions, regardless of appointment
length, appointing unit or source of funding.
II. Date: August 25, 2004


1. Policy:
Trip reports. Each employee traveling on Professional Development travel funding is
required to complete a trip report relating the information and experience gained in
the course of travel. The employee must provide the trip report to the supervisor
responsible for authorizing reimbursement. The employee must also post the trip
report to the "Trip Reports" public folder stored on the library's communication
server. The authorizing manager is responsible for seeing that trip reports are
completed as required for all authorized travel.
Date: August 25, 2004




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs