Digital Library of the Caribbean : a user-centric model for technology development in collaborative digitization projects

C# Code for application reference in article ( .zip )

Article ( .pdf )

Digital Library of the Caribbean :
a user-centric model for technology development in
collaborative digitization projects

Mark Sullivan
Marilyn N. Ochoa


Abstract:
Purpose The paper aims to describe the technological approach used by the Digital Library of the
Caribbean project to build an international collaborative library across many separate institutions with
varying degrees of expertise, technological abilities, and motivations.
Design methodology/approach Several methodologies were employed which helped build the
successful collaborative effort. A freely-distributed digitization toolkit was developed to assist with
tracking and digitization of the distributed resources. In addition, on-site training was provided for most
partners and was open for anyone in the area to attend. Progress and sustainability were ensured by
making this development process integral with the University of Florida Digital Collections initiative and by
continuing to perform usability studies on the web presence. The strong emphasis on branding and
presenting the same data within differently branded interfaces in the web presence encouraged greater
participation of international partners.
Findings- Digital Library of the Caribbean (dLOC), a joint project of the University of Florida, University of
Virgin Islands and Florida International University in partnership with institutions in the Caribbean and
circum-Caribbean, serves as the access point for scholars, students and citizens of interdisciplinary
Caribbean and circum-Caribbean research, gathers together a critical mass of cultural, historical and
research materials originally held in archives, libraries and private collections. This unique digital library
provides content submitted directly from dLOC partners and members and allows users to browse
materials or search the text through multilingual interfaces. Each of these items are generated in
distributed manner by each of the partners, and then submitted to a central server. This pattern has
worked quite well in a region where constant internet access is often lacking. An emphasis on partner
needs, such as institutional branding, has greatly contributed to the success of this project. Involving
partners in the process of usability studies, as well as conducting internal usability studies, have also
assisted in the creation of a successful project web interface.
Practical implications (if applicable) One concrete practical implication is the availability of
a standards-based digitization toolkit, which can be freely used by non-members as well as
members. Findings also suggest a course for digital library development in collaborative ventures. In
addition, this case shows the benefits of the recurring, iterative process of performing usability studies
during the development phase.
Originality / Value This paper has value for anyone looking to build collaboration in under-represented
regions.
Article Type: Case study
Keywords: Caribbean, digital libraries, sustainability, usability, collection management, digital cultural
content services


Digital Library of the Caribbean (dLOC), a joint project of the University of Florida, University of
Virgin Islands and Florida International University in partnership with institutions in the
Caribbean and circum-Caribbean, serves as the access point for scholars, students and citizens
of interdisciplinary Caribbean and circum-Caribbean research, and gathers together a critical
mass of cultural, historical and research materials originally held in archives, libraries and
private collections. This unique digital library provides content submitted directly from dLOC









partners and members and allows users to browse materials or search the text through
multilingual interfaces. Each of these items are generated in distributed manner by each of the
partners, and then submitted to a central server. This pattern has worked quite well in a region
where constant internet access is often lacking. An emphasis on partner needs, such
as institutional branding, has greatly contributed to the success of this project. Involving
partners in the process of usability studies, as well as conducting internal usability studies, have
also assisted in the creation of a successful project web interface.


Background
The Digital Library of the Caribbean (dLOC) is a collaborative digitization project from and about
the Caribbean. dLOC was established by a committee of librarians, scholars, and archivists at a
meeting held in San Juan, Puerto Rico in July, 2004. The goal of dLOC is to build a cooperative
digital library among partners within the Caribbean and circum-Caribbean, thus providing
scholars, students, and citizens around the world with open on-line access to Caribbean
cultural, historical and scientific materials. Collections include newspapers, photographs,
archives of Caribbean leaders and governments, official historical documents, and historic and
contemporary maps. dLOC's multi-institutional collaborative model is bolstered by asking
partner institutions to contribute to dLOC's evolution and success through both digital resource
creation and through a shared governance model. dLOC is comprised of educational, research,
governmental and non-governmental institutions aligned for the purpose of facilitating efficient
access to electronic collections about the Caribbean. As such, membership in dLOC is open to
archives, libraries, and museums; associations, organizations, and research centers; and
publishers. Administered by Florida International University in partnership with the University of
the Virgin Islands and the University of Florida, dLOC's technical infrastructure is provided by
the University of Florida.


Early development of the technical infrastructure for this project promoted a user-centric model.
Users were split into three main categories project partners, web users, and scholars. The
needs of each group were identified and reviewed throughout the life of the project. Through
this continual systematic analysis, the project has flourished and the number of partners and
use statistics have steadily increased.


For any successful collaborative effort, the concerns and needs for current and potential
partners must be addressed. In dLOC, each partner contributes a portion of their materials









directly to a central server for hosting and for preservation purposes. These materials are freely
served over the Internet. The needs of the partners included issues related to both the
submission method and the public interface.


Digital Resource Creation
Digital resource are created and submitted by each partner. By learning digitization techniques
ranging from scanning to metadata creation, each partner builds local capacity, while
contributing to a central project. An analysis of the problem resulted in identification of several
needs. Due to limitations on the Internet access and bandwidth throughout the Caribbean,
partners needed a digitization toolkit which could work off-line, only connecting to perform the
final submission of the resource to the central server. For many of the partners, the dLOC
project represented their first foray into digitization. So, partners needed a toolkit which was
easy to use and which required the least amount of training. In addition, for the training to be
more widely applicable, the toolkit would need to be standards-based.


Digitization with the dLOC Toolkit breaks the digitization process into five steps. The first step is
the creation of the bibliographic metadata, or basic description of the item. The second step is
the actual scanning of the item. After scanning, an item is automatically prepared during the
third step. The user reviews the images and adds structural metadata in the fourth step.
Finally, the user's built digital resource is submitted via FTP.


Digital Resource Creation Metadata
The dLOC Toolkit was developed as the metadata creation and submission tool. It is based on
the Metadata Encoding Transmission Standard (METS) and the bibliographic metadata is
stored in the Metadata Object Description Schema (MODS) within the METS file, although the
toolkit can also read METS files with Dublin Core as the bibliographic metadata scheme. The
toolkit assists with metadata creation by providing a customizable metadata creation and editing
tool. Prior to digitization, partners enter the basic metadata about the item. All of the metadata
is stored in a METS file in a folder automatically created for each resource.


A set of default metadata is established for each institution and project. This includes subject
keywords, holding locations, and spatial and geographic information. Each toolkit is installed
with the default metadata established for the Caribbean project, and additional default metadata











is added which is institution-specific. When each item is created, this default metadata is applied

before the user begins to manually enter the specific metadata for the individual item.



The metadata entry is divided into two portions. The first screen (Figure 1) has general

information about the resource, including titles, resource type, language of the text, creator,

publisher, etc.. During digitization training, partners are trained to create this metadata from the

in-hand original. Basic standards are discussed, although ultimately it is up to the local

institution to set their guidelines.










1. Deribe Ite 3. Prepre for ulity Submit Item
QItem QE: Control

General Information Subjects and Keywords
Title Information
Title: L'Ett Civi: Ce qu'l faut savoir

Series Title: JANAD'HA Liaison


Add Alternate Title


Add Uniform Title


Physical Description
Resource Type:

Physical Description:

Language:

Identifier:

Holding Location:


Name Authorities
Creator:

Publisher:

Publication Date:


Book

Bulletin semestrial; de 24 pages

French +]

vol 22-2 I Type: I

|Archives Nationales dHaiti


Archives Nationales d'Haiti

ANAD'HA

Aot 1999


iHeA ger l Exit ) i 14 Back Save ) forward N)



Figure 1: Adding the general metadata to the resource.











The second metadata entry screen, seen in Figure 2, includes the subject keywords and notes

about the resource. Partners are encouraged to enter as much spatial and geographic

information as possible. Abstracts are particularly encouraged for photographic material.




0 Dgitl Lbray f te Cribean- igiizaionTooki


1. Describe 2. ScanIte


General Information Subjects and Keywords
Subjects
Subject Keyword: Iaibbean

I Etat civil haitien

Geographic Coverage: Iport-au-prince
Haiti

Coordinates: Latitude


3. Prepare for
Qc


4). Ctality
Control


5. Submit Item


I ilExit BackSave Forward )


Figure 2: Adding subject keywords and notes to the resource.



The default template is kept as simple as possible, due to the varying degrees of expertise with

the partners. Within the different institutions which contribute to dLOC, the amount of exposure

to cataloging and metadata rules ranges from little experience to a working knowledge of

AACR2 rules for cataloging. This template is easily customizable from the preferences section

of the toolkit and can include a much larger percentage of the complete MODS and MARC

standard. The default set of elements includes the elements in Table 1 below.


] Scheme: I

] Scheme: .. +


Lngitude
Longitude: I


Notes
Abstract:


Language: French vi
document d'Etat civil relatant les procedures de declaration de naissance de reconnaissance de marriage de divorce et
de decks


Note: I Bulletin periodique pr6par6 pat les Archives Nationales d'Haiti


(Funding) Digitized with funding from the Digital Library of the Caribbean grant awarded by TICFIA.


I


I J











Title (main) Resource Type Holding Location
Title (series) Physical Description Creator
Title (alternate) Language Publisher
Title (uniform) Identifiers Publication Date
Subject Keywords Geographic Coverage Coordinates
Abstract (summary) Notes
Table 1: Metadata elements included in default template.


The bibliographic metadata entered by the technician is submitted with the completed digital
resource in the last step. Once the item is loaded, the item is flagged for metadata review.
Central catalogers review the metadata for each item and add additional metadata as required.
None of the originally entered data is removed, but additional information and translations may
be provided.


Digital Resource Creation Scanning
Once the item has been described, the toolkit prompts the partner to either scan the item in
hand, or to copy the scanned images into a folder automatically created for the resource.
During the training sessions, partners are taught how to scan their items and instructed in the
specifications recommended for their archival TIFF files, including color space and resolution.
The toolkit also accepts PDF files and several forms of audio-video files. Once acceptable
file(s) have been added to the resource folder, the user can go to the next step.


Digital Resource Creation Quality Control
The third step is a fully automated process to prepare the resource for quality control review and
subsequent submission. Each image file is processed to include several derivative files. Two
derivative files are created for mounting on the web server: a JPEG2000 file for the zoomable
image server and a simple JPEG file. Thumbnails for each image are also created, to be used
in thumbnail views in the online library. An addition JPEG file is created for each page to assist
with quality control in the next step. All of the information about the archival TIFF files is also
added to the growing METS file.


The fourth step, shown in Figure 3, allows the user to perform quality control on the images,
image correction, and enter the structural metadata about their item. All of the page images










are shown together as thumbnails. At this point, the user can visually inspect each image. The

toolkit allows the users to switch the order of the pages, or delete incorrect images. The user
can also name each page and create a table of contents for the resource by entering division
information. Many standard division names are provided, but users can also select 'Chapter'
and provide the actual name of a chapter. All of the entered information is collected in the
METS file and will be used to provide an online table of contents in the final web version. If an
institution chooses not to enter page names, a default is used, numbering from Page 1 through
all of the pages. Likewise, an institution may choose not to add the division information for an
item. In that case, no table of contents will appear online for the item.















jgs--- m^ ^^^^ ^^^ ^^^^ ^^^^ ^^K. 11^ ^^^^ ^^^^ ^^^^ ^^^ ^^^ J--- ^^^^ ^-- ------ --- ^ Bk
4. quality Control 5. Submit Item






















Hein Exit ) Back ) Save Forward


Figure 3: Performing quality control on the resource.












Each image can be edited inside the toolkit, as shown in Figure 4. This includes rotating the

image, cropping the image, and adjusting the color levels and balance. Once editing is

complete, the changes are applied to the archival TIFF and all of the image derivatives are

recreated.


0I gfdto oki C S *O* S: 0000 i0.4m ( 30 pi


Package Page Image Edt View Help
Page: |Page 1 Division Type: ]Front Matter v, B

SaveEdts DoneEditing Tool: A' Font: Arial *12 *









In
#:j


o I m


pe K)


k B 0 It XI ml
smooth B I M ot W lose V Cormplete

~I9


Guillaume Nicolas Fabre GEFFRARD
Le vlnnaire a qu rIon doot
les "ArchlW Nationale d'Half
ce6es le 20 aot 1860


H 1 0 H


Figure 4: Editing images in the dLOC Toolkit.


Digital Resource Creation Submission

The final step submits the item to an online library via FTP. The user can enter their own FTP

information, including server name and username and password. If a dLOC License Code has


IMI










been provided to unlock the dLOC specific features, it will automatically be sent to the dLOC
servers.


Digital Resource Creation Customization
Through the course of this work, another unexpected user-base was discovered. As scholars in
other fields began to hear about the dLOC capacity-building model and submission toolkit, it
became apparent that there was a wider need for a simple, standard-based, offline digital
resource toolkit. Already several other projects have begun to employ the digitization toolkit
which was originally created for dLOC. To facilitate their use, "themes" were added to the
toolkit. This allows the interface to be customized for each different project. (See Figure 5 for
the view of an additional theme.) In addition, the user can select the language for the toolkit.
English, Spanish, and French are currently supported, with hopes to add Dutch, Haitian Creole,
and Papiamento in the near future.






qa$ HavOaY GLLECTIONS

Basic Advanced


EXIT


UF olF r ThItor N
The Foundation for The Cator, Nalon


Figure 5: Theme customization in the dLOC Toolkit.


What would you like to do today?
Start a new item

Finish uncompleted items

Edit item from dLOC

Review all items

View working directory ( C:\DLOC\

Edit preferences









Digital Resource Creation Usability
As the developer took part in the training which was provided to the partners, the partners in
turn provided a very useful metric back to the developer. This unusual amount of developer-
user interaction helped to steer the development of the toolkit. In fact, the initial release of the
toolkit was much less user-friendly and often left partners confused on their next step. After
about two years of use and training, a new toolkit was developed and released. Subsequent
training on the new toolkit proved the increased level of usability and ease. This feedback,
while short of an actual supervised usability study, was critical to the success of the toolkit and
subsequently the project.


Digital Resource Creation Future Enhancements
Work continues on the toolkit with new changes and releases expected shortly. The new
version of the toolkit will add more interoperability with the MARC21 standard, allowing import
and export of records in MARC XML and MARC21 formats. In addition, the C# source code is
targeted to be released under the GNU license by summer of 2009. The toolkit is currently
available freely for download at the dLOC project web site ( http://www.dLOC.com ).


dLOC Web Interface
The dLOC web interface (shown in Figure 6) developed simultaneously and in conjunction with
the metadata toolkit, based on the requirements of the grant, needs of the users, and continued
input from usability studies. The foundation for the dLOC interface is the technology employed
by the University of Florida Digital Collections (UFDC) which is a METS-based presentation
layer over a Greenstone core. This allows dLOC to take advantage of all the work put into
testing UFDC and adds to the general sustainability of the dLOC web interface. Attempts to
improve the overall delivery system for dLOC focused on two groups: partners and users. The
partners needs were generally identified as branding. For the users, several usability studies
were performed.


From discussions with partners, it was obvious that they were willing to contribute their time and
resources to a collaborative digitization effort, but didn't want their individual contributions lost
among all the others. While dLOC already had branding in the form of wordmarks when viewing
an individual resource, that didn't appear to be quite sufficient. To solve this issue, two new
concepts were added to the UFDC technology. Interfaces were added to allow the web page to
take on a very different look and feel, often one more like the partner's institutional web pages.










Institutional pages were also added, to allow browsing and searching of only those materials

added by a particular institution. The combination of institutional pages and interfaces allows

dLOC to present very different appearances to the same collection of items. (See Figure 7 for

an example of an interface on an institutional page.) This ability is very attractive to both

existing partners and potential partners. In fact, dLOC itself is an interface, which allows all of

the dLOC pages to be branded with Digital Library of the Caribbean, despite existing in the

same server as the general University of Florida collections.


english espailol frangais


Dig-ial Lbrar of th Carbbea ---


mrLIMSTEW ITiEM


Search Collection: I G


Welcome! iBienvenidos! Bienvenue!

The Digital Library of the Caribbean (dLOC) is a cooperative digital library
for resources from and about the Caribbean and circum-Caribbean. dLOC
provides access to digitized versions of Caribbean cultural, historical and
research materials currently held in archives, libraries, and private
collections.

For more information on eligibility requirements for partnership and
conditions of membership in dLOC, please see our New Member
Application (in .pdf format).

Read more about us.

The Digital Library of the Caribbean (dLOC) is issuing a call for partners in a new effort to ensure preservation of and increase
access to newspapers in the Caribbean.



Figure 6: Digital Library of the Caribbean main web page.


r ADVANCED SEAR


ASBOUT DL-


PARTNER COLLECTIONS


L SHOW SLIBCOLLE





































Figure 7: Institutional page with an interface to mimic a different web presence.


Usability testing
Researchers have access to a plethora of content online from which to search. Some users may
not understand the importance of using library-provided resources, and instead may choose to
use a website solely based on how easy it is to search and navigate. Considering the
competition for getting users to return to a given online resource, institutions that produce digital
libraries such as UF must develop resources that are user-centered and are easy to use and
navigate as well.


Developers consider user expectations in their design process to create a user-centered and
appealing resource to which a user will return. Usability testing is a method to identify
expectations and has had a role in design and implementation of computers, interfaces and
other technology products for decades (Nielsen, 15). According to the U.S. Department of
Health & Human Services Usability.gov site, usability refers to how well users can learn and use
a product to achieve their goals and how satisfied they are with that process; usability means
that people who use the product can do so quickly and easily to accomplish their tasks (Dumas









and Redish 885). Usability measures the quality of a user's experience when interacting with a
product or system, a website or software application-"the extent to which a product can be
used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction
in a specified context of user" (ISO 9241-11). Overall, usability refers to a resource that:


* Is easy to learn-intuitive where basic tasks can be accomplished soon after seeing the
resource for the first time
* Has efficiency of use-how fast can experienced users accomplish desired tasks
* Is easy to remember-the resource can be used efficiently the next time a user sees it
* Is satisfying to use
* Meets user and organization's needs, objectives and expectations


Because usability testing encourages users to take ownership of the resource and to continue to
use it and share it, UF has engaged in testing since 2005 throughout the development or
customization of electronic resources. From February 2007 to April 2008, various collections
within the UFDC framework were tested, with findings discussed in two separate usability test
reports (Ochoa 2007, 2008). The tests sought to examine how well the UFDC organized its 30
collections, and how the UFDC compared to similar resources in terms of usability. Testing
conducted by a public services librarian at UF with experience in usability testing (usability
consultant) occurred before the full launch of UFDC in April 2007, and again as part of the
iterative design process to improve UFDC and its subcollections.


During this year of testing, key areas of concern identified prior to testing included:
* Are the UFDC search pages-all collections, basic and advance-intuitive?
* Do the result pages provide the type of information a participant needs?
* Are the item viewer navigation functions appropriate to ensure easy navigation and ease of
use?
Further, the usability consultant ensured that general navigability and features or functionality
resulting from use of certain formats beyond text were tested. Examples are images and serials
found in the Baldwin Library of Historical Children's Literature and the Florida Digital Newspaper
Library.


Along with these considerations, identifying the intended use of UFDC to be for research
purposes, along with the type of user were essential in defining the type of the testing to be









done. Since this digital library was useful for discovering hidden research resources, the
potential users of this resource include researchers, university scholars and the everyday public;
therefore, testing university participants as a representative sample of the user population, using
multiple testing methods would be appropriate to identify the issues of UFDC. Once testing was
determined to require direct contact with end users, a human testing protocol was issued prior
to recruitment/scheduling of test participants and testing. At the University of Florida the test
protocol includes an informed consent and all survey instruments, including pre- and post-test
questionnaires.


Scenario-based and focus group testing were used in UFDC testing. Because even minor
interface details can affect the way a participant uses the resource, using a very narrow
approach, i.e. scenario-based tasks to target a specific feature of the resource, was effective to
identify how to improve the resource; each step, response and behavior was carefully noted and
recorded by a note taker. Having the participant talk aloud during the testing provided additional
information about both how and why users complete tasks in the manner they do. This type of
testing provided the added benefit for developers to learn more about the thought process
behind using the interface, and to identify what is not intuitive about a particular feature. In
addition, focus groups were used to obtain supplemental data about errors in completing tasks
and other overall user impressions.


The data collected through focus group sessions and through individual scenario-based
sessions demonstrated how well the UFDC matched participant needs and expectations.
Patterns in the common mistakes towards completing tasks, noted user perspectives and
comments, and conclusions drawn from overall behavior were considered when listing the
problems for future redesign. Overall, the feedback from testing resulted in the change to the
interface, result pages and item viewer to improve navigation and retrieval. Some modifications
throughout the 2007 to 2008 period included:
* Introductory/explanatory information on the resource as a whole and individual
subcollections
* Better navigation including linking back to the result lists
* More intuitive design for selecting collections for a federated search
* Ease of access to resources from Result pages especially for serials
* More visible and more intuitive functions of the Item viewer navigation









Usability Testing of dLOC
In early Summer 2008, dLOC was tested as a part of the broader iterative design process for
UFDC. While dLOC benefited from previous UFDC usability testing and subsequent
modification, the uniqueness of dLOC as a multi-lingual resource and the requirements of the
grant through which dLOC was funded, warranted testing of the subcollection.


The testing strategy for dLOC incorporated past testing experiences, but the development of the
plan also relied on discussions with the dLOC Coordinator at Florida International University.
With the coordinator's help, the usability consultant identified some of the goals of dLOC testing:
* To consider the multi-site interface for the Spanish, French and English interfaces of dLOC.
* To justify additional costs for translation
* To identify needed interface problems, including use of certain terminology
* To optimize already implemented features
* To enable marketing and familiarity with dLOC among Caribbean partners and potential
partners


In addition to identifying these goals, the intended use of dLOC and its audience were
necessary to plan the strategy for dLOC testing. dLOC testing required direct contact with
researchers, university scholars and the everyday public in the US and in the Caribbean and
circum-Caribbean. The loosely formed usability team (dLOC Coordinator and usability
consultant/librarian) felt university participants in the U.S. alone would only provide one
perspective to identify issues and improve dLOC; usability data collected from dLOC partners
would be useful.


The team identified factors of location, language, internet connectivity and staffing as possible
complications to the usability testing of dLOC. While general UFDC usability testing was
conducted cheaply and easily-due to testing location on the UF campus, potential onsite users
and equipment readily available-the dLOC testing proved more challenging. The assumed
heavy use by Caribbean and circum-Caribbean researchers forced the team to look outside of
UF and FlU for test participants. This audience would require locating and contacting users
who would use the content, and who would likely be non-English or multilingual speakers with
unpredictable internet availability. Based on these considerations possible options for testing
included:









* Incorporating usability training and/or testing across the Caribbean during already-planned
UFDC Toolkit training so that each partner location can conduct usability testing
* Conducting usability testing at an event which already brought dLOC partners and
Caribbean library staff together, and asking for usability testing assistance


The first option was not selected because the Toolkit training session agenda was already
extensive. No dedicated usability specialist was hired to travel and provide training or testing at
regular intervals within the Caribbean; the UF librarian who worked on usability testing did so as
a small percentage of her position. Further, asking facilitators at each location to test was
thrown out at this stage; few partners were familiar with usability testing and would first need to
experience it themselves and be trained before trying to conduct the test. Plus, the dLOC
coordinator would need time and resources to negotiate this new role/commitment by partners.


The second option of testing at an event was selected with evaluation of dLOC conducted at the
annual Association of Caribbean University, Research and Institutional Libraries (ACURIL),
XXXVIII Conference, Montego Bay, Jamaica, in June 2008. Because testing requires direct
contact with end users, this location/event was useful because it allowed the dLOC coordinator,
designer, and usability consultant an opportunity to work with a variety of actual users and
partners of dLOC, without additional cost of travel to anyone. The session allowed users to
share their experiences, comments and attitudes about dLOC and to identify problems with
using the resource. It encouraged a sense of goodwill and ownership for the resource by
Caribbean partners, ultimately resulting in more use and a better reputation for dLOC and
UFDC as a whole.


The research methodology of dLOC testing involved scenario-based testing and a focus group
session to evaluate online participant search behavior and attitude. Two sets of test questions
were used. All task questions were given in writing to aid in preventing misinterpretation.


Participants completed pre- and post- test questionnaires to identify user experience using
online library resources and satisfaction with using dLOC. Then all participants participated in
an open group talk aloud session of dLOC which consisted of 3 questions. All attendees
worked together in this session to complete the structured exercises. Participants were then
split into groups to answer a set of questions focusing on basic and advanced search interfaces,
multi-lingual interfaces, result pages and item viewer; these examples and guided exercises









were based on real content and reference questions directed to UF and FlU about dLOC.
During the group test session, the facilitators (dLOC coordinator and usability consultant) asked
participants to talk out loud about their process to complete the question and what they
expected would happen; they also observed user behavior and identified users' errors.
Responses were written down. Finally, the focus group consisted of a discussion on experience
with dLOC, guided by the post-test questionnaire submitted.


Overall, onsite testing was a useful experiment with usability testing for multilingual users. It
identified problems with terminology, use of images or icons for navigation, and the use of a
tabbed approached to reach various views of the resource found.


Future of dLOC Usability
Usability testing will continue to be a critical part of the design process. Having the input of
Caribbean partners and their users can offer some new and different perspectives from past
UFDC usability testers. In addition, having assistance from partners to join the dLOC usability
team was extended for possible multi-site assessment to further guide the development for
dLOC. Since the workshop provided the theory and application of usability testing, partners
have been given a framework for evaluating electronic resources in general. By evaluating
dLOC themselves, participants would already be familiar with the testing methodology for dLOC.


The future for dLOC usability stems from an idea that a UF or FlU-led dLOC Assessment
program can be created to centrally develop questionnaires and test instruments for the
resource and to provide training. Challenges of off-site usability testing remain, however: how
to provide training of usability testing for partners, site limitations (including internet connectivity
and staffing), and development of a single, overarching human testing protocol.


About the Authors


Mark Sullivan works as a developer for the Systems Department at the University of Florida
Libraries supporting the Digital Library Center. He has been integral to several recent grants at
UF and is the chief architect of the dLOC toolkit and UF's Digital Collections
(http://www.uflib.ufl.edu/ufdc/) He joined the dLOC project early and has taken part in every
training session since its inception.









Marilyn N. Ochoa is currently the Assistant Head of the Education Library at the University of
Florida. Prior to this position, she served as the digital services librarian for the UF Humanities
and Social Sciences Services Department.




References

Dumas, Joseph and Janice (Ginny) Redish. 2002. Usability in practice: formative usability
evaluations evolution and revolution Conference on Human Factors in Computing
Systems Minneapolis, Minnesota, USA 885 890 Usability in Practice Session.

International Standards Organization, "Ergonomic requirements for office work with visual
display terminals (VDTs) -- Part 11: Guidance on usability," ISO 9241-11:1998, March
13, 1998
83> (February 2, 2005).

International Standards Organization, "Human-centred design processes for interactive
systems,"
ISO 13407:1999, July 1999
97> (February 2, 2005).

Krug, Steve. 2006. Don't make me think!: a common sense approach to Web usability.
Berkeley, Calif: New Riders Pub.

Nielsen, Jakob. 1993. Usability engineering. Boston: Academic Press.
U.S. Department of Health & Human Services, "Usability.gov,"
(February 16, 2005).

Ochoa, Marilyn. 2008. Usability report for the University of Florida Digital Collections (UFDC).
Gainesville, Fla.: George A. Smathers Libraries, University of Florida


Ochoa, Marilyn, Usability Test Report for Florida Digital Newspaper Library, UFDC. Gainesville,
Fla.: George A. Smathers Libraries, University of Florida
.


Association of Caribbean University, Research and Institutional Libraries, XXXVIII Conference,
Montego Bay, Jamaica, Workshop: "Usability Testing: A User-Centered Approach to
Improve Electronic Resource Design." June 2, 2008, jointly with Brooke Wooldrige.

Association of Caribbean University, Research and Institutional Libraries, XXXVIII Conference,
Montego Bay, Jamaica, IT-SIG: "Strategy to Increase Effectiveness of Library Electronic
Resources: Designing Usability Testing." June 4, 2008.

PAGE 1

Digital Library of the Caribbean : a user centric model for technology development in collaborative digitization projects Mark Sullivan Marilyn N. Ochoa Abstract: Purpose The paper aims to describe the technological approach used by the Digital Library of the Caribbean project to build an international collaborative library a cross many separate institutions with varying degrees of expertise, technological abilities, and motivations. Design methodology/approach Several methodologies were e mployed which helped build the successful collaborative effort. A freely distribute d digitization toolkit was developed to assist with tracking and digitization of the distributed resources. In addition, on site training was provided for most partners and was open for anyone in the area to attend. Progress and sustainability were ensu red by making this development process integral with the University of Florida Digital Collections initiative and by continuing to perform usability studies on the web presence. The strong emphasis on branding and presenting the same data within different ly branded interfaces in the web presence encourage d greater participation of international partners. Findings Digital Library of the Caribbean (dLOC), a joint project of the University of Florida, University of Virgin Islands and Florida International U niversity in partnership with institutions in the Caribbean and circum Caribbean, serves as the access point for scholars, students and citizens of interdisciplinary Caribbean and circum Caribbean research, gathers together a critical mass of cultural, his torical and research materials originally held in archives, libraries and private collections. This unique digital library provides content submitted directly from dLOC partners and members and allows users to browse materials or search the text through m ultilingual interfaces. Each of these items are generated in distributed manner by each of the partners, and then submitted to a central server. This pattern has worked quite well in a region where constant internet access is often lacking. An emphasis on partner needs, such as institutional branding, has greatly contributed to the success of this project. Involving partners in the process of usability studies, as well as conducting internal usability studies, have also assisted in the creation of a suc cessful project web interface. Practical implications (if applicable) One concrete practical implication is the availability of a standards based digitization toolkit which can be freely used by non members as well as members. F indings also suggest a course for digital library development in collaborative ventures. In addition, this case shows the benefits of the recurring, iterative process of performing usability studies during the development phase. Originality / Value This paper has value for an yone looking to build collaboration in under represented regions. Article Type: Case study Keywords: Caribbean, digital libraries, sustainability, usability, collection management, digital cultural content services D igital Library of the Caribbean (dLO C), a joint project of the University of Florida, University of Virgin Islands and Florida International University in partnership with institutions in the Caribbean and circum Caribbean, serves as the access point for scholars, students and citizens of in terdisciplinary Caribbean and circum Caribbean research, and gathers together a critical mass of cultural, historical and research materials originally held in archives, libraries and private collections. This unique digital library provides content submi tted directly from dLOC

PAGE 2

partners and members and allows users to browse materials or search the text through multilingual interfaces. Each of these items are generated in distributed manner by each of the partners, and then submitted to a central server. This pattern has worked quite well in a region where constant internet access is often lacking. An emphasis on partner needs, such as institutional branding, has greatly contributed to the success of this project. Involving partners in the process of us ability studies, as well as conducting internal usability studies, have also assisted in the creation of a successful project web interface. Background The Digital Library of the Caribbean ( dLOC ) is a collaborative digitization project from and about the Caribbean dLOC was established by a committee of librarians, scholars, and archivists at a meeting held in San Juan, Puerto Rico in July, 2004. The goal of dLOC is to build a cooperative digital library among partners within the Caribbean and circum Cari bbean, thus providing scholars, students, and citizens around the world with open on line access to Caribbean cultural, historical and scientific materials. Collections include newspapers, photographs, archives of Caribbean leaders and governments, offic ial historical documents, and historic and contemporary maps. dLOC 's multi institutional collaborative model is bolstered by asking partner institutions to contribute to dLOC 's evolution and success through both digital resource creation and through a sha red governance model. dLOC is comprised of educational, research, governmental and non governmental institutions aligned for the purpose of facilitating efficient access to electronic collections about the Caribbean. As such, membership in dLOC is open to archives, libraries, and museums; associations, organizations, and research centers; and publishers. Administered by Florida International University in partnership with the University of the Virgin Islands and the University of Florida, dLOC 's technical infrastructure is provided by the University of Florida. Early development of the technical infrastructure for this project promoted a user centric model. Users were split into three main categories project partners, web users, and scholars. The n eeds of each group were identified and reviewed throughout the life of the project. Through this continual systematic analy sis, the project has flourished and the number of partners and use statistics have steadily increased. For any successful collabora tive effort, the concerns and needs for current and pote ntial partners must be addressed In dLOC each partner contributes a portion of their materials

PAGE 3

directly to a central server for hosting and for preservation purposes. These materials are freely se rved over the I nternet. The needs of the partners included issues related to both the submission method and the public interface Digital Resource Creation Digital resource are created and submitted by each partner By learning digitization techniques r anging from scanning to metadata creation, each partner builds local capacity, while contributing to a central project. An analysis of the problem resulted in identification of several needs. Due to limitations on the I nternet access and bandwidth throug hout the Caribbean, partners needed a digitization toolkit which could work off line, only connecting to perform the final submission of the resource to the central server. For many of the partners, the dLOC project represented their first foray into digi tization. So, partners needed a toolkit which was easy to use and which required the least amount of training. In addition, for the training to be more widely applicable, the toolkit would need to be standards based. Digitization with the dLOC Toolkit breaks the digitization process into five steps. The first step is the creation of the bibliographic metadata, or basic description of the item. The second step is the actual scanning of the item. After scanning, an item is automatically prepared durin g the third step. The user reviews the images and adds structural metadata in the fourth step. Digital Resource Creation Metadata The dLOC T oolkit was developed as the metadata creatio n and submission tool. It is based on the Metadata Encoding Transmission Standard (METS) and the bibl iographic metadata is stored in the Metadata Object Description Schema ( MODS ) with in the METS file, although the toolkit can also read METS files with Dub lin Core as the bibliographic metadata scheme. The toolkit assists with metadata creation by providing a customizable metadata creation and editing tool. Prior to digitization, partners enter the basic metadata about the item. All of the metadata is sto red in a METS file in a folder automatically created for each resource. A set of default metadata is established for each institution and project. This includes subject keywords, holding locations, and spatial and geographic information. Each toolkit is installed with the default metadata established for the Caribbean project, and additional default metadata

PAGE 4

is added which is institution specific. When each item is created, t his default metadata is applied before the user begins to manually enter the sp ecific metadata for the individual item. The metadata entry is divided into two portions. The first screen (Figure 1) has general information about the resource, including titles, resource type, language of the text, creator, publisher, etc.. During dig itization training, partners are trained to create this metadata from the in hand original. Basic standards are discussed, although ultimately it is up to the local institution to set their guidelines. Figure 1: Adding the general metadata to the resou rce.

PAGE 5

The second metadata entry screen, seen in Figure 2, includes the subject keywords and notes about the resource Partners are encouraged to enter as much spatial and geographic information as possible. Abstracts are particularly encouraged for photo graphic material. Figure 2: Adding subject keywords and notes to the resource. The default template is kept as simple as possible, due to the varying degrees of expertise with the partners. Within the different institutions which contribute to dLOC the amount of exposure to cata loging and metadata rules ranges from little experience to a working knowledge of AACR2 rules for cataloging This template is easily customizable from the preferences section of the toolkit and can include a much larger per centage of the complete MODS and MARC standard. The default set of elements includes the elements in Table 1 below.

PAGE 6

Title (main) Resource Type Holding Location Title (series) Physical Description Creator Title (alternate) Language Publisher Title (uni form) Identifiers Publication Date Subject Keywords Geographic Coverage Coordinates Abstract (summary) Notes Table 1: Metadata elements included in default template. The bibliographic metadata entered by the technician is submitted with the complete d digital resource in the last step. Once the item is loaded, the item is flagged for metadata review. Central c atalogers review the metadata for each item and add additional metadata as required. None of the originally entered data is removed, but additi onal information and translations may be provided. Digital Resource Creation Scanning Once the item has been described, the toolkit prompts the partner to either scan the item in hand, or to copy the scanned images into a folder automatically created fo r the resource. During the training sessions, partners are taught how to scan their items and instructed in the specifications recommended for their archival TIFF files, including color space and resolution. The toolkit also accepts PDF files and several forms of audio video files. Once acceptable file (s) ha ve been added to the resource folder, the user can go to the next step. Digital Resource Creation Quality Control The third step is a fully automated process to prepare the resource for quality con trol review and subsequent submission. Each image file is processed to include several derivative files. Two derivative files are created for mounting on the web server: a JPEG2000 file for the zoomable image server and a simple JPEG file. Thumbnails fo r each image are also created, to be used in thumbnail views in the online library An addition JPEG file is created for each page to assist with quality control in the next step. All of the information about the archival TIFF files is also added to the growing METS file. The fourth step shown in Figure 3, allows the user to perform quality control on the images, image correction, and enter the structural metadata about their item. All of the page images

PAGE 7

are shown together as thumbnails. At this poin t, the user can visually inspect each image The toolkit allows the users to switch the order of the pages, or delete incorrect images. The user can also name each page and create a table of contents for the resource by entering division information. Ma and provide the actual name of a chapter. All of the entered information is collected in the METS file and will be used to provide an online table of contents in the final web ve rsion. If an institution chooses not to enter page names, a default is used, numbering from Page 1 through all of the pages. Likewise, an institution may choose not to add the division information for an item. In that case, no table of contents will app ear online for the item. Figure 3: Performing quality control on the resource.

PAGE 8

Each image can be edited inside the toolkit as shown in Figure 4. This includes rotat ing the image cropping the image and adjusting the color levels and balance Once e diting is complete, the changes are applied to the archival TIFF a nd all of the image derivatives are recreated. Figure 4: Editing images in the dLOC Toolkit. Digital Resource Creation Submission The final step submits the item to an online library v ia FTP. The user can enter their own FTP information, including server name and username and password. If a dLOC License Code has

PAGE 9

been provided to unlock the dLOC specific fearures it will automatically be sent to the dLOC servers. Digital Resource C reation Customization Through the course of this work, another unexpected user base was discovered As scholars in other fields began to hear about the dLOC capacity building model and submission toolkit, it became apparent that there was a wider need f or a simple, standard based, offline digital resource toolkit. Already several other projects have begun to employ the digitization toolkit which was originally created for dLOC toolkit. This allows t he interface to be customized for each different project. (See Figure 5 for the view of an additional theme.) In addition, the user can select the language for the toolkit. English, Spanish, and French are currently supported, with hopes to add Dutch, H aitian Creole, and Papiamento in the near future. Figure 5: Theme customization in the dLOC Toolkit.

PAGE 10

Digital Resource Creation Usability As the developer took part in the training which was provided to the partners, the partners in turn provided a v ery useful metric back to the developer. This unusual amount of developer user interaction helped to steer the development of the toolkit. In fact, the initial release of the toolkit was much less user friendly and often left partners confused on their n ext step After about two years of use and training, a new toolkit was developed and released. Subsequent training on the new toolkit proved the increased level of usability and ease. This feedback while short of an actual supervised usability study, w as critical to the success of the toolkit and subsequently the project. Digital Resource Creation Future Enhancements Work continues on the toolkit with new changes and releases expected shortly. The new version of the toolkit will add more interoperab ility with the MARC21 standard, allowing import and export of records in MARC XML and MARC21 formats. In addition, the C# source code is targeted to be released under the GNU license by summer of 2009. The toolkit is currently available freely for downlo ad at the dLOC project web site ( http://www. dLOC .com ). dLOC Web Interface The dLOC web interface (shown in Figure 6) developed simultaneously and in conjunction with the metadata toolkit, based on the requirements of the grant, needs of the users, and continued input from usability studies. The foundation for the dLOC interface i s the technology employed by the University of Florida Digital Collections (UFDC) which is a METS based presentation layer over a Greenstone core. This allow s dLOC to take advantage of all the work put into testing UFDC and adds to the general sustainability of the dLOC web interface. Attempts to improve the overall delivery system for dLOC focused on two groups: partners and users. The par tners needs were generally identified as branding. For the users, several usability studies were performed. From discussions with partners it was obvious that they were willing to contribute their time and resources to a collaborative digitization eff among all the others. While dLOC already had branding in the form of wordmarks when viewing an individual resource concepts were added to the UFDC technology. Interfaces were added to allow the web page to

PAGE 11

Institutional pages were also added, to allow browsing and searching of o nly those materials added by a particular institution. The combination of institutional pages and interfaces allow s dLOC to present very different appearances to the same collection of items (See Figure 7 for an example of an interface on an institution al page.) This ability i s very attractive to both existing partners and potential partners. In fact, dLOC itself is an interface, which allows all of the dLOC pages to be branded with Digital Library of the Caribbean, despite existing in the same server as the general University of Florida collections. Figure 6: Digital Library of the Caribbean main web page.

PAGE 12

Figure 7: Institutional page with an interface to mimic a different web presence. Usability testing Researchers have access to a plethora o f content online from which to search. Some users may not understand the importance of using library provided resources and instead may choose to use a web site solely based on how easy it is to search and navigate. Considering the competition for getting users to return to a given online resource, institutions that produce digital libraries such as UF must develop resources that are user centered and are easy to use and navigate as well. Developers consider user expectations in their design process to c reate a user centered and appealing resource to which a user will return. Usability testing is a method to identify expectations and has had a role in design and implementation of computers, interfaces and other technology products for decades (Nielsen, 1 5). According to the U.S. Department of Health & Human Services Usability.gov site, usability refers to how well users can learn and use a product to achieve their goals and how satisfied they are with that process; usability means that people who use the product can do so quickly and easily to accomplish their tasks (Dumas

PAGE 13

and Redish 885). Usability measures the quality of a user's experience when interacting with a product or system, a website or software application e used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction 11). Overall, usability refers to a resource that: Is easy to learn intuitive where basic tasks can be accompli shed soon after seeing the resource for the first time Has efficiency of use how fast can experienced users accomplish desired tasks Is easy to remember the resource can be used efficiently the next time a user sees it Is satisfying to use Meets user and Because usability testing encourages users to take ownership of the resource and to continue to use it and share it, UF has engaged in testing since 2005 throughout the development or customization of ele ctronic resources. From February 2007 to April 2008, various collections within the UFDC framework were tested, with findings discussed in two separate usability test reports (Ochoa 2007, 2008). The tests sought to examine how well the UFDC organized its 30 collections and how the UFDC compared to similar resources in terms of usability. Testing conducted by a public services librarian at UF with experience in usability testing (usability consultant) occurred before the full launch of UFDC in April 2007 and again as part of the iterative design process to improve UFDC and its subcollections. During this year of testing, key areas of concern identified prior to testing included: Are the UFDC search pages all collections, basic and advance intuitive? D o the result pages provide the type of information a participant needs? Are the item viewer navigation functions appropriate to ensure easy navigation and ease of use? Further, the usability consultant ensured that general navigability and features or func tionality resulting from use of certain formats beyond text were tested. Examples are images and serials found in the Baldwin Library of Historical Children's Literature and the Florida Digital Newspaper Library Along with these considerations, identif ying the intended use of UFDC to be for research purposes, along with the type of user were essential in defining the type of the testing to be

PAGE 14

done. Since this digital library was useful for discovering hidden research resources the potential users of t his resource include researchers, university scholars and the everyday public ; t herefore, testing university participants as a representative sample of the user population, using multiple testing methods would be appropriate to identify the issues of UFDC Once testing was determined to require direct contact with end users, a human testing protocol was issued prior to recruitment/scheduling of test participants and testing. At the University of Florida the test protocol includes an informed consent and all survey instruments, including pre and post test questionnaires. Scenario based and focus group testing were used in UFDC testing. Because even minor interface details can affect the way a participant uses the resource, using a very narrow approach, i.e. scenario based tasks to target a specific feature of the resource, was effective to identify how to improve the resource; each step, response and behavior was carefully noted and recorded by a note taker. Having the participant talk aloud during the testing provided additional information about both how and why users complete tasks in the manner they do. This type of testing provided the added benefit for developers to learn more about the thought process behind using the interface and to identify w hat is not intuitive about a particular feature. In addition, focus groups were used to obtain supplemental data about errors in completing tasks and other overall user impressions. The data collected through focus group sessions and through individual s cenario based sessions demonstrated how well the UFDC matched participant needs and expectations. Patterns in the common mistakes towards completing tasks, noted user perspectives and comments, and conclusions drawn from overall behavior were considered w hen listing the problems for future redesign. Overall, the feedback from testing resulted in the change to the interface, result pages and item viewer to improve navigation and retrieval. Some modifications throughout the 2007 to 2008 period included: In troductory/explanatory information on the resource as a whole and individual subcollections Better navigation including linking back to the result lists More intuitive design for selecting collections for a federated search Ease of access to resources fro m Result pages especially for serials More visible and more intuitive functions of the Item viewer navigation

PAGE 15

Usability Testing of dLOC In early Summer 2008, dLOC was tested as a part of the broader iterative design process for UFDC. While dLOC benefited from previous UFDC usability testing and subsequent modification, the uniqueness of dLOC as a multi lingual resource and the requirements of the grant through which dLOC was funded, warranted testing of the subcollection. The testing strategy for dLOC incorporated past testing experiences but the development of the plan also relied on discussions with the dLOC Coordinator at Florida International University. ting: To consider the multi site interface for the Spanish, French and English interfaces of dLOC. To justify additional costs for translation To identify needed interface problems, including use of certain terminology To optimize already implemented f eatures To enable marketing and familiarity with dLOC among Caribbean partners and potential partners In addition to identifying these goals, the intended use of dLOC and its audience were necessary to plan the strategy for dLOC testing. dLOC testing req uired direct contact with researchers, university scholars and the everyday public in the US and in the Caribbean and circum Caribbean. The loosely formed usability team (dLOC Coordinator and usability consultant/librarian) felt university participants in the U S alone would only provide one perspective to identify issues and improve dLOC; usability data collected from dLOC partners would be useful. The team identified factors of location, language, internet connectivity and staffing as possible complica tions to the usability testing of dLOC. While general UFDC usability testing was conducted cheaply and easily due to testing location on the UF campus, potential onsite users and equipment readily available the dLOC testing proved more challenging. The a ssumed heavy use by Caribbean and circum Caribbean researchers forced the team to look outside of UF and FIU for test participants. This audience would require locating and contacting users who would use the content and who would likely be non English or multilingual speakers with unpredictable internet availability. Based on these considerations possible options for testing included:

PAGE 16

Incorporating usability training and/or testing across the Caribbean during already planned UFDC Toolkit training so that each partner location can conduct usability testing Conducting usability testing at an event which already br ought dLOC partners and Caribbean library staff together and asking for usability testing assistance The first option was not selected because the Toolkit training session agenda was already extensive. No dedicated usability specialist was hired to travel and provide training or testing at regular intervals within the Caribbean; the UF librarian who worked on usability testing did so as a small percentage of her position. Further, asking facilitators at each location to test was thrown out at this stage; few partners were familiar with usability testing and would first need to experience it themselves and be trained before trying to conduct the test. Plus, the dLOC coordinator would need time and resources to negotiate this new role/commitment by partners. The second option of testing at an event was selected with evaluation of dLOC conducted at the annual Association of Caribbean University, Research and Institutional Libraries (ACURIL), XXXVIII Conference, Montego Bay, Jamaica, in June 2008. Because testing requires direct contact with end users this location/event was useful because it allowed the dLOC coordinator, designer, and usability consultant an opportunity to work with a variety of actual users and partners of dLOC, without additional cost of travel to anyone. The session allowed users to share their experiences, comments and attitudes about dLOC and to identify problems with using the resource. It encouraged a sense of goodwill and ownership for the resource by Caribbean partners, ultimately resulting in more use and a better reputation for dLOC and UFDC as a whole. The research methodology of dLOC testing involved scenario based testing and a focus group session to evaluate online participant search behavior and attitude. Two sets of test questions were used. All task questions were given in writing to aid in preventing misinterpretation. Participants completed pre and post test questionnaires to identify user experience using online library resources and satisfaction with using dLOC. Then all participants participated in an open group talk aloud session of dLOC which consisted of 3 questions. All attendees worked together i n this session to complete the structured exercises. Participants were then split into groups to answer a set of questions focusing on basic and advanced search interfaces multi lingual interfaces result pages and item viewer; these examples and guided exercises

PAGE 17

were based on real content and reference questions directed to UF and FIU about dLOC. During the group test session, the facilitators (dLOC coordinator and usability consultant) asked participants to talk out loud about their process to complete the question and what they Responses were written down. Finally, the focus group consisted of a discussion on experience with dLOC, guided by the post test questionnai re submitted. Overall onsite testing was a useful experiment with usability testing for multilingual users. It identified problems with terminology, use of images or icons for navigation, and the use of a tabbed approached to reach various views of the resource found. Future of dLOC Usability Usability testing will continue to be a critical part of the design process. Having the input of Caribbean partners and their users can offer some new and different perspectives from past UFDC usability testers. In addition, having assistance from partners to join the dLOC usability team was extended for possible multi site assessment to further guide the development for dLOC. Since the workshop provided the theory and application of usability testing partners have been given a framework for evaluating electronic resources in general. By evaluating dLOC themselves, participants would already be familiar with the testing methodology for dLOC. The future for dLOC usability stems from an idea that a UF or FIU le d dLOC Assessment program can be created to centrally develop questionnaires and test instruments for the resource and to provide training. C hallenges of off site usability testing remain however : how to provide training of usability testing for partner s site limitations (including internet connectivity and staffing) and development of a single, overarching human testing protocol. About the Authors Mark Sullivan works as a developer for the Systems Department at the University of Florida Libraries supporting the Digital Library Center. He has been integral to several recent grants at ( http://www.uflib.ufl.edu/ufdc/ ) He j oined the dLOC project early and has taken part in every training session since its inception.

PAGE 18

Marilyn N. Ochoa is currently the Assistant Head of the Education Library at the University of Florida. Prior to this position, she served as the digital servi ces librarian for the UF Humanities and Social Sciences Services Department. References Dumas, Joseph and Janice (Ginny) Redish. 2002. Usability in practice: formative usability evaluations ev olution and revolution Conference on Human Factors in Computing Systems Minneapolis, Minnesota, USA 885 890 Usability in Practice Session. display terminals (VDTs ) -11:1998, March 13, 1998 (February 2, 2005). centred design processes for i nteractive ISO 13407:1999, July 1999 (February 2, 2005). Krug, Steve. 2006. Don't make me think!: a common sense approach to Web usability. Berkeley, Calif : New Riders Pub. Nielsen, Jakob. 1993. Usability engineering. Boston: Academic Press. (February 16, 2005). Ochoa, Marilyn. 2008. Usability re port for the University of Florida Digital Collections (UFDC). Gainesville, Fla.: George A. Smathers Libraries, University of Florida < http://www.uflib.ufl.edu/ufdc/?b=UF00073771&v=00001 > O choa, Marilyn, Usability Test Report for Florida Digital Newspaper Library, UFDC. Gainesville, Fla.: George A. Smathers Libraries, University of Florida < http://www.uflib.ufl.edu/ufdc/?b=U F00083039&v=00001 >. Association of Caribbean University, Research and Institutional Libraries, XXXVIII Conference, Centered Approach to with Brooke Wooldrige. Association of Caribbean University, Research and Institutional Libraries, XXXVIII Conference, Montego Bay, Jamaica, IT Increase Effectiveness of Library Electronic 2008.